├── .gitignore ├── save-sd18 ├── fonts │ ├── cmunbi.otf │ ├── cmunbx.otf │ ├── cmunrm.otf │ ├── cmunsi.otf │ ├── cmunss.otf │ ├── cmunsx.otf │ ├── cmunti.otf │ ├── cmuntt.otf │ └── cmunbxo.otf ├── img │ ├── ML-Schema.png │ ├── metalevel.png │ └── DMOPLogisticRegression.png ├── css │ ├── rash.css │ └── lncs.css ├── bib │ └── biblio.bib └── index.html ├── README.md ├── MLSchemaExample2.ttl ├── MLSchemaSandbox.ttl ├── MLSchemaExample1.ttl ├── dmop-dmexperiment-iris.ttl ├── AlgorithmImplementationExecution.owl └── MLSchema.ttl /.gitignore: -------------------------------------------------------------------------------- 1 | .DS_Store 2 | *~ 3 | 4 | catalog-*.xml -------------------------------------------------------------------------------- /save-sd18/fonts/cmunbi.otf: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ML-Schema/core/HEAD/save-sd18/fonts/cmunbi.otf -------------------------------------------------------------------------------- /save-sd18/fonts/cmunbx.otf: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ML-Schema/core/HEAD/save-sd18/fonts/cmunbx.otf -------------------------------------------------------------------------------- /save-sd18/fonts/cmunrm.otf: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ML-Schema/core/HEAD/save-sd18/fonts/cmunrm.otf -------------------------------------------------------------------------------- /save-sd18/fonts/cmunsi.otf: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ML-Schema/core/HEAD/save-sd18/fonts/cmunsi.otf -------------------------------------------------------------------------------- /save-sd18/fonts/cmunss.otf: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ML-Schema/core/HEAD/save-sd18/fonts/cmunss.otf -------------------------------------------------------------------------------- /save-sd18/fonts/cmunsx.otf: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ML-Schema/core/HEAD/save-sd18/fonts/cmunsx.otf -------------------------------------------------------------------------------- /save-sd18/fonts/cmunti.otf: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ML-Schema/core/HEAD/save-sd18/fonts/cmunti.otf -------------------------------------------------------------------------------- /save-sd18/fonts/cmuntt.otf: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ML-Schema/core/HEAD/save-sd18/fonts/cmuntt.otf -------------------------------------------------------------------------------- /save-sd18/fonts/cmunbxo.otf: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ML-Schema/core/HEAD/save-sd18/fonts/cmunbxo.otf -------------------------------------------------------------------------------- /save-sd18/img/ML-Schema.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ML-Schema/core/HEAD/save-sd18/img/ML-Schema.png -------------------------------------------------------------------------------- /save-sd18/img/metalevel.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ML-Schema/core/HEAD/save-sd18/img/metalevel.png -------------------------------------------------------------------------------- /save-sd18/img/DMOPLogisticRegression.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ML-Schema/core/HEAD/save-sd18/img/DMOPLogisticRegression.png -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # ML Schema 2 | 3 | ML-Schema is a collaborative, community effort with a mission to develop, maintain, and promote standard schemas for data mining and machine learning algorithms, datasets, and experiment 4 | 5 | See the wiki for more details: https://github.com/ML-Schema/core/wiki 6 | 7 | **Guidelines** 8 | * Open GitHub issues for any kind of discussion. Selected items will be discussed during telco's. For instance, see how this works for schema.org 9 | * Store all schema in http://www.w3.org/TR/turtle/. This you can load in Protege, Jena,... if you want to edit them locally. 10 | * Use the GitHub Wiki for information relevant to the group and users: https://github.com/ML-Schema/core/wiki 11 | * Include structure (inheritance, value bounds) so arrive at a non-ambiguous schema. 12 | * We use namespace 'mls' and will register this at PURL asap. 13 | * If there is an issue to discuss, first create a GitHub issue, and then label it as 'discuss in call' 14 | Conference calls 15 | 16 | **Conference calls** 17 | We do a conference call on Hangout every two weeks, on Monday 13:30 - 14:30 (European time zone) 18 | * Next calls 2015: Nov 9, Nov 23, Dec 7, Dec 21 19 | * Next calls 2016: Jan 4, Jan 18, Feb 1, ... 20 | 21 | **Google group** 22 | To receive updates about ML Schema, or contact everybody, please join our Google group: https://groups.google.com/forum/#!forum/mlw3c 23 | 24 | When you join, please send a short introduction of yourself, e.g. who are you, what is your interest in ML schema, what would you like to contribute,... 25 | -------------------------------------------------------------------------------- /MLSchemaExample2.ttl: -------------------------------------------------------------------------------- 1 | @prefix : . 2 | @prefix dc: . 3 | @prefix owl: . 4 | @prefix rdf: . 5 | @prefix xml: . 6 | @prefix xsd: . 7 | @prefix rdfs: . 8 | @base . 9 | 10 | rdf:type owl:Ontology ; 11 | 12 | owl:versionIRI ; 13 | 14 | owl:imports , 15 | . 16 | 17 | 18 | ################################################################# 19 | # 20 | # Annotation properties 21 | # 22 | ################################################################# 23 | 24 | 25 | ### http://purl.org/dc/elements/1.1/dateSubmitted 26 | 27 | dc:dateSubmitted rdf:type owl:AnnotationProperty . 28 | 29 | 30 | 31 | ### http://purl.org/dc/elements/1.1/licence 32 | 33 | dc:licence rdf:type owl:AnnotationProperty . 34 | 35 | 36 | 37 | ### http://purl.org/dc/elements/1.1/source 38 | 39 | dc:source rdf:type owl:AnnotationProperty . 40 | 41 | 42 | 43 | 44 | 45 | ################################################################# 46 | # 47 | # Classes 48 | # 49 | ################################################################# 50 | 51 | 52 | ### http://purl.org/dc/elements/1.1/BibliographicResource 53 | 54 | dc:BibliographicResource rdf:type owl:Class . 55 | 56 | 57 | 58 | ### http://purl.org/dc/elements/1.1/LicenceDocument 59 | 60 | dc:LicenceDocument rdf:type owl:Class . 61 | 62 | 63 | 64 | ### http://www.w3.org/ns/mls#Study 65 | 66 | rdfs:subClassOf . 67 | 68 | 69 | 70 | 71 | 72 | ################################################################# 73 | # 74 | # Individuals 75 | # 76 | ################################################################# 77 | 78 | 79 | ### http://example.org/MLSchemaExample2#CC_by_3.0 80 | 81 | :CC_by_3.0 rdf:type dc:LicenceDocument , 82 | owl:NamedIndividual . 83 | 84 | 85 | 86 | ### http://example.org/MLSchemaExample2#EPSRC 87 | 88 | :EPSRC rdf:type , 89 | owl:NamedIndividual ; 90 | 91 | :grant1 . 92 | 93 | 94 | 95 | ### http://example.org/MLSchemaExample2#article1 96 | 97 | :article1 rdf:type dc:BibliographicResource , 98 | owl:NamedIndividual ; 99 | 100 | rdfs:label "Article: Multi Task Learning with a Natural Metric for Quantitative Structure Activity Relationship Learning" . 101 | 102 | 103 | 104 | ### http://example.org/MLSchemaExample2#grant1 105 | 106 | :grant1 rdf:type , 107 | owl:NamedIndividual ; 108 | 109 | "EP/K030582/1" ; 110 | 111 | :meta-qsar_project . 112 | 113 | 114 | 115 | ### http://example.org/MLSchemaExample2#meta-qsar_project 116 | 117 | :meta-qsar_project rdf:type owl:NamedIndividual , 118 | . 119 | 120 | 121 | 122 | ### http://example.org/MLSchemaExample2#mtl_dataset 123 | 124 | :mtl_dataset rdf:type owl:NamedIndividual , 125 | ; 126 | 127 | dc:source "http://www.openml.org/s/3" ; 128 | 129 | dc:dateSubmitted "06/09/15" ; 130 | 131 | dc:licence :CC_by_3.0 . 132 | 133 | 134 | 135 | ### http://example.org/MLSchemaExample2#study1 136 | 137 | :study1 rdf:type owl:NamedIndividual , 138 | ; 139 | 140 | :meta-qsar_project ; 141 | 142 | :mtl_dataset . 143 | 144 | 145 | 146 | 147 | ### Generated by the OWL API (version 3.5.1) http://owlapi.sourceforge.net 148 | 149 | -------------------------------------------------------------------------------- /save-sd18/css/rash.css: -------------------------------------------------------------------------------- 1 | /* 2 | rash.css - Version 0.5, December 28, 2016 3 | by Silvio Peroni 4 | 5 | This work is licensed under a Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/). 6 | You are free to: 7 | * Share - copy and redistribute the material in any medium or format 8 | * Adapt - remix, transform, and build upon the material 9 | for any purpose, even commercially. 10 | 11 | The licensor cannot revoke these freedoms as long as you follow the license terms. 12 | 13 | Under the following terms: 14 | * Attribution - You must give appropriate credit, provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use. 15 | */ 16 | 17 | @namespace mathml url(http://www.w3.org/1998/Math/MathML); 18 | 19 | /* Font */ 20 | body { font-size: 12pt; } 21 | /* /END Font */ 22 | 23 | /* Footnote counters */ 24 | body { counter-reset: fn_pointer; } 25 | section[class=doc-endnotes] { counter-reset: fn; } 26 | section[class=doc-endnotes] > section[class=doc-endnote]:before { 27 | counter-increment: fn; 28 | content: counter(fn); 29 | float: left; 30 | padding-right: 10px; 31 | } 32 | /* /END Footnote counters */ 33 | 34 | /* Figure, table and formula */ 35 | img { 36 | max-width:90%; 37 | height:auto; 38 | } 39 | 40 | table { 41 | max-width:90%; 42 | } 43 | 44 | figure { 45 | margin-top: 2em; 46 | margin-bottom: 2em; 47 | text-align:center; 48 | } 49 | 50 | figure > p:not(mathml|math) { 51 | margin-bottom:0px; 52 | } 53 | 54 | figure img:not([class=math]), figure svg { 55 | border: 1px solid black; 56 | } 57 | 58 | figure > pre { 59 | text-align:left; 60 | } 61 | 62 | figcaption { 63 | margin-top:5px; 64 | } 65 | 66 | table { 67 | margin: 0 auto; 68 | } 69 | 70 | td , th { 71 | border:1px solid black; 72 | padding:5px; 73 | } 74 | 75 | td p { 76 | margin-bottom:0px; 77 | } 78 | 79 | th { 80 | background-color: #606060; 81 | color: white; 82 | text-align:center; 83 | } 84 | 85 | .rash-math div { 86 | display: inline !important; 87 | } 88 | /* /END Figure, table and formula */ 89 | 90 | /* Heading counters */ 91 | body { counter-reset: h1; } 92 | h1 { counter-reset: h2; } 93 | h2 { counter-reset: h3; } 94 | h3 { counter-reset: h4; } 95 | h4 { counter-reset: h5; } 96 | h5 { counter-reset: h6; } 97 | 98 | section:not([class=doc-abstract]):not([class=doc-bibliography]):not([class=doc-acknowledgements]):not([class=doc-endnotes]) h1:before { 99 | counter-increment: h1; 100 | content: counter(h1) ". "; 101 | } 102 | 103 | section:not([class=doc-abstract]):not([class=doc-bibliography]):not([class=doc-acknowledgements]):not([class=doc-endnotes]) h2:before{ 104 | counter-increment: h2; 105 | content: counter(h1) "." counter(h2) ". "; 106 | } 107 | 108 | section:not([class=doc-abstract]):not([class=doc-bibliography]):not([class=doc-acknowledgements]):not([class=doc-endnotes]) h3:before{ 109 | counter-increment: h3; 110 | content: counter(h1) "." counter(h2) "." counter(h3) ". "; 111 | } 112 | 113 | section:not([class=doc-abstract]):not([class=doc-bibliography]):not([class=doc-acknowledgements]):not([class=doc-endnotes]) h4:before{ 114 | counter-increment: h4; 115 | content: counter(h1) "." counter(h2) "." counter(h3) "." counter(h4) ". "; 116 | } 117 | 118 | section:not([class=doc-abstract]):not([class=doc-bibliography]):not([class=doc-acknowledgements]):not([class=doc-endnotes]) h5:before{ 119 | counter-increment: h5; 120 | content: counter(h1) "." counter(h2) "." counter(h3) "." counter(h4) "." counter(h5) ". "; 121 | } 122 | 123 | section:not([class=doc-abstract]):not([class=doc-bibliography]):not([class=doc-acknowledgements]):not([class=doc-endnotes]) h6:before{ 124 | counter-increment: h6; 125 | content: counter(h1) "." counter(h2) "." counter(h3) "." counter(h4) "." counter(h5) "." counter(h6) ". "; 126 | } 127 | /* /END Heading counters */ 128 | 129 | /* Header */ 130 | .title { 131 | font-size: 300%; 132 | } 133 | /* /END Header */ 134 | 135 | /* Footer */ 136 | html { 137 | position: relative; 138 | min-height: 100%; 139 | } 140 | 141 | body { 142 | /* Margin bottom by footer height */ 143 | margin-bottom: 60px; 144 | } 145 | 146 | .footer { 147 | position: fixed; 148 | bottom: 0; 149 | width: 100%; 150 | /* Set the fixed height of the footer here */ 151 | height: 60px; 152 | background-color: #f5f5f5; 153 | padding: 20px; 154 | } 155 | 156 | .footer span , .footer div { 157 | margin-left: 30px; 158 | } 159 | 160 | #layoutselection { 161 | margin-left: 0px; 162 | margin-right: 5px; 163 | } 164 | 165 | .footer div { 166 | margin-top:-5px; 167 | } 168 | 169 | .footer p { 170 | display: inline; 171 | } 172 | /* /END Footer */ 173 | 174 | /* Not cited */ 175 | .notcited { 176 | color: red; 177 | cursor:help; 178 | } 179 | /* /END Not cited */ 180 | 181 | /* Sections */ 182 | body > section { 183 | padding-right:15px; 184 | padding-left:15px; 185 | margin-right:auto; 186 | margin-left:auto; 187 | } 188 | 189 | @media (min-width:768px){ 190 | body > section { 191 | width:750px; 192 | } 193 | } 194 | 195 | @media (min-width:992px){ 196 | body > section { 197 | width:970px; 198 | } 199 | } 200 | 201 | @media (min-width:1200px){ 202 | body > section { 203 | width:1170px; 204 | } 205 | } 206 | 207 | body > section:after{ 208 | display:table; 209 | content:" "; 210 | clear:both; 211 | } 212 | /* /END Sections */ 213 | 214 | /* Error */ 215 | .error { 216 | color: red; 217 | font-weight:bold; 218 | } 219 | .error:before { 220 | content: "[["; 221 | } 222 | .error:after { 223 | content: "]]"; 224 | } 225 | /* /END Error */ 226 | 227 | /* Reference list */ 228 | section[class=doc-bibliography] > ul { 229 | list-style-type: decimal; 230 | } 231 | /* /END Reference list */ 232 | 233 | /* ### Rules for printing the document */ 234 | @media print { 235 | @page { 236 | size: A4; 237 | margin: 20mm 0mm 20mm 0mm; 238 | } 239 | 240 | html, body { 241 | width: 210mm; 242 | height: 297mm; 243 | } 244 | 245 | p { 246 | font-size: 10pt; 247 | text-align:justify; 248 | } 249 | 250 | pre { 251 | text-align:left; 252 | } 253 | 254 | figcaption , figure > p { 255 | text-align:center; 256 | } 257 | 258 | .title { 259 | font-size: 16pt !important; 260 | } 261 | 262 | h1 { 263 | font-size: 14pt !important; 264 | } 265 | 266 | h2 { 267 | font-size: 12pt !important; 268 | } 269 | 270 | h3 { 271 | font-size: 11pt !important; 272 | } 273 | 274 | h4, h5, h6 { 275 | font-size: 10pt !important; 276 | } 277 | 278 | pre { 279 | font-size: 8pt !important; 280 | } 281 | 282 | blockquote { 283 | font-size: 10pt !important; 284 | } 285 | 286 | address { 287 | font-size: 12pt !important; 288 | } 289 | 290 | table { 291 | max-width:100%; 292 | font-size: 80%; 293 | } 294 | } -------------------------------------------------------------------------------- /MLSchemaSandbox.ttl: -------------------------------------------------------------------------------- 1 | @prefix : . 2 | @prefix mls: . 3 | @prefix owl: . 4 | @prefix rdf: . 5 | @prefix xml: . 6 | @prefix xsd: . 7 | @prefix rdfs: . 8 | @prefix skos: . 9 | @prefix terms: . 10 | @prefix protege: . 11 | @base . 12 | 13 | rdf:type owl:Ontology ; 14 | 15 | owl:imports . 16 | 17 | 18 | ################################################################# 19 | # 20 | # Individuals 21 | # 22 | ################################################################# 23 | 24 | 25 | ### http://example.org#TenFoldCrossValidation 26 | 27 | :TenFoldCrossValidation rdf:type owl:NamedIndividual , 28 | mls:EvaluationProcedure . 29 | 30 | 31 | 32 | ### http://example.org#classEntropy 33 | 34 | :classEntropy rdf:type owl:NamedIndividual , 35 | mls:DatasetCharacteristic . 36 | 37 | 38 | 39 | ### http://example.org#credit-a 40 | 41 | :credit-a rdf:type owl:NamedIndividual , 42 | mls:Dataset ; 43 | 44 | mls:hasQuality :classEntropy , 45 | :decisionStumpAUC , 46 | :defaultAccuracy , 47 | :j48_001_Stump , 48 | :numberOfClasses , 49 | :numberOfFeatures , 50 | :numberOfInstances , 51 | :numberOfMissingValues . 52 | 53 | 54 | 55 | ### http://example.org#decisionStumpAUC 56 | 57 | :decisionStumpAUC rdf:type owl:NamedIndividual , 58 | mls:DatasetCharacteristic . 59 | 60 | 61 | 62 | ### http://example.org#defaultAccuracy 63 | 64 | :defaultAccuracy rdf:type owl:NamedIndividual , 65 | mls:DatasetCharacteristic . 66 | 67 | 68 | 69 | ### http://example.org#evaluationSpecification1 70 | 71 | :evaluationSpecification1 rdf:type owl:NamedIndividual , 72 | mls:EvaluationSpecification ; 73 | 74 | mls:hasPart :TenFoldCrossValidation , 75 | :predictiveAccuracy . 76 | 77 | 78 | 79 | ### http://example.org#j48_001_Stump 80 | 81 | :j48_001_Stump rdf:type owl:NamedIndividual , 82 | mls:DatasetCharacteristic . 83 | 84 | 85 | 86 | ### http://example.org#logisticRegression 87 | 88 | :logisticRegression rdf:type owl:NamedIndividual , 89 | mls:Algorithm . 90 | 91 | 92 | 93 | ### http://example.org#modelEvaluation100241 94 | 95 | :modelEvaluation100241 rdf:type owl:NamedIndividual , 96 | mls:ModelEvaluation ; 97 | 98 | mls:specifiedBy :predictiveAccuracy ; 99 | 100 | mls:hasValue 0.8478 . 101 | 102 | 103 | 104 | ### http://example.org#numberOfClasses 105 | 106 | :numberOfClasses rdf:type owl:NamedIndividual , 107 | mls:DatasetCharacteristic . 108 | 109 | 110 | 111 | ### http://example.org#numberOfFeatures 112 | 113 | :numberOfFeatures rdf:type owl:NamedIndividual , 114 | mls:DatasetCharacteristic . 115 | 116 | 117 | 118 | ### http://example.org#numberOfInstances 119 | 120 | :numberOfInstances rdf:type owl:NamedIndividual , 121 | mls:DatasetCharacteristic . 122 | 123 | 124 | 125 | ### http://example.org#numberOfMissingValues 126 | 127 | :numberOfMissingValues rdf:type owl:NamedIndividual , 128 | mls:DatasetCharacteristic . 129 | 130 | 131 | 132 | ### http://example.org#predictiveAccuracy 133 | 134 | :predictiveAccuracy rdf:type owl:NamedIndividual , 135 | mls:EvaluationMeasure . 136 | 137 | 138 | 139 | ### http://example.org#run100241 140 | 141 | :run100241 rdf:type owl:NamedIndividual , 142 | mls:Run ; 143 | 144 | mls:executes :wekaLogistic ; 145 | 146 | mls:hasInput :credit-a ; 147 | 148 | mls:hasOutput :modelEvaluation100241 , 149 | :wekaLogisticModel100241 ; 150 | 151 | mls:realizes :task29 . 152 | 153 | 154 | 155 | ### http://example.org#task29 156 | 157 | :task29 rdf:type owl:NamedIndividual , 158 | mls:Task ; 159 | 160 | mls:definedOn :credit-a , 161 | :evaluationSpecification1 . 162 | 163 | 164 | 165 | ### http://example.org#wekaLogistic 166 | 167 | :wekaLogistic rdf:type owl:NamedIndividual , 168 | mls:Implementation ; 169 | 170 | mls:hasHyperParameter :wekaLogisticC , 171 | :wekaLogisticDoNotCheckCapabilities , 172 | :wekaLogisticM , 173 | :wekaLogisticOutputDebugInfo , 174 | :wekaLogisticR ; 175 | 176 | mls:implements :logisticRegression . 177 | 178 | 179 | 180 | ### http://example.org#wekaLogisticC 181 | 182 | :wekaLogisticC rdf:type owl:NamedIndividual , 183 | mls:HyperParameter . 184 | 185 | 186 | 187 | ### http://example.org#wekaLogisticDoNotCheckCapabilities 188 | 189 | :wekaLogisticDoNotCheckCapabilities rdf:type owl:NamedIndividual , 190 | mls:HyperParameter . 191 | 192 | 193 | 194 | ### http://example.org#wekaLogisticM 195 | 196 | :wekaLogisticM rdf:type owl:NamedIndividual , 197 | mls:HyperParameter . 198 | 199 | 200 | 201 | ### http://example.org#wekaLogisticMSetting29 202 | 203 | :wekaLogisticMSetting29 rdf:type owl:NamedIndividual , 204 | mls:HyperParameterSetting ; 205 | 206 | mls:specifiedBy :wekaLogisticM ; 207 | 208 | mls:hasValue -1 . 209 | 210 | 211 | 212 | ### http://example.org#wekaLogisticModel100241 213 | 214 | :wekaLogisticModel100241 rdf:type owl:NamedIndividual , 215 | mls:Model . 216 | 217 | 218 | 219 | ### http://example.org#wekaLogisticOutputDebugInfo 220 | 221 | :wekaLogisticOutputDebugInfo rdf:type owl:NamedIndividual , 222 | mls:HyperParameter . 223 | 224 | 225 | 226 | ### http://example.org#wekaLogisticR 227 | 228 | :wekaLogisticR rdf:type owl:NamedIndividual , 229 | mls:HyperParameter . 230 | 231 | 232 | 233 | ### http://example.org#wekaLogisticRSetting29 234 | 235 | :wekaLogisticRSetting29 rdf:type owl:NamedIndividual , 236 | mls:HyperParameterSetting ; 237 | 238 | mls:specifiedBy :wekaLogisticR ; 239 | 240 | mls:hasValue "1.0E-8"^^xsd:float . 241 | 242 | 243 | 244 | 245 | ### Generated by the OWL API (version 4.1.3.20151118-2017) https://github.com/owlcs/owlapi 246 | 247 | -------------------------------------------------------------------------------- /MLSchemaExample1.ttl: -------------------------------------------------------------------------------- 1 | @prefix : . 2 | @prefix mls: . 3 | @prefix owl: . 4 | @prefix rdf: . 5 | @prefix xml: . 6 | @prefix xsd: . 7 | @prefix rdfs: . 8 | @prefix skos: . 9 | @prefix terms: . 10 | @prefix protege: . 11 | @base . 12 | 13 | rdf:type owl:Ontology ; 14 | 15 | owl:versionIRI ; 16 | 17 | owl:imports . 18 | 19 | 20 | ################################################################# 21 | # 22 | # Individuals 23 | # 24 | ################################################################# 25 | 26 | 27 | ### http://example.org/MLSchemaExample1#Weka 28 | 29 | :Weka rdf:type owl:NamedIndividual , 30 | mls:Software ; 31 | 32 | mls:hasPart :wekaLogistic . 33 | 34 | 35 | 36 | ### http://example.org/MLSchemaExample1#credit-a 37 | 38 | :credit-a rdf:type owl:NamedIndividual , 39 | mls:Dataset ; 40 | 41 | mls:hasQuality :defaultAccuracy , 42 | :numberOfFeatures , 43 | :numberOfInstances . 44 | 45 | 46 | 47 | ### http://example.org/MLSchemaExample1#defaultAccuracy 48 | 49 | :defaultAccuracy rdf:type owl:NamedIndividual , 50 | mls:DatasetCharacteristic ; 51 | 52 | mls:hasValue "0.56"^^xsd:float . 53 | 54 | 55 | 56 | ### http://example.org/MLSchemaExample1#evaluationSpecification1 57 | 58 | :evaluationSpecification1 rdf:type owl:NamedIndividual , 59 | mls:EvaluationSpecification ; 60 | 61 | mls:hasPart :predictiveAccuracy ; 62 | 63 | mls:defines :task29 ; 64 | 65 | mls:hasPart :tenFoldCrossValidation . 66 | 67 | 68 | 69 | ### http://example.org/MLSchemaExample1#logisticRegression 70 | 71 | :logisticRegression rdf:type owl:NamedIndividual , 72 | mls:Algorithm . 73 | 74 | 75 | 76 | ### http://example.org/MLSchemaExample1#modelEvaluation100241 77 | 78 | :modelEvaluation100241 rdf:type owl:NamedIndividual , 79 | mls:ModelEvaluation ; 80 | 81 | mls:hasValue 0.8478 ; 82 | 83 | mls:specifiedBy :predictiveAccuracy . 84 | 85 | 86 | 87 | ### http://example.org/MLSchemaExample1#numberOfFeatures 88 | 89 | :numberOfFeatures rdf:type owl:NamedIndividual , 90 | mls:DatasetCharacteristic ; 91 | 92 | mls:hasValue "16"^^xsd:long . 93 | 94 | 95 | 96 | ### http://example.org/MLSchemaExample1#numberOfInstances 97 | 98 | :numberOfInstances rdf:type owl:NamedIndividual , 99 | mls:DatasetCharacteristic ; 100 | 101 | mls:hasValue "690"^^xsd:long . 102 | 103 | 104 | 105 | ### http://example.org/MLSchemaExample1#predictiveAccuracy 106 | 107 | :predictiveAccuracy rdf:type owl:NamedIndividual , 108 | mls:EvaluationMeasure . 109 | 110 | 111 | 112 | ### http://example.org/MLSchemaExample1#run100241 113 | 114 | :run100241 rdf:type owl:NamedIndividual , 115 | mls:Run ; 116 | 117 | mls:hasInput :credit-a ; 118 | 119 | mls:realizes :logisticRegression ; 120 | 121 | mls:hasOutput :modelEvaluation100241 ; 122 | 123 | mls:achieves :task29 ; 124 | 125 | mls:executes :wekaLogistic ; 126 | 127 | mls:hasInput :wekaLogisticMSetting29 ; 128 | 129 | mls:hasOutput :wekaLogisticModel100241 ; 130 | 131 | mls:hasInput :wekaLogisticRSetting29 . 132 | 133 | 134 | 135 | ### http://example.org/MLSchemaExample1#task29 136 | 137 | :task29 rdf:type owl:NamedIndividual , 138 | mls:Task ; 139 | 140 | mls:definedOn :credit-a , 141 | :evaluationSpecification1 . 142 | 143 | 144 | 145 | ### http://example.org/MLSchemaExample1#tenFoldCrossValidation 146 | 147 | :tenFoldCrossValidation rdf:type owl:NamedIndividual , 148 | mls:EvaluationProcedure . 149 | 150 | 151 | 152 | ### http://example.org/MLSchemaExample1#wekaLogistic 153 | 154 | :wekaLogistic rdf:type owl:NamedIndividual , 155 | mls:Implementation ; 156 | 157 | mls:implements :logisticRegression ; 158 | 159 | mls:hasHyperParameter :wekaLogisticC , 160 | :wekaLogisticDoNotCheckCapabilities , 161 | :wekaLogisticM , 162 | :wekaLogisticOutputDebugInfo , 163 | :wekaLogisticR . 164 | 165 | 166 | 167 | ### http://example.org/MLSchemaExample1#wekaLogisticC 168 | 169 | :wekaLogisticC rdf:type owl:NamedIndividual , 170 | mls:HyperParameter . 171 | 172 | 173 | 174 | ### http://example.org/MLSchemaExample1#wekaLogisticDoNotCheckCapabilities 175 | 176 | :wekaLogisticDoNotCheckCapabilities rdf:type owl:NamedIndividual , 177 | mls:HyperParameter . 178 | 179 | 180 | 181 | ### http://example.org/MLSchemaExample1#wekaLogisticM 182 | 183 | :wekaLogisticM rdf:type owl:NamedIndividual , 184 | mls:HyperParameter . 185 | 186 | 187 | 188 | ### http://example.org/MLSchemaExample1#wekaLogisticMSetting29 189 | 190 | :wekaLogisticMSetting29 rdf:type owl:NamedIndividual , 191 | mls:HyperParameterSetting ; 192 | 193 | mls:hasValue -1 ; 194 | 195 | mls:specifiedBy :wekaLogisticM . 196 | 197 | 198 | 199 | ### http://example.org/MLSchemaExample1#wekaLogisticModel100241 200 | 201 | :wekaLogisticModel100241 rdf:type owl:NamedIndividual , 202 | mls:Model . 203 | 204 | 205 | 206 | ### http://example.org/MLSchemaExample1#wekaLogisticOutputDebugInfo 207 | 208 | :wekaLogisticOutputDebugInfo rdf:type owl:NamedIndividual , 209 | mls:HyperParameter . 210 | 211 | 212 | 213 | ### http://example.org/MLSchemaExample1#wekaLogisticR 214 | 215 | :wekaLogisticR rdf:type owl:NamedIndividual , 216 | mls:HyperParameter . 217 | 218 | 219 | 220 | ### http://example.org/MLSchemaExample1#wekaLogisticRSetting29 221 | 222 | :wekaLogisticRSetting29 rdf:type owl:NamedIndividual , 223 | mls:HyperParameterSetting ; 224 | 225 | mls:hasValue "1.0E-8"^^xsd:float ; 226 | 227 | mls:specifiedBy :wekaLogisticR . 228 | 229 | 230 | 231 | 232 | ### Generated by the OWL API (version 3.5.1) http://owlapi.sourceforge.net 233 | 234 | -------------------------------------------------------------------------------- /save-sd18/bib/biblio.bib: -------------------------------------------------------------------------------- 1 | @TechReport{mlschema:201610, 2 | author = "Diego Esteves and Agnieszka Lawrynowicz and Pance Panov and 3 | Larisa N. Soldatova and Tommaso Soru and Joaquin Vanschoren", 4 | title = "ML Schema Core Specification", 5 | month = oct, 6 | note = "http://www.w3.org/2016/10/mls/", 7 | year = "2016", 8 | institution = "W3C", 9 | } 10 | 11 | 12 | @article{DBLP:journals/ws/KeetLdKNPSH15, 13 | author = {C. Maria Keet and 14 | Agnieszka Lawrynowicz and 15 | Claudia d'Amato and 16 | Alexandros Kalousis and 17 | Phong Nguyen and 18 | Ra{\'{u}}l Palma and 19 | Robert Stevens and 20 | Melanie Hilario}, 21 | title = {The Data Mining OPtimization Ontology}, 22 | journal = {J. Web Sem.}, 23 | volume = {32}, 24 | pages = {43--53}, 25 | year = {2015}, 26 | url = {http://dx.doi.org/10.1016/j.websem.2015.01.001}, 27 | doi = {10.1016/j.websem.2015.01.001}, 28 | timestamp = {Tue, 04 Aug 2015 13:45:56 +0200}, 29 | biburl = {http://dblp.uni-trier.de/rec/bib/journals/ws/KeetLdKNPSH15}, 30 | bibsource = {dblp computer science bibliography, http://dblp.org} 31 | } 32 | 33 | @article{DBLP:journals/datamine/PanovSD14, 34 | author = {Pance Panov and 35 | Larisa N. Soldatova and 36 | Saso Dzeroski}, 37 | title = {Ontology of core data mining entities}, 38 | journal = {Data Min. Knowl. Discov.}, 39 | volume = {28}, 40 | number = {5-6}, 41 | pages = {1222--1265}, 42 | year = {2014}, 43 | url = {http://dx.doi.org/10.1007/s10618-014-0363-0}, 44 | doi = {10.1007/s10618-014-0363-0}, 45 | timestamp = {Tue, 02 Sep 2014 16:25:40 +0200}, 46 | biburl = {http://dblp.uni-trier.de/rec/bib/journals/datamine/PanovSD14}, 47 | bibsource = {dblp computer science bibliography, http://dblp.org} 48 | } 49 | 50 | @article{DBLP:journals/ml/VanschorenBPH12, 51 | author = {Joaquin Vanschoren and 52 | Hendrik Blockeel and 53 | Bernhard Pfahringer and 54 | Geoffrey Holmes}, 55 | title = {Experiment databases - {A} new way to share, organize and learn from 56 | experiments}, 57 | journal = {Machine Learning}, 58 | volume = {87}, 59 | number = {2}, 60 | pages = {127--158}, 61 | year = {2012}, 62 | url = {http://dx.doi.org/10.1007/s10994-011-5277-0}, 63 | doi = {10.1007/s10994-011-5277-0}, 64 | timestamp = {Fri, 20 Apr 2012 14:04:56 +0200}, 65 | biburl = {http://dblp.uni-trier.de/rec/bib/journals/ml/VanschorenBPH12}, 66 | bibsource = {dblp computer science bibliography, http://dblp.org} 67 | } 68 | 69 | @inproceedings{DBLP:conf/i-semantics/EstevesMNSUAL15, 70 | author = {Diego Esteves and 71 | Diego Moussallem and 72 | Ciro Baron Neto and 73 | Tommaso Soru and 74 | Ricardo Usbeck and 75 | Markus Ackermann and 76 | Jens Lehmann}, 77 | title = {{MEX} vocabulary: a lightweight interchange format for machine learning 78 | experiments}, 79 | booktitle = {Proceedings of the 11th International Conference on Semantic Systems, 80 | {SEMANTICS} 2015, Vienna, Austria, September 15-17, 2015}, 81 | pages = {169--176}, 82 | year = {2015}, 83 | url = {http://doi.acm.org/10.1145/2814864.2814883}, 84 | doi = {10.1145/2814864.2814883}, 85 | timestamp = {Wed, 23 Dec 2015 12:57:36 +0100}, 86 | biburl = {http://dblp.uni-trier.de/rec/bib/conf/i-semantics/EstevesMNSUAL15}, 87 | bibsource = {dblp computer science bibliography, http://dblp.org} 88 | } 89 | 90 | @article{lebo2013prov, 91 | title={Prov-o: The prov ontology}, 92 | author={Lebo, Timothy and Sahoo, Satya and McGuinness, Deborah and Belhajjame, Khalid and Cheney, James and Corsar, David and Garijo, Daniel and Soiland-Reyes, Stian and Zednik, Stephan and Zhao, Jun}, 93 | journal={W3C Recommendation}, 94 | volume={30}, 95 | year={2013} 96 | } 97 | @book{Arp:2015:BOB:2846229, 98 | author = {Arp, Robert and Smith, Barry and Spear, Andrew D.}, 99 | title = {Building Ontologies with Basic Formal Ontology}, 100 | year = {2015}, 101 | isbn = {0262527812, 9780262527811}, 102 | publisher = {The MIT Press}, 103 | } 104 | @article{10.1371/journal.pone.0154556, 105 | author = {Bandrowski, Anita AND Brinkman, Ryan AND Brochhausen, Mathias AND Brush, Matthew H. AND Bug, Bill AND Chibucos, Marcus C. AND Clancy, Kevin AND Courtot, Mélanie AND Derom, Dirk AND Dumontier, Michel AND Fan, Liju AND Fostel, Jennifer AND Fragoso, Gilberto AND Gibson, Frank AND Gonzalez-Beltran, Alejandra AND Haendel, Melissa A. AND He, Yongqun AND Heiskanen, Mervi AND Hernandez-Boussard, Tina AND Jensen, Mark AND Lin, Yu AND Lister, Allyson L. AND Lord, Phillip AND Malone, James AND Manduchi, Elisabetta AND McGee, Monnie AND Morrison, Norman AND Overton, James A. AND Parkinson, Helen AND Peters, Bjoern AND Rocca-Serra, Philippe AND Ruttenberg, Alan AND Sansone, Susanna-Assunta AND Scheuermann, Richard H. AND Schober, Daniel AND Smith, Barry AND Soldatova, Larisa N. AND Stoeckert, Jr., Christian J. AND Taylor, Chris F. AND Torniai, Carlo AND Turner, Jessica A. AND Vita, Randi AND Whetzel, Patricia L. AND Zheng, Jie}, 106 | journal = {PLoS ONE}, 107 | publisher = {Public Library of Science}, 108 | title = {The Ontology for Biomedical Investigations}, 109 | year = {2016}, 110 | month = {04}, 111 | volume = {11}, 112 | url = {http://dx.doi.org/10.1371%2Fjournal.pone.0154556}, 113 | pages = {1-19}, 114 | abstract = {

