├── .gitignore ├── LICENSE ├── MANIFEST.in ├── README.md ├── code └── dimsim │ ├── build │ └── lib │ │ └── dimsim │ │ └── dimsim.py │ └── dimsim │ └── dimsim.py ├── dimsim ├── __init__.py ├── core │ ├── __init__.py │ └── model.py ├── data │ ├── __init__.py │ ├── pinyin_to_simplified.pickle │ └── pinyin_to_traditional.pickle ├── tests │ ├── __init__.py │ └── test_dimsim.py └── utils │ ├── __init__.py │ ├── maps.py │ ├── pinyin.py │ └── utils.py ├── setup.cfg └── setup.py /.gitignore: -------------------------------------------------------------------------------- 1 | # Byte-compiled / optimized / DLL files 2 | __pycache__/ 3 | *.py[cod] 4 | *$py.class 5 | 6 | # C extensions 7 | *.so 8 | 9 | # Distribution / packaging 10 | .Python 11 | build/ 12 | develop-eggs/ 13 | dist/ 14 | downloads/ 15 | eggs/ 16 | .eggs/ 17 | lib/ 18 | lib64/ 19 | parts/ 20 | sdist/ 21 | var/ 22 | wheels/ 23 | *.egg-info/ 24 | .installed.cfg 25 | *.egg 26 | MANIFEST 27 | 28 | # PyInstaller 29 | # Usually these files are written by a python script from a template 30 | # before PyInstaller builds the exe, so as to inject date/other infos into it. 31 | *.manifest 32 | *.spec 33 | 34 | # Installer logs 35 | pip-log.txt 36 | pip-delete-this-directory.txt 37 | 38 | # Unit test / coverage reports 39 | htmlcov/ 40 | .tox/ 41 | .coverage 42 | .coverage.* 43 | .cache 44 | nosetests.xml 45 | coverage.xml 46 | *.cover 47 | .hypothesis/ 48 | .pytest_cache/ 49 | 50 | # Translations 51 | *.mo 52 | *.pot 53 | 54 | # Django stuff: 55 | *.log 56 | local_settings.py 57 | db.sqlite3 58 | 59 | # Flask stuff: 60 | instance/ 61 | .webassets-cache 62 | 63 | # Scrapy stuff: 64 | .scrapy 65 | 66 | # Sphinx documentation 67 | docs/_build/ 68 | 69 | # PyBuilder 70 | target/ 71 | 72 | # Jupyter Notebook 73 | .ipynb_checkpoints 74 | 75 | # pyenv 76 | .python-version 77 | 78 | # celery beat schedule file 79 | celerybeat-schedule 80 | 81 | # SageMath parsed files 82 | *.sage.py 83 | 84 | # Environments 85 | .env 86 | .venv 87 | env/ 88 | venv/ 89 | ENV/ 90 | env.bak/ 91 | venv.bak/ 92 | 93 | # Spyder project settings 94 | .spyderproject 95 | .spyproject 96 | 97 | # Rope project settings 98 | .ropeproject 99 | 100 | # mkdocs documentation 101 | /site 102 | 103 | # mypy 104 | .mypy_cache/ 105 | 106 | # local test files 107 | test_image_utils/ 108 | .idea/ 109 | 110 | # other 111 | maxfw/.DS_Store 112 | .vscode/settings.json 113 | .DS_Store 114 | -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- 1 | Apache License 2 | Version 2.0, January 2004 3 | http://www.apache.org/licenses/ 4 | 5 | TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION 6 | 7 | 1. Definitions. 8 | 9 | "License" shall mean the terms and conditions for use, reproduction, 10 | and distribution as defined by Sections 1 through 9 of this document. 11 | 12 | "Licensor" shall mean the copyright owner or entity authorized by 13 | the copyright owner that is granting the License. 14 | 15 | "Legal Entity" shall mean the union of the acting entity and all 16 | other entities that control, are controlled by, or are under common 17 | control with that entity. For the purposes of this definition, 18 | "control" means (i) the power, direct or indirect, to cause the 19 | direction or management of such entity, whether by contract or 20 | otherwise, or (ii) ownership of fifty percent (50%) or more of the 21 | outstanding shares, or (iii) beneficial ownership of such entity. 22 | 23 | "You" (or "Your") shall mean an individual or Legal Entity 24 | exercising permissions granted by this License. 25 | 26 | "Source" form shall mean the preferred form for making modifications, 27 | including but not limited to software source code, documentation 28 | source, and configuration files. 29 | 30 | "Object" form shall mean any form resulting from mechanical 31 | transformation or translation of a Source form, including but 32 | not limited to compiled object code, generated documentation, 33 | and conversions to other media types. 34 | 35 | "Work" shall mean the work of authorship, whether in Source or 36 | Object form, made available under the License, as indicated by a 37 | copyright notice that is included in or attached to the work 38 | (an example is provided in the Appendix below). 39 | 40 | "Derivative Works" shall mean any work, whether in Source or Object 41 | form, that is based on (or derived from) the Work and for which the 42 | editorial revisions, annotations, elaborations, or other modifications 43 | represent, as a whole, an original work of authorship. For the purposes 44 | of this License, Derivative Works shall not include works that remain 45 | separable from, or merely link (or bind by name) to the interfaces of, 46 | the Work and Derivative Works thereof. 47 | 48 | "Contribution" shall mean any work of authorship, including 49 | the original version of the Work and any modifications or additions 50 | to that Work or Derivative Works thereof, that is intentionally 51 | submitted to Licensor for inclusion in the Work by the copyright owner 52 | or by an individual or Legal Entity authorized to submit on behalf of 53 | the copyright owner. For the purposes of this definition, "submitted" 54 | means any form of electronic, verbal, or written communication sent 55 | to the Licensor or its representatives, including but not limited to 56 | communication on electronic mailing lists, source code control systems, 57 | and issue tracking systems that are managed by, or on behalf of, the 58 | Licensor for the purpose of discussing and improving the Work, but 59 | excluding communication that is conspicuously marked or otherwise 60 | designated in writing by the copyright owner as "Not a Contribution." 61 | 62 | "Contributor" shall mean Licensor and any individual or Legal Entity 63 | on behalf of whom a Contribution has been received by Licensor and 64 | subsequently incorporated within the Work. 65 | 66 | 2. Grant of Copyright License. Subject to the terms and conditions of 67 | this License, each Contributor hereby grants to You a perpetual, 68 | worldwide, non-exclusive, no-charge, royalty-free, irrevocable 69 | copyright license to reproduce, prepare Derivative Works of, 70 | publicly display, publicly perform, sublicense, and distribute the 71 | Work and such Derivative Works in Source or Object form. 72 | 73 | 3. Grant of Patent License. Subject to the terms and conditions of 74 | this License, each Contributor hereby grants to You a perpetual, 75 | worldwide, non-exclusive, no-charge, royalty-free, irrevocable 76 | (except as stated in this section) patent license to make, have made, 77 | use, offer to sell, sell, import, and otherwise transfer the Work, 78 | where such license applies only to those patent claims licensable 79 | by such Contributor that are necessarily infringed by their 80 | Contribution(s) alone or by combination of their Contribution(s) 81 | with the Work to which such Contribution(s) was submitted. If You 82 | institute patent litigation against any entity (including a 83 | cross-claim or counterclaim in a lawsuit) alleging that the Work 84 | or a Contribution incorporated within the Work constitutes direct 85 | or contributory patent infringement, then any patent licenses 86 | granted to You under this License for that Work shall terminate 87 | as of the date such litigation is filed. 88 | 89 | 4. Redistribution. You may reproduce and distribute copies of the 90 | Work or Derivative Works thereof in any medium, with or without 91 | modifications, and in Source or Object form, provided that You 92 | meet the following conditions: 93 | 94 | (a) You must give any other recipients of the Work or 95 | Derivative Works a copy of this License; and 96 | 97 | (b) You must cause any modified files to carry prominent notices 98 | stating that You changed the files; and 99 | 100 | (c) You must retain, in the Source form of any Derivative Works 101 | that You distribute, all copyright, patent, trademark, and 102 | attribution notices from the Source form of the Work, 103 | excluding those notices that do not pertain to any part of 104 | the Derivative Works; and 105 | 106 | (d) If the Work includes a "NOTICE" text file as part of its 107 | distribution, then any Derivative Works that You distribute must 108 | include a readable copy of the attribution notices contained 109 | within such NOTICE file, excluding those notices that do not 110 | pertain to any part of the Derivative Works, in at least one 111 | of the following places: within a NOTICE text file distributed 112 | as part of the Derivative Works; within the Source form or 113 | documentation, if provided along with the Derivative Works; or, 114 | within a display generated by the Derivative Works, if and 115 | wherever such third-party notices normally appear. The contents 116 | of the NOTICE file are for informational purposes only and 117 | do not modify the License. You may add Your own attribution 118 | notices within Derivative Works that You distribute, alongside 119 | or as an addendum to the NOTICE text from the Work, provided 120 | that such additional attribution notices cannot be construed 121 | as modifying the License. 