Publication D - Abstracts of IEICE TRANSACTIONS on Information and Systems (Japanese Edition)Vol.J95-DNo.6pp.1390-1400 Publication Date: 2012/06/01 Online ISSN: 1881-0225 DOI: Print ISSN: 1880-4535 Type of Manuscript: PAPER Category: Keyword: dialogue agent, non-task-oriented, learning to rank,
Full Text(in Japanese): PDF(537KB)>>
Summary: Computerized dialogue agents have recently been actively investigated in various fields. There is a great demand not only for task-oriented dialogue agents such as reservation services but also for non-task-oriented ones such as chatbots. This paper presents a statistical responce method for non-task-oriented dialogue agents based on learning to rank. Learning to rank is a statistical learning technology for sorting objects. Our method ranks candidate utterances which are prepared beforehand, in order of suitability for a response, and selects the top utterance.