The Ontology for Biomedical Investigations (OBI) is an ontology that provides terms with precisely defined meanings to describe all aspects of how investigations in the biological and medical domains are conducted. OBI re-uses ontologies that provide a representation of biomedical knowledge from the Open Biological and Biomedical Ontologies (OBO) project and adds the ability to describe how this knowledge was derived. We here describe the state of OBI and several applications that are using it, such as adding semantic expressivity to existing databases, building data entry forms, and enabling interoperability between knowledge resources. OBI covers all phases of the investigation process, such as planning, execution and reporting. It represents information and material entities that participate in these processes, as well as roles and functions. Prior to OBI, it was not possible to use a single internally consistent resource that could be applied to multiple types of experiments for these applications. OBI has made this possible by creating terms for entities involved in biological and medical investigations and by importing parts of other biomedical ontologies such as GO, Chemical Entities of Biological Interest (ChEBI) and Phenotype Attribute and Trait Ontology (PATO) without altering their meaning. OBI is being used in a wide range of projects covering genomics, multi-omics, immunology, and catalogs of services. OBI has also spawned other ontologies (Information Artifact Ontology) and methods for importing parts of ontologies (Minimum information to reference an external ontology term (MIREOT)). The OBI project is an open cross-disciplinary collaborative effort, encompassing multiple research communities from around the globe. To date, OBI has created 2366 classes and 40 relations along with textual and formal definitions. The OBI Consortium maintains a web resource (http://obi-ontology.org) providing details on the people, policies, and issues being addressed in association with OBI. The current release of OBI is available at http://purl.obolibrary.org/obo/obi.owl.

