arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2007.09878cs.CLcs.IR
收藏

QA在书籍上令人沮丧的硬证据检索

Frustratingly Hard Evidence Retrieval for QA Over Books

Rensselaer Polytechnic Institute(伦斯勒理工学院);IBM Research(IBM研究院);Columbia University(哥伦比亚大学);IBM Watson(IBM沃森)

Xiangyang Mou, Mo Yu, Bingsheng Yao, Chenghao Yang, Xiaoxiao Guo, Saloni Potdar, Hui Su

摘要:

近年来,问答(QA)的研究取得了很大的进展,但对叙事性书籍故事问答这一特殊问题的研究还不够深入。我们制定BookQA作为一个开放领域的QA任务,其类似的依赖证据检索。我们进一步研究如何国家的最先进的开放领域的QA方法可以帮助BookQA。除了在NarrativeQA基准上达到最先进的水平外,我们的研究还揭示了在书籍中进行证据检索的困难,并进行了大量的实验和分析-这需要未来努力开发BookQA中证据检索的新解决方案。

英文摘要:

A lot of progress has been made to improve question answering (QA) in recent years, but the special problem of QA over narrative book stories has not been explored in-depth. We formulate BookQA as an open-domain QA task given its similar dependency on evidence retrieval. We further investigate how state-of-the-art open-domain QA approaches can help BookQA. Besides achieving state-of-the-art on the NarrativeQA benchmark, our study also reveals the difficulty of evidence retrieval in books with a wealth of experiments and analysis - which necessitates future effort on novel solutions for evidence retrieval in BookQA.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