├── LICENSE.txt
├── README.md
└── json
├── emotion-detection-dev.json
├── emotion-detection-trn.json
└── emotion-detection-tst.json
/LICENSE.txt:
--------------------------------------------------------------------------------
1 | Copyright 2017, Emory University
2 |
3 | Licensed under the Apache License, Version 2.0 (the "License");
4 | you may not use this file except in compliance with the License.
5 | You may obtain a copy of the License at
6 |
7 | http://www.apache.org/licenses/LICENSE-2.0
8 |
9 | Unless required by applicable law or agreed to in writing, software
10 | distributed under the License is distributed on an "AS IS" BASIS,
11 | WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
12 | See the License for the specific language governing permissions and
13 | limitations under the License.
--------------------------------------------------------------------------------
/README.md:
--------------------------------------------------------------------------------
1 | # Emotion Detection
2 |
3 | Emotion Detection aims to classify a fine-grained emotion for each utterance in multiparty dialogue.
4 | Our annotation is based on the primary emotions in the Feeling Wheel (Willcox, 1982).
5 | We must admit that the inter-annotator agreement of this annotation is not the greatest; we welcome any contribution from the community to improve the annotation quality.
6 | This task is a part of the [Character Mining](../../../character-mining) project led by the [Emory NLP](http://nlp.mathcs.emory.edu) research group.
7 |
8 |
9 |
10 |
11 |
12 |
13 | ## Dataset
14 |
15 | Each utterance is annotated with one of the seven emotions, *sad*, *mad*, *scared*, *powerful*, *peaceful*, *joyful*, and *neutral*, that are the primary emotions in the Feeling Wheel.
16 |
17 | * Latest release: [v1.0](https://github.com/emorynlp/emotion-detection/archive/emotion-detection-1.0.tar.gz).
18 | * [Release notes](https://github.com/emorynlp/emotion-detection/releases).
19 |
20 | ## Statistics
21 |
22 | The following episodes are used for the training, development, and evaluation sets:
23 |
24 | * Train (TRN): [s01\_e02, s01\_e03, s01\_e04, s01\_e05, s01\_e06, s01\_e07, s01\_e08, s01\_e09, s01\_e11, s01\_e12, s01\_e13, s01\_e14, s01\_e16, s01\_e17, s01\_e18, s01\_e19, s01\_e21, s01\_e22, s01\_e23, s01\_e24, s02\_e01, s02\_e02, s02\_e03, s02\_e04, s02\_e05, s02\_e06, s02\_e07, s02\_e09, s02\_e11, s02\_e12, s02\_e13, s02\_e14, s02\_e15, s02\_e16, s02\_e17, s02\_e18, s02\_e19, s02\_e21, s02\_e22, s02\_e24, s03\_e02, s03\_e03, s03\_e04, s03\_e05, s03\_e06, s03\_e07, s03\_e10, s03\_e11, s03\_e12, s03\_e13, s03\_e14, s03\_e15, s03\_e16, s03\_e17, s03\_e18, s03\_e19, s03\_e22, s03\_e23, s03\_e24, s03\_e25, s04\_e03, s04\_e04, s04\_e05, s04\_e07, s04\_e08, s04\_e09, s04\_e11, s04\_e12, s04\_e13, s04\_e14, s04\_e15, s04\_e16, s04\_e18, s04\_e19, s04\_e22, s04\_e23, s04\_e24]
25 | * Development (DEV): [s01\_e15, s01\_e20, s02\_e10, s02\_e20, s03\_e01, s03\_e09, s03\_e21, s04\_e01, s04\_e06, s04\_e10, s04\_e21]
26 | * Evaluation (TST): [s01\_e01, s01\_e10, s02\_e08, s02\_e23, s03\_e08, s03\_e20, s04\_e02, s04\_e17, s04\_e20]
27 |
28 | | Dataset | Episodes | Scenes | Utterances |
29 | |:-------:|---------:|-------:|-----------:|
30 | | TRN | 77 | 713 | 9,934 |
31 | | DEV | 11 | 99 | 1,344 |
32 | | TST | 9 | 85 | 1,328 |
33 | | Total | 97 | 897 | 12,606 |
34 |
35 | | Dataset | Neutral | Joyful | Peaceful | Powerful | Scared | Mad | Sad | Total |
36 | |:-------:|--------:|-------:|---------:|---------:|-------:|------:|----:|-------:|
37 | | TRN | 3,034 | 2,184 | 900 | 784 | 1,285 | 1,076 | 671 | 9,934 |
38 | | DEV | 393 | 289 | 132 | 134 | 178 | 143 | 75 | 1,344 |
39 | | TST | 349 | 282 | 159 | 145 | 182 | 113 | 98 | 1,328 |
40 | | Total | 3,776 | 2,755 | 1,191 | 1,063 | 1,645 | 1,332 | 844 | 12,606 |
41 |
42 | ## Annotation
43 |
44 | Each utterance has the field `emotion`.
45 | Three utterances in the following example are annotated with the emotions of *Neutral*, *Joyful*, and *Powerful*, respectively.
46 |
47 | ```json
48 | {
49 | "utterance_id": "s01_e02_c01_u002",
50 | "speakers": ["Joey Tribbiani"],
51 | "transcript": "Yeah, right!.......Y'serious?",
52 | "tokens": [
53 | ["Yeah", ",", "right", "!"],
54 | ["......."],
55 | ["Y'serious", "?"]
56 | ],
57 | "emotion": "Neutral"
58 | },
59 | {
60 | "utterance_id": "s01_e02_c01_u003",
61 | "speakers": ["Phoebe Buffay"],
62 | "transcript": "Oh, yeah!",
63 | "tokens": [
64 | ["Oh", ",", "yeah", "!"]
65 | ],
66 | "emotion": "Joyful"
67 | },
68 | {
69 | "utterance_id": "s01_e02_c01_u004",
70 | "speakers": ["Rachel Green"],
71 | "transcript": "Everything you need to know is in that first kiss.",
72 | "tokens": [
73 | ["Everything", "you", "need", "to", "know", "is", "in", "that", "first", "kiss", "."]
74 | ],
75 | "emotion": "Powerful"
76 | }
77 | ```
78 |
79 | ## Citation
80 |
81 | * [Emotion Detection on TV Show Transcripts with Sequence-based Convolutional Neural Networks](https://arxiv.org/abs/1708.04299). Sayyed Zahiri and Jinho D. Choi. In The AAAI Workshop on Affective Content Analysis, AFFCON'18, 2018.
82 |
83 |
84 | ## Contact
85 |
86 | * [Jinho D. Choi](http://www.mathcs.emory.edu/~choi).
87 |
--------------------------------------------------------------------------------