}, 115 | number = {4}, 116 | doi = {10.1371/journal.pone.0154556} 117 | } 118 | 119 | 120 | 121 | -------------------------------------------------------------------------------- /save-sd18/css/lncs.css: -------------------------------------------------------------------------------- 1 | /* 2 | lncs.css - Version 0.3, February 17, 2016 3 | by Silvio Peroni 4 | 5 | based on 6 | 7 | lncs.css - Version dated February 2, 2015 8 | by Sarven Capadisli http://csarven.ca/#i 9 | in the context of the Linked Research Project 10 | (https://github.com/csarven/linked-research) 11 | 12 | This work is licensed under a Apache License, Version 2.0 (http://www.apache.org/licenses/LICENSE-2.0.html) 13 | and follows the typographical specifications introduced in the LNCS template styles 14 | (http://www.springer.com/computer/lncs?SGWID=0-164-6-793341-0). 15 | 16 | Computer Modern Unicode fonts are from http://canopus.iacp.dvo.ru/~panov/cm-unicode/ with X11 License. 17 | */ 18 | 19 | /* General settings */ 20 | @font-face{ 21 | font-family:"CMU Serif"; 22 | src:local("CMU Serif"), url("../fonts/cmunrm.otf"); 23 | } 24 | 25 | @font-face{ 26 | font-family:"CMU Serif Bold"; 27 | src:local("CMU Serif Bold"), url("../fonts/cmunbx.otf"); 28 | font-weight:bold; 29 | } 30 | 31 | @font-face{ 32 | font-family:"CMU Serif Italic"; 33 | src:local("CMU Serif Italic"), url("../fonts/cmunti.otf"); 34 | font-style:italic; 35 | } 36 | 37 | @font-face{ 38 | font-family:"CMU Serif Bold Italic"; 39 | src:local("CMU Serif Bold Italic"), url("../fonts/cmunbi.otf"); 40 | font-weight:bold; 41 | font-style:italic; 42 | } 43 | 44 | @font-face{ 45 | font-family:"CMU Serif Monospace"; 46 | src:local("CMU Serif Monospace"), url("../fonts/cmuntt.otf"); 47 | } 48 | 49 | @page { 50 | size:A4; 51 | margin-top: 4.3cm; 52 | margin-bottom: 6.0cm; 53 | } 54 | 55 | body{ 56 | font-size:10pt !important; 57 | margin:0 auto !important; 58 | padding:0 !important; 59 | width:122mm !important; 60 | height:193mm !important; 61 | color:#000000; 62 | font-family:"CMU Serif", serif !important; 63 | position:relative; 64 | margin:0 auto; 65 | padding:2em 0; 66 | line-height:1.231; 67 | } 68 | 69 | section{ 70 | text-align:justify; 71 | } 72 | 73 | *{ 74 | margin:0; 75 | padding:0; 76 | } 77 | /* /END General settings */ 78 | 79 | 80 | /* Inlines */ 81 | strong{ 82 | font-family:"CMU Serif Bold", serif !important; 83 | } 84 | 85 | em{ 86 | font-family:"CMU Serif Italic", serif !important; 87 | } 88 | 89 | h1 em, 90 | h2 em, 91 | h3 em, 92 | strong em, 93 | em strong{ 94 | font-family:"CMU Serif Bold Italic", serif !important; 95 | } 96 | /* /END Inlines */ 97 | 98 | 99 | /* Links */ 100 | a{ 101 | text-decoration:none; 102 | color:#000; 103 | word-break:break-all; 104 | } 105 | /* /END Links */ 106 | 107 | 108 | /* Codes */ 109 | code, 110 | pre{ 111 | font-family:"CMU Serif Monospace", monospace !important; 112 | font-size:8pt; 113 | } 114 | 115 | pre{ 116 | font-size:8pt; 117 | word-break:break-all; 118 | white-space:pre-wrap; 119 | margin:1em 0; 120 | } 121 | /* /END Codes */ 122 | 123 | 124 | /* Paragraphs */ 125 | p{ 126 | text-indent:1em; 127 | } 128 | 129 | section p:first-of-type{ 130 | text-indent:0; 131 | } 132 | /* /END Paragraphs */ 133 | 134 | 135 | /* Title of the article */ 136 | header h1{ 137 | font-size:14pt; 138 | } 139 | /* /END Title of the article */ 140 | 141 | 142 | /* Headings */ 143 | h1{ 144 | text-align:left; 145 | } 146 | 147 | h1, 148 | h2{ 149 | overflow:hidden; 150 | font-family:"CMU Serif Bold", serif !important; 151 | } 152 | 153 | h1{ 154 | font-size:12pt; 155 | margin-top:1.5em; 156 | margin-bottom:0.9em; 157 | } 158 | 159 | h2{ 160 | font-size:10pt; 161 | margin-top:1.35em; 162 | margin-bottom:0.75em; 163 | } 164 | 165 | h3, 166 | h4, 167 | h5, 168 | h6{ 169 | display:inline; 170 | font-size:10pt; 171 | } 172 | 173 | h3:after{ 174 | font-weight:bold; 175 | content:". "; 176 | } 177 | 178 | h4, 179 | h5, 180 | h6{ 181 | font-weight:normal; 182 | } 183 | 184 | h4:after, 185 | h5:after, 186 | h6:after{ 187 | content:". "; 188 | } 189 | 190 | h3 + p, 191 | h4 + p, 192 | h5 + p, 193 | h6 + p{ 194 | display:inline; 195 | } 196 | /* /END Headings */ 197 | 198 | 199 | /* Abstract */ 200 | section[role=doc-abstract]{ 201 | font-size:9pt; 202 | margin:0 2.3em 1em; 203 | margin-bottom:2.75em; 204 | } 205 | 206 | section[role=doc-abstract] h1, 207 | section[role=doc-abstract] > p:first-of-type{ 208 | display:inline; 209 | } 210 | 211 | section[role=doc-abstract] h1:after{ 212 | content:". "; 213 | } 214 | /* /END Abstract */ 215 | 216 | 217 | /* Front matter */ 218 | .affiliation { 219 | font-size:9pt !important; 220 | } 221 | 222 | .affiliation *, 223 | section[role=doc-abstract] h1, 224 | section[role=doc-abstract] p{ 225 | font-size:0.93em; 226 | } 227 | 228 | .keywords strong:after{ 229 | content:": "; 230 | } 231 | 232 | .keywords ul{ 233 | display:inline; 234 | margin-left: 0px; 235 | } 236 | 237 | .keywords li:after{ 238 | content:", "; 239 | } 240 | 241 | .keywords li:last-of-type:after{ 242 | content:""; 243 | } 244 | 245 | .keywords li code{ 246 | font-family:"CMU Serif", serif !important; 247 | } 248 | 249 | .keywords{ 250 | text-indent:0px; 251 | margin-top:0.5cm; 252 | margin-bottom:2.75em; 253 | } 254 | 255 | .keywords li{ 256 | display:inline; 257 | list-style-type:none; 258 | } 259 | 260 | .title{ 261 | text-align:center; 262 | } 263 | 264 | .acm_subject_categories, 265 | .acm_general_terms{ 266 | display:none; 267 | } 268 | 269 | .authors{ 270 | margin:0.7cm auto 2.5em -3%; 271 | text-align:center; 272 | width:106%; 273 | line-height:1; 274 | font-style:normal; 275 | } 276 | 277 | .authors strong { 278 | font-weight:normal; 279 | } 280 | 281 | .authors strong:after { 282 | content:", "; 283 | } 284 | 285 | .authors strong:last-of-type:after { 286 | content:""; 287 | } 288 | 289 | .email{ 290 | font-family:"CMU Serif Monospace", monospace !important; 291 | font-size:8pt; 292 | } 293 | /* /END Front matter */ 294 | 295 | /* Images */ 296 | img, svg{ 297 | border:0; 298 | page-break-inside:avoid; 299 | max-width:100%; 300 | height:auto; 301 | } 302 | figure img, figure svg { 303 | vertical-align:middle; 304 | } 305 | /* /END Images */ 306 | 307 | 308 | /* Tables */ 309 | figure[role=tablebox] { 310 | margin-top:1em; 311 | margin-bottom:1em; 312 | } 313 | 314 | table{ 315 | border-collapse:collapse; 316 | width:100%; 317 | margin-bottom:1em; 318 | border-top:2px solid #000; 319 | border-bottom:2px solid #000; 320 | font-size:8pt; 321 | word-break:break-all; 322 | } 323 | 324 | th, 325 | td{ 326 | padding-top:0.25em; 327 | padding-bottom:0.25em; 328 | padding-right:1em; 329 | text-align:left; 330 | } 331 | 332 | tr{ 333 | vertical-align:top; 334 | border-bottom:1px solid #000; 335 | } 336 | 337 | tr:last-child{ 338 | border-bottom:0; 339 | } 340 | 341 | th:last-child, 342 | td:last-child{ 343 | padding-right:0; 344 | } 345 | 346 | table p { 347 | text-indent:0; 348 | } 349 | /* /END Tables */ 350 | 351 | 352 | /* Heading counters */ 353 | body{ 354 | counter-reset:h1; 355 | } 356 | h1{ 357 | counter-reset:h2; 358 | } 359 | 360 | section:not([role=doc-abstract]):not([role=doc-bibliography]):not([role=doc-acknowledgements]):not([role=doc-endnotes]) h1:before{ 361 | counter-increment:h1; 362 | content:counter(h1); 363 | margin-right:1.5em; 364 | } 365 | 366 | section:not([role=doc-abstract]):not([role=doc-bibliography]):not([role=doc-acknowledgements]):not([role=doc-endnotes]) h2:before{ 367 | counter-increment:h2; 368 | content:counter(h1) "." counter(h2); 369 | margin-right:1em; 370 | } 371 | /* /END Heading counters */ 372 | 373 | 374 | /* Lists */ 375 | ul, 376 | ol{ 377 | margin-left:1em; 378 | margin-top:0.5em; 379 | margin-bottom:0.5em; 380 | } 381 | 382 | ul ul, 383 | ol ol, 384 | ul ol, 385 | ol ul{ 386 | margin-top:0.2em; 387 | margin-bottom:0.2em; 388 | } 389 | 390 | li p:first-child{ 391 | display:inline; 392 | } 393 | /* /END Lists */ 394 | 395 | /* Figure (any) */ 396 | figure { 397 | margin-top:1em; 398 | margin-bottom:1em; 399 | position:relative; 400 | } 401 | 402 | figure > p { 403 | text-align:center; 404 | } 405 | /* /END Figure (any) */ 406 | 407 | /* Listing */ 408 | figure > pre { 409 | border: 1px solid black; 410 | padding: 3px; 411 | } 412 | /* /END Listing */ 413 | 414 | /* Captions */ 415 | figcaption{ 416 | text-align:center; 417 | font-size:9pt; 418 | } 419 | /* /END Captions */ 420 | 421 | /* Footnotes */ 422 | section[role=doc-endnotes] .hidden-print { 423 | display: none; 424 | } 425 | 426 | section[role=doc-endnotes] { 427 | counter-reset: fn; 428 | font-size:9pt; 429 | } 430 | 431 | section[role=doc-endnotes] > section[role=doc-endnote]:before { 432 | counter-increment: fn; 433 | content: counter(fn); 434 | float: left; 435 | padding-right: 10px; 436 | } 437 | /* /END Footnotes */ 438 | 439 | /* Footer */ 440 | html{ 441 | position:relative; 442 | min-height:100%; 443 | } 444 | 445 | section[role=doc-endnotes] { 446 | /* Margin bottom by footer height */ 447 | margin-bottom:90px; 448 | } 449 | 450 | .footer{ 451 | background-color:#F5F5F5; 452 | bottom:0px; 453 | box-sizing:border-box; 454 | color:#333; 455 | display:block; 456 | font-family:"Helvetica Neue", Helvetica, Arial, sans-serif !important; 457 | font-size:12px; 458 | height:50px; 459 | line-height:22.85px; 460 | padding-bottom:5px; 461 | padding-left:5px; 462 | padding-right:5px; 463 | padding-top:5px; 464 | position:fixed; 465 | width:122mm; 466 | } 467 | 468 | .footer span, 469 | .footer div{ 470 | margin-left:7px; 471 | } 472 | 473 | .footer div{ 474 | float:right; 475 | } 476 | 477 | .footer div button{ 478 | height:100%; 479 | } 480 | 481 | #layoutselection{ 482 | margin-left:0px; 483 | margin-right:5px; 484 | } 485 | 486 | .footer div{ 487 | position:relative; 488 | vertical-align:middle; 489 | display:inline; 490 | } 491 | 492 | .footer > p { 493 | float:left; 494 | width:60%; 495 | padding:0; 496 | text-align:left; 497 | } 498 | 499 | .footer p span:first-of-type { 500 | display:block; 501 | margin-left:-4px; 502 | } 503 | 504 | .footer div button{ 505 | background-color:#FFF; 506 | background-image:none; 507 | border-bottom-color:#CCC; 508 | border-bottom-left-radius:4px; 509 | border-bottom-right-radius:4px; 510 | border-bottom-style:solid; 511 | border-bottom-width:1px; 512 | border-image-outset:0 0 0 0; 513 | border-image-repeat:stretch stretch; 514 | border-image-slice:100% 100% 100% 100%; 515 | border-image-source:none; 516 | border-image-width:1 1 1 1; 517 | border-left-color:#CCC; 518 | border-left-style:solid; 519 | border-left-width:1px; 520 | border-right-color:#CCC; 521 | border-right-style:solid; 522 | border-right-width:1px; 523 | border-top-color:#CCC; 524 | border-top-left-radius:4px; 525 | border-top-right-radius:4px; 526 | border-top-style:solid; 527 | border-top-width:1px; 528 | box-sizing:border-box; 529 | color:#333; 530 | cursor:pointer; 531 | display:block; 532 | float:right; 533 | font-feature-settings:normal; 534 | font-kerning:auto; 535 | font-language-override:normal; 536 | font-size-adjust:none; 537 | font-stretch:normal; 538 | font-style:normal; 539 | font-synthesis:weight style; 540 | font-variant:normal; 541 | font-variant-alternates:normal; 542 | font-variant-caps:normal; 543 | font-variant-east-asian:normal; 544 | font-variant-ligatures:normal; 545 | font-variant-numeric:normal; 546 | font-variant-position:normal; 547 | font-weight:400; 548 | line-height:10px; 549 | margin-bottom:0px; 550 | margin-left:0px; 551 | margin-right:0px; 552 | margin-top:0px; 553 | overflow:visible; 554 | overflow-x:visible; 555 | overflow-y:visible; 556 | padding-bottom:4px; 557 | padding-left:5px; 558 | padding-right:5px; 559 | padding-top:4px; 560 | position:relative; 561 | text-align:center; 562 | text-transform:none; 563 | vertical-align:middle; 564 | white-space:nowrap; 565 | -moz-border-bottom-colors:none; 566 | -moz-border-left-colors:none; 567 | -moz-border-right-colors:none; 568 | -moz-border-top-colors:none; 569 | -moz-user-select:none; 570 | } 571 | 572 | .footer div button .caret{ 573 | border-bottom-color:#333; 574 | border-bottom-style:solid; 575 | border-bottom-width:4px; 576 | border-left-color:transparent; 577 | border-left-style:solid; 578 | border-left-width:4px; 579 | border-right-color:transparent; 580 | border-right-style:solid; 581 | border-right-width:4px; 582 | border-top-color:#333; 583 | border-top-style:none; 584 | border-top-width:0px; 585 | box-sizing:border-box; 586 | color:#333; 587 | content:""; 588 | cursor:pointer; 589 | display:inline-block; 590 | font-feature-settings:normal; 591 | font-kerning:auto; 592 | font-language-override:normal; 593 | font-size-adjust:none; 594 | font-stretch:normal; 595 | font-style:normal; 596 | font-synthesis:weight style; 597 | font-variant:normal; 598 | font-variant-alternates:normal; 599 | font-variant-caps:normal; 600 | font-variant-east-asian:normal; 601 | font-variant-ligatures:normal; 602 | font-variant-numeric:normal; 603 | font-variant-position:normal; 604 | font-weight:400; 605 | height:4px; 606 | line-height:20px; 607 | margin-left:0px; 608 | text-align:center; 609 | text-transform:none; 610 | vertical-align:middle; 611 | white-space:nowrap; 612 | width:8px; 613 | } 614 | 615 | .footer div ul li { 616 | padding: 2px 5px 2px 5px; 617 | } 618 | 619 | .footer div ul li a { 620 | display: block; 621 | } 622 | 623 | .footer div ul li:hover, .footer div button:hover { 624 | background-color:#D8D8D8; 625 | } 626 | 627 | .footer div ul{ 628 | background-clip:padding-box; 629 | background-color:#FFF; 630 | border-bottom-color:rgba(0, 0, 0, 0.15); 631 | border-bottom-left-radius:4px; 632 | border-bottom-right-radius:4px; 633 | border-bottom-style:solid; 634 | border-bottom-width:1px; 635 | border-image-outset:0 0 0 0; 636 | border-image-repeat:stretch stretch; 637 | border-image-slice:100% 100% 100% 100%; 638 | border-image-source:none; 639 | border-image-width:1 1 1 1; 640 | border-left-color:rgba(0, 0, 0, 0.15); 641 | border-left-style:solid; 642 | border-left-width:1px; 643 | border-right-color:rgba(0, 0, 0, 0.15); 644 | border-right-style:solid; 645 | border-right-width:1px; 646 | border-top-color:rgba(0, 0, 0, 0.15); 647 | border-top-left-radius:4px; 648 | border-top-right-radius:4px; 649 | border-top-style:solid; 650 | border-top-width:1px; 651 | bottom:100%; 652 | box-shadow:rgba(0, 0, 0, 0.176) 0px 6px 12px 0px; 653 | box-sizing:border-box; 654 | display:none; 655 | float:left; 656 | left:0px; 657 | list-style-image:none; 658 | list-style-position:outside; 659 | list-style-type:none; 660 | margin-bottom:6px; 661 | margin-left:11px; 662 | margin-right:0px; 663 | min-width:147px; 664 | padding-bottom:5px; 665 | padding-left:0px; 666 | padding-right:0px; 667 | padding-top:5px; 668 | position:absolute; 669 | text-align:left; 670 | top:auto; 671 | z-index:1000; 672 | -moz-border-bottom-colors:none; 673 | -moz-border-left-colors:none; 674 | -moz-border-right-colors:none; 675 | -moz-border-top-colors:none; 676 | } 677 | 678 | .footer p{ 679 | display:inline; 680 | } 681 | /* /END Footer */ 682 | 683 | /* Formulas */ 684 | .rash-math div { 685 | display: inline !important; 686 | } 687 | /* /END formulas */ 688 | 689 | /* Reference list */ 690 | section[role=doc-bibliography] > ul { 691 | list-style-type: decimal; 692 | } 693 | /* /END Reference list */ 694 | 695 | 696 | @media print{ 697 | .footer{ 698 | display:none; 699 | } 700 | 701 | h1, h2{ 702 | page-break-after:avoid; 703 | } 704 | 705 | h1, 706 | p, 707 | ul, 708 | ol, 709 | li, 710 | pre, 711 | code{ 712 | orphans:3; 713 | widows:3; 714 | } 715 | } 716 | 717 | -------------------------------------------------------------------------------- /dmop-dmexperiment-iris.ttl: -------------------------------------------------------------------------------- 1 | @prefix : . 2 | @prefix owl: . 3 | @prefix rdf: . 4 | @prefix xml: . 5 | @prefix xsd: . 6 | @prefix DMEX: . 7 | @prefix DMOP: . 8 | @prefix rdfs: . 9 | @prefix ExtendedDnS: . 10 | @prefix RMOperators: . 11 | @prefix WekaOperators: . 12 | @base . 13 | 14 | rdf:type owl:Ontology ; 15 | 16 | rdfs:comment "DMEX.owl: Sandbox illustrating the representation of a DM experiment with its associated workflow, dataset, results and metadata. " . 17 | 18 | 19 | ################################################################# 20 | # 21 | # Annotation properties 22 | # 23 | ################################################################# 24 | 25 | 26 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#definition 27 | 28 | DMOP:definition rdf:type owl:AnnotationProperty . 29 | 30 | 31 | 32 | 33 | 34 | ################################################################# 35 | # 36 | # Object Properties 37 | # 38 | ################################################################# 39 | 40 | 41 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#executes 42 | 43 | DMOP:executes rdf:type owl:ObjectProperty . 44 | 45 | 46 | 47 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#hasDistinctValue 48 | 49 | DMOP:hasDistinctValue rdf:type owl:ObjectProperty . 50 | 51 | 52 | 53 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#hasFeature 54 | 55 | DMOP:hasFeature rdf:type owl:ObjectProperty . 56 | 57 | 58 | 59 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#hasFirstSubprocess 60 | 61 | DMOP:hasFirstSubprocess rdf:type owl:ObjectProperty . 62 | 63 | 64 | 65 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#hasInput 66 | 67 | DMOP:hasInput rdf:type owl:ObjectProperty . 68 | 69 | 70 | 71 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#hasOutput 72 | 73 | DMOP:hasOutput rdf:type owl:ObjectProperty . 74 | 75 | 76 | 77 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#hasParameterSetting 78 | 79 | DMOP:hasParameterSetting rdf:type owl:ObjectProperty . 80 | 81 | 82 | 83 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#hasTargetFeature 84 | 85 | DMOP:hasTargetFeature rdf:type owl:ObjectProperty . 86 | 87 | 88 | 89 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#setsValueOf 90 | 91 | DMOP:setsValueOf rdf:type owl:ObjectProperty . 92 | 93 | 94 | 95 | ### http://www.loa-cnr.it/ontologies/ExtendedDnS.owl#direct-successor 96 | 97 | ExtendedDnS:direct-successor rdf:type owl:ObjectProperty . 98 | 99 | 100 | 101 | 102 | 103 | ################################################################# 104 | # 105 | # Data properties 106 | # 107 | ################################################################# 108 | 109 | 110 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#hasDataValue 111 | 112 | DMOP:hasDataValue rdf:type owl:DatatypeProperty . 113 | 114 | 115 | 116 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#hasNumberOfSupportVectors 117 | 118 | DMOP:hasNumberOfSupportVectors rdf:type owl:DatatypeProperty . 119 | 120 | 121 | 122 | 123 | 124 | ################################################################# 125 | # 126 | # Classes 127 | # 128 | ################################################################# 129 | 130 | 131 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#C4.5CrispTreeModel 132 | 133 | DMOP:C4.5CrispTreeModel rdf:type owl:Class . 134 | 135 | 136 | 137 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#CategoricalFeature 138 | 139 | DMOP:CategoricalFeature rdf:type owl:Class . 140 | 141 | 142 | 143 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#CategoricalFeatureValue 144 | 145 | DMOP:CategoricalFeatureValue rdf:type owl:Class . 146 | 147 | 148 | 149 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#CategoricalLabeledDataSet 150 | 151 | DMOP:CategoricalLabeledDataSet rdf:type owl:Class . 152 | 153 | 154 | 155 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#ContinuousFeature 156 | 157 | DMOP:ContinuousFeature rdf:type owl:Class . 158 | 159 | 160 | 161 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#DM-Experiment 162 | 163 | DMOP:DM-Experiment rdf:type owl:Class . 164 | 165 | 166 | 167 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#DM-Operation 168 | 169 | DMOP:DM-Operation rdf:type owl:Class . 170 | 171 | 172 | 173 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#DM-Process 174 | 175 | DMOP:DM-Process rdf:type owl:Class . 176 | 177 | 178 | 179 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#DM-Workflow 180 | 181 | DMOP:DM-Workflow rdf:type owl:Class . 182 | 183 | 184 | 185 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#DataTable 186 | 187 | DMOP:DataTable rdf:type owl:Class . 188 | 189 | 190 | 191 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#ModelEvaluationMeasure 192 | 193 | DMOP:ModelEvaluationMeasure rdf:type owl:Class . 194 | 195 | 196 | 197 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#OpParameterSetting 198 | 199 | DMOP:OpParameterSetting rdf:type owl:Class . 200 | 201 | 202 | 203 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#RM-OperatorParameter 204 | 205 | DMOP:RM-OperatorParameter rdf:type owl:Class . 206 | 207 | 208 | 209 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#RapidMinerOperator 210 | 211 | DMOP:RapidMinerOperator rdf:type owl:Class . 212 | 213 | 214 | 215 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#Weka-OperatorParameter 216 | 217 | DMOP:Weka-OperatorParameter rdf:type owl:Class . 218 | 219 | 220 | 221 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/DMOP.owl#WekaOperator 222 | 223 | DMOP:WekaOperator rdf:type owl:Class . 224 | 225 | 226 | 227 | 228 | 229 | ################################################################# 230 | # 231 | # Individuals 232 | # 233 | ################################################################# 234 | 235 | 236 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DCHARS.owl#Iris 237 | 238 | DCHARS:Iris rdf:type owl:NamedIndividual . 239 | 240 | 241 | 242 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#AvgPerfMeasure3 243 | 244 | :AvgPerfMeasure3 rdf:type DMOP:ModelEvaluationMeasure , 245 | owl:NamedIndividual . 246 | 247 | 248 | 249 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Exp3-xvalFSC-IG-Weka-J48-Iris 250 | 251 | :Exp3-xvalFSC-IG-Weka-J48-Iris rdf:type DMOP:DM-Experiment , 252 | owl:NamedIndividual ; 253 | 254 | DMOP:hasFirstSubprocess :Proc3 . 255 | 256 | 257 | 258 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#FSC-Infogain-J48-Xval-Wf 259 | 260 | :FSC-Infogain-J48-Xval-Wf rdf:type DMOP:DM-Workflow , 261 | owl:NamedIndividual . 262 | 263 | 264 | 265 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Iris-MainTable 266 | 267 | :Iris-MainTable rdf:type DMOP:DataTable , 268 | owl:NamedIndividual ; 269 | 270 | DMOP:hasNumberOfSupportVectors 13 ; 271 | 272 | DMOP:hasFeature :Iris-PetalLength , 273 | :Iris-PetalWidth , 274 | :Iris-SepalLength , 275 | :Iris-SepalWidth ; 276 | 277 | DMOP:hasTargetFeature :Iris-Species . 278 | 279 | 280 | 281 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Iris-PetalLength 282 | 283 | :Iris-PetalLength rdf:type DMOP:ContinuousFeature , 284 | owl:NamedIndividual . 285 | 286 | 287 | 288 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Iris-PetalWidth 289 | 290 | :Iris-PetalWidth rdf:type DMOP:ContinuousFeature , 291 | owl:NamedIndividual . 292 | 293 | 294 | 295 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Iris-SepalLength 296 | 297 | :Iris-SepalLength rdf:type DMOP:ContinuousFeature , 298 | owl:NamedIndividual . 299 | 300 | 301 | 302 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Iris-SepalWidth 303 | 304 | :Iris-SepalWidth rdf:type DMOP:ContinuousFeature , 305 | owl:NamedIndividual . 306 | 307 | 308 | 309 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Iris-Setosa 310 | 311 | :Iris-Setosa rdf:type DMOP:CategoricalFeatureValue , 312 | owl:NamedIndividual . 313 | 314 | 315 | 316 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Iris-Species 317 | 318 | :Iris-Species rdf:type DMOP:CategoricalFeature , 319 | owl:NamedIndividual ; 320 | 321 | DMOP:hasDistinctValue :Iris-Setosa , 322 | :Iris-Versicolor , 323 | :Iris-Virginica . 324 | 325 | 326 | 327 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Iris-Trn-FSubset3.i 328 | 329 | :Iris-Trn-FSubset3.i rdf:type DMOP:CategoricalLabeledDataSet , 330 | owl:NamedIndividual . 331 | 332 | 333 | 334 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Iris-Trn-FeatureWeights-3.1.fold-i 335 | 336 | :Iris-Trn-FeatureWeights-3.1.fold-i rdf:type DMOP:DataTable , 337 | owl:NamedIndividual . 338 | 339 | 340 | 341 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Iris-Trn3.i 342 | 343 | :Iris-Trn3.i rdf:type DMOP:CategoricalLabeledDataSet , 344 | owl:NamedIndividual . 345 | 346 | 347 | 348 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Iris-Tst-FSubset3.i 349 | 350 | :Iris-Tst-FSubset3.i rdf:type DMOP:CategoricalLabeledDataSet , 351 | owl:NamedIndividual . 352 | 353 | 354 | 355 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Iris-Tst-Predictions-3.1.fold-i 356 | 357 | :Iris-Tst-Predictions-3.1.fold-i rdf:type DMOP:DataTable , 358 | owl:NamedIndividual . 359 | 360 | 361 | 362 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Iris-Tst3.i 363 | 364 | :Iris-Tst3.i rdf:type DMOP:CategoricalLabeledDataSet , 365 | owl:NamedIndividual . 366 | 367 | 368 | 369 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Iris-Versicolor 370 | 371 | :Iris-Versicolor rdf:type DMOP:CategoricalFeatureValue , 372 | owl:NamedIndividual . 373 | 374 | 375 | 376 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Iris-Virginica 377 | 378 | :Iris-Virginica rdf:type DMOP:CategoricalFeatureValue , 379 | owl:NamedIndividual . 380 | 381 | 382 | 383 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#J48Model3.i 384 | 385 | :J48Model3.i rdf:type DMOP:C4.5CrispTreeModel , 386 | owl:NamedIndividual . 387 | 388 | 389 | 390 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#J48Model3Final 391 | 392 | :J48Model3Final rdf:type DMOP:C4.5CrispTreeModel , 393 | owl:NamedIndividual . 394 | 395 | 396 | 397 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Opex3-TrainFinalModel 398 | 399 | :Opex3-TrainFinalModel rdf:type DMOP:DM-Operation , 400 | owl:NamedIndividual ; 401 | 402 | DMOP:hasInput DCHARS:Iris ; 403 | 404 | DMOP:hasOutput :J48Model3Final . 405 | 406 | 407 | 408 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Opex3-Xval 409 | 410 | :Opex3-Xval rdf:type DMOP:DM-Operation , 411 | owl:NamedIndividual ; 412 | 413 | DMOP:hasOutput :AvgPerfMeasure3 ; 414 | 415 | ExtendedDnS:direct-successor :Opex3-TrainFinalModel ; 416 | 417 | DMOP:hasParameterSetting :Opset3 ; 418 | 419 | DMOP:hasFirstSubprocess :Proc3.i ; 420 | 421 | DMOP:executes RMOperators:RM-X-Validation . 422 | 423 | 424 | 425 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Opex3.i.1-WeightByInfogain 426 | 427 | :Opex3.i.1-WeightByInfogain rdf:type DMOP:DM-Operation , 428 | owl:NamedIndividual ; 429 | 430 | DMOP:hasOutput :Iris-Trn-FeatureWeights-3.1.fold-i , 431 | :Iris-Trn3.i ; 432 | 433 | DMOP:hasInput :Iris-Trn3.i ; 434 | 435 | ExtendedDnS:direct-successor :Opex3.i.2-SelectByWeights ; 436 | 437 | DMOP:executes RMOperators:RM-Weight_by_Information_Gain . 438 | 439 | 440 | 441 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Opex3.i.2-SelectByWeights 442 | 443 | :Opex3.i.2-SelectByWeights rdf:type DMOP:DM-Operation , 444 | owl:NamedIndividual ; 445 | 446 | DMOP:hasInput :Iris-Trn-FeatureWeights-3.1.fold-i ; 447 | 448 | DMOP:hasOutput :Iris-Tst-FSubset3.i ; 449 | 450 | DMOP:hasInput :Iris-Tst3.i ; 451 | 452 | ExtendedDnS:direct-successor :Opex3.i.3-J48 , 453 | :Opex3.i.4-SelectByWeights ; 454 | 455 | DMOP:executes RMOperators:RM-Select_by_Weights . 456 | 457 | 458 | 459 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Opex3.i.3-J48 460 | 461 | :Opex3.i.3-J48 rdf:type DMOP:DM-Operation , 462 | owl:NamedIndividual ; 463 | 464 | DMOP:hasInput :Iris-Trn-FSubset3.i ; 465 | 466 | DMOP:hasOutput :J48Model3.i ; 467 | 468 | ExtendedDnS:direct-successor :Opex3.i.5-ApplyModel ; 469 | 470 | DMOP:hasParameterSetting :Opset3.i.1 ; 471 | 472 | DMOP:executes WekaOperators:Weka-J48 . 473 | 474 | 475 | 476 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Opex3.i.4-SelectByWeights 477 | 478 | :Opex3.i.4-SelectByWeights rdf:type DMOP:DM-Operation , 479 | owl:NamedIndividual ; 480 | 481 | DMOP:hasOutput :Iris-Trn-FSubset3.i ; 482 | 483 | DMOP:hasInput :Iris-Trn-FeatureWeights-3.1.fold-i ; 484 | 485 | DMOP:hasOutput :Iris-Trn-FeatureWeights-3.1.fold-i ; 486 | 487 | DMOP:hasInput :Iris-Trn3.i ; 488 | 489 | ExtendedDnS:direct-successor :Opex3.i.5-ApplyModel ; 490 | 491 | DMOP:executes RMOperators:RM-Select_by_Weights . 