Dataset Card for ReDIAL

Dataset Summary

This is an adapted version of the original redial dataset, for supporting different tasks in our project RecBot. The redial dataset provides over 10,000 conversations centered around movie recommendations. It was released in the paper "Towards Deep Conversational Recommendations" at NeurIPS 2018.

Supported Tasks and Leaderboards

1. Sentiment Analysis: Use the SA config for sentiment analysis.
2. Recommendation: Use the autorec config for recommendation task.
3. Conversational recommendation: Use the rec config for conversational recommendation task.

Languages

English

Dataset Structure

Data Instances

SA

An example of 'test' looks as follows.

{
  "movieId": 111776,
  "movieName": "Super Troopers",
  "messages": [
    "Hi I am looking for a movie like @111776",
    "You should watch @151656",
    "Is that a great one? I have never seen it. I have seen @192131\nI mean @134643",
    "Yes @151656 is very funny and so is @94688",
    "It sounds like I need to check them out",
    "yes you will enjoy them",
    "I appreciate your time. I will need to check those out. Are there any others you would recommend?",
    "yes @101794",
    "Thank you i will watch that too",
    "and also @91481",
    "Thanks for the suggestions.",
    "you are welcome\nand also @124771",
    "thanks goodbye"
  ],
  "senders": [1, -1, 1, -1, 1, -1, 1, -1, 1, -1, 1, -1, 1],
  "form": [0, 1, 1, 0, 1, 1]
}

rec

An example of 'test' looks as follows.

{
  'movieIds': [111776, 91481, 151656, 134643, 192131, 124771, 94688, 101794],
  'messages': ['Hi I am looking for a movie like @111776',
    'You should watch @151656',
    'Is that a great one? I have never seen it. I have seen @192131\nI mean @134643',
    'Yes @151656 is very funny and so is @94688',
    'It sounds like I need to check them out',
    'yes you will enjoy them',
    'I appreciate your time. I will need to check those out. Are there any others you would recommend?',
    'yes @101794',
    'Thank you i will watch that too',
     'and also @91481',
    'Thanks for the suggestions.',
    'you are welcome\nand also @124771',
    'thanks goodbye'],
  'senders': [1, -1, 1, -1, 1, -1, 1, -1, 1, -1, 1, -1, 1]
}

autorec

An example of 'test' looks as follows.

{
  "movieIds": [
    111776,
    151656,
    134643,
    192131,
    94688
  ],
  "ratings": [
    1.0,
    1.0,
    1.0,
    1.0,
    1.0
  ]
}

Data Fields

SA

• movieId: the movie's ID in the MovieLens dataset.
• movieName: the movie's name.
• messages: a list of string. The conversation messages related to the movie. Note that one conversation can contain mutiple movies. The conversation messages are repeated for each movie as a sample.
• senders: a list of 1 or -1. It has the same length of messages. Each element indicates the message at the same index is from the initiatorWorker (with 1) or the respondentWorkerId (with -1).
• form: a list generated by: [init_q[movieId]["suggested"], init_q[movieId]["seen"], init_q[movieId]["liked"], resp_q[movieId]["suggested"], resp_q[movieId]["seen"], resp_q[movieId]["liked"]. init_q is the initiator questions in the conversation. resp_q is the respondent questions in the conversation.

rec

• movieIds: a list of movie ids in a conversation.
• messages: a list of string. see config SA for detail.
• senders: a list of 1 or -1. see config SA for detail.

autorec:

• movieIds: a list of movie ids in a conversation.
• ratings: a list of 0 or 1. It has the same length as movieIds. Each element indicates the inititator's "liked" value for the movie.

Dataset Creation

Curation Rationale

[More Information Needed]

Source Data

Initial Data Collection and Normalization

[More Information Needed]

Who are the source language producers?

[More Information Needed]

Annotations

Annotation process

[More Information Needed]

Who are the annotators?

[More Information Needed]

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

[More Information Needed]

Citation Information

[More Information Needed]

Contributions

[More Information Needed]