MeetingBank-QA-Summary is a dataset released by Microsoft in 2024, focusing on question answering and summary generation for meeting transcripts. The related paper is "LLMLingua-2: Data Distillation for Efficient and Faithful Task-Agnostic Prompt Compression", which aims to evaluate the performance of compressed meeting transcripts in downstream tasks.

The dataset contains 862 meeting transcripts from the MeetingBank test set, with a data size of approximately 13.4 MB, and includes summaries and high-quality question-answer pairs generated by GPT-4.

Dataset Composition

The dataset includes the following main fields and structure:

• idx: Sample index.
• prompt: Meeting transcript text, used as contextual input.
• QA_pairs: A list containing question-answer pairs, each of which includes:
• answer: The answer to the question.
• question: A question generated based on the transcript text.
• summary: The generated summary.
• gpt4_summary: A summary generated by GPT-4.

Citation

@inproceedings{pan2024llmlingua2,
    title={LLMLingua-2: Data Distillation for Efficient and Faithful Task-Agnostic Prompt Compression},
    author={Zhuoshi Pan and Qianhui Wu and Huiqiang Jiang and Menglin Xia and Xufang Luo and Jue Zhang and Qingwei Lin and Victor Rühle and Yuqing Yang and Chin-Yew Lin and H. Vicky Zhao and Lili Qiu and Dongmei Zhang},
    year={2024},
    booktitle = {Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics},
    publisher = {Association for Computational Linguistics}
}