492 | 493 | 494 | 495 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Opex3.i.5-ApplyModel 496 | 497 | :Opex3.i.5-ApplyModel rdf:type DMOP:DM-Operation , 498 | owl:NamedIndividual ; 499 | 500 | DMOP:hasInput :Iris-Tst-FSubset3.i ; 501 | 502 | DMOP:hasOutput :Iris-Tst-FSubset3.i , 503 | :Iris-Tst-Predictions-3.1.fold-i ; 504 | 505 | DMOP:hasInput :J48Model3.i ; 506 | 507 | ExtendedDnS:direct-successor :Opex3.i.6-Performance ; 508 | 509 | DMOP:executes RMOperators:RM-Apply_Model . 510 | 511 | 512 | 513 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Opex3.i.6-Performance 514 | 515 | :Opex3.i.6-Performance rdf:type DMOP:DM-Operation , 516 | owl:NamedIndividual ; 517 | 518 | DMOP:hasInput :Iris-Tst-FSubset3.i , 519 | :Iris-Tst-Predictions-3.1.fold-i ; 520 | 521 | DMOP:hasOutput :PerfMeasure-3.1.fold-i ; 522 | 523 | DMOP:executes RMOperators:RM-Performance . 524 | 525 | 526 | 527 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Opset3 528 | 529 | :Opset3 rdf:type DMOP:OpParameterSetting , 530 | owl:NamedIndividual ; 531 | 532 | DMOP:hasDataValue 10 ; 533 | 534 | DMOP:setsValueOf RMOperators:RM-X-Validation-number_of_validations . 535 | 536 | 537 | 538 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Opset3.i.1 539 | 540 | :Opset3.i.1 rdf:type DMOP:OpParameterSetting , 541 | owl:NamedIndividual ; 542 | 543 | DMOP:hasDataValue "0.25"^^xsd:float ; 544 | 545 | DMOP:setsValueOf WekaOperators:Weka-J48-C . 546 | 547 | 548 | 549 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#PerfMeasure-3.1.fold-i 550 | 551 | :PerfMeasure-3.1.fold-i rdf:type DMOP:ModelEvaluationMeasure , 552 | owl:NamedIndividual . 553 | 554 | 555 | 556 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Proc3 557 | 558 | :Proc3 rdf:type DMOP:DM-Process , 559 | owl:NamedIndividual ; 560 | 561 | DMOP:hasInput DCHARS:Iris ; 562 | 563 | DMOP:hasOutput :AvgPerfMeasure3 ; 564 | 565 | DMOP:executes :FSC-Infogain-J48-Xval-Wf ; 566 | 567 | DMOP:hasOutput :J48Model3Final ; 568 | 569 | DMOP:hasFirstSubprocess :Opex3-Xval . 570 | 571 | 572 | 573 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#Proc3.i 574 | 575 | :Proc3.i rdf:type DMOP:DM-Process , 576 | owl:NamedIndividual ; 577 | 578 | DMOP:hasInput :Iris-Trn3.i , 579 | :Iris-Tst3.i ; 580 | 581 | DMOP:hasFirstSubprocess :Opex3.i.1-WeightByInfogain ; 582 | 583 | DMOP:hasOutput :PerfMeasure-3.1.fold-i . 584 | 585 | 586 | 587 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#svalC-Weka-J48-Iris-Wf-1 588 | 589 | :svalC-Weka-J48-Iris-Wf-1 rdf:type DMOP:DM-Workflow , 590 | owl:NamedIndividual . 591 | 592 | 593 | 594 | ### http://www.e-LICO.eu/ontologies/dmo/DMOP/DMEX.owl#xvalC-Weka-J48-Iris-Wf-2 595 | 596 | :xvalC-Weka-J48-Iris-Wf-2 rdf:type DMOP:DM-Workflow , 597 | owl:NamedIndividual . 598 | 599 | 600 | 601 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/RMOperators.owl#RM-Apply_Model 602 | 603 | RMOperators:RM-Apply_Model rdf:type DMOP:RapidMinerOperator , 604 | owl:NamedIndividual . 605 | 606 | 607 | 608 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/RMOperators.owl#RM-Performance 609 | 610 | RMOperators:RM-Performance rdf:type DMOP:RapidMinerOperator , 611 | owl:NamedIndividual . 612 | 613 | 614 | 615 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/RMOperators.owl#RM-Select_by_Weights 616 | 617 | RMOperators:RM-Select_by_Weights rdf:type DMOP:RapidMinerOperator , 618 | owl:NamedIndividual . 619 | 620 | 621 | 622 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/RMOperators.owl#RM-Weight_by_Information_Gain 623 | 624 | RMOperators:RM-Weight_by_Information_Gain rdf:type DMOP:RapidMinerOperator , 625 | owl:NamedIndividual . 626 | 627 | 628 | 629 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/RMOperators.owl#RM-X-Validation 630 | 631 | RMOperators:RM-X-Validation rdf:type DMOP:RapidMinerOperator , 632 | owl:NamedIndividual . 633 | 634 | 635 | 636 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/RMOperators.owl#RM-X-Validation-number_of_validations 637 | 638 | RMOperators:RM-X-Validation-number_of_validations rdf:type DMOP:RM-OperatorParameter , 639 | owl:NamedIndividual . 640 | 641 | 642 | 643 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/WekaOperators.owl#Weka-J48 644 | 645 | WekaOperators:Weka-J48 rdf:type DMOP:WekaOperator , 646 | owl:NamedIndividual . 647 | 648 | 649 | 650 | ### http://www.e-lico.eu/ontologies/dmo/DMOP/WekaOperators.owl#Weka-J48-C 651 | 652 | WekaOperators:Weka-J48-C rdf:type DMOP:Weka-OperatorParameter , 653 | owl:NamedIndividual . 654 | 655 | 656 | 657 | 658 | 659 | ################################################################# 660 | # 661 | # Annotations 662 | # 663 | ################################################################# 664 | 665 | 666 | DMOP:RegressionTask DMOP:definition "RegressionTask: a PredictiveModelingTask in which the target or dependent variable is continuous-valued." . 667 | 668 | 669 | 670 | DMOP:StructuredPredictionTask DMOP:definition "StructuredPredictionTask: a PredictiveModelingTask in which the target or dependent variable represents a complex structure (e.g., a vector, graph, probability distribution) rather than a single categorical (classification) or real (regression) value. " . 671 | 672 | 673 | 674 | DMOP:ClassificationTask DMOP:definition "ClassificationTask: is a PredictiveModelingTask in which the target or dependent variable takes a finite set of discrete values called classes." . 675 | 676 | 677 | 678 | 679 | 680 | -------------------------------------------------------------------------------- /AlgorithmImplementationExecution.owl: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | ]> 14 | 15 | 16 | 26 | 27 | 1.0 28 | Algorithm-Implementation-Execution ontology design pattern 29 | EN 30 | ML-Schema is a collaborative, community effort with a mission to develop, maintain, and promote standard schemas for data mining and machine learning algorithms, datasets, and experiments 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 45 | 46 | 47 | 48 | 49 | 50 | 51 | 52 | OWL 2 53 | OWL 2 54 | 55 | 56 | 57 | 58 | 59 | 60 | 61 | 62 | 63 | 64 | 65 | 66 | 67 | 1.0 68 | 1.0 69 | 70 | 71 | 72 | 73 | 74 | 75 | 76 | 77 | 78 | 79 | 80 | 81 | 82 | 83 | 84 | 85 | 86 | 87 | 88 | 89 | 90 | 91 | 92 | 93 | 94 | 95 | 96 | 97 | 98 | 99 | 100 | Which algorithm is implemented by this implementation? 101 | What are the implementations of this algorithm? 102 | Which implementation is executed? 103 | What are the parameters of this implementation? 104 | What are the parameter settings of particular parameters in this execution? 105 | What is the input to this implementation execution? 106 | What is the output produced by the this implementation execution? 107 | What algorithm does this execution realize? 108 | What task does this execution achieves? 109 | What is the duration of this execution? 110 | What is the input this task is defined on? 111 | 112 | 113 | 114 | 115 | 116 | 117 | 118 | http://ontologydesignpatterns.org/wiki/Submissions:TimeInterval 119 | 120 | 121 | 122 | 123 | 124 | 125 | 126 | The intent of the design pattern is to model algorithm specifications, their implementations and executions. This includes also parameters of implementations, settings of the parameters for the execution, and inputs the execution consumes (e.g., data) and outputs the execution produces (e.g., models, reports). 127 | 128 | 129 | 130 | 131 | 132 | 133 | 134 | ML Schema, see: https://www.w3.org/community/ml-schema/ 135 | ML Schema, see: https://www.w3.org/community/ml-schema/ 136 | 137 | 138 | 139 | 140 | 141 | 142 | 143 | Consider a scenario in machine learning (ML) domain. 144 | The scenario deals with a machine learning task realization and it is based on an example derived from the OpenML portal (http://www.openml.org). 145 | There is an ML Task :task29 which is a supervised classification task defined on the dataset :credit-a. 146 | This task is achieved by the Execution :run100241 which executes the Implementation :wekaLogistic of the Algorithm :logisticRegression which 147 | this execution realizes. 148 | The Implementation :wekaLogistic has five hyperparameters (Parameter): :wekaLogisticC, :wekaLogisticDoNotCheckCapabilities, :wekaLogisticM, :wekaLogisticOutputDebugInfo, :wekaLogisticR. The values of two of these hyperparameters are set. The hyperparameter :wekaLogisticM has value set to -1 (expressed via the ParameterSetting :wekaLogisticMSetting29), and the hyperparameter :wekaLogisticR that has its value set to "1.0E-8"^^xsd:float (expressed via the ParameterSetting :wekaLogisticRSetting29). 149 | The Execution :run100241 has on Input the :credit-a dataset and the parameter settings and its Output is the ML model :wekaLogisticModel100241. 150 | 151 | 152 | 153 | 154 | 155 | 156 | 157 | 158 | 159 | 160 | 167 | 168 | 169 | 170 | 171 | 172 | 173 | 174 | 175 | 176 | 177 | 184 | 185 | 186 | 187 | 188 | 189 | 190 | 191 | 192 | 193 | 194 | 195 | 196 | 197 | definedOn 198 | A relation between a task and either the data or an evaluation specification pertinent to this task. 199 | A relation between a task and either the data or an evaluation specification pertinent to this task. 200 | 201 | 202 | 203 | 204 | 205 | 206 | 207 | 208 | A relation between a run and an implemantation that is being executed during the run. 209 | A relation between a run and an implemantation that is being executed during the run. 210 | 211 | 212 | 213 | 214 | 215 | 216 | 217 | 218 | 219 | 220 | 221 | 222 | 223 | 224 | hasInput 225 | A relation between a run and data that is taken as input to the run. 226 | A relation between a run and data that is taken as input to the run. 227 | 228 | 229 | 230 | 231 | 232 | 233 | 234 | 235 | 236 | A relation between a run and either a model or model evaluation that is produced on it’s output. 237 | A relation between a run and either a model or model evaluation that is produced on it’s output. 238 | 239 | 240 | 241 | 242 | 243 | 244 | 245 | 246 | hasHyperParameter 247 | A relation between an implementation of a machine learning algorithm and its hyperparameter. 248 | A relation between an implementation of a machine learning algorithm and its hyperparameter. 249 | 250 | 251 | 252 | 253 | 254 | 255 | 256 | 257 | hasPart 258 | A relation which represents a part-whole relationship holding between an entity and its part. 259 | A relation which represents a part-whole relationship holding between an entity and its part. 260 | 261 | 262 | 263 | 264 | 265 | 266 | 267 | implements 268 | A relation between an information entity and a specification that it conforms to. 269 | A relation between an information entity and a specification that it conforms to. 270 | 271 | 272 | 273 | 274 | 275 | 276 | 277 | A relation between a run and a task, where the run realizes specifications formulated by the task. 278 | A relation between a run and a task, where the run realizes specifications formulated by the task. 279 | 280 | 281 | 282 | 283 | 284 | 285 | 286 | 287 | A relation between an entity and the information content entity that specifies it. 288 | A relation between an entity and the information content entity that specifies it. 289 | 290 | 291 | 292 | 293 | 300 | 301 | 302 | 303 | 304 | 305 | 306 | 307 | 308 | 309 | 310 | 317 | 318 | 319 | 320 | 321 | 322 | 323 | 324 | 325 | The algorithm regardless software implementation. 326 | The algorithm regardless software implementation. 327 | 328 | 329 | 330 | 331 | 332 | 333 | 334 | 335 | 336 | 337 | 338 | 339 | 340 | 341 | 342 | 343 | 344 | 345 | 346 | 347 | 348 | 349 | 350 | 351 | 352 | 353 | 354 | 355 | 356 | 357 | 358 | 359 | 360 | 361 | 362 | 363 | 364 | 365 | Run is an execution of an implementation on a machine (computer). It is limited in time (has a start and end point), can be successful or failed. 366 | Run is an execution of an implementation on a machine (computer). It is limited in time (has a start and end point), can be successful or failed. 367 | 368 | 369 | 370 | 371 | 372 | 373 | 374 | 375 | 376 | 377 | 378 | 379 | 380 | 381 | 382 | 383 | 384 | 385 | 386 | 387 | Implementation is an executable implementation of a machine learning algorithm, a script, or a workflow. It is versioned, and sometimes belongs to a library (e.g. WEKA). 388 | Implementation is an executable implementation of a machine learning algorithm, a script, or a workflow. It is versioned, and sometimes belongs to a library (e.g. WEKA). 389 | 390 | 391 | 392 | 393 | 394 | 395 | 396 | 397 | 398 | 399 | 400 | 401 | 402 | 403 | Data is a data item composed of data examples and it may be of a various level of granularity and complexity. 404 | With regard to granularity, it can be a whole dataset (for instance, one main table and possibly other tables), or only a single table, or only a feature (e.g., a column of a table), or only an instance (e.g., row of a table), or a single feature-value pair. 405 | With regard to complexity, data examples are characterized by their datatype, which may be arbitrarily complex (e.g., instead of a table it can be an arbitrary graph). 406 | Data is a data item composed of data examples and it may be of a various level of granularity and complexity. 407 | With regard to granularity, it can be a whole dataset (for instance, one main table and possibly other tables), or only a single table, or only a feature (e.g., a column of a table), or only an instance (e.g., row of a table), or a single feature-value pair. 408 | With regard to complexity, data examples are characterized by their datatype, which may be arbitrarily complex (e.g., instead of a table it can be an arbitrary graph). 409 | 410 | 411 | 412 | 413 | 414 | 415 | 416 | 417 | Model is a generalization of a set of training data able to predict values for unseen instances. It is an output from an execution of a data mining algorithm implementation. Models have a dual nature. They can be treated as data structures and as such represented, stored and manipulated. On the other hand, they act as functions and are executed, taking as input data examples and giving as output the result of applying the function to a data example. Models can also be divided into global or local ones. A global model has global coverage of a data set, i.e., it generalizes the whole data set. A local model, such as a pattern set, is a set of local hypotheses, i.e. each applies to a limited region of the data set. 418 | Model is a generalization of a set of training data able to predict values for unseen instances. It is an output from an execution of a data mining algorithm implementation. Models have a dual nature. They can be treated as data structures and as such represented, stored and manipulated. On the other hand, they act as functions and are executed, taking as input data examples and giving as output the result of applying the function to a data example. Models can also be divided into global or local ones. A global model has global coverage of a data set, i.e., it generalizes the whole data set. A local model, such as a pattern set, is a set of local hypotheses, i.e. each applies to a limited region of the data set. 419 | 420 | 421 | 422 | 423 | 424 | 425 | 426 | 427 | Hyperparameter is a prior parameter of an implementation, i.e., a parameter which is set before its execution (e.g. C, the complexity parameter, in weka.SMO). 428 | Hyperparameter is a prior parameter of an implementation, i.e., a parameter which is set before its execution (e.g. C, the complexity parameter, in weka.SMO). 429 | 430 | 431 | 432 | 433 | 434 | 435 | 436 | 437 | 438 | 439 | 440 | 441 | 442 | 443 | 444 | 445 | 446 | 447 | 448 | 449 | HyperParameterSetting is an entity which connects a hyperparameter and its value that is being set before an implementation execution. 450 | HyperParameterSetting is an entity which connects a hyperparameter and its value that is being set before an implementation execution. 451 | 452 | 453 | 454 | 455 | 456 | 457 | 458 | Process 459 | 460 | 461 | 462 | 463 | 464 | 465 | 466 | 467 | 468 | 469 | 470 | 471 | 472 | 473 | Task is a formal description of a process that needs to be completed (e.g. based on inputs and outputs). A Task is any piece of work that needs to be addressed in the data mining process. In ML Schema, it is defined based on data and on the evaluation specification. 474 | Task is a formal description of a process that needs to be completed (e.g. based on inputs and outputs). A Task is any piece of work that needs to be addressed in the data mining process. In ML Schema, it is defined based on data and on the evaluation specification. 475 | 476 | 477 | 478 | 479 | 486 | 487 | 488 | Contributor 489 | 1999-07-02 490 | 2008-01-14 491 | A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document "DCMI Metadata Terms" (http://dublincore.org/documents/dcmi-terms/) for an explanation. 492 | An entity responsible for making contributions to the resource. 493 | Examples of a Contributor include a person, an organization, or a service. Typically, the name of a Contributor should be used to indicate the entity. 494 | 495 | 496 | 497 | 498 | Publisher 499 | 1999-07-02 500 | 2008-01-14 501 | A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document "DCMI Metadata Terms" (http://dublincore.org/documents/dcmi-terms/) for an explanation. 502 | An entity responsible for making the resource available. 503 | Examples of a Publisher include a person, an organization, or a service. Typically, the name of a Publisher should be used to indicate the entity. 504 | 505 | 506 | 507 | 508 | Relation 509 | 1999-07-02 510 | 2008-01-14 511 | A related resource. 512 | A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document "DCMI Metadata Terms" (http://dublincore.org/documents/dcmi-terms/) for an explanation. 513 | Recommended best practice is to identify the related resource by means of a string conforming to a formal identification system. 514 | 515 | 516 | 517 | 518 | Title 519 | 1999-07-02 520 | 2008-01-14 521 | A name given to the resource. 522 | A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document "DCMI Metadata Terms" (http://dublincore.org/documents/dcmi-terms/) for an explanation. 523 | 524 | 525 | 526 | 527 | Source 528 | 1999-07-02 529 | 2008-01-14 530 | A related resource from which the described resource is derived. 531 | A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document "DCMI Metadata Terms" (http://dublincore.org/documents/dcmi-terms/) for an explanation. 532 | The described resource may be derived from the related resource in whole or in part. Recommended best practice is to identify the related resource by means of a string conforming to a formal identification system. 533 | 534 | 535 | 536 | 537 | Language 538 | 1999-07-02 539 | 2008-01-14 540 | A language of the resource. 541 | A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document "DCMI Metadata Terms" (http://dublincore.org/documents/dcmi-terms/) for an explanation. 542 | Recommended best practice is to use a controlled vocabulary such as RFC 4646 [RFC4646]. 543 | 544 | 545 | 546 | 547 | 548 | 2012-06-14 549 | Dublin Core Metadata Element Set, Version 1.1 550 | 551 | 552 | 553 | Date 554 | 1999-07-02 555 | 2008-01-14 556 | A point or period of time associated with an event in the lifecycle of the resource. 557 | A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document "DCMI Metadata Terms" (http://dublincore.org/documents/dcmi-terms/) for an explanation. 558 | Date may be used to express temporal information at any level of granularity. Recommended best practice is to use an encoding scheme, such as the W3CDTF profile of ISO 8601 [W3CDTF]. 559 | 560 | 561 | 562 | 563 | Coverage 564 | 1999-07-02 565 | 2008-01-14 566 | A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document "DCMI Metadata Terms" (http://dublincore.org/documents/dcmi-terms/) for an explanation. 567 | Spatial topic and spatial applicability may be a named place or a location specified by its geographic coordinates. Temporal topic may be a named period, date, or date range. A jurisdiction may be a named administrative entity or a geographic place to which the resource applies. Recommended best practice is to use a controlled vocabulary such as the Thesaurus of Geographic Names [TGN]. Where appropriate, named places or time periods can be used in preference to numeric identifiers such as sets of coordinates or date ranges. 568 | The spatial or temporal topic of the resource, the spatial applicability of the resource, or the jurisdiction under which the resource is relevant. 569 | 570 | 571 | 572 | 573 | Subject 574 | 1999-07-02 575 | 2012-06-14 576 | A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document "DCMI Metadata Terms" (http://dublincore.org/documents/dcmi-terms/) for an explanation. 577 | The topic of the resource. 578 | Typically, the subject will be represented using keywords, key phrases, or classification codes. Recommended best practice is to use a controlled vocabulary. 579 | 580 | 581 | 582 | 583 | Type 584 | 1999-07-02 585 | 2008-01-14 586 | A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document "DCMI Metadata Terms" (http://dublincore.org/documents/dcmi-terms/) for an explanation. 587 | Recommended best practice is to use a controlled vocabulary such as the DCMI Type Vocabulary [DCMITYPE]. To describe the file format, physical medium, or dimensions of the resource, use the Format element. 588 | The nature or genre of the resource. 589 | 590 | 591 | 592 | 593 | Description 594 | 1999-07-02 595 | 2008-01-14 596 | A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document "DCMI Metadata Terms" (http://dublincore.org/documents/dcmi-terms/) for an explanation. 597 | An account of the resource. 598 | Description may include but is not limited to: an abstract, a table of contents, a graphical representation, or a free-text account of the resource. 599 | 600 | 601 | 602 | 603 | Identifier 604 | 1999-07-02 605 | 2008-01-14 606 | A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document "DCMI Metadata Terms" (http://dublincore.org/documents/dcmi-terms/) for an explanation. 607 | An unambiguous reference to the resource within a given context. 608 | Recommended best practice is to identify the resource by means of a string conforming to a formal identification system. 609 | 610 | 611 | 612 | 613 | Format 614 | 1999-07-02 615 | 2008-01-14 616 | A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document "DCMI Metadata Terms" (http://dublincore.org/documents/dcmi-terms/) for an explanation. 617 | Examples of dimensions include size and duration. Recommended best practice is to use a controlled vocabulary such as the list of Internet Media Types [MIME]. 618 | The file format, physical medium, or dimensions of the resource. 619 | 620 | 621 | 622 | 623 | Rights 624 | 1999-07-02 625 | 2008-01-14 626 | A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document "DCMI Metadata Terms" (http://dublincore.org/documents/dcmi-terms/) for an explanation. 627 | Information about rights held in and over the resource. 628 | Typically, rights information includes a statement about various property rights associated with the resource, including intellectual property rights. 629 | 630 | 631 | 632 | 633 | Creator 634 | 1999-07-02 635 | 2008-01-14 636 | A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document "DCMI Metadata Terms" (http://dublincore.org/documents/dcmi-terms/) for an explanation. 637 | An entity primarily responsible for making the resource. 638 | Examples of a Creator include a person, an organization, or a service. Typically, the name of a Creator should be used to indicate the entity. 639 | 640 | 641 | 642 | 643 | 644 | 645 | 646 | 647 | 648 | -------------------------------------------------------------------------------- /MLSchema.ttl: -------------------------------------------------------------------------------- 1 | @prefix : . 2 | @prefix dct: . 3 | @prefix owl: . 4 | @prefix rdf: . 5 | @prefix xml: . 6 | @prefix xsd: . 7 | @prefix rdfs: . 8 | @base . 9 | 10 | rdf:type owl:Ontology ; 11 | owl:versionIRI ; 12 | dct:description "ML-Schema is a collaborative, community effort with a mission to develop, maintain, and promote standard schemas for data mining and machine learning algorithms, datasets, and experiments" ; 13 | dct:hasVersion 1.0 ; 14 | dct:license ; 15 | dct:title "Machine Learning Schema" . 16 | 17 | ################################################################# 18 | # Annotation properties 19 | ################################################################# 20 | 21 | ### http://purl.org/dc/terms/description 22 | dct:description rdf:type owl:AnnotationProperty . 23 | 24 | 25 | ### http://purl.org/dc/terms/hasVersion 26 | dct:hasVersion rdf:type owl:AnnotationProperty . 27 | 28 | 29 | ### http://purl.org/dc/terms/issued 30 | dct:issued rdf:type owl:AnnotationProperty . 31 | 32 | 33 | ### http://purl.org/dc/terms/license 34 | dct:license rdf:type owl:AnnotationProperty . 35 | 36 | 37 | ### http://purl.org/dc/terms/modified 38 | dct:modified rdf:type owl:AnnotationProperty . 39 | 40 | 41 | ### http://purl.org/dc/terms/publisher 42 | dct:publisher rdf:type owl:AnnotationProperty . 43 | 44 | 45 | ### http://purl.org/dc/terms/title 46 | dct:title rdf:type owl:AnnotationProperty . 47 | 48 | 49 | ### http://www.w3.org/2004/02/skos/core#note 50 | rdf:type owl:AnnotationProperty . 51 | 52 | 53 | ################################################################# 54 | # Datatypes 55 | ################################################################# 56 | 57 | ### http://www.w3.org/2001/XMLSchema#date 58 | xsd:date rdf:type rdfs:Datatype . 59 | 60 | 61 | ################################################################# 62 | # Object Properties 63 | ################################################################# 64 | 65 | ### http://www.w3.org/ns/mls#achieves 66 | :achieves rdf:type owl:ObjectProperty ; 67 | rdfs:domain :Run ; 68 | rdfs:range :Task ; 69 | dct:description "A relation between a run and a task, where the run achieves specifications formulated by the task." ; 70 | rdfs:comment "A relation between a run and a task, where the run achieves specifications formulated by the task." . 71 | 72 | 73 | ### http://www.w3.org/ns/mls#definedOn 74 | :definedOn rdf:type owl:ObjectProperty ; 75 | rdfs:subPropertyOf owl:topObjectProperty ; 76 | owl:inverseOf :defines ; 77 | rdfs:domain :Task ; 78 | rdfs:range [ rdf:type owl:Class ; 79 | owl:unionOf ( :Data 80 | :EvaluationSpecification 81 | ) 82 | ] ; 83 | dct:description "A relation between a task and either the data or an evaluation specification pertinent to this task." ; 84 | rdfs:comment "A relation between a task and either the data or an evaluation specification pertinent to this task." ; 85 | rdfs:label "definedOn"@en . 86 | 87 | 88 | ### http://www.w3.org/ns/mls#defines 89 | :defines rdf:type owl:ObjectProperty . 90 | 91 | 92 | ### http://www.w3.org/ns/mls#executes 93 | :executes rdf:type owl:ObjectProperty ; 94 | rdfs:subPropertyOf owl:topObjectProperty ; 95 | rdfs:domain :Run ; 96 | rdfs:range :Implementation ; 97 | dct:description "A relation between a run and an implemantation that is being executed during the run." ; 98 | rdfs:comment "A relation between a run and an implemantation that is being executed during the run." . 99 | 100 | 101 | ### http://www.w3.org/ns/mls#hasHyperParameter 102 | :hasHyperParameter rdf:type owl:ObjectProperty ; 103 | rdfs:subPropertyOf owl:topObjectProperty ; 104 | rdfs:domain :Implementation ; 105 | rdfs:range :HyperParameter ; 106 | dct:description "A relation between an implementation of a machine learning algorithm and its hyperparameter." ; 107 | rdfs:comment "A relation between an implementation of a machine learning algorithm and its hyperparameter." ; 108 | rdfs:label "hasHyperParameter"@en . 109 | 110 | 111 | ### http://www.w3.org/ns/mls#hasInput 112 | :hasInput rdf:type owl:ObjectProperty ; 113 | rdfs:subPropertyOf owl:topObjectProperty ; 114 | rdfs:domain :Run ; 115 | rdfs:range [ rdf:type owl:Class ; 116 | owl:unionOf ( :Data 117 | :HyperParameterSetting 118 | ) 119 | ] ; 120 | dct:description "A relation between a run and data that is taken as input to the run." ; 121 | rdfs:comment "A relation between a run and data that is taken as input to the run." ; 122 | rdfs:label "hasInput"@en . 123 | 124 | 125 | ### http://www.w3.org/ns/mls#hasOutput 126 | :hasOutput rdf:type owl:ObjectProperty ; 127 | rdfs:domain :Run ; 128 | rdfs:range [ rdf:type owl:Class ; 129 | owl:unionOf ( :Model 130 | :ModelEvaluation 131 | ) 132 | ] ; 133 | dct:description "A relation between a run and either a model or model evaluation that is produced on it’s output." ; 134 | rdfs:comment "A relation between a run and either a model or model evaluation that is produced on it’s output." . 