CoQA (Conversational Question Answering)
127k Questions With Answers, 8k Conversations About Text From Seven Domains.
By Huggingface Hub [source]
About this dataset
CoQA is an impactful and large-scale dataset of conversations, questions, and answers related to passages from seven diverse domains. This collection consists of an impressive 127,000 questions along with the answers provided by 8,000 conversations. What sets CoQA apart from other question-answering datasets is that the questions asked were conversational in nature. Each passage comes with its own set of answered queries, plus corresponding evidence emphasized in the accompanying text. With all this considered, CoQA offers a wealth of possibilities for researchers and people alike as it presents a strong compilation of data ideal for constructing various conversation/question-answering systems alike. As such this dataset can serve as a resource point not only to solve existing challenges but also stand as a platform to spur innovation within question-answering technologies moving forward
More Datasets
For more datasets, click here.
Featured Notebooks
- 🚨 Your notebook can be here! 🚨!
How to use the dataset
How to Use the CoQA Kaggle Dataset
Welcome to the world of conversational question answering! The CoQA Kaggle dataset is a great resource for those interested in building their own conversational question answering system. Here is a guide on how to take advantage of this dataset.
Research Ideas
- Capturing natural language understanding by mapping questions to relevant portions in a passage.
- Developing intelligent systems that can provide proper answers within a conversational state while taking into account the context of the conversation.
- Creating models that are capable of interactively responding to users’ inquiries using relevant evidence from the dataset's variety of domains
Acknowledgements
If you use this dataset in your research, please credit the original authors.
Data Source
License
License: CC0 1.0 Universal (CC0 1.0) - Public Domain Dedication
No Copyright - You can copy, modify, distribute and perform the work, even for commercial purposes, all without asking permission. See Other Information.
Columns
File: validation.csv
Column name |
Description |
source |
The domain from which the conversation or question-answer pair is from. (String) |
story |
The text passage from which questions were asked and answered. (String) |
answers |
The concise answer response. (String) |
File: train.csv
Column name |
Description |
source |
The domain from which the conversation or question-answer pair is from. (String) |
story |
The text passage from which questions were asked and answered. (String) |
answers |
The concise answer response. (String) |
Acknowledgements
If you use this dataset in your research, please credit the original authors.
If you use this dataset in your research, please credit Huggingface Hub.