122 | 123 | You may add Your own copyright statement to Your modifications and 124 | may provide additional or different license terms and conditions 125 | for use, reproduction, or distribution of Your modifications, or 126 | for any such Derivative Works as a whole, provided Your use, 127 | reproduction, and distribution of the Work otherwise complies with 128 | the conditions stated in this License. 129 | 130 | 5. Submission of Contributions. Unless You explicitly state otherwise, 131 | any Contribution intentionally submitted for inclusion in the Work 132 | by You to the Licensor shall be under the terms and conditions of 133 | this License, without any additional terms or conditions. 134 | Notwithstanding the above, nothing herein shall supersede or modify 135 | the terms of any separate license agreement you may have executed 136 | with Licensor regarding such Contributions. 137 | 138 | 6. Trademarks. This License does not grant permission to use the trade 139 | names, trademarks, service marks, or product names of the Licensor, 140 | except as required for reasonable and customary use in describing the 141 | origin of the Work and reproducing the content of the NOTICE file. 142 | 143 | 7. Disclaimer of Warranty. Unless required by applicable law or 144 | agreed to in writing, Licensor provides the Work (and each 145 | Contributor provides its Contributions) on an "AS IS" BASIS, 146 | WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or 147 | implied, including, without limitation, any warranties or conditions 148 | of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A 149 | PARTICULAR PURPOSE. You are solely responsible for determining the 150 | appropriateness of using or redistributing the Work and assume any 151 | risks associated with Your exercise of permissions under this License. 152 | 153 | 8. Limitation of Liability. In no event and under no legal theory, 154 | whether in tort (including negligence), contract, or otherwise, 155 | unless required by applicable law (such as deliberate and grossly 156 | negligent acts) or agreed to in writing, shall any Contributor be 157 | liable to You for damages, including any direct, indirect, special, 158 | incidental, or consequential damages of any character arising as a 159 | result of this License or out of the use or inability to use the 160 | Work (including but not limited to damages for loss of goodwill, 161 | work stoppage, computer failure or malfunction, or any and all 162 | other commercial damages or losses), even if such Contributor 163 | has been advised of the possibility of such damages. 164 | 165 | 9. Accepting Warranty or Additional Liability. While redistributing 166 | the Work or Derivative Works thereof, You may choose to offer, 167 | and charge a fee for, acceptance of support, warranty, indemnity, 168 | or other liability obligations and/or rights consistent with this 169 | License. However, in accepting such obligations, You may act only 170 | on Your own behalf and on Your sole responsibility, not on behalf 171 | of any other Contributor, and only if You agree to indemnify, 172 | defend, and hold each Contributor harmless for any liability 173 | incurred by, or claims asserted against, such Contributor by reason 174 | of your accepting any such warranty or additional liability. 175 | 176 | END OF TERMS AND CONDITIONS 177 | 178 | APPENDIX: How to apply the Apache License to your work. 179 | 180 | To apply the Apache License to your work, attach the following 181 | boilerplate notice, with the fields enclosed by brackets "[]" 182 | replaced with your own identifying information. (Don't include 183 | the brackets!) The text should be enclosed in the appropriate 184 | comment syntax for the file format. We also recommend that a 185 | file or class name and description of purpose be included on the 186 | same "printed page" as the copyright notice for easier 187 | identification within third-party archives. 188 | 189 | Copyright [yyyy] [name of copyright owner] 190 | 191 | Licensed under the Apache License, Version 2.0 (the "License"); 192 | you may not use this file except in compliance with the License. 193 | You may obtain a copy of the License at 194 | 195 | http://www.apache.org/licenses/LICENSE-2.0 196 | 197 | Unless required by applicable law or agreed to in writing, software 198 | distributed under the License is distributed on an "AS IS" BASIS, 199 | WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 200 | See the License for the specific language governing permissions and 201 | limitations under the License. 202 | -------------------------------------------------------------------------------- /MANIFEST.in: -------------------------------------------------------------------------------- 1 | include dimsim/data/*.* -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # DimSim - A Chinese Soundex Library (Python version) 2 | 3 | DimSim is a library developed by the Scalable Knowledge Intelligence team at IBM Almaden Research Center as part of the [SystemT](https://researcher.watson.ibm.com/researcher/view_group.php?id=1264) project. 4 | 5 | The PyPi project page can be found [here](https://pypi.org/project/dimsim/). It was created in collaboration with IBM Center for Open-Source Data and AI Technologies ([CODAIT](http://codait.org)). 6 | 7 | ## Overview 8 | We provide a phonetic algorithm for indexing Chinese characters by sound. The technical details can be found in the following [paper](http://aclweb.org/anthology/K18-1043): 9 | 10 | Min Li, Marina Danilevsky, Sara Noeman and Yunyao Li. *DIMSIM: An Accurate Chinese Phonetic Similarity Algorithm based on Learned High Dimensional Encoding*. CoNLL 2018. 11 | 12 | In this library, we provide a pre-trained model that can perform the following functions, in compliance with the phonetic principles of Mandarin Chinese as guided by the Romanization defined in [ISO 7098:2015](https://www.iso.org/standard/61420.html): 13 | - Given two Chinese phrases (of the same length), return the phonetic distance of the input phrases. Optionally you can feed in pinyin strings of Chinese phrases too. 14 | - Given a Chinese phrase, return its top-k similar (phoentically) Chinese phrases. 15 | 16 | 17 | 18 | ## How to install 19 | 20 | **Dependencies**: 21 | - [pypinyin](https://github.com/mozillazg/python-pinyin): used for translating Chinese characters into their correponding pinyins. 22 | 23 | There are two ways to install this library: 24 | - Install from PyPi 25 | 26 | ```shell 27 | pip install dimsim 28 | ``` 29 | - Download the source code by cloning this repo and compile it yourself. 30 | 31 | ```shell 32 | git clone git@github.com:System-T/DimSim.git 33 | 34 | cd DimSim/ 35 | 36 | pip install -e . 37 | ``` 38 | 39 | ## How to use 40 | Once you have the package installed you can use it for the two functions as shown below. 41 | 42 | - Computing phonetic distance of two Chinese phrases. The optional argument `pinyin` (False by default) can be used to provide a pinyin string list directly. See example usage below. 43 | 44 | ```python 45 | import dimsim 46 | 47 | dist = dimsim.get_distance("大侠","大虾") 48 | 0.0002380952380952381 49 | 50 | dist = dimsim.get_distance("大侠","大人") 51 | 25.001417183349876 52 | 53 | dist = dimsim.get_distance(['da4','xia2'],['da4','xia1']], pinyin=True) 54 | 0.0002380952380952381 55 | 56 | dist = dimsim.get_distance(['da4','xia2'],['da4','ren2']], pinyin=True) 57 | 25.001417183349876 58 | 59 | ``` 60 | *** 61 | - Return top-k phonetically similar phrases of a given Chinese phrase. Two parameters: 62 | - **mode** controls the character type of the returned Chinese phrases, where 'simplified' represents simplified Chinese and 'traditional' represents traditional Chinese. 63 | - **theta** controls the size of search space for the candidate phrases. 