135 | 136 | 137 | ### http://www.w3.org/ns/mls#hasPart 138 | :hasPart rdf:type owl:ObjectProperty ; 139 | dct:description "A relation which represents a part-whole relationship holding between an entity and its part." ; 140 | rdfs:comment "A relation which represents a part-whole relationship holding between an entity and its part." ; 141 | rdfs:label "hasPart"@en . 142 | 143 | 144 | ### http://www.w3.org/ns/mls#hasQuality 145 | :hasQuality rdf:type owl:ObjectProperty ; 146 | dct:description "A relation between entities and their various characteristics." ; 147 | rdfs:comment "A relation between entities and their various characteristics." . 148 | 149 | 150 | ### http://www.w3.org/ns/mls#implements 151 | :implements rdf:type owl:ObjectProperty ; 152 | rdfs:domain :InformationEntity ; 153 | rdfs:range :InformationEntity ; 154 | dct:description "A relation between an information entity and a specification that it conforms to." ; 155 | rdfs:comment "A relation between an information entity and a specification that it conforms to." ; 156 | rdfs:label "implements"@en . 157 | 158 | 159 | ### http://www.w3.org/ns/mls#realizes 160 | :realizes rdf:type owl:ObjectProperty ; 161 | rdfs:subPropertyOf owl:topObjectProperty ; 162 | rdfs:domain :Run ; 163 | rdfs:range :Algorithm ; 164 | dct:description "A relation between a run and an algorithm, where the run realizes specifications formulated by the algorithm." ; 165 | rdfs:comment "A relation between a run and an algorithm, where the run realizes specifications formulated by the algorithm." . 166 | 167 | 168 | ### http://www.w3.org/ns/mls#specifiedBy 169 | :specifiedBy rdf:type owl:ObjectProperty ; 170 | rdfs:range :InformationEntity ; 171 | dct:description "A relation between an entity and the information content entity that specifies it." ; 172 | rdfs:comment "A relation between an entity and the information content entity that specifies it." . 173 | 174 | 175 | ################################################################# 176 | # Data properties 177 | ################################################################# 178 | 179 | ### http://www.w3.org/ns/mls#hasValue 180 | :hasValue rdf:type owl:DatatypeProperty . 181 | 182 | 183 | ################################################################# 184 | # Classes 185 | ################################################################# 186 | 187 | ### http://www.w3.org/ns/mls#Algorithm 188 | :Algorithm rdf:type owl:Class ; 189 | rdfs:subClassOf :InformationEntity ; 190 | dct:description "The algorithm regardless software implementation." ; 191 | rdfs:comment "The algorithm regardless software implementation." . 192 | 193 | 194 | ### http://www.w3.org/ns/mls#Data 195 | :Data rdf:type owl:Class ; 196 | rdfs:subClassOf :InformationEntity , 197 | [ rdf:type owl:Restriction ; 198 | owl:onProperty :hasQuality ; 199 | owl:someValuesFrom :DataCharacteristic 200 | ] ; 201 | dct:description """Data is a data item composed of data examples and it may be of a various level of granularity and complexity. 202 | With regard to granularity, it can be a whole dataset (for instance, one main table and possibly other tables), or only a single table, or only a feature (e.g., a column of a table), or only an instance (e.g., row of a table), or a single feature-value pair. 203 | With regard to complexity, data examples are characterized by their datatype, which may be arbitrarily complex (e.g., instead of a table it can be an arbitrary graph).""" ; 204 | rdfs:comment """Data is a data item composed of data examples and it may be of a various level of granularity and complexity. 205 | With regard to granularity, it can be a whole dataset (for instance, one main table and possibly other tables), or only a single table, or only a feature (e.g., a column of a table), or only an instance (e.g., row of a table), or a single feature-value pair. 206 | With regard to complexity, data examples are characterized by their datatype, which may be arbitrarily complex (e.g., instead of a table it can be an arbitrary graph).""" . 207 | 208 | 209 | ### http://www.w3.org/ns/mls#DataCharacteristic 210 | :DataCharacteristic rdf:type owl:Class ; 211 | rdfs:subClassOf :Quality ; 212 | dct:description """DataCharacteristic is a distinguishing quality or property that distinguish one data from another. 213 | Such characteristics are often statistical ones (e.g., the number of instances 214 | or the number of features of a data set). They may be also informationtheoretic measures (e.g., class entropy of a categorical data set) or geometric measures of data complexity (e.g., the highest discriminatory power of any single feature in the data set).""" ; 215 | rdfs:comment """DataCharacteristic is a distinguishing quality or property that distinguish one data from another. 216 | Such characteristics are often statistical ones (e.g., the number of instances 217 | or the number of features of a data set). They may be also informationtheoretic measures (e.g., class entropy of a categorical data set) or geometric measures of data complexity (e.g., the highest discriminatory power of any single feature in the data set).""" . 218 | 219 | 220 | ### http://www.w3.org/ns/mls#Dataset 221 | :Dataset rdf:type owl:Class ; 222 | rdfs:subClassOf :Data ; 223 | owl:disjointWith :Feature . 224 | 225 | 226 | ### http://www.w3.org/ns/mls#DatasetCharacteristic 227 | :DatasetCharacteristic rdf:type owl:Class ; 228 | rdfs:subClassOf :DataCharacteristic ; 229 | owl:disjointWith :FeatureCharacteristic . 230 | 231 | 232 | ### http://www.w3.org/ns/mls#EvaluationMeasure 233 | :EvaluationMeasure rdf:type owl:Class ; 234 | rdfs:subClassOf :InformationEntity ; 235 | dct:description "EvaluationMeasure is a measure to assess the performance of the model generated by the process that realizes the task. Examples are predictive accuracy or f-measure." ; 236 | rdfs:comment "EvaluationMeasure is a measure to assess the performance of the model generated by the process that realizes the task. Examples are predictive accuracy or f-measure." . 237 | 238 | 239 | ### http://www.w3.org/ns/mls#EvaluationProcedure 240 | :EvaluationProcedure rdf:type owl:Class ; 241 | rdfs:subClassOf :InformationEntity ; 242 | dct:description "EvaluationProcedure is a technique to evaluate machine learning models. Examples are cross-validation and leave-one-out." ; 243 | rdfs:comment "EvaluationProcedure is a technique to evaluate machine learning models. Examples are cross-validation and leave-one-out." . 244 | 245 | 246 | ### http://www.w3.org/ns/mls#EvaluationSpecification 247 | :EvaluationSpecification rdf:type owl:Class ; 248 | rdfs:subClassOf :InformationEntity , 249 | [ rdf:type owl:Restriction ; 250 | owl:onProperty :defines ; 251 | owl:someValuesFrom :Task 252 | ] , 253 | [ rdf:type owl:Restriction ; 254 | owl:onProperty :hasPart ; 255 | owl:someValuesFrom :EvaluationMeasure 256 | ] , 257 | [ rdf:type owl:Restriction ; 258 | owl:onProperty :hasPart ; 259 | owl:someValuesFrom :EvaluationProcedure 260 | ] ; 261 | dct:description "EvaluationSpecification is a specification of an evaluation measure and an evaluation procedure to assess the performance of the process that realizes the task." ; 262 | rdfs:comment "EvaluationProcedure is a technique to evaluate machine learning models. Examples are cross-validation and leave-one-out." . 263 | 264 | 265 | ### http://www.w3.org/ns/mls#Experiment 266 | :Experiment rdf:type owl:Class ; 267 | rdfs:subClassOf :Process , 268 | [ rdf:type owl:Restriction ; 269 | owl:onProperty :hasPart ; 270 | owl:someValuesFrom :Run 271 | ] ; 272 | dct:description "Experiment is a collection of runs." ; 273 | rdfs:comment "Experiment is a collection of runs." . 274 | 275 | 276 | ### http://www.w3.org/ns/mls#Feature 277 | :Feature rdf:type owl:Class ; 278 | rdfs:subClassOf :Data . 279 | 280 | 281 | ### http://www.w3.org/ns/mls#FeatureCharacteristic 282 | :FeatureCharacteristic rdf:type owl:Class ; 283 | rdfs:subClassOf :DataCharacteristic . 284 | 285 | 286 | ### http://www.w3.org/ns/mls#HyperParameter 287 | :HyperParameter rdf:type owl:Class ; 288 | rdfs:subClassOf :InformationEntity ; 289 | dct:description "Hyperparameter is a prior parameter of an implementation, i.e., a parameter which is set before its execution (e.g. C, the complexity parameter, in weka.SMO)." ; 290 | rdfs:comment "Hyperparameter is a prior parameter of an implementation, i.e., a parameter which is set before its execution (e.g. C, the complexity parameter, in weka.SMO)." . 291 | 292 | 293 | ### http://www.w3.org/ns/mls#HyperParameterSetting 294 | :HyperParameterSetting rdf:type owl:Class ; 295 | rdfs:subClassOf :InformationEntity , 296 | [ rdf:type owl:Restriction ; 297 | owl:onProperty :specifiedBy ; 298 | owl:someValuesFrom :HyperParameter 299 | ] , 300 | [ rdf:type owl:Restriction ; 301 | owl:onProperty :hasValue ; 302 | owl:someValuesFrom rdfs:Literal 303 | ] ; 304 | dct:description "HyperParameterSetting is an entity which connects a hyperparameter and its value that is being set before an implementation execution." ; 305 | rdfs:comment "HyperParameterSetting is an entity which connects a hyperparameter and its value that is being set before an implementation execution." . 306 | 307 | 308 | ### http://www.w3.org/ns/mls#Implementation 309 | :Implementation rdf:type owl:Class ; 310 | rdfs:subClassOf :InformationEntity , 311 | [ rdf:type owl:Restriction ; 312 | owl:onProperty :hasHyperParameter ; 313 | owl:someValuesFrom :HyperParameter 314 | ] , 315 | [ rdf:type owl:Restriction ; 316 | owl:onProperty :hasQuality ; 317 | owl:someValuesFrom :ImplementationCharacteristic 318 | ] , 319 | [ rdf:type owl:Restriction ; 320 | owl:onProperty :implements ; 321 | owl:someValuesFrom :Algorithm 322 | ] ; 323 | dct:description "Implementation is an executable implementation of a machine learning algorithm, a script, or a workflow. It is versioned, and sometimes belongs to a library (e.g. WEKA)." ; 324 | rdfs:comment "Implementation is an executable implementation of a machine learning algorithm, a script, or a workflow. It is versioned, and sometimes belongs to a library (e.g. WEKA)." . 325 | 326 | 327 | ### http://www.w3.org/ns/mls#ImplementationCharacteristic 328 | :ImplementationCharacteristic rdf:type owl:Class ; 329 | rdfs:subClassOf :Quality ; 330 | dct:description "ImplementationCharacteristic is a distinguishing quality or property that distinguish one implementation from another." ; 331 | rdfs:comment "ImplementationCharacteristic is a distinguishing quality or property that distinguish one implementation from another." . 332 | 333 | 334 | ### http://www.w3.org/ns/mls#InformationEntity 335 | :InformationEntity rdf:type owl:Class ; 336 | owl:disjointWith :Process . 337 | 338 | 339 | ### http://www.w3.org/ns/mls#Model 340 | :Model rdf:type owl:Class ; 341 | rdfs:subClassOf :InformationEntity , 342 | [ rdf:type owl:Restriction ; 343 | owl:onProperty :hasQuality ; 344 | owl:someValuesFrom :ModelCharacteristic 345 | ] ; 346 | dct:description "Model is a generalization of a set of training data able to predict values for unseen instances. It is an output from an execution of a data mining algorithm implementation. Models have a dual nature. They can be treated as data structures and as such represented, stored and manipulated. On the other hand, they act as functions and are executed, taking as input data examples and giving as output the result of applying the function to a data example. Models can also be divided into global or local ones. A global model has global coverage of a data set, i.e., it generalizes the whole data set. A local model, such as a pattern set, is a set of local hypotheses, i.e. each applies to a limited region of the data set." ; 347 | rdfs:comment "Model is a generalization of a set of training data able to predict values for unseen instances. It is an output from an execution of a data mining algorithm implementation. Models have a dual nature. They can be treated as data structures and as such represented, stored and manipulated. On the other hand, they act as functions and are executed, taking as input data examples and giving as output the result of applying the function to a data example. Models can also be divided into global or local ones. A global model has global coverage of a data set, i.e., it generalizes the whole data set. A local model, such as a pattern set, is a set of local hypotheses, i.e. each applies to a limited region of the data set." . 348 | 349 | 350 | ### http://www.w3.org/ns/mls#ModelCharacteristic 351 | :ModelCharacteristic rdf:type owl:Class ; 352 | rdfs:subClassOf :Quality ; 353 | dct:description "ModelCharacteristic is a distinguishing quality or property that distinguish one model from another. An example model characetristic may be interpretabilty or a complexity of the model." ; 354 | rdfs:comment "ModelCharacteristic is a distinguishing quality or property that distinguish one model from another. An example model characetristic may be interpretabilty or a complexity of the model." . 355 | 356 | 357 | ### http://www.w3.org/ns/mls#ModelEvaluation 358 | :ModelEvaluation rdf:type owl:Class ; 359 | rdfs:subClassOf :InformationEntity , 360 | [ rdf:type owl:Restriction ; 361 | owl:onProperty :specifiedBy ; 362 | owl:someValuesFrom :EvaluationMeasure 363 | ] , 364 | [ rdf:type owl:Restriction ; 365 | owl:onProperty :hasValue ; 366 | owl:someValuesFrom rdfs:Literal 367 | ] ; 368 | dct:description "ModelEvaluation is a setting of a value of the performance measure specified by the evaluation specification. It connects a measure specification with its value." ; 369 | rdfs:comment "ModelEvaluation is a setting of a value of the performance measure specified by the evaluation specification. It connects a measure specification with its value." . 370 | 371 | 372 | ### http://www.w3.org/ns/mls#Process 373 | :Process rdf:type owl:Class ; 374 | owl:disjointWith :Quality ; 375 | rdfs:label "Process"@en . 376 | 377 | 378 | ### http://www.w3.org/ns/mls#Quality 379 | :Quality rdf:type owl:Class . 380 | 381 | 382 | ### http://www.w3.org/ns/mls#Run 383 | :Run rdf:type owl:Class ; 384 | rdfs:subClassOf :Process , 385 | [ rdf:type owl:Restriction ; 386 | owl:onProperty :achieves ; 387 | owl:someValuesFrom :Task 388 | ] , 389 | [ rdf:type owl:Restriction ; 390 | owl:onProperty :executes ; 391 | owl:someValuesFrom :Implementation 392 | ] , 393 | [ rdf:type owl:Restriction ; 394 | owl:onProperty :hasInput ; 395 | owl:someValuesFrom :Data 396 | ] , 397 | [ rdf:type owl:Restriction ; 398 | owl:onProperty :hasInput ; 399 | owl:someValuesFrom :HyperParameterSetting 400 | ] , 401 | [ rdf:type owl:Restriction ; 402 | owl:onProperty :hasOutput ; 403 | owl:someValuesFrom :Model 404 | ] , 405 | [ rdf:type owl:Restriction ; 406 | owl:onProperty :hasOutput ; 407 | owl:someValuesFrom :ModelEvaluation 408 | ] , 409 | [ rdf:type owl:Restriction ; 410 | owl:onProperty :realizes ; 411 | owl:someValuesFrom :Algorithm 412 | ] ; 413 | dct:description "Run is an execution of an implementation on a machine (computer). It is limited in time (has a start and end point), can be successful or failed." ; 414 | rdfs:comment "Run is an execution of an implementation on a machine (computer). It is limited in time (has a start and end point), can be successful or failed." . 415 | 416 | 417 | ### http://www.w3.org/ns/mls#Software 418 | :Software rdf:type owl:Class ; 419 | rdfs:subClassOf :InformationEntity , 420 | [ rdf:type owl:Restriction ; 421 | owl:onProperty :hasPart ; 422 | owl:someValuesFrom :Implementation 423 | ] ; 424 | dct:description "Software is implemented computer programs, procedures, scripts or rules with associated documentation, possibly constituting an organized environment, stored in read/write memory for the purpose of being executed within a computer system." ; 425 | rdfs:comment "Software is implemented computer programs, procedures, scripts or rules with associated documentation, possibly constituting an organized environment, stored in read/write memory for the purpose of being executed within a computer system." . 426 | 427 | 428 | ### http://www.w3.org/ns/mls#Study 429 | :Study rdf:type owl:Class ; 430 | rdfs:subClassOf :Process , 431 | [ rdf:type owl:Restriction ; 432 | owl:onProperty :hasPart ; 433 | owl:someValuesFrom :Experiment 434 | ] ; 435 | dct:description "Study is a collection of runs that belong together to do some kind of analysis on its results. This analysis can be general or very specific (e.g. a hypothesis test). Can be linked to files, data, that belong to it." ; 436 | rdfs:comment "Study is a collection of runs that belong together to do some kind of analysis on its results. This analysis can be general or very specific (e.g. a hypothesis test). Can be linked to files, data, that belong to it." . 437 | 438 | 439 | ### http://www.w3.org/ns/mls#Task 440 | :Task rdf:type owl:Class ; 441 | rdfs:subClassOf :InformationEntity , 442 | [ rdf:type owl:Restriction ; 443 | owl:onProperty :definedOn ; 444 | owl:someValuesFrom :Data 445 | ] ; 446 | dct:description "Task is a formal description of a process that needs to be completed (e.g. based on inputs and outputs). A Task is any piece of work that needs to be addressed in the data mining process. In ML Schema, it is defined based on data." ; 447 | rdfs:comment "Task is a formal description of a process that needs to be completed (e.g. based on inputs and outputs). A Task is any piece of work that needs to be addressed in the data mining process. In ML Schema, it is defined based on data." . 448 | 449 | 450 | ################################################################# 451 | # Annotations 452 | ################################################################# 453 | 454 | dct:publisher ; 455 | dct:title "Dublin Core Metadata Element Set, Version 1.1"@en ; 456 | dct:modified "2012-06-14"^^xsd:date . 457 | 458 | 459 | rdfs:label "Contributor"@en ; 460 | dct:hasVersion ; 461 | "A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document \"DCMI Metadata Terms\" (http://dublincore.org/documents/dcmi-terms/) for an explanation."@en ; 462 | rdfs:comment "An entity responsible for making contributions to the resource."@en ; 463 | dct:description "Examples of a Contributor include a person, an organization, or a service. Typically, the name of a Contributor should be used to indicate the entity."@en ; 464 | dct:modified "2008-01-14"^^xsd:date ; 465 | dct:issued "1999-07-02"^^xsd:date ; 466 | rdfs:isDefinedBy . 467 | 468 | 469 | dct:hasVersion ; 470 | rdfs:isDefinedBy ; 471 | dct:modified "2008-01-14"^^xsd:date ; 472 | "A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document \"DCMI Metadata Terms\" (http://dublincore.org/documents/dcmi-terms/) for an explanation."@en ; 473 | rdfs:comment "The spatial or temporal topic of the resource, the spatial applicability of the resource, or the jurisdiction under which the resource is relevant."@en ; 474 | dct:description "Spatial topic and spatial applicability may be a named place or a location specified by its geographic coordinates. Temporal topic may be a named period, date, or date range. A jurisdiction may be a named administrative entity or a geographic place to which the resource applies. Recommended best practice is to use a controlled vocabulary such as the Thesaurus of Geographic Names [TGN]. Where appropriate, named places or time periods can be used in preference to numeric identifiers such as sets of coordinates or date ranges."@en ; 475 | dct:issued "1999-07-02"^^xsd:date ; 476 | rdfs:label "Coverage"@en . 477 | 478 | 479 | dct:modified "2008-01-14"^^xsd:date ; 480 | rdfs:label "Creator"@en ; 481 | rdfs:comment "An entity primarily responsible for making the resource."@en ; 482 | rdfs:isDefinedBy ; 483 | dct:description "Examples of a Creator include a person, an organization, or a service. Typically, the name of a Creator should be used to indicate the entity."@en ; 484 | dct:hasVersion ; 485 | "A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document \"DCMI Metadata Terms\" (http://dublincore.org/documents/dcmi-terms/) for an explanation."@en ; 486 | dct:issued "1999-07-02"^^xsd:date . 487 | 488 | 489 | dct:issued "1999-07-02"^^xsd:date ; 490 | "A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document \"DCMI Metadata Terms\" (http://dublincore.org/documents/dcmi-terms/) for an explanation."@en ; 491 | dct:modified "2008-01-14"^^xsd:date ; 492 | rdfs:isDefinedBy ; 493 | rdfs:comment "A point or period of time associated with an event in the lifecycle of the resource."@en ; 494 | dct:hasVersion ; 495 | dct:description "Date may be used to express temporal information at any level of granularity. Recommended best practice is to use an encoding scheme, such as the W3CDTF profile of ISO 8601 [W3CDTF]."@en ; 496 | rdfs:label "Date"@en . 497 | 498 | 499 | dct:description "Description may include but is not limited to: an abstract, a table of contents, a graphical representation, or a free-text account of the resource."@en ; 500 | rdfs:isDefinedBy ; 501 | dct:issued "1999-07-02"^^xsd:date ; 502 | "A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document \"DCMI Metadata Terms\" (http://dublincore.org/documents/dcmi-terms/) for an explanation."@en ; 503 | rdfs:comment "An account of the resource."@en ; 504 | dct:modified "2008-01-14"^^xsd:date ; 505 | rdfs:label "Description"@en ; 506 | dct:hasVersion . 507 | 508 | 509 | dct:issued "1999-07-02"^^xsd:date ; 510 | "A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document \"DCMI Metadata Terms\" (http://dublincore.org/documents/dcmi-terms/) for an explanation."@en ; 511 | dct:modified "2008-01-14"^^xsd:date ; 512 | rdfs:label "Format"@en ; 513 | rdfs:comment "The file format, physical medium, or dimensions of the resource."@en ; 514 | dct:hasVersion ; 515 | rdfs:isDefinedBy ; 516 | dct:description "Examples of dimensions include size and duration. Recommended best practice is to use a controlled vocabulary such as the list of Internet Media Types [MIME]."@en . 517 | 518 | 519 | dct:hasVersion ; 520 | dct:description "Recommended best practice is to identify the resource by means of a string conforming to a formal identification system. "@en ; 521 | rdfs:isDefinedBy ; 522 | rdfs:comment "An unambiguous reference to the resource within a given context."@en ; 523 | dct:issued "1999-07-02"^^xsd:date ; 524 | dct:modified "2008-01-14"^^xsd:date ; 525 | rdfs:label "Identifier"@en ; 526 | "A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document \"DCMI Metadata Terms\" (http://dublincore.org/documents/dcmi-terms/) for an explanation."@en . 527 | 528 | 529 | dct:modified "2008-01-14"^^xsd:date ; 530 | dct:issued "1999-07-02"^^xsd:date ; 531 | "A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document \"DCMI Metadata Terms\" (http://dublincore.org/documents/dcmi-terms/) for an explanation."@en ; 532 | rdfs:label "Language"@en ; 533 | dct:hasVersion ; 534 | dct:description "Recommended best practice is to use a controlled vocabulary such as RFC 4646 [RFC4646]."@en ; 535 | rdfs:isDefinedBy ; 536 | rdfs:comment "A language of the resource."@en ; 537 | rdfs:seeAlso . 538 | 539 | 540 | dct:hasVersion ; 541 | "A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document \"DCMI Metadata Terms\" (http://dublincore.org/documents/dcmi-terms/) for an explanation."@en ; 542 | dct:issued "1999-07-02"^^xsd:date ; 543 | dct:modified "2008-01-14"^^xsd:date ; 544 | rdfs:label "Publisher"@en ; 545 | dct:description "Examples of a Publisher include a person, an organization, or a service. Typically, the name of a Publisher should be used to indicate the entity."@en ; 546 | rdfs:isDefinedBy ; 547 | rdfs:comment "An entity responsible for making the resource available."@en . 548 | 549 | 550 | dct:hasVersion ; 551 | rdfs:label "Relation"@en ; 552 | dct:description "Recommended best practice is to identify the related resource by means of a string conforming to a formal identification system. "@en ; 553 | dct:modified "2008-01-14"^^xsd:date ; 554 | dct:issued "1999-07-02"^^xsd:date ; 555 | "A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document \"DCMI Metadata Terms\" (http://dublincore.org/documents/dcmi-terms/) for an explanation."@en ; 556 | rdfs:comment "A related resource."@en ; 557 | rdfs:isDefinedBy . 558 | 559 | 560 | dct:hasVersion ; 561 | rdfs:isDefinedBy ; 562 | "A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document \"DCMI Metadata Terms\" (http://dublincore.org/documents/dcmi-terms/) for an explanation."@en ; 563 | dct:description "Typically, rights information includes a statement about various property rights associated with the resource, including intellectual property rights."@en ; 564 | dct:issued "1999-07-02"^^xsd:date ; 565 | rdfs:label "Rights"@en ; 566 | dct:modified "2008-01-14"^^xsd:date ; 567 | rdfs:comment "Information about rights held in and over the resource."@en . 568 | 569 | 570 | dct:hasVersion ; 571 | "A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document \"DCMI Metadata Terms\" (http://dublincore.org/documents/dcmi-terms/) for an explanation."@en ; 572 | rdfs:comment "A related resource from which the described resource is derived."@en ; 573 | rdfs:isDefinedBy ; 574 | dct:modified "2008-01-14"^^xsd:date ; 575 | dct:description "The described resource may be derived from the related resource in whole or in part. Recommended best practice is to identify the related resource by means of a string conforming to a formal identification system."@en ; 576 | dct:issued "1999-07-02"^^xsd:date ; 577 | rdfs:label "Source"@en . 578 | 579 | 580 | dct:description "Typically, the subject will be represented using keywords, key phrases, or classification codes. Recommended best practice is to use a controlled vocabulary."@en ; 581 | "A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document \"DCMI Metadata Terms\" (http://dublincore.org/documents/dcmi-terms/) for an explanation."@en ; 582 | dct:issued "1999-07-02"^^xsd:date ; 583 | rdfs:label "Subject"@en ; 584 | rdfs:comment "The topic of the resource."@en ; 585 | rdfs:isDefinedBy ; 586 | dct:modified "2012-06-14"^^xsd:date ; 587 | dct:hasVersion . 588 | 589 | 590 | rdfs:comment "A name given to the resource."@en ; 591 | dct:modified "2008-01-14"^^xsd:date ; 592 | "A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document \"DCMI Metadata Terms\" (http://dublincore.org/documents/dcmi-terms/) for an explanation."@en ; 593 | dct:issued "1999-07-02"^^xsd:date ; 594 | rdfs:label "Title"@en ; 595 | dct:hasVersion ; 596 | rdfs:isDefinedBy . 597 | 598 | 599 | rdfs:comment "The nature or genre of the resource."@en ; 600 | dct:hasVersion ; 601 | rdfs:label "Type"@en ; 602 | rdfs:isDefinedBy ; 603 | dct:description "Recommended best practice is to use a controlled vocabulary such as the DCMI Type Vocabulary [DCMITYPE]. To describe the file format, physical medium, or dimensions of the resource, use the Format element."@en ; 604 | dct:issued "1999-07-02"^^xsd:date ; 605 | dct:modified "2008-01-14"^^xsd:date ; 606 | "A second property with the same name as this property has been declared in the dcterms: namespace (http://purl.org/dc/terms/). See the Introduction to the document \"DCMI Metadata Terms\" (http://dublincore.org/documents/dcmi-terms/) for an explanation."@en . 607 | 608 | 609 | ################################################################# 610 | # General axioms 611 | ################################################################# 612 | 613 | [ rdf:type owl:AllDisjointClasses ; 614 | owl:members ( :Algorithm 615 | :Data 616 | :EvaluationMeasure 617 | :EvaluationProcedure 618 | :EvaluationSpecification 619 | :HyperParameter 620 | :HyperParameterSetting 621 | :Implementation 622 | :Model 623 | :ModelEvaluation 624 | :Software 625 | :Task 626 | ) 627 | ] . 628 | 629 | 630 | [ rdf:type owl:AllDisjointClasses ; 631 | owl:members ( :DataCharacteristic 632 | :ImplementationCharacteristic 633 | :ModelCharacteristic 634 | ) 635 | ] . 636 | 637 | 638 | [ rdf:type owl:AllDisjointClasses ; 639 | owl:members ( :Experiment 640 | :Run 641 | :Study 642 | ) 643 | ] . 644 | 645 | 646 | ### Generated by the OWL API (version 4.5.9.2019-02-01T07:24:44Z) https://github.com/owlcs/owlapi 647 | -------------------------------------------------------------------------------- /save-sd18/index.html: -------------------------------------------------------------------------------- 1 | 2 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | ML Schema: Exposing the Semantics of Machine Learning with Schemas and Ontologies 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | 44 | 46 | 47 | 49 | 50 | 52 | 54 | 55 | 56 | 57 | 58 | 59 | 60 | 61 | 62 | 63 | 64 | 65 | 66 | 67 | 68 | 69 | 70 | 71 | 72 |
73 |

