Beyond slot-filling: Dialogue System for Intellectual Cooperation

3 March 2017
March 3, 2017

Time: h. 12:00 pm
Location: Room Ofek, Polo scientifico e tecnologico "Fabio Ferrari", Building Povo 1, via Sommarive 5, Povo (Trento)

Speaker

Dr. Koichiro Yoshino, Nara Institute of Science and Technology

Abstract

Dialogue systems have been developed to fulfill users’ demands as a considerate secretary by using defined frames that describe knowledges to help humans, especially as task-oriented dialogue systems. However, it is difficult for existing dialogue systems to respond to users’ queries that are beyond the descriptive capacities of frames, because such frames are defined by considering possible action space of the system. 
On the other hand, keyword search or question answering systems covers the limitation of the frame-based system, however, it is difficult for these systems to behave as the frame-based system. Toward the system that can go beyond these limitations, we focus on predicate-argument structure (PAS), a semantic representation, to make the system possible to cooperate with users in tasks of information navigation and attentive listening. 
The system is able not only to capture the intention of the user by using the semantic representation but also to cooperate with users. The system dynamically extends their knowledge, which is represented in PAS, by integrating their original knowledge and extracted knowledge through a dialogue, to realize the intellectual cooperative dialogue.

About the Speaker

Koichiro Yoshino received his B.A. degree in 2009 from Keio University, M.S. degree in informatics in 2011, and Ph.D. degree in informatics in 2014 from Kyoto University, respectively. From 2014 to 2015, he was a research fellow (PD) of Japan Society for Promotion of Science at Academic Center for Computer and Media Studies, Kyoto University. Currently, he is an Assistant Professor in the Graduate School of Information Science, Nara Institute of Science and Technology.
His research interests include spoken language processing, especially on spoken dialogue system, syntactic and semantic parsing, and language modeling. 
Dr. Koichiro Yoshino received the JSAI SIG-research award in 2013. He is a member of IEEE, ACL, IPSJ, and ANLP.

Contact person regarding this talk: prof. Giuseppe Riccardi, giuseppe.riccardi [at] unitn.it