64 | ```python 65 | import dimsim 66 | 67 | candidates = dimsim.get_candidates("大侠", mode="simplified", theta=1) 68 | 69 | ['打下', '大虾', '大侠'] 70 | 71 | candidates = dimsim.get_candidates("粉丝", mode="traditinoal", theta=1) 72 | ['門市', '分時', '焚屍', '粉飾', '粉絲'] 73 | ``` 74 | 75 | *** 76 | Return Pinyin RCode 77 | ```python 78 | rcode = getRCode("海底小纵队") 79 | "hai'di'xiao'zong'dui" 80 | 81 | rcode = getRCode("海底小中队") 82 | "hai'di'xiao'zong'dui" 83 | ``` 84 | 85 | ## Citation 86 | 87 | Please cite the library by referencing the published paper: 88 | ``` 89 | @InProceedings{K18-1043, 90 | author = {Li, Min and Danilevsky, Marina and Noeman, Sara and Li, Yunyao}, 91 | title = {{DIMSIM:} An Accurate Chinese Phonetic Similarity Algorithm Based on Learned High Dimensional Encoding}, 92 | booktitle = {Proceedings of the 22nd Conference on Computational Natural Language Learning}, 93 | year = {2018}, 94 | publisher = {Association for Computational Linguistics}, 95 | pages = {444-453}, 96 | location = {Brussels, Belgium}, 97 | url = {http://aclweb.org/anthology/K18-1043} 98 | } 99 | ``` 100 | -------------------------------------------------------------------------------- /code/dimsim/build/lib/dimsim/dimsim.py: -------------------------------------------------------------------------------- 1 | 2 | # coding: utf-8 3 | 4 | # In[656]: 5 | 6 | 7 | from pypinyin import pinyin, lazy_pinyin, Style 8 | import sys 9 | import math 10 | import itertools 11 | import pickle 12 | import os 13 | 14 | ''' 15 | @liushaohua 16 | 本更改仅限个人使用，仅仅使用编码方式，对拼音进行一个统一编码 17 | ''' 18 | 19 | 20 | # In[657]: 21 | 22 | 23 | work_dir = os.path.abspath(__file__) 24 | indexOfLastSlash = work_dir.rfind("/") 25 | work_dir = work_dir[0:indexOfLastSlash]+"/" 26 | 27 | 28 | # In[658]: 29 | 30 | 31 | sfile = open(work_dir+'pinyin_to_simplified.pickle', 'rb') 32 | 33 | # dump information to that file 34 | pinyin_to_simplified = pickle.load(sfile) 35 | sfile.close() 36 | 37 | tfile = open(work_dir+'pinyin_to_traditional.pickle', 'rb') 38 | pinyin_to_traditional = pickle.load(tfile) 39 | tfile.close() 40 | 41 | 42 | # In[659]: 43 | 44 | 45 | def to_pinyin(utterance): 46 | length = len(utterance) 47 | translated = [] 48 | pinyin_encodings = pinyin(utterance, style=Style.TONE2) 49 | for i in range(length): 50 | currPinyin = pinyin_encodings[i][0] 51 | # print("{} translates to {}".format(currPinyin, putToneToEnd(currPinyin))) 52 | translated.append(putToneToEnd(currPinyin)) 53 | return translated 54 | 55 | def putToneToEnd(input_pinyin): 56 | if len(input_pinyin) is 1: 57 | return input_pinyin + '1' 58 | tone_index = 0 59 | tone = '1' 60 | for index, character in enumerate(input_pinyin): 61 | if character in ("1","2","3","4"): 62 | tone_index = index 63 | tone = input_pinyin[index] 64 | break; 65 | if tone_index is 0: 66 | return input_pinyin + "5" 67 | return input_pinyin[0:index] + input_pinyin[index+1:] + tone 68 | 69 | def get_distance(utterance1, utterance2): 70 | if(len(utterance1) is not len(utterance2)): 71 | print("the two inputs do not have the same length") 72 | return sys.float_info.max 73 | else: 74 | u1 = to_pinyin(utterance1) 75 | u2 = to_pinyin(utterance2) 76 | 77 | la = [] 78 | lb = [] 79 | for py in u1: 80 | la.append(Pinyin(py)) 81 | for py in u2: 82 | lb.append(Pinyin(py)) 83 | 84 | 85 | res = 0.0 86 | numDiff = 0 87 | tot = len(utterance1)*2.1 88 | for i in range (len(utterance1)): 89 | apy = la[i] 90 | bpy = lb[i] 91 | 92 | if (apy is None) or (bpy is None): 93 | print("!Error {},{}".format(la, lb)) 94 | 95 | res += getEditDistanceClose_TwoDCode(apy, bpy) 96 | 97 | if apy.consonant is not bpy.consonant: 98 | numDiff+=1 99 | 100 | if not(str(apy.vowel) == str(bpy.vowel)): 101 | numDiff+=1 102 | 103 | if apy.tone is not bpy.tone: 104 | numDiff+=0.01; 105 | 106 | diffRatio = (numDiff)/tot; 107 | a = 0 108 | if diffRatio is 0: 109 | a=1 110 | return res*diffRatio; 111 | 112 | def getEditDistanceClose_TwoDCode(a, b): 113 | res = 0 114 | try: 115 | if (a is None) or (b is None): 116 | print("Error:pinyin({},{})".format(a.toString(),b.toString())) 117 | return res 118 | 119 | twoDcode_consonant_a = consonantMap_TwoDCode[a.consonant] 120 | twoDcode_consonant_b = consonantMap_TwoDCode[b.consonant] 121 | 122 | cDis = abs(getDistance_TwoDCode(twoDcode_consonant_a, twoDcode_consonant_b)) 123 | 124 | twoDcode_vowel_a = vowelMap_TwoDCode[a.vowel] 125 | twoDcode_vowel_b = vowelMap_TwoDCode[b.vowel] 126 | 127 | vDis = abs(getDistance_TwoDCode(twoDcode_vowel_a, twoDcode_vowel_b)) 128 | 129 | hcDis = getSimDisFromHardcodMap(a,b) 130 | 131 | res = min((cDis+vDis),hcDis) + 1.0*abs(a.tone-b.tone)/10 132 | 133 | except: 134 | print("Error pinyin {}{}".format(a.toString(), b.toString())) 135 | raise 136 | return res 137 | 138 | def getSimDisFromHardcodMap(a, b): 139 | try: 140 | simPy = hardcodeMap[a.toStringNoTone()] 141 | if simPy is not None: 142 | if simPy is b.toStringNoTone(): 143 | return 2.0 144 | else: 145 | simPy=hardcodeMap[b.toStringNoTone()] 146 | if simPy is not None and simPy is a.toStringNoTone(): 147 | return 2.0 148 | return sys.float_info.max 149 | except: 150 | return sys.float_info.max 151 | 152 | 153 | def getDistance_TwoDCode(X, Y): 154 | x1, x2 = X 155 | y1, y2 = Y 156 | 157 | x1d = abs(x1-y1) 158 | x2d = abs(x2-y2) 159 | 160 | return math.sqrt( x1d**2 + x2d**2) 161 | 162 | 163 | consonantMap_TwoDCode ={ 164 | "b":(1.0,0.5), 165 | "p":(1.0,1.5), 166 | 167 | "g":(7.0,0.5), 168 | "k":(7.0,1.5), 169 | "h":(7.0,3.0), 170 | "f":(7.0,4.0), 171 | 172 | "d":(12.0,0.5), 173 | "t":(12.0,1.5), 174 | 175 | "n":(22.5,0.5), 176 | "l":(22.5,1.5), 177 | "r":( 22.5,2.5), 178 | 179 | 180 | "zh":(30,1.7), 181 | "z":(30,1.5), 182 | "j":(30.0,0.5), 183 | 184 | "ch":(31,1.7), 185 | "c":(31,1.5), 186 | "q":(31.0,0.5), 187 | 188 | "sh":(33,3.7), 189 | "s":(33,3.5), 190 | "x":(33,2.5), 191 | 192 | 193 | "m":(50.0,3.5), 194 | 195 | "y":(40.0,0.0), 196 | "w":(40,5.0), 197 | 198 | "":(99999.0,99999.0) 199 | } 200 | 201 | 202 | # In[662]: 203 | 204 | 205 | vowelMap_TwoDCode = { 206 | "a":(1.0,0.0), 207 | "an":(1.0,1.0), 208 | "ang":(1.0,1.5), 209 | 210 | 211 | "ia":(0.0,0.0), 212 | "ian":(0.0,1.0), 213 | "iang":(0.0,1.5), 214 | 215 | "ua":(2.0,0.0), 216 | "uan":(2.0,1.0), 217 | "uang":(2.0,1.5), 218 | "u:an":(2.0,1.0), 219 | 220 | 221 | "ao":(5.0,0.0), 222 | "iao":(5.0,1.5), 223 | 224 | "ai":(8.0,0.0), 225 | "uai":(8.0,1.5), 226 | 227 | 228 | 229 | "o":(20,0.0), 230 | "io":(20,2.5), 231 | "iou":(20,4), 232 | "iu":(20,4), 233 | "ou":(20,5.5), 234 | "uo":(20,6.0), 235 | 236 | "ong":(20,8.0), 237 | "iong":(20,9.5), 238 | 239 | 240 | "er":(41,1), 241 | "e":(41,0.0), 242 | 243 | "u:e":(40,5.0), 244 | "ve":(40,5.0), 245 | "ue":(40,5.0), 246 | "ie":(40,4.5), 247 | "ei":(40,4.0), 248 | "uei":(40,3.0), 249 | "ui":(40,3.0), 250 | 251 | "en":(42,0.5), 252 | "eng":(42,1.0), 253 | 254 | "uen":(43,0.5), 255 | "un":(43,0.5), 256 | "ueng":(43,1.0), 257 | 258 | 259 | "i":(60,1.0), 260 | "in":(60,2.5), 261 | "ing":(60,3.0), 262 | 263 | "u:":(61,1.0), 264 | "v":(61,1.0), 265 | "u:n":(61,2.5), 266 | "vn":(61,2.5), 267 | 268 | "u":(80,0.0), 269 | 270 | "":(99999.0,99999.0) 271 | } 272 | 273 | # 这里根据拼音相似的计算规则，对元音、辅音统一编码, 先卡严格一点, 274 | # (a, b) b值小于0.5的，编成一个，按这套相似计算规则，0.5也算是相似的 275 | # 如果这么干，可能会太泛了，导致线上效果太差 276 | consonantMap_RCode ={ 277 | "b":"b", 278 | "p":"p", 279 | 280 | "g":"g", 281 | "k":"k", 282 | "h":"h", 283 | "f":"f", 284 | 285 | "d":"d", 286 | "t":"t", 287 | 288 | "n":"l", 289 | "l":"l", 290 | "r":"r", 291 | 292 | 293 | "zh":"z", 294 | "z":"z", 295 | "j":"j", 296 | 297 | "ch":"c", 298 | "c":"c", 299 | "q":"q", 300 | 301 | "sh":"s", 302 | "s":"s", 303 | "x":"x", 304 | 305 | 306 | "m":"m", 307 | 308 | "y":"y", 309 | "w":"w", 310 | 311 | "":"C" 312 | } 313 | 314 | # 对于元音，(a, b) b值小于等于0.5，编码成一个 315 | vowelMap_RCode = { 316 | "a":"a", 317 | "an":"an", 318 | "ang":"an", 319 | 320 | 321 | "ia":"ia", 322 | "ian":"ian", 323 | "iang":"ian", 324 | 325 | "ua":"ua", 326 | "uan":"uan", 327 | "uang":"uan", 328 | "u:an":"u:an", 329 | 330 | 331 | "ao":"ao", 332 | "iao":"iao", 333 | 334 | "ai":"ai", 335 | "uai":"uai", 336 | 337 | 338 | 339 | "o":"o", 340 | "io":"io", 341 | "iou":"iou", 342 | "iu":"iu", 343 | "ou":"ou", 344 | "uo":"ou", 345 | 346 | "ong":"ong", 347 | "iong":"iong", 348 | 349 | 350 | "er":"er", 351 | "e":"e", 352 | 353 | "u:e":"ve", 354 | "ve":"ve", 355 | "ue":"ve", 356 | "ie":"ie", 357 | "ei":"ie", 358 | "uei":"uei", 359 | "ui":"ui", 360 | 361 | "en":"en", 362 | "eng":"en", 363 | 364 | "uen":"un", 365 | "un":"un", 366 | "ueng":"un", 367 | 368 | 369 | "i":"i", 370 | "in":"in", 371 | "ing":"in", 372 | 373 | "u:":"v", 374 | "v":"v", 375 | "u:n":"vn", 376 | "vn":"vn", 377 | 378 | "u":"u", 379 | 380 | "":"V" 381 | } 382 | 383 | # 对于数字来说，不做拼音转换，直接跳过 384 | rcode_ignore_set = set("0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZⅠ") 385 | 386 | # In[663]: 387 | 388 | 389 | consonantList = ["b", "p", "m", "f", "d", "t", "n", "l", "g", "k","h", "j", "q", "x", "zh", "ch", "sh", "r", "z", "c", "s","y", "w"] 390 | 391 | 392 | # In[664]: 393 | 394 | 395 | vowelList = ["a", "o", "e", "i", "u", "v","u:","er", "ao","ai", "ou","ei", "ia", "iao", "iu", "iou","ie", "ui","uei","ua","uo","uai", "u:e","ve", "an", "en", "in", "un","uen", "vn","u:n","ian","uan", "u:an","van", "ang", "eng", "ing", "ong","iang","iong","uang","ueng"] 396 | 397 | 398 | # In[665]: 399 | 400 | 401 | class Pinyin: 402 | consonantList = ["b", "p", "m", "f", "d", "t", "n", "l", "g", "k", "h", "j", "q", "x", "zh", "ch", "sh", "r", "z", "c", "s","y", "w"] 403 | vowelList = ["a", "o", "e", "i", "u", "v","u:","er", "ao","ai", "ou","ei", "ia", "iao", "iu", "iou","ie", "ui","uei","ua","uo","uai", "u:e","ve", "an", "en", "in", "un","uen", "vn","u:n","ian","uan", "u:an","van", "ang", "eng", "ing", "ong","iang","iong","uang","ueng"] 404 | 405 | def __init__(self, pinyinstr): 406 | self.tone = int(pinyinstr[-1]) 407 | self.locp = pinyinstr[0:-1].lower() 408 | self.consonant, self.vowel = self.parseConsonant(self.locp) 409 | # print("before rewriting consonant={}, vowel={}, locp={}, tone={}".format(self.consonant, 410 | # self.vowel, 411 | # self.locp, 412 | # self.tone)) 413 | self.pinyinRewrite() 414 | # print("after rewriting consonant={}, vowel={}, locp={}, tone={}".format(self.consonant, 415 | # self.vowel, 416 | # self.locp, 417 | # self.tone)) 418 | 419 | def parseConsonant(self, pinyin): 420 | for consonant in consonantList: 421 | if pinyin.startswith(consonant): 422 | return (consonant, pinyin[len(consonant):]) 423 | # it's a vowel without consonant 424 | if pinyin in vowelList: 425 | return None, pinyin.lower() 426 | 427 | print("Invalid Pinyin, please check!") 