74 | ML Schema: Exposing the Semantics of Machine Learning with Schemas and Ontologies 75 |

76 |
77 | 78 | Gustavo Correa Publio 79 | 80 | 85 |
86 | 87 | AKSW, Institute of Computer Science, University of Leipzig, Germany 88 | 89 |
90 |
91 | 92 | Diego Esteves 93 | 94 | 99 |
100 | 101 | SDA Research, University of Bonn, Germany 102 | 103 |
104 |
105 | 106 | Agnieszka Ławrynowicz 107 | 108 | 113 |
114 | 115 | Faculty of Computing, Poznan University of Technology, Poland 116 | 117 |
118 |
119 | 120 | Panče Panov 121 | 122 | 127 |
128 | 129 | Department of Knowledge Technologies, Jožef Stefan Institute, Ljubljana, Slovenia 130 | 131 |
132 |
133 | 134 | Larisa Soldatova 135 | 136 | 141 |
142 | 143 | Brunel University, London, United Kingdom 144 | 145 |
146 |
147 | 148 | Tommaso Soru 149 | 150 | 155 |
156 | 157 | AKSW, Institute of Computer Science, University of Leipzig, Germany 158 | 159 |
160 |
161 | 162 | Joaquin Vanschoren 163 | 164 | 169 |
170 | 171 | Eindhoven University of Technology, Netherlands 172 | 173 |
174 |

