File size: 6,850 Bytes
92376a0
 
 
 
 
 
 
c2af003
92376a0
c2af003
92376a0
 
 
 
 
 
 
 
 
 
 
1e7eda4
a54c15b
7150e59
 
 
 
 
 
 
 
 
 
 
 
396368c
 
7150e59
 
 
 
396368c
 
 
 
 
7150e59
 
 
 
 
 
 
 
 
 
92376a0
 
fa52ba3
92376a0
 
 
 
1e7eda4
92376a0
 
 
1e7eda4
 
92376a0
 
 
 
 
 
 
 
 
 
 
 
 
b1e6ad5
92376a0
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
b1e6ad5
 
 
7150e59
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
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
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
---
annotations_creators:
- crowdsourced
- machine-generated
language_creators:
- crowdsourced
- found
language:
- en
license:
- unknown
multilinguality:
- monolingual
size_categories:
- 10K<n<100K
source_datasets:
- original
task_categories:
- question-answering
task_ids:
- multiple-choice-qa
paperswithcode_id: mc-taco
pretty_name: MC-TACO
dataset_info:
  features:
  - name: sentence
    dtype: string
  - name: question
    dtype: string
  - name: answer
    dtype: string
  - name: label
    dtype:
      class_label:
        names:
          '0': 'no'
          '1': 'yes'
  - name: category
    dtype:
      class_label:
        names:
          '0': Event Duration
          '1': Event Ordering
          '2': Frequency
          '3': Typical Time
          '4': Stationarity
  config_name: plain_text
  splits:
  - name: test
    num_bytes: 1785553
    num_examples: 9442
  - name: validation
    num_bytes: 713023
    num_examples: 3783
  download_size: 2385137
  dataset_size: 2498576
---

# Dataset Card for MC-TACO

## Table of Contents
- [Dataset Description](#dataset-description)
  - [Dataset Summary](#dataset-summary)
  - [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards)
  - [Languages](#languages)
- [Dataset Structure](#dataset-structure)
  - [Data Instances](#data-instances)
  - [Data Fields](#data-fields)
  - [Data Splits](#data-splits)
- [Dataset Creation](#dataset-creation)
  - [Curation Rationale](#curation-rationale)
  - [Source Data](#source-data)
  - [Annotations](#annotations)
  - [Personal and Sensitive Information](#personal-and-sensitive-information)
- [Considerations for Using the Data](#considerations-for-using-the-data)
  - [Social Impact of Dataset](#social-impact-of-dataset)
  - [Discussion of Biases](#discussion-of-biases)
  - [Other Known Limitations](#other-known-limitations)
- [Additional Information](#additional-information)
  - [Dataset Curators](#dataset-curators)
  - [Licensing Information](#licensing-information)
  - [Citation Information](#citation-information)
  - [Contributions](#contributions)

## Dataset Description

- **Homepage:** [MC-TACO](https://cogcomp.seas.upenn.edu/page/resource_view/125)
- **Repository:** [Github repository](https://github.com/CogComp/MCTACO)
- **Paper:** ["Going on a vacation" takes longer than "Going for a walk": A Study of Temporal Commonsense Understanding](https://arxiv.org/abs/1909.03065)
- **Leaderboard:** [AI2 Leaderboard](https://leaderboard.allenai.org/mctaco)

### Dataset Summary

MC-TACO (Multiple Choice TemporAl COmmonsense) is a dataset of 13k question-answer pairs that require temporal commonsense comprehension. A system receives a sentence providing context information, a question designed to require temporal commonsense knowledge, and multiple candidate answers. More than one candidate answer can be plausible.

### Supported Tasks and Leaderboards

The task is framed as binary classification: givent he context, the question, and the candidate answer, the task is to determine whether the candidate answer is plausible ("yes") or not ("no").

Performance is measured using two metrics:

- Exact Match -- the average number of questions for which all the candidate answers are predicted correctly.
- F1 -- is slightly more relaxed than EM. It measures the overlap between one’s predictions and the ground truth, by computing the geometric mean of Precision and Recall.

### Languages

The text in the dataset is in English. The associated BCP-47 code is `en`.

## Dataset Structure

### Data Instances

An example looks like this:

```
{
  "sentence": "However, more recently, it has been suggested that it may date from earlier than Abdalonymus' death.",
  "question": "How often did Abdalonymus die?",
  "answer": "every two years",
  "label": "no",
  "category": "Frequency",
}
```

### Data Fields

All fields are strings:
- `sentence`: a sentence (or context) on which the question is based
- `question`: a question querying some temporal commonsense knowledge
- `answer`: a potential answer to the question (all lowercased)
- `label`: whether the answer is a correct. "yes" indicates the answer is correct/plaussible, "no" otherwise
- `category`: the temporal category the question belongs to (among "Event Ordering", "Event Duration", "Frequency", "Stationarity", and "Typical Time")

### Data Splits

The development set contains 561 questions and 3,783 candidate answers. The test set contains 1,332 questions and 9,442 candidate answers.

From the original repository:

*Note that there is no training data, and we provide the dev set as the only source of supervision. The rationale is that we believe a successful system has to bring in a huge amount of world knowledge and derive commonsense understandings prior to the current task evaluation. We therefore believe that it is not reasonable to expect a system to be trained solely on this data, and we think of the development data as only providing a definition of the task.*

## Dataset Creation

### Curation Rationale

MC-TACO is used as a testbed to study the temporal commonsense understanding on NLP systems.

### Source Data

From the original paper:

*The context sentences are randomly selected from [MultiRC](https://www.aclweb.org/anthology/N18-1023/) (from each of its 9 domains). For each sentence, we use crowdsourcing on Amazon Mechanical Turk to collect questions and candidate answers (both correct and wrong ones).*

#### Initial Data Collection and Normalization

[More Information Needed]

#### Who are the source language producers?

[More Information Needed]

### Annotations

From the original paper:

*To ensure the quality of the results, we limit the annotations to native speakers and use qualification tryouts.*

#### Annotation process

The crowdsourced construction/annotation of the dataset follows 4 steps described in Section 3 of the [paper](https://arxiv.org/abs/1909.03065): question generation, question verification, candidate answer expansion and answer labeling.

#### Who are the annotators?

Paid crowdsourcers.

### Personal and Sensitive Information

[More Information Needed]

## Considerations for Using the Data

### Social Impact of Dataset

[More Information Needed]

### Discussion of Biases

[More Information Needed]

### Other Known Limitations

[More Information Needed]

## Additional Information

### Dataset Curators

[More Information Needed]

### Licensing Information

Unknwon

### Citation Information

```
@inproceedings{ZKNR19,
    author = {Ben Zhou, Daniel Khashabi, Qiang Ning and Dan Roth},
    title = {“Going on a vacation” takes longer than “Going for a walk”: A Study of Temporal Commonsense Understanding },
    booktitle = {EMNLP},
    year = {2019},
}
```

### Contributions

Thanks to [@VictorSanh](https://github.com/VictorSanh) for adding this dataset.