428 | return None, None 429 | 430 | def toStringNoTone(self): 431 | return "{}{}".format(self.consonant, self.vowel) 432 | 433 | def toStringWithTone(self): 434 | return "{}{}{}".format(self.consonant, self.vowel, self.tone) 435 | 436 | def toString(self): 437 | return "{}{}{}".format(self.consonant, self.vowel, self.tone) 438 | 439 | def pinyinRewrite(self): 440 | import re 441 | yVowels = {"u","ue","uan","un","u:","u:e","u:an","u:n"} 442 | tconsonant = {"j","g","x"} 443 | if 'v' in self.vowel: 444 | self.vowel = self.vowel.replace("v", "u:") 445 | 446 | if self.consonant is None or self.consonant is "": 447 | self.consonant = "" 448 | return 449 | if self.consonant is "y": 450 | if self.vowel in yVowels: 451 | if "u:" not in self.vowel: 452 | self.vowel = self.vowel.replace("u","u:") 453 | else: 454 | self.vowel="i"+self.vowel 455 | regex = re.compile("i+") 456 | self.vowel = self.vowel.replace("iii","i") 457 | self.vowel = self.vowel.replace("ii","i") 458 | self.consonant="" 459 | 460 | if self.consonant is "w": 461 | self.vowel="u"+self.vowel; 462 | self.vowel=self.vowel.replace("uuu","u") 463 | self.vowel=self.vowel.replace("uu","u") 464 | self.consonant = "" 465 | 466 | if (self.consonant in tconsonant) and (self.vowel is "u") or (self.vowel is "v"): 467 | self.vowel="u:" 468 | 469 | if self.vowel is "iou": 470 | self.vowel = "iu" 471 | 472 | if self.vowel is "uei": 473 | self.vowel = "ui" 474 | 475 | if self.vowel is "uen": 476 | self.vowel = "un" 477 | 478 | 479 | 480 | 481 | 482 | 483 | # In[666]: 484 | 485 | 486 | hardcodeMap = { 487 | "hua":"fa", 488 | "fa":"hua", 489 | "huan":"fan", 490 | "fan":"huan", 491 | "hui":"fei", 492 | "jie":"zhe", 493 | "kou":"ke", 494 | "gou":"ge", 495 | "zhong":"zen", 496 | "san":"shang" 497 | } 498 | 499 | 500 | # In[667]: 501 | 502 | 503 | consonantMap = { 504 | "b":1.0, 505 | "p":2.0, 506 | 507 | "m":11.0, 508 | "f":12.0, 509 | 510 | "d":21.0, 511 | "t":22.0, 512 | 513 | "n":31.0, 514 | "l":31.0, 515 | "r":32.0, 516 | 517 | "g":41.0, 518 | "k":42.0, 519 | "h":43.0, 520 | 521 | "j":46.0, 522 | "q":47.0, 523 | "x":48.0, 524 | 525 | "z":61.0, 526 | "c":62.0, 527 | 528 | "zh":71.0, 529 | "ch":72.0, 530 | 531 | "sh":81.0, 532 | "s":82.0, 533 | 534 | "y":90.0, 535 | "w":100.0, 536 | 537 | "":99999.0, 538 | "__v":99999.0 539 | } 540 | 541 | 542 | # In[668]: 543 | 544 | 545 | vowelMap = { 546 | "ia":0.0, 547 | "a":2.0, 548 | "ai":3.0, 549 | "uai":4.0, 550 | "iao":6.0, 551 | "ao":7.0, 552 | 553 | "uan":10.0, 554 | "an":11.0, 555 | "ang":12.0, 556 | "ian":14.0, 557 | "iang":15.0, 558 | "uang":17.0, 559 | "ua":18.0, 560 | 561 | "o":21.0, 562 | "io":22.0, 563 | "ou":23.0, 564 | "uo":24.0, 565 | "ong":26.0, 566 | "iong":27.0, 567 | 568 | "e":31.0, 569 | "ei":33.0, 570 | "ie":34.0, 571 | "er":37.0, 572 | 573 | "ve":40.0, 574 | "ue":40.0, 575 | "u:e":40.0, 576 | 577 | "en":43.0, 578 | "eng":44.0, 579 | 580 | "uen":45.0, 581 | "ueng":45.0, 582 | 583 | "u:en":42.0, 584 | "ven":42.0, 585 | 586 | "i":50.0, 587 | "u:":51.0, 588 | "v":51.0, 589 | "u:n":53.0, 590 | "vn":53.0, 591 | "u:an":55.0, 592 | "v:an":55.0, 593 | 594 | "in":53.0, 595 | "ing":55.0, 596 | 597 | "u":60.0, 598 | "ui":63.0, 599 | "uei":63.0, 600 | "iu":64.0, 601 | "iou":64.0, 602 | "un":66.0, 603 | 604 | "":99999.0, 605 | "__v":99999.0 606 | } 607 | 608 | 609 | # ### Get Pinyin Candidates that are close to an input pinyin 610 | 611 | # In[644]: 612 | 613 | 614 | doubleConsonantsMap = {} 615 | doubleVowelsMap = {} 616 | 617 | def getClosePinyinCandids(word, theta=2): 618 | res = [] 619 | word_pinyin = to_pinyin(word) 620 | word_py = Pinyin(word_pinyin[0]) 621 | 622 | cCandids = getConsonantCandids(theta, word_py) 623 | for i in range(len(cCandids)): 624 | if cCandids[i] == word_py.consonant or cCandids[i] == '__v': 625 | continue 626 | for j in range(1,5,1): 627 | newPy = cCandids[i]+word_py.vowel+str(j) 628 | res.append(Pinyin(newPy)) 629 | 630 | vCandids = getVowelCandids(theta, word_py) 631 | for i in range(len(vCandids)): 632 | for j in range(1,5,1): 633 | if word_py.consonant is None: 634 | # print(word_py.toStringWithTone(),"has none consonant") 635 | newPy = vCandids[i]+str(j) 636 | else: 637 | newPy = word_py.consonant+vCandids[i]+str(j) 638 | res.append(Pinyin(newPy)) 639 | return res 640 | 641 | 642 | 643 | def getConsonantCandids(theta, word_py): 644 | populateDoubleConsonantsMap() 645 | res = [] 646 | curCode = 0 647 | if word_py.consonant is None: 648 | orgCode = consonantMap["__v"] 649 | else: 650 | orgCode = consonantMap[word_py.consonant] 651 | for i in range(int(orgCode-theta), int(orgCode+theta), 1): 652 | if float(i) in doubleConsonantsMap: 653 | cand = doubleConsonantsMap[float(i)] 654 | if cand is not None: 655 | res += cand 656 | return res 657 | 658 | 659 | def getVowelCandids(theta, word_py): 660 | populateDoubleVowelsMap() 661 | res = [] 662 | curCode = 0 663 | orgCode = vowelMap[word_py.vowel] 664 | for i in range(int(orgCode-theta), int(orgCode+theta), 1): 665 | if float(i) in doubleVowelsMap: 666 | cand = doubleVowelsMap[float(i)] 667 | if cand is not None: 668 | res += cand 669 | return res 670 | 671 | def populateDoubleConsonantsMap(): 672 | if len(doubleConsonantsMap) is not 0: 673 | return 674 | hmCdouble = consonantMap 675 | for consonant in hmCdouble: 676 | if hmCdouble[consonant] not in doubleConsonantsMap: 677 | doubleConsonantsMap[hmCdouble[consonant]] = [] 678 | 679 | doubleConsonantsMap[hmCdouble[consonant]].append(consonant) 680 | 681 | def populateDoubleVowelsMap(): 682 | if len(doubleVowelsMap) is not 0: 683 | return 684 | hmVdouble = vowelMap 685 | for vowel in hmVdouble: 686 | if hmVdouble[vowel] not in doubleVowelsMap: 687 | doubleVowelsMap[hmVdouble[vowel]] = [] 688 | 689 | doubleVowelsMap[hmVdouble[vowel]].append(vowel) 690 | 691 | 692 | def getCandidates(sentence, mode="simplified", theta=1): 693 | candidates = [] 694 | words_candidates = [] 695 | for word in sentence: 696 | candid = getClosePinyinCandids(word, theta) 697 | words_candidates.append(candid) 698 | all_combinations = itertools.product(*words_candidates) 699 | counter = 0 700 | for combination in all_combinations: 701 | counter+=1 702 | searchKey = "" 703 | for i in combination: 704 | searchKey = searchKey + i.toStringWithTone().replace("None","") + " " 705 | if mode is "simplified": 706 | if searchKey.strip() in pinyin_to_simplified: 707 | candidates+=pinyin_to_simplified[searchKey.strip()] 708 | else: 709 | if searchKey.strip() in pinyin_to_traditional: 710 | candidates+=pinyin_to_traditional[searchKey.strip()] 711 | return candidates 712 | 713 | 714 | # 编码成RCode，这个可以根据需要来增加或减少泛化 715 | def getRCode(sentence): 716 | rcode = [] 717 | pinyin = [] 718 | for wd in sentence: 719 | if wd in rcode_ignore_set: 720 | pinyin.append(wd) 721 | else: 722 | pinyin.append(to_pinyin(wd)[0]) 723 | for py in pinyin: 724 | if py in rcode_ignore_set: 725 | rcode.append(py) 726 | else: 727 | new_py = Pinyin(py) 728 | code = consonantMap_RCode[new_py.consonant] + vowelMap_RCode[new_py.vowel] 729 | rcode.append(code) 730 | 731 | return "'".join(rcode) 732 | 733 | 734 | 735 | 736 | 737 | -------------------------------------------------------------------------------- /code/dimsim/dimsim/dimsim.py: -------------------------------------------------------------------------------- 1 | 2 | # coding: utf-8 3 | 4 | # In[656]: 5 | 6 | 7 | from pypinyin import pinyin, lazy_pinyin, Style 8 | import sys 9 | import math 10 | import itertools 11 | import pickle 12 | import os 13 | 14 | ''' 15 | @liushaohua 16 | 本更改仅限个人使用，仅仅使用编码方式，对拼音进行一个统一编码 17 | ''' 18 | 19 | 20 | # In[657]: 21 | 22 | 23 | work_dir = os.path.abspath(__file__) 24 | indexOfLastSlash = work_dir.rfind("/") 25 | work_dir = work_dir[0:indexOfLastSlash]+"/" 26 | 27 | 28 | # In[658]: 29 | 30 | 31 | sfile = open(work_dir+'pinyin_to_simplified.pickle', 'rb') 32 | 33 | # dump information to that file 34 | pinyin_to_simplified = pickle.load(sfile) 35 | sfile.close() 36 | 37 | tfile = open(work_dir+'pinyin_to_traditional.pickle', 'rb') 38 | pinyin_to_traditional = pickle.load(tfile) 39 | tfile.close() 40 | 41 | 42 | # In[659]: 43 | 44 | 45 | def to_pinyin(utterance): 46 | length = len(utterance) 47 | translated = [] 48 | pinyin_encodings = pinyin(utterance, style=Style.TONE2) 49 | for i in range(length): 50 | currPinyin = pinyin_encodings[i][0] 51 | # print("{} translates to {}".format(currPinyin, putToneToEnd(currPinyin))) 52 | translated.append(putToneToEnd(currPinyin)) 53 | return translated 54 | 55 | def putToneToEnd(input_pinyin): 56 | if len(input_pinyin) is 1: 57 | return input_pinyin + '1' 58 | tone_index = 0 59 | tone = '1' 60 | for index, character in enumerate(input_pinyin): 61 | if character in ("1","2","3","4"): 62 | tone_index = index 63 | tone = input_pinyin[index] 64 | break; 65 | if tone_index is 0: 66 | return input_pinyin + "5" 67 | return input_pinyin[0:index] + input_pinyin[index+1:] + tone 68 | 69 | def get_distance(utterance1, utterance2): 70 | if(len(utterance1) is not len(utterance2)): 71 | print("the two inputs do not have the same length") 72 | return sys.float_info.max 73 | else: 74 | u1 = to_pinyin(utterance1) 75 | u2 = to_pinyin(utterance2) 76 | 77 | la = [] 78 | lb = [] 79 | for py in u1: 80 | la.append(Pinyin(py)) 81 | for py in u2: 82 | lb.append(Pinyin(py)) 83 | 84 | 85 | res = 0.0 86 | numDiff = 0 87 | tot = len(utterance1)*2.1 88 | for i in range (len(utterance1)): 89 | apy = la[i] 90 | bpy = lb[i] 91 | 92 | if (apy is None) or (bpy is None): 93 | print("!Error {},{}".format(la, lb)) 94 | 95 | res += getEditDistanceClose_TwoDCode(apy, bpy) 96 | 97 | if apy.consonant is not bpy.consonant: 98 | numDiff+=1 99 | 100 | if not(str(apy.vowel) == str(bpy.vowel)): 101 | numDiff+=1 102 | 103 | if apy.tone is not bpy.tone: 104 | numDiff+=0.01; 105 | 106 | diffRatio = (numDiff)/tot; 107 | a = 0 108 | if diffRatio is 0: 109 | a=1 110 | return res*diffRatio; 111 | 112 | def getEditDistanceClose_TwoDCode(a, b): 113 | res = 0 114 | try: 115 | if (a is None) or (b is None): 116 | print("Error:pinyin({},{})".format(a.toString(),b.toString())) 117 | return res 118 | 119 | twoDcode_consonant_a = consonantMap_TwoDCode[a.consonant] 120 | twoDcode_consonant_b = consonantMap_TwoDCode[b.consonant] 121 | 122 | cDis = abs(getDistance_TwoDCode(twoDcode_consonant_a, twoDcode_consonant_b)) 123 | 124 | twoDcode_vowel_a = vowelMap_TwoDCode[a.vowel] 125 | twoDcode_vowel_b = vowelMap_TwoDCode[b.vowel] 126 | 127 | vDis = abs(getDistance_TwoDCode(twoDcode_vowel_a, twoDcode_vowel_b)) 128 | 129 | hcDis = getSimDisFromHardcodMap(a,b) 130 | 131 | res = min((cDis+vDis),hcDis) + 1.0*abs(a.tone-b.tone)/10 132 | 133 | except: 134 | print("Error pinyin {}{}".format(a.toString(), b.toString())) 135 | raise 136 | return res 137 | 138 | def getSimDisFromHardcodMap(a, b): 139 | try: 140 | simPy = hardcodeMap[a.toStringNoTone()] 141 | if simPy is not None: 142 | if simPy is b.toStringNoTone(): 143 | return 2.0 144 | else: 145 | simPy=hardcodeMap[b.toStringNoTone()] 146 | if simPy is not None and simPy is a.toStringNoTone(): 147 | return 2.0 148 | return sys.float_info.max 149 | except: 150 | return sys.float_info.max 151 | 152 | 153 | def getDistance_TwoDCode(X, Y): 154 | x1, x2 = X 155 | y1, y2 = Y 156 | 157 | x1d = abs(x1-y1) 158 | x2d = abs(x2-y2) 159 | 160 | return math.sqrt( x1d**2 + x2d**2) 161 | 162 | 163 | consonantMap_TwoDCode ={ 164 | "b":(1.0,0.5), 165 | "p":(1.0,1.5), 166 | 167 | "g":(7.0,0.5), 168 | "k":(7.0,1.5), 169 | "h":(7.0,3.0), 170 | "f":(7.0,4.0), 171 | 172 | "d":(12.0,0.5), 173 | "t":(12.0,1.5), 174 | 175 | "n":(22.5,0.5), 176 | "l":(22.5,1.5), 177 | "r":( 22.5,2.5), 178 | 179 | 180 | "zh":(30,1.7), 181 | "z":(30,1.5), 182 | "j":(30.0,0.5), 183 | 184 | "ch":(31,1.7), 185 | "c":(31,1.5), 186 | "q":(31.0,0.5), 187 | 188 | "sh":(33,3.7), 189 | "s":(33,3.5), 190 | "x":(33,2.5), 191 | 192 | 193 | "m":(50.0,3.5), 194 | 195 | "y":(40.0,0.0), 196 | "w":(40,5.0), 197 | 198 | "":(99999.0,99999.0) 199 | } 200 | 201 | 202 | # In[662]: 203 | 204 | 205 | vowelMap_TwoDCode = { 206 | "a":(1.0,0.0), 207 | "an":(1.0,1.0), 208 | "ang":(1.0,1.5), 209 | 210 | 211 | "ia":(0.0,0.0), 212 | "ian":(0.0,1.0), 213 | "iang":(0.0,1.5), 214 | 215 | "ua":(2.0,0.0), 216 | "uan":(2.0,1.0), 217 | "uang":(2.0,1.5), 218 | "u:an":(2.0,1.0), 219 | 220 | 221 | "ao":(5.0,0.0), 222 | "iao":(5.0,1.5), 223 | 224 | "ai":(8.0,0.0), 225 | "uai":(8.0,1.5), 226 | 227 | 228 | 229 | "o":(20,0.0), 230 | "io":(20,2.5), 231 | "iou":(20,4), 232 | "iu":(20,4), 233 | "ou":(20,5.5), 234 | "uo":(20,6.0), 235 | 236 | "ong":(20,8.0), 237 | "iong":(20,9.5), 238 | 239 | 240 | "er":(41,1), 241 | "e":(41,0.0), 242 | 243 | "u:e":(40,5.0), 244 | "ve":(40,5.0), 245 | "ue":(40,5.0), 246 | "ie":(40,4.5), 247 | "ei":(40,4.0), 248 | "uei":(40,3.0), 249 | "ui":(40,3.0), 250 | 251 | "en":(42,0.5), 252 | "eng":(42,1.0), 253 | 254 | "uen":(43,0.5), 255 | "un":(43,0.5), 256 | "ueng":(43,1.0), 257 | 258 | 259 | "i":(60,1.0), 260 | "in":(60,2.5), 261 | "ing":(60,3.0), 262 | 263 | "u:":(61,1.0), 264 | "v":(61,1.0), 265 | "u:n":(61,2.5), 266 | "vn":(61,2.5), 267 | 268 | "u":(80,0.0), 269 | 270 | "":(99999.0,99999.0) 271 | } 272 | 273 | # 这里根据拼音相似的计算规则，对元音、辅音统一编码, 先卡严格一点, 274 | # (a, b) b值小于0.5的，编成一个，按这套相似计算规则，0.5也算是相似的 275 | # 如果这么干，可能会太泛了，导致线上效果太差 276 | consonantMap_RCode ={ 277 | "b":"b", 278 | "p":"p", 279 | 280 | "g":"g", 281 | "k":"k", 282 | "h":"h", 283 | "f":"f", 284 | 285 | "d":"d", 286 | "t":"t", 287 | 288 | "n":"l", 289 | "l":"l", 290 | "r":"r", 291 | 292 | 293 | "zh":"z", 294 | "z":"z", 295 | "j":"j", 296 | 297 | "ch":"c", 298 | "c":"c", 299 | "q":"q", 300 | 301 | "sh":"s", 302 | "s":"s", 303 | "x":"x", 304 | 305 | 306 | "m":"m", 307 | 308 | "y":"y", 309 | "w":"w", 310 | 311 | "":"C" 312 | } 313 | 314 | # 对于元音，(a, b) b值小于等于0.5，编码成一个 315 | vowelMap_RCode = { 316 | "a":"a", 317 | "an":"an", 318 | "ang":"an", 319 | 320 | 321 | "ia":"ia", 322 | "ian":"ian", 323 | "iang":"ian", 324 | 325 | "ua":"ua", 326 | "uan":"uan", 327 | "uang":"uan", 328 | "u:an":"u:an", 329 | 330 | 331 | "ao":"ao", 332 | "iao":"iao", 333 | 334 | "ai":"ai", 335 | "uai":"uai", 336 | 337 | 338 | 339 | "o":"o", 340 | "io":"io", 341 | "iou":"iou", 342 | "iu":"iu", 343 | "ou":"ou", 344 | "uo":"ou", 345 | 346 | "ong":"ong", 347 | "iong":"iong", 348 | 349 | 350 | "er":"er", 351 | "e":"e", 352 | 353 | "u:e":"ve", 354 | "ve":"ve", 355 | "ue":"ve", 356 | "ie":"ie", 357 | "ei":"ie", 358 | "uei":"uei", 359 | "ui":"ui", 360 | 361 | "en":"en", 362 | "eng":"en", 363 | 364 | "uen":"un", 365 | "un":"un", 366 | "ueng":"un", 367 | 368 | 369 | "i":"i", 370 | "in":"in", 371 | "ing":"in", 372 | 373 | "u:":"v", 374 | "v":"v", 375 | "u:n":"vn", 376 | "vn":"vn", 377 | 378 | "u":"u", 379 | 380 | "":"V" 381 | } 382 | 383 | # 对于数字来说，不做拼音转换，直接跳过 384 | rcode_ignore_set = set("0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZⅠ") 385 | 386 | # In[663]: 387 | 388 | 389 | consonantList = ["b", "p", "m", "f", "d", "t", "n", "l", "g", "k","h", "j", "q", "x", "zh", "ch", "sh", "r", "z", "c", "s","y", "w"] 390 | 391 | 392 | # In[664]: 393 | 394 | 395 | vowelList = ["a", "o", "e", "i", "u", "v","u:","er", "ao","ai", "ou","ei", "ia", "iao", "iu", "iou","ie", "ui","uei","ua","uo","uai", "u:e","ve", "an", "en", "in", "un","uen", "vn","u:n","ian","uan", "u:an","van", "ang", "eng", "ing", "ong","iang","iong","uang","ueng"] 396 | 397 | 398 | # In[665]: 399 | 400 | 401 | class Pinyin: 402 | consonantList = ["b", "p", "m", "f", "d", "t", "n", "l", "g", "k", "h", "j", "q", "x", "zh", "ch", "sh", "r", "z", "c", "s","y", "w"] 403 | vowelList = ["a", "o", "e", "i", "u", "v","u:","er", "ao","ai", "ou","ei", "ia", "iao", "iu", "iou","ie", "ui","uei","ua","uo","uai", "u:e","ve", "an", "en", "in", "un","uen", "vn","u:n","ian","uan", "u:an","van", "ang", "eng", "ing", "ong","iang","iong","uang","ueng"] 404 | 405 | def __init__(self, pinyinstr): 406 | self.tone = int(pinyinstr[-1]) 407 | self.locp = pinyinstr[0:-1].lower() 408 | self.consonant, self.vowel = self.parseConsonant(self.locp) 409 | # print("before rewriting consonant={}, vowel={}, locp={}, tone={}".format(self.consonant, 410 | # self.vowel, 411 | # self.locp, 412 | # self.tone)) 413 | self.pinyinRewrite() 414 | # print("after rewriting consonant={}, vowel={}, locp={}, tone={}".format(self.consonant, 415 | # self.vowel, 416 | # self.locp, 417 | # self.tone)) 418 | 419 | def parseConsonant(self, pinyin): 420 | for consonant in consonantList: 421 | if pinyin.startswith(consonant): 422 | return (consonant, pinyin[len(consonant):]) 423 | # it's a vowel without consonant 424 | if pinyin in vowelList: 425 | return None, pinyin.lower() 426 | 427 | print("Invalid Pinyin, please check!") 