175 | 176 | Keywords 177 | 178 |

    179 |
  • 180 | 181 | Machine Learning 182 | 183 |
    • 184 |
  • 185 | 186 | Interoperability 187 | 188 |
    • 189 |
  • 190 | 191 | Provenance 192 | 193 |
    • 194 |
  • 195 | 196 | Metadata 197 | 198 |
    • 199 |
  • 200 | 201 | ML-Schema 202 | 203 |
    • 204 |
  • 205 | 206 | MEX 207 | 208 |
    • 209 |
  • 210 | 211 | Expose 212 | 213 |
    • 214 |
  • 215 | 216 | OntoDM 217 | 218 |
    • 219 |
  • 220 | 221 | DMOP 222 | 223 |
    • 224 |
  • 225 | 226 | OpenML 227 | 228 |
    • 229 |
  • 230 | 231 | WASOTA 232 | 233 |
    • 234 |
  • 235 | 236 | METArchive 237 | 238 |
    • 239 |
240 |
241 | 242 |
243 |

244 | Abstract 245 |

246 |

247 | The ML Schema, proposed by the W3C Machine Learning Schema Community Group, is a top-level ontology that provides a set of classes, 248 | properties, and restrictions for representing and interchanging information on machine learning algorithms, datasets, 249 | and experiments. It can be easily extended and specialized and it is also mapped to other more domain-specific ontologies 250 | developed in the area of machine learning and data mining. In this paper, we present the ML Schema and argue that exposing semantics 251 | of machine learning algorithms, models, and experiments may pave the way to better interpretability and to achieve the full interoperability of experiments. 252 |

253 |
254 | 255 |
256 |

257 | Introduction 258 |

259 |

260 | Complex machine learning models have recently achieved great successes in many predictive tasks. Despite their successes, 261 | a major problem is that they are often hard to interpret, which may affect their safeness and the level of trust 262 | of their users. 263 |

264 |

265 | The problem of interpretability is one of the key research issues in the area of knowledge engineering and the Semantic Web 266 | community, which deals with making the semantics of various phenomena explicit. In this community, the problem 267 | of dealing with trust and traceability has gained a major interest last years and resulted in proposals and uptake 268 | of models such as the provenance ontology PROV-O 269 | 270 | 271 | [6]. 272 |

273 |

274 | In line with those efforts, in this position paper, we present the ML Schema 275 | 276 | 277 | 1 278 | 279 | 280 | 281 | 282 | [1], developed within the W3C Machine Learning Schema Community Group 283 | 284 | 285 | 2. It is a simply shared schema that provides a set of classes, properties, and restrictions that can be used 286 | to represent and interchange information on Machine Learning (ML) algorithms, datasets, and experiments. It can 287 | be easily specialized to create new classes and properties and it is also mapped to more specific ontologies 288 | and vocabularies on machine learning 289 | 290 | 291 | 292 | [2] 293 | 294 | 295 | 296 | 297 | [3] 298 | 299 | 300 | 301 | 303 | [4] 304 | 305 | 306 | 307 | 308 | [5]. These ontologies, in turn, contain terms for representing 309 | experiment configurations (e.g., characteristics and properties of 310 | ML datasets, algorithms, and models) either in a more detailed or 311 | more abstract level. 312 |

