Social Network Extraction from the Web
Relations among people are important for many intelligent systems. In my research, a social network, especially a collaboration network of researchers, is extracted automatically from the Web. The algorithm distinguishes the relations into co-authorship, co-participation to the same project, co-affiliation, and co-attendance to the same conferences. This project, funded by Ministry of Education, Culture, Sports, Science and Technology, begins from 2003 and published several papers in WWW, AAAI, and IJCAI conferences.
- Yutaka Matsuo, Masahiro Hamasaki, Yoshiyuki Nakamura, Takuichi Nishimura, Koiti Hasida, Hideyuki Takeda, Junichiro Mori, Danushka Bollegara, and Mitsuru Ishizuka: Spinning Multiple Social Network for