428 | return None, None 429 | 430 | def toStringNoTone(self): 431 | return "{}{}".format(self.consonant, self.vowel) 432 | 433 | def toStringWithTone(self): 434 | return "{}{}{}".format(self.consonant, self.vowel, self.tone) 435 | 436 | def toString(self): 437 | return "{}{}{}".format(self.consonant, self.vowel, self.tone) 438 | 439 | def pinyinRewrite(self): 440 | import re 441 | yVowels = {"u","ue","uan","un","u:","u:e","u:an","u:n"} 442 | tconsonant = {"j","g","x"} 443 | if 'v' in self.vowel: 444 | self.vowel = self.vowel.replace("v", "u:") 445 | 446 | if self.consonant is None or self.consonant is "": 447 | self.consonant = "" 448 | return 449 | if self.consonant is "y": 450 | if self.vowel in yVowels: 451 | if "u:" not in self.vowel: 452 | self.vowel = self.vowel.replace("u","u:") 453 | else: 454 | self.vowel="i"+self.vowel 455 | regex = re.compile("i+") 456 | self.vowel = self.vowel.replace("iii","i") 457 | self.vowel = self.vowel.replace("ii","i") 458 | self.consonant="" 459 | 460 | if self.consonant is "w": 461 | self.vowel="u"+self.vowel; 462 | self.vowel=self.vowel.replace("uuu","u") 463 | self.vowel=self.vowel.replace("uu","u") 464 | self.consonant = "" 465 | 466 | if (self.consonant in tconsonant) and (self.vowel is "u") or (self.vowel is "v"): 467 | self.vowel="u:" 468 | 469 | if self.vowel == "iou": 470 | self.vowel = "iu" 471 | 472 | if self.vowel == "uei": 473 | self.vowel = "ui" 474 | 475 | if self.vowel == "uen": 476 | self.vowel = "un" 477 | 478 | 479 | 480 | 481 | 482 | 483 | # In[666]: 484 | 485 | 486 | hardcodeMap = { 487 | "hua":"fa", 488 | "fa":"hua", 489 | "huan":"fan", 490 | "fan":"huan", 491 | "hui":"fei", 492 | "jie":"zhe", 493 | "kou":"ke", 494 | "gou":"ge", 495 | "zhong":"zen", 496 | "san":"shang" 497 | } 498 | 499 | 500 | # In[667]: 501 | 502 | 503 | consonantMap = { 504 | "b":1.0, 505 | "p":2.0, 506 | 507 | "m":11.0, 508 | "f":12.0, 509 | 510 | "d":21.0, 511 | "t":22.0, 512 | 513 | "n":31.0, 514 | "l":31.0, 515 | "r":32.0, 516 | 517 | "g":41.0, 518 | "k":42.0, 519 | "h":43.0, 520 | 521 | "j":46.0, 522 | "q":47.0, 523 | "x":48.0, 524 | 525 | "z":61.0, 526 | "c":62.0, 527 | 528 | "zh":71.0, 529 | "ch":72.0, 530 | 531 | "sh":81.0, 532 | "s":82.0, 533 | 534 | "y":90.0, 535 | "w":100.0, 536 | 537 | "":99999.0, 538 | "__v":99999.0 539 | } 540 | 541 | 542 | # In[668]: 543 | 544 | 545 | vowelMap = { 546 | "ia":0.0, 547 | "a":2.0, 548 | "ai":3.0, 549 | "uai":4.0, 550 | "iao":6.0, 551 | "ao":7.0, 552 | 553 | "uan":10.0, 554 | "an":11.0, 555 | "ang":12.0, 556 | "ian":14.0, 557 | "iang":15.0, 558 | "uang":17.0, 559 | "ua":18.0, 560 | 561 | "o":21.0, 562 | "io":22.0, 563 | "ou":23.0, 564 | "uo":24.0, 565 | "ong":26.0, 566 | "iong":27.0, 567 | 568 | "e":31.0, 569 | "ei":33.0, 570 | "ie":34.0, 571 | "er":37.0, 572 | 573 | "ve":40.0, 574 | "ue":40.0, 575 | "u:e":40.0, 576 | 577 | "en":43.0, 578 | "eng":44.0, 579 | 580 | "uen":45.0, 581 | "ueng":45.0, 582 | 583 | "u:en":42.0, 584 | "ven":42.0, 585 | 586 | "i":50.0, 587 | "u:":51.0, 588 | "v":51.0, 589 | "u:n":53.0, 590 | "vn":53.0, 591 | "u:an":55.0, 592 | "v:an":55.0, 593 | 594 | "in":53.0, 595 | "ing":55.0, 596 | 597 | "u":60.0, 598 | "ui":63.0, 599 | "uei":63.0, 600 | "iu":64.0, 601 | "iou":64.0, 602 | "un":66.0, 603 | 604 | "":99999.0, 605 | "__v":99999.0 606 | } 607 | 608 | 609 | # ### Get Pinyin Candidates that are close to an input pinyin 610 | 611 | # In[644]: 612 | 613 | 614 | doubleConsonantsMap = {} 615 | doubleVowelsMap = {} 616 | 617 | def getClosePinyinCandids(word, theta=2): 618 | res = [] 619 | word_pinyin = to_pinyin(word) 620 | word_py = Pinyin(word_pinyin[0]) 621 | 622 | cCandids = getConsonantCandids(theta, word_py) 623 | for i in range(len(cCandids)): 624 | if cCandids[i] == word_py.consonant or cCandids[i] == '__v': 625 | continue 626 | for j in range(1,5,1): 627 | newPy = cCandids[i]+word_py.vowel+str(j) 628 | res.append(Pinyin(newPy)) 629 | 630 | vCandids = getVowelCandids(theta, word_py) 631 | for i in range(len(vCandids)): 632 | for j in range(1,5,1): 633 | if word_py.consonant is None: 634 | # print(word_py.toStringWithTone(),"has none consonant") 635 | newPy = vCandids[i]+str(j) 636 | else: 637 | newPy = word_py.consonant+vCandids[i]+str(j) 638 | res.append(Pinyin(newPy)) 639 | return res 640 | 641 | 642 | 643 | def getConsonantCandids(theta, word_py): 644 | populateDoubleConsonantsMap() 645 | res = [] 646 | curCode = 0 647 | if word_py.consonant is None: 648 | orgCode = consonantMap["__v"] 649 | else: 650 | orgCode = consonantMap[word_py.consonant] 651 | for i in range(int(orgCode-theta), int(orgCode+theta), 1): 652 | if float(i) in doubleConsonantsMap: 653 | cand = doubleConsonantsMap[float(i)] 654 | if cand is not None: 655 | res += cand 656 | return res 657 | 658 | 659 | def getVowelCandids(theta, word_py): 660 | populateDoubleVowelsMap() 661 | res = [] 662 | curCode = 0 663 | orgCode = vowelMap[word_py.vowel] 664 | for i in range(int(orgCode-theta), int(orgCode+theta), 1): 665 | if float(i) in doubleVowelsMap: 666 | cand = doubleVowelsMap[float(i)] 667 | if cand is not None: 668 | res += cand 669 | return res 670 | 671 | def populateDoubleConsonantsMap(): 672 | if len(doubleConsonantsMap) is not 0: 673 | return 674 | hmCdouble = consonantMap 675 | for consonant in hmCdouble: 676 | if hmCdouble[consonant] not in doubleConsonantsMap: 677 | doubleConsonantsMap[hmCdouble[consonant]] = [] 678 | 679 | doubleConsonantsMap[hmCdouble[consonant]].append(consonant) 680 | 681 | def populateDoubleVowelsMap(): 682 | if len(doubleVowelsMap) is not 0: 683 | return 684 | hmVdouble = vowelMap 685 | for vowel in hmVdouble: 686 | if hmVdouble[vowel] not in doubleVowelsMap: 687 | doubleVowelsMap[hmVdouble[vowel]] = [] 688 | 689 | doubleVowelsMap[hmVdouble[vowel]].append(vowel) 690 | 691 | 692 | def getCandidates(sentence, mode="simplified", theta=1): 693 | candidates = [] 694 | words_candidates = [] 695 | for word in sentence: 696 | candid = getClosePinyinCandids(word, theta) 697 | words_candidates.append(candid) 698 | all_combinations = itertools.product(*words_candidates) 699 | counter = 0 700 | for combination in all_combinations: 701 | counter+=1 702 | searchKey = "" 703 | for i in combination: 704 | searchKey = searchKey + i.toStringWithTone().replace("None","") + " " 705 | if mode is "simplified": 706 | if searchKey.strip() in pinyin_to_simplified: 707 | candidates+=pinyin_to_simplified[searchKey.strip()] 708 | else: 709 | if searchKey.strip() in pinyin_to_traditional: 710 | candidates+=pinyin_to_traditional[searchKey.strip()] 711 | return candidates 712 | 713 | 714 | # 编码成RCode，这个可以根据需要来增加或减少泛化 715 | def getRCode(sentence): 716 | rcode = [] 717 | pinyin = [] 718 | for wd in sentence: 719 | if wd in rcode_ignore_set: 720 | pinyin.append(wd) 721 | else: 722 | pinyin.append(to_pinyin(wd)[0]) 723 | for py in pinyin: 724 | if py in rcode_ignore_set: 725 | rcode.append(py) 726 | else: 727 | new_py = Pinyin(py) 728 | code = consonantMap_RCode[new_py.consonant] + vowelMap_RCode[new_py.vowel] 729 | rcode.append(code) 730 | 731 | return "'".join(rcode) 732 | 733 | 734 | 735 | 736 | 737 | -------------------------------------------------------------------------------- /dimsim/__init__.py: -------------------------------------------------------------------------------- 1 | from .core.model import get_distance 2 | from .core.model import get_candidates -------------------------------------------------------------------------------- /dimsim/core/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/System-T/DimSim/77261afc51e5d0995b5905b9ed1c45317082d613/dimsim/core/__init__.py -------------------------------------------------------------------------------- /dimsim/core/model.py: -------------------------------------------------------------------------------- 1 | from pypinyin import pinyin, lazy_pinyin, Style 2 | import sys 3 | import math 4 | import itertools 5 | import os 6 | 7 | from dimsim.utils.pinyin import Pinyin, load_pinyin_to_simplified, load_pinyin_to_traditional 8 | from dimsim.utils.utils import get_edit_distance_close_2d_code, to_pinyin 9 | from dimsim.utils.maps import vowelMap, consonantMap 10 | 11 | doubleConsonantsMap = {} 12 | doubleVowelsMap = {} 13 | 14 | pinyin_to_simplified = load_pinyin_to_simplified() 15 | pinyin_to_traditional = load_pinyin_to_traditional() 16 | 17 | def get_distance(utterance1, utterance2, pinyin=False): 18 | ''' 19 | Calculates the distances between embeddings of two Chinese words. 20 | input: 21 | utterance1, utterance2: utf-8 strings for Chinese words or 22 | pinyin strings. 23 | pinyin : Boolean - indicates if words are in Chinese or Pinyin. 