313 |

314 | Despite recent efforts to achieve a high level of interoperability 315 | of ML experiments we still run into problems 316 | created due to the existence of different ML platforms: each of 317 | those has a specific conceptualization or schema for representing 318 | data and metadata (Figure 1, items 3 and 4: vertical 319 | interoperability). This scenario leads to an extra coding-effort 320 | (Figure 1: item 2) 321 | to achieve both the desired interoperability and a 322 | better provenance level as well as a more automatized environment 323 | for obtaining the generated results. To reduce this gap, the 324 | aforementioned ML vocabularies and ontologies have been proposed 325 | (e.g., MEX and Expose) (Figure 1: item 5). 326 | 327 | The gap can be further significantly reduced by achieving 328 | interoperability among state-of-the-art (SOTA) schemata of those 329 | resources (Figure 1: item 5), 330 | i.e., achieving the horizontal interoperability (Figure 1: item 6). 331 | Therefore, different groups of researchers could exchange SOTA 332 | metadata files in a transparent manner via web services, e.g.: 333 | from OntoDM and MEX (MLSchema.Schema data = mlschema.convert('myfile.ttl', MLSchema.Ontology.OntoDM, MLSchema.Ontology.MEX)). 334 | Furthermore, the canonical format also directly benefits different environments, such as 335 | ML ecosystems (e.g. OpenML 336 | [10]) 337 | and ML Metadata Repositories (e.g. WASOTA 338 | [9]) 339 | which can benefit on the mappings of a shared standard. 340 |

341 |
342 |

343 | ML Schema Core classes. 344 |

345 |
346 | 347 | Figure 1. 348 | 349 | Vertical and Horizontal Interoperability across ML Environments. 350 |
351 |
352 |
353 |
354 |

355 | The ML Schema 356 |

357 |

358 | In 359 | 360 | Figure 2, we depict the classes and the relationships between them, representing the ML Schema. The schema contains 361 | classes for representing different aspects of machine learning. This includes representations data, datasets 362 | and data/dataset characteristics. Next, it includes representations of algorithms, implementations, parameters 363 | of implementations and software. Furthermore, the schema contains representations of models, model characteristics 364 | and model evaluations. Finally, the schema has the ability to represent machine learning experiments with different 365 | granularity. This includes representations of runs of implementations of algorithms on the lowest level and representation 366 | of studies at the highest level. 367 | 368 |

369 |

370 | ML Schema Core classes. 371 |

372 |
373 | 374 | Figure 2. 375 | 376 | ML Schema Core classes. Boxes represent classes. Arrows without filled heads represent properties, arrows with empty heads 377 | represent subclass relations, and arrows with diamonds represent part-of relations. 378 |
379 |
380 |
381 |

Machine Learning Ontologies

382 |

383 | In this section, we present the relationship of the ML Schema to other proposed ontologies and vocabularies for the domain 384 | of machine learning. The development of ML Schema was highly influenced by, initially independent, 385 | research of several groups on modeling the domain of machine learning. Due to this the classes and relations 386 | present in the ML Schema re-appear in the current ML ontologies and vocabularies. 387 |

388 |

389 | In 390 | 391 | Table 1, 392 | we present the mapping between the terms present in the ML Schema and the current ML ontologies and 393 | vocabularies. Below, we describe each of the mentioned ontology/vocabulary. 394 |

395 | 396 |

397 |

398 | 399 | 400 | 401 | 402 | 403 | 404 | 405 | 406 | 407 | 412 | 417 | 422 | 427 | 432 | 433 | 434 | 439 | 444 | 449 | 454 | 459 | 460 | 461 | 466 | 471 | 476 | 481 | 486 | 487 | 488 | 493 | 498 | 503 | 508 | 513 | 514 | 515 | 520 | 525 | 530 | 535 | 540 | 541 | 542 | 547 | 552 | 557 | 562 | 567 | 568 | 569 | 574 | 579 | 584 | 589 | 594 | 595 | 596 | 601 | 606 | 611 | 616 | 621 | 622 | 623 | 628 | 633 | 638 | 643 | 648 | 649 | 650 | 655 | 660 | 665 | 670 | 675 | 676 | 677 | 682 | 687 | 692 | 697 | 702 | 703 | 704 | 709 | 714 | 719 | 724 | 729 | 730 | 731 | 736 | 741 | 746 | 751 | 756 | 757 | 758 | 763 | 768 | 773 | 778 | 783 | 784 | 785 | 790 | 795 | 800 | 805 | 810 | 811 | 812 | 817 | 822 | 827 | 831 | 836 | 837 | 838 | 843 | 848 | 853 | 858 | 863 | 864 | 865 | 870 | 875 | 880 | 885 | 890 | 891 | 892 | 897 | 902 | 907 | 912 | 917 | 918 | 919 | 924 | 929 | 934 | 939 | 944 | 945 | 946 | 951 | 956 | 961 | 966 | 971 | 972 |
ML SchemaOntoDM-coreDMOPExposeMEX Vocabulary
408 |

409 | Task 410 |

411 | 		413 |

414 | Data mining task 415 |

416 | 		418 |

419 | DM-Task 420 |

421 | 		423 |

424 | Task 425 |

426 | 		428 |

429 | mexcore:ExperimentConfiguration 430 |

431 |
435 |

436 | Algorithm 437 |

438 | 		440 |

441 | Data mining algorithm 442 |

443 | 		445 |

446 | DM-Algorithm 447 |

448 | 		450 |

451 | Algorithm 452 |

453 | 		455 |

456 | mexalgo:Algorithm 457 |

458 |
462 |

463 | Software 464 |

465 | 		467 |

468 | Data mining software 469 |

470 | 		472 |

473 | DM-Software 474 |

475 | 		477 |

478 | N/A 479 |

480 | 		482 |

483 | mexalgo:Tool 484 |

485 |
489 |

490 | Implementation 491 |

492 | 		494 |

495 | Data mining algorithm implementation 496 |

497 | 		499 |

500 | DM-Operator 501 |

502 | 		504 |

505 | Algorithm implementation 506 |

507 | 		509 |

510 | N/A 511 |

512 |
516 |

517 | HyperParameter 518 |

519 | 		521 |

522 | Parameter 523 |

524 | 		526 |

527 | Parameter 528 |

529 | 		531 |

532 | Parameter 533 |

534 | 		536 |

537 | mexalgo:HyperParameter 538 |

539 |
543 |

544 | HyperParameterSetting 545 |

546 | 		548 |

549 | Parameter setting 550 |

551 | 		553 |

554 | OpParameterSetting 555 |

556 | 		558 |

559 | Parameter setting 560 |

561 | 		563 |

564 | N/A 565 |

566 |
570 |

571 | Study 572 |

573 | 		575 |

576 | Investigation 577 |

578 | 		580 |

581 | N/A 582 |

583 | 		585 |

586 | N/A 587 |

588 | 		590 |

591 | mexcore:Experiment 592 |

593 |
597 |

598 | Experiment 599 |

600 | 		602 |

603 | N/A 604 |

605 | 		607 |

608 | DM-Experiment 609 |

610 | 		612 |

613 | Experiment 614 |

615 | 		617 |

618 | N/A 619 |

620 |
624 |

625 | Run 626 |

627 | 		629 |

630 | Data mining algorithm execution 631 |

632 | 		634 |

635 | DM-Operation 636 |

637 | 		639 |

640 | Algorithm execution 641 |

642 | 		644 |

645 | mexcore:Execution 646 |

647 |
651 |

652 | Data 653 |

654 | 		656 |

657 | Data item 658 |

659 | 		661 |

662 | DM-Data 663 |

664 | 		666 |

667 | N/A 668 |

669 | 		671 |

672 | mexcore:Example 673 |

674 |
678 |

679 | Dataset 680 |

681 | 		683 |

684 | DM dataset 685 |

686 | 		688 |

689 | DataSet 690 |

691 | 		693 |

694 | Dataset 695 |

696 | 		698 |

699 | mexcore:Dataset 700 |

701 |
705 |

706 | Feature 707 |

708 | 		710 |

711 | N/A 712 |

713 | 		715 |

716 | Feature 717 |

718 | 		720 |

721 | N/A 722 |

723 | 		725 |

726 | mexcore:Feature 727 |

728 |
732 |

733 | DataCharacteristic 734 |

735 | 		737 |

738 | Data specification 739 |

740 | 		742 |

743 | DataCharacteristic 744 |

745 | 		747 |

748 | Dataset specification 749 |

750 | 		752 |

753 | N/A 754 |

755 |
759 |

760 | DatasetCharacteristic 761 |

762 | 		764 |

765 | Dataset specification 766 |

767 | 		769 |

770 | DataSetCharacteristic 771 |

772 | 		774 |

775 | Data quality 776 |

777 | 		779 |

780 | N/A 781 |

782 |
786 |

787 | FeatureCharacteristic 788 |

789 | 		791 |

792 | Feature specification 793 |

794 | 		796 |

797 | FeatureCharacteristic 798 |

799 | 		801 |

802 | N/A 803 |

804 | 		806 |

807 | N/A 808 |

809 |
813 |

814 | Model 815 |

816 | 		818 |

819 | Generalization 820 |

821 | 		823 |

824 | DM-Hypothesis 825 |

826 | 		828 |

829 | Model

830 | 		832 |

833 | mexcore:Model 834 |

835 |
839 |

840 | ModelCharacteristic 841 |

842 | 		844 |

845 | Generalization quality 846 |

847 | 		849 |

850 | HypothesisCharacteristic 851 |

852 | 		854 |

855 | Model Structure, Parameter, ... 856 |

857 | 		859 |

860 | N/A 861 |

862 |
866 |

867 | ModelEvaluation 868 |

869 | 		871 |

872 | Generalization evaluation 873 |

874 | 		876 |

877 | ModelPerformance 878 |

879 | 		881 |

882 | Evaluation 883 |

884 | 		886 |

887 | N/A 888 |

889 |
893 |

894 | EvaluationMeasure 895 |

896 | 		898 |

899 | Evaluation datum 900 |

901 | 		903 |

904 | ModelEvaluationMeasure 905 |

906 | 		908 |

909 | Evaluation measure 910 |

911 | 		913 |

914 | mexperf:PerformanceMeasure 915 |

916 |
920 |

921 | EvaluationSpecification 922 |

923 | 		925 |

926 | N/A 927 |

928 | 		930 |

931 | N/A 932 |

933 | 		935 |

936 | N/A 937 |

938 | 		940 |

941 | N/A 942 |

943 |
947 |

948 | EvaluationProcedure 949 |

950 | 		952 |

953 | Evaluation algorithm 954 |

955 | 		957 |

958 | ModelEvaluationAlgorithm 959 |

960 | 		962 |

963 | Performance Estimation 964 |

965 | 		967 |

968 | N/A 969 |

970 |
973 |
974 | 975 | Table 1. 976 | The mapping between terms of ML Schema and the different ML/DM ontologies and vocabularies.
977 |
978 |

979 | 980 |

981 | The OntoDM-core ontology. 982 | The OntoDM-core ontology has been designed to provide generic representations of principle entities in the area of data mining 983 | 984 | 985 | 986 | [3]. In one of the preliminary versions of the ontology, the authors decided to align the proposed ontology 987 | with the Ontology of Biomedical Investigations (OBI) 988 | [8] and consequently with the Basic Formal Ontology (BFO) 989 | [7] 990 | at the top level, in terms of top-level 991 | classes and the set of relations. That was beneficial for structuring the domain in a more elegant way 992 | and the basic differentiation of information entities, implementation entities, and processual entities. 993 | In this context, the authors proposed a horizontal description structure that includes three layers: 994 | a specification layer, an implementation layer, and an application layer. The specification layer, in 995 | general, contains information entities. In the domain of data mining, example classes are data mining 996 | task and data mining algorithm. The implementation layer, in general, contains qualities and entities that 997 | are realized in a process, such as parameters and implementations of algorithms. The application layer 998 | contains processual classes, such as the execution of the data mining algorithm. 999 |

1000 | 1001 |

1002 | The Exposé ontology. 1003 | The main goal of Exposé 1004 | [4] is to describe (and reason about) machine learning experiments. 1005 | It is built on top of OntoDM, as well as top-level ontologies from bio-informatics. It is currently used 1006 | in OpenML 1007 | 3 1008 | [10], 1009 | as a way to structure data (e.g. database design) and share data (APIs). MLSchema will be used to export 1010 | all OpenML data as linked open data (in RDF). 1011 |

1012 | 1013 |

1014 | The DMOP ontology. 1015 | The Data Mining OPtimization Ontology (DMOP) 1016 | 1017 | 1018 | [2] 1019 | 1020 | 1021 | has been developed with a primary use 1022 | case in meta-mining, that is meta-learning extended to an analysis of full DM processes. At the level 1023 | of both single algorithms and more complex workflows, it follows a very similar modeling pattern as described 1024 | in the MLSchema. To support meta-mining, DMOP contains a taxonomy of algorithms used in DM processes 1025 | which are described in detail in terms of their underlying assumptions, cost functions, optimization 1026 | strategies, generated models or pattern sets, and other properties. Such a ''glass box'' approach which 1027 | makes explicit internal algorithm characteristics allows meta-learners using DMOP to generalize over 1028 | algorithms and their properties, including those algorithms which were not used for training meta-learners. 1029 |

1030 | 1031 |

1032 | DMOP also contains sub-taxonomies of ML models and provides a vocabulary to describe their properties and characteristics, 1033 | e.g. model structures, model complexity measures, and parameters. 1034 |

1035 | 1036 |

1037 | The MEX Vocabulary. 1038 | The MEX vocabulary 1039 | [5] 1040 | has been designed to reuse existing ontologies (i.e., PROV-O 1041 | [6] 1042 | for representing basic machine learning experiment configuration and its outcomes). The aim is not to describe a complete data-mining process, which can be modeled by more complex and semantically refined structures. Instead, MEX is designed 1043 | to provide a simple and lightweight vocabulary for exchanging basic machine learning metadata in order to achieve 1044 | a high level of interoperability. Moreover, the schema aims to serve as a basis for data management of ML outcomes in the context of WASOTA [9]. 1045 | 1046 |

1047 | 1048 |
1049 |
1050 |

1051 | Conclusions 1052 |

1053 |

1054 | In this position paper, we have presented ML Schema, a comprehensive schema for modeling ML domain. 1055 |

1056 | 1057 |

1058 | The ML Schema aligns more fine-grained ontologies and vocabularies, some of which contain detailed vocabulary for representing meta-data on ML models. 1059 |

1060 | 1061 |

1062 | The vocabulary and axiomatization included in those resources may be used to make explicit the semantics of ML models, making 1063 | them better interpretable for human users. 1064 |

1065 |
1066 |
1067 |

1068 | Acknowledgements. 1069 | Gustavo Correa Publio acknowledges the support of the Smart 1070 | Data Web BMWi project (GA-01MD15010B) and CNPq Foundation, Ministry of Education of Brazil 1071 | (scholarship 201808/2015-3). Diego Esteves acknowledges the 1072 | support of CAPES Foundation, Ministry of Education of 1073 | Brazil (10179/2013-5). Agnieszka Ławrynowicz 1074 | acknowledges the support from the National Science Centre, 1075 | Poland, within the grant number 2014/13/D/ST6/02076. 1076 |

1077 |
1078 | 1079 |
1080 |

1081 | References 1082 |

1083 |
    1084 | 1085 |
  • 1087 |

    Diego Esteves, Agnieszka Lawrynowicz, Pance Panov, Larisa N. Soldatova, Tommaso Soru, Joaquin Vanschoren 1088 | (2016). ML Schema Core Specification. 1089 | https://www.w3.org/community/ml-schema/ Last Accessed in February 1st 2018.

    1090 |
    • 1091 | 1092 |
  • 1094 |

    C. Maria Keet, Agnieszka Lawrynowicz, Claudia d'Amato, Alexandros Kalousis, Phong Nguyen, Raul Palma, 1095 | Robert Stevens, Melanie Hilario (2015). The Data Mining OPtimization Ontology. J. Web Sem.

    1096 |
    • 1097 | 1098 |
  • 1100 |

    Pance Panov, Larisa N. Soldatova, Saso Dzeroski (2014). Ontology of core data mining entities. Data 1101 | Min. Knowl. Discov.

    1102 |
    • 1103 | 1104 |
  • 1106 |

    Joaquin Vanschoren, Hendrik Blockeel, Bernhard Pfahringer, Geoffrey Holmes (2012). Experiment databases 1107 | - A new way to share, organize and learn from experiments. Machine Learning.

    1108 |
    • 1109 | 1110 |
  • 1112 |

    Diego Esteves, Diego Moussallem, Ciro Baron Neto, Tommaso Soru, Ricardo Usbeck, Markus Ackermann, 1113 | Jens Lehmann (2015). MEX vocabulary: a lightweight interchange format for machine learning 1114 | experiments. Proceedings of the 11th International Conference on Semantic Systems, SEMANTiCS 1115 | 2015, Vienna, Austria, September 15-17, 2015.

    1116 |
    • 1117 | 1118 |
  • 1120 |

    Lebo T., Sahoo S., McGuinness D., Belhajjame K., Cheney J., Corsar D., Garijo D., Soiland-Reyes S., 1121 | Zednik S., Zhao J. (2013). Prov-o: The prov ontology. W3C Recommendation.

    1122 |
    • 1123 | 1124 |
  • 1126 |

    Arp R., Smith B., Spear A. (2015). Building Ontologies with Basic Formal Ontology. The MIT Press.

    1127 |
    • 1128 | 1129 |
  • 1131 |

    Bandrowski A. et. al. (2016). The Ontology for Biomedical Investigations. PLoS ONE. 11(4):1–19

    1132 |
    • 1133 | 1134 |
  • 1136 |

    Neto, C. B., Esteves, D., Soru, T., Moussallem, D., Valdestilhas, A., & Marx, E. (2016). WASOTA: What Are the States Of The Art?. In SEMANTiCS (Posters, Demos, SuCCESS).

    1137 |
    • 1138 | 1139 |
  • 1141 |

    Vanschoren, J., Van Rijn, J. N., Bischl, B., & Torgo, L. (2014). OpenML: networked science in machine learning. ACM SIGKDD Explorations Newsletter, 15(2), 49-60.

    1142 |
    • 1143 | 1144 |
1145 |
1146 | 1147 |
1148 |

1149 | Footnotes 1150 |

1151 |
1152 |

1153 | ML Schema: 1154 | 1155 | http://ml-schema.github.io/documentation/ 1156 | 1157 | 1158 | 1159 | [back] 1160 | 1161 | 1162 |

1163 |
1164 |
1165 |

1166 | W3C ML Schema Community Group: 1167 | 1168 | https://www.w3.org/community/ml-schema/ 1169 | 1170 | 1171 | 1172 | [back] 1173 | 1174 | 1175 |

1176 |
1177 |
1178 |

1179 | OpenML: 1180 | 1181 | http://www.openml.org/ 1182 | 1183 | 1184 | 1185 | [back] 1186 | 1187 | 1188 |

1189 |
1190 |
1191 | 1192 | 1193 | 1194 | --------------------------------------------------------------------------------