24 | output: 25 | distance - float. 26 | ''' 27 | assert (len(utterance1) == len(utterance2)),"The two inputs do not have the same length" 28 | 29 | if not pinyin: 30 | u1 = to_pinyin(utterance1) 31 | u2 = to_pinyin(utterance2) 32 | 33 | else: 34 | u1 = utterance1 35 | u2 = utterance2 36 | 37 | la = [] 38 | lb = [] 39 | for py in u1: 40 | la.append(Pinyin(py)) 41 | for py in u2: 42 | lb.append(Pinyin(py)) 43 | 44 | 45 | res = 0.0 46 | numDiff = 0 47 | tot = len(utterance1)*2.1 48 | for i in range (len(utterance1)): 49 | apy = la[i] 50 | bpy = lb[i] 51 | 52 | if (apy is None) or (bpy is None): 53 | raise Exception("!Empty Pinyin {},{}".format(la, lb)) 54 | res += get_edit_distance_close_2d_code(apy, bpy) 55 | 56 | if apy.consonant is not bpy.consonant: 57 | numDiff+=1 58 | 59 | if not(str(apy.vowel) == str(bpy.vowel)): 60 | numDiff+=1 61 | 62 | if apy.tone is not bpy.tone: 63 | numDiff+=0.01 64 | 65 | diffRatio = (numDiff)/tot 66 | return res*diffRatio 67 | 68 | 69 | def get_candidates(sentence, mode="simplified", theta=1): 70 | ''' 71 | Gets similar sounding words / candidates based on embeddings. 72 | inputs: 73 | sentence - utf-8 string with the Chinese words. 74 | outputs: 75 | candidates - a list containing utf-8 string Chinese words. 76 | ''' 77 | candidates = [] 78 | words_candidates = [] 79 | for word in sentence: 80 | candid = _get_close_pinyin_candids(word, theta) 81 | words_candidates.append(candid) 82 | all_combinations = itertools.product(*words_candidates) 83 | counter = 0 84 | for combination in all_combinations: 85 | counter+=1 86 | searchKey = "" 87 | for i in combination: 88 | searchKey = searchKey + i.toStringWithTone().replace("None","") + " " 89 | if mode is "simplified": 90 | if searchKey.strip() in pinyin_to_simplified: 91 | candidates+=pinyin_to_simplified[searchKey.strip()] 92 | else: 93 | if searchKey.strip() in pinyin_to_traditional: 94 | candidates+=pinyin_to_traditional[searchKey.strip()] 95 | return candidates 96 | 97 | 98 | def _get_close_pinyin_candids(word, theta=2): 99 | res = [] 100 | word_pinyin = to_pinyin(word) 101 | word_py = Pinyin(word_pinyin[0]) 102 | 103 | cCandids = _get_consonant_candids(theta, word_py) 104 | for i in range(len(cCandids)): 105 | if cCandids[i] == word_py.consonant: 106 | continue 107 | for j in range(1,5,1): 108 | newPy = cCandids[i]+word_py.vowel+str(j) 109 | res.append(Pinyin(newPy)) 110 | 111 | vCandids = _get_vowel_candids(theta, word_py) 112 | for i in range(len(vCandids)): 113 | for j in range(1,5,1): 114 | if word_py.consonant is None: 115 | newPy = vCandids[i]+str(j) 116 | else: 117 | newPy = word_py.consonant+vCandids[i]+str(j) 118 | res.append(Pinyin(newPy)) 119 | return res 120 | 121 | def _get_consonant_candids(theta, word_py): 122 | _populate_double_consonants_map() 123 | res = [] 124 | if word_py.consonant is None: 125 | orgCode = consonantMap["__v"] 126 | else: 127 | orgCode = consonantMap[word_py.consonant] 128 | for i in range(int(orgCode-theta), int(orgCode+theta), 1): 129 | if float(i) in doubleConsonantsMap: 130 | cand = doubleConsonantsMap[float(i)] 131 | if cand is not None: 132 | res += cand 133 | return res 134 | 135 | 136 | def _get_vowel_candids(theta, word_py): 137 | _populate_double_vowels_map() 138 | res = [] 139 | orgCode = vowelMap[word_py.vowel] 140 | for i in range(int(orgCode-theta), int(orgCode+theta), 1): 141 | if float(i) in doubleVowelsMap: 142 | cand = doubleVowelsMap[float(i)] 143 | if cand is not None: 144 | res += cand 145 | return res 146 | 147 | def _populate_double_consonants_map(): 148 | if len(doubleConsonantsMap) is not 0: 149 | return 150 | hmCdouble = consonantMap 151 | for consonant in hmCdouble: 152 | if hmCdouble[consonant] not in doubleConsonantsMap: 153 | doubleConsonantsMap[hmCdouble[consonant]] = [] 154 | 155 | doubleConsonantsMap[hmCdouble[consonant]].append(consonant) 156 | 157 | def _populate_double_vowels_map(): 158 | if len(doubleVowelsMap) is not 0: 159 | return 160 | hmVdouble = vowelMap 161 | for vowel in hmVdouble: 162 | if hmVdouble[vowel] not in doubleVowelsMap: 163 | doubleVowelsMap[hmVdouble[vowel]] = [] 164 | 165 | doubleVowelsMap[hmVdouble[vowel]].append(vowel) 166 | -------------------------------------------------------------------------------- /dimsim/data/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/System-T/DimSim/77261afc51e5d0995b5905b9ed1c45317082d613/dimsim/data/__init__.py -------------------------------------------------------------------------------- /dimsim/data/pinyin_to_simplified.pickle: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/System-T/DimSim/77261afc51e5d0995b5905b9ed1c45317082d613/dimsim/data/pinyin_to_simplified.pickle -------------------------------------------------------------------------------- /dimsim/data/pinyin_to_traditional.pickle: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/System-T/DimSim/77261afc51e5d0995b5905b9ed1c45317082d613/dimsim/data/pinyin_to_traditional.pickle -------------------------------------------------------------------------------- /dimsim/tests/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/System-T/DimSim/77261afc51e5d0995b5905b9ed1c45317082d613/dimsim/tests/__init__.py -------------------------------------------------------------------------------- /dimsim/tests/test_dimsim.py: -------------------------------------------------------------------------------- 1 | # -*- coding: utf-8 -*- 2 | 3 | # Standard libs 4 | import io 5 | import os 6 | # Dependencies 7 | import pytest 8 | 9 | # The module to test 10 | from dimsim.core.model import get_distance, get_candidates 11 | 12 | def test_distance_near(): 13 | dist = get_distance(u'大侠',u'大虾') 14 | assert dist == 0.0002380952380952381 15 | 16 | def test_distance_far(): 17 | dist = get_distance(u'大侠',u'大人') 18 | assert dist == 25.001417183349876 19 | 20 | def test_distance_pinyin(): 21 | dist = get_distance(['da4','xia2'],['da4','xia1'],pinyin=True) 22 | assert dist == 0.0002380952380952381 23 | 24 | def test_invalid_input(): 25 | pytest.raises(AssertionError, get_distance, u'大侠', u'大') 26 | 27 | def test_get_candidates_simplified(): 28 | candidates = get_candidates(u'大侠', mode='simplified', theta=1) 29 | for c in candidates: 30 | assert c in [u'打下', u'大虾', u'大侠'] 31 | 32 | def test_get_candidates_traditional(): 33 | candidates = get_candidates(u'粉丝', mode='traditional', theta=1) 34 | for c in candidates: 35 | assert c in [u'門市', u'分時', u'焚屍', u'粉飾', u'粉絲'] 36 | 37 | if __name__ == '__main__': 38 | pytest.main([__file__]) 39 | -------------------------------------------------------------------------------- /dimsim/utils/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/System-T/DimSim/77261afc51e5d0995b5905b9ed1c45317082d613/dimsim/utils/__init__.py -------------------------------------------------------------------------------- /dimsim/utils/maps.py: -------------------------------------------------------------------------------- 1 | consonantMap_TwoDCode ={ 2 | "b":(1.0,0.5), 3 | "p":(1.0,1.5), 4 | 5 | "g":(7.0,0.5), 6 | "k":(7.0,1.5), 7 | "h":(7.0,3.0), 8 | "f":(7.0,4.0), 9 | 10 | "d":(12.0,0.5), 11 | "t":(12.0,1.5), 12 | 13 | "n":(22.5,0.5), 14 | "l":(22.5,1.5), 15 | "r":( 22.5,2.5), 16 | 17 | 18 | "zh":(30,1.7), 19 | "z":(30,1.5), 20 | "j":(30.0,0.5), 21 | 22 | "ch":(31,1.7), 23 | "c":(31,1.5), 24 | "q":(31.0,0.5), 25 | 26 | "sh":(33,3.7), 27 | "s":(33,3.5), 28 | "x":(33,2.5), 29 | 30 | 31 | "m":(50.0,3.5), 32 | 33 | "y":(40.0,0.0), 34 | "w":(40,5.0), 35 | 36 | "":(99999.0,99999.0) 37 | } 38 | 39 | vowelMap_TwoDCode = { 40 | "a":(1.0,0.0), 41 | "an":(1.0,1.0), 42 | "ang":(1.0,1.5), 43 | 44 | 45 | "ia":(0.0,0.0), 46 | "ian":(0.0,1.0), 47 | "iang":(0.0,1.5), 48 | 49 | "ua":(2.0,0.0), 50 | "uan":(2.0,1.0), 51 | "uang":(2.0,1.5), 52 | "u:an":(2.0,1.0), 53 | 54 | 55 | "ao":(5.0,0.0), 56 | "iao":(5.0,1.5), 57 | 58 | "ai":(8.0,0.0), 59 | "uai":(8.0,1.5), 60 | 61 | 62 | 63 | "o":(20,0.0), 64 | "io":(20,2.5), 65 | "iou":(20,4), 66 | "iu":(20,4), 67 | "ou":(20,5.5), 68 | "uo":(20,6.0), 69 | 70 | "ong":(20,8.0), 71 | "iong":(20,9.5), 72 | 73 | 74 | "er":(41,1), 75 | "e":(41,0.0), 76 | 77 | "u:e":(40,5.0), 78 | "ve":(40,5.0), 79 | "ue":(40,5.0), 80 | "ie":(40,4.5), 81 | "ei":(40,4.0), 82 | "uei":(40,3.0), 83 | "ui":(40,3.0), 84 | 85 | "en":(42,0.5), 86 | "eng":(42,1.0), 87 | 88 | "uen":(43,0.5), 89 | "un":(43,0.5), 90 | "ueng":(43,1.0), 91 | 92 | 93 | "i":(60,1.0), 94 | "in":(60,2.5), 95 | "ing":(60,3.0), 96 | 97 | "u:":(61,1.0), 98 | "v":(61,1.0), 99 | "u:n":(61,2.5), 100 | "vn":(61,2.5), 101 | 102 | "u":(80,0.0), 103 | 104 | "":(99999.0,99999.0) 105 | } 106 | 107 | consonantList = ["b", "p", "m", "f", "d", "t", "n", "l", "g", "k","h", "j", "q", "x", "zh", "ch", "sh", "r", "z", "c", "s","y", "w"] 108 | 109 | vowelList = ["a", "o", "e", "i", "u", "v","u:","er", "ao","ai", "ou","ei", "ia", "iao", "iu", "iou","ie", "ui","uei","ua","uo","uai", "u:e","ve", "an", "en", "in", "un","uen", "vn","u:n","ian","uan", "u:an","van", "ang", "eng", "ing", "ong","iang","iong","uang","ueng"] 110 | 111 | hardcodeMap = { 112 | "hua":"fa", 113 | "fa":"hua", 114 | "huan":"fan", 115 | "fan":"huan", 116 | "hui":"fei", 117 | "jie":"zhe", 118 | "kou":"ke", 119 | "gou":"ge", 120 | "zhong":"zen", 121 | "san":"shang" 122 | } 123 | 124 | consonantMap = { 125 | "b":1.0, 126 | "p":2.0, 127 | 128 | "m":11.0, 129 | "f":12.0, 130 | 131 | "d":21.0, 132 | "t":22.0, 133 | 134 | "n":31.0, 135 | "l":31.0, 136 | "r":32.0, 137 | 138 | "g":41.0, 139 | "k":42.0, 140 | "h":43.0, 141 | 142 | "j":46.0, 143 | "q":47.0, 144 | "x":48.0, 145 | 146 | "z":61.0, 147 | "c":62.0, 148 | 149 | "zh":71.0, 150 | "ch":72.0, 151 | 152 | "sh":81.0, 153 | "s":82.0, 154 | 155 | "y":90.0, 156 | "w":100.0, 157 | 158 | "":99999.0, 159 | "__v":99999.0 160 | } 161 | 162 | vowelMap = { 163 | "ia":0.0, 164 | "a":2.0, 165 | "ai":3.0, 166 | "uai":4.0, 167 | "iao":6.0, 168 | "ao":7.0, 169 | 170 | "uan":10.0, 171 | "an":11.0, 172 | "ang":12.0, 173 | "ian":14.0, 174 | "iang":15.0, 175 | "uang":17.0, 176 | "ua":18.0, 177 | 178 | "o":21.0, 179 | "io":22.0, 180 | "ou":23.0, 181 | "uo":24.0, 182 | "ong":26.0, 183 | "iong":27.0, 184 | 185 | "e":31.0, 186 | "ei":33.0, 187 | "ie":34.0, 188 | "er":37.0, 189 | 190 | "ve":40.0, 191 | "ue":40.0, 192 | "u:e":40.0, 193 | 194 | "en":43.0, 195 | "eng":44.0, 196 | 197 | "uen":45.0, 198 | "ueng":45.0, 199 | 200 | "u:en":42.0, 201 | "ven":42.0, 202 | 203 | "i":50.0, 204 | "u:":51.0, 205 | "v":51.0, 206 | "u:n":53.0, 207 | "vn":53.0, 208 | "u:an":55.0, 209 | "v:an":55.0, 210 | 211 | "in":53.0, 212 | "ing":55.0, 213 | 214 | "u":60.0, 215 | "ui":63.0, 216 | "uei":63.0, 217 | "iu":64.0, 218 | "iou":64.0, 219 | "un":66.0, 220 | 221 | "":99999.0, 222 | "__v":99999.0 223 | } 224 | -------------------------------------------------------------------------------- /dimsim/utils/pinyin.py: -------------------------------------------------------------------------------- 1 | import os 2 | import pickle 3 | 4 | from pypinyin import pinyin, lazy_pinyin, Style 5 | from dimsim.utils import maps 6 | 7 | class Pinyin: 8 | consonantList = maps.consonantList 9 | vowelList = maps.vowelList 10 | 11 | def __init__(self, pinyinstr): 12 | self.tone = int(pinyinstr[-1]) 13 | self.locp = pinyinstr[0:-1].lower() 14 | self.consonant, self.vowel = self.parseConsonant(self.locp) 15 | self.pinyinRewrite() 16 | 17 | def parseConsonant(self, pinyin): 18 | for consonant in self.consonantList: 19 | if pinyin.startswith(consonant): 20 | return (consonant, pinyin[len(consonant):]) 21 | # it's a vowel without consonant 22 | if pinyin in self.vowelList: 23 | return None, pinyin.lower() 24 | 25 | print("Invalid Pinyin, please check!") 26 | return None, None 27 | 28 | def toStringNoTone(self): 29 | return "{}{}".format(self.consonant, self.vowel) 30 | 31 | def toStringWithTone(self): 32 | return "{}{}{}".format(self.consonant, self.vowel, self.tone) 33 | 34 | def toString(self): 35 | return "{}{}{}".format(self.consonant, self.vowel, self.tone) 36 | 37 | def pinyinRewrite(self): 38 | import re 39 | yVowels = {"u","ue","uan","un","u:","u:e","u:an","u:n"} 40 | tconsonant = {"j","g","x"} 41 | if 'v' in self.vowel: 42 | self.vowel = self.vowel.replace("v", "u:") 43 | 44 | if self.consonant is None or self.consonant is "": 45 | self.consonant = "" 46 | return 47 | if self.consonant is "y": 48 | if self.vowel in yVowels: 49 | if "u:" not in self.vowel: 50 | self.vowel = self.vowel.replace("u","u:") 51 | else: 52 | self.vowel="i"+self.vowel 53 | regex = re.compile("i+") 54 | self.vowel = self.vowel.replace("iii","i") 55 | self.vowel = self.vowel.replace("ii","i") 56 | self.consonant="" 57 | 58 | if self.consonant is "w": 59 | self.vowel="u"+self.vowel 60 | self.vowel=self.vowel.replace("uuu","u") 61 | self.vowel=self.vowel.replace("uu","u") 62 | self.consonant = "" 63 | 64 | if (self.consonant in tconsonant) and (self.vowel is "u") or (self.vowel is "v"): 65 | self.vowel="u:" 66 | 67 | if self.vowel is "iou": 68 | self.vowel = "iu" 69 | 70 | if self.vowel is "uei": 71 | self.vowel = "ui" 72 | 73 | if self.vowel is "uen": 74 | self.vowel = "un" 75 | 76 | def load_pinyin_to_simplified(): 77 | 78 | curr_dir, _ = os.path.split(__file__) 79 | root_dir, _ = os.path.split(curr_dir) 80 | DATA_PATH = os.path.join(root_dir, "data", "pinyin_to_simplified.pickle") 81 | sfile = open(DATA_PATH, 'rb') 82 | pinyin_to_simplified = pickle.load(sfile) 83 | sfile.close() 84 | return pinyin_to_simplified 85 | 86 | 87 | def load_pinyin_to_traditional(): 88 | 89 | curr_dir, _ = os.path.split(__file__) 90 | root_dir, _ = os.path.split(curr_dir) 91 | DATA_PATH = os.path.join(root_dir, "data", "pinyin_to_traditional.pickle") 92 | tfile = open(DATA_PATH, 'rb') 93 | pinyin_to_traditional = pickle.load(tfile) 94 | tfile.close() 95 | return pinyin_to_traditional 96 | -------------------------------------------------------------------------------- /dimsim/utils/utils.py: -------------------------------------------------------------------------------- 1 | import sys 2 | import math 3 | from pypinyin import pinyin, lazy_pinyin, Style 4 | 5 | from dimsim.utils.maps import consonantMap_TwoDCode, vowelMap_TwoDCode, hardcodeMap 6 | 7 | def to_pinyin(utterance): 8 | length = len(utterance) 9 | translated = [] 10 | pinyin_encodings = pinyin(utterance, style=Style.TONE2) 11 | for i in range(length): 12 | currPinyin = pinyin_encodings[i][0] 13 | translated.append(put_tone_to_end(currPinyin)) 14 | return translated 15 | 16 | def put_tone_to_end(input_pinyin): 17 | if len(input_pinyin) is 1: 18 | return input_pinyin + '1' 19 | tone_index = 0 20 | tone = '1' 21 | for index, character in enumerate(input_pinyin): 22 | if character in ("1","2","3","4"): 23 | tone_index = index 24 | tone = input_pinyin[index] 25 | break 26 | if tone_index is 0: 27 | return input_pinyin + "5" 28 | return input_pinyin[0:index] + input_pinyin[index+1:] + tone 29 | 30 | def get_edit_distance_close_2d_code(a, b): 31 | res = 0 32 | try: 33 | if (a is None) or (b is None): 34 | print("Error:pinyin({},{})".format(a.toString(),b.toString())) 35 | return res 36 | 37 | twoDcode_consonant_a = consonantMap_TwoDCode[a.consonant] 38 | twoDcode_consonant_b = consonantMap_TwoDCode[b.consonant] 39 | 40 | cDis = abs(get_distance_2d_code(twoDcode_consonant_a, twoDcode_consonant_b)) 41 | 42 | twoDcode_vowel_a = vowelMap_TwoDCode[a.vowel] 43 | twoDcode_vowel_b = vowelMap_TwoDCode[b.vowel] 44 | 45 | vDis = abs(get_distance_2d_code(twoDcode_vowel_a, twoDcode_vowel_b)) 46 | 47 | hcDis = get_sim_dis_from_hardcod_map(a,b) 48 | 49 | res = min((cDis+vDis),hcDis) + 1.0*abs(a.tone-b.tone)/10 50 | 51 | except: 52 | raise Exception("Error pinyin {}{}".format(a.toString(), b.toString())) 53 | return res 54 | 55 | def get_sim_dis_from_hardcod_map(a, b): 56 | try: 57 | simPy = hardcodeMap[a.toStringNoTone()] 58 | if simPy is not None: 59 | if simPy is b.toStringNoTone(): 60 | return 2.0 61 | else: 62 | simPy=hardcodeMap[b.toStringNoTone()] 63 | if simPy is not None and simPy is a.toStringNoTone(): 64 | return 2.0 65 | return sys.float_info.max 66 | except: 67 | return sys.float_info.max 68 | 69 | def get_distance_2d_code(X, Y): 70 | x1, x2 = X 71 | y1, y2 = Y 72 | 73 | x1d = abs(x1-y1) 74 | x2d = abs(x2-y2) 75 | 76 | return math.sqrt( x1d**2 + x2d**2) -------------------------------------------------------------------------------- /setup.cfg: -------------------------------------------------------------------------------- 1 | [metadata] 2 | description-file = README.md -------------------------------------------------------------------------------- /setup.py: -------------------------------------------------------------------------------- 1 | from setuptools import setup 2 | 3 | with open("README.md", "r") as fh: 4 | long_description = fh.read() 5 | 6 | setup(name='dimsim', 7 | version='0.2.2', 8 | description='Python implementation of the Chinese soundex project DimSim', 9 | long_description=long_description, 10 | long_description_content_type="text/markdown", 11 | author='IBM SystemT, IBM CODAIT', 12 | author_email='qian.kun@ibm.com, karthik.muthuraman@ibm.com, ihjhuo@ibm.com, frreiss@us.ibm.com', 13 | url='https://github.com/System-T/DimSim', 14 | packages=['dimsim', 'dimsim.core', 'dimsim.utils', 'dimsim.data'], 15 | package_data={'':['dimsim/data/pinyin_to_simplified.pickle','dimsim/data/pinyin_to_traditional.pickle']}, 16 | include_package_data=True, 17 | classifiers=['License :: OSI Approved :: Apache Software License'], 18 | install_requires=[ 19 | 'pypinyin', 20 | ], 21 | test_suite='nose.collector', 22 | tests_require=['nose'] 23 | ) 24 | --------------------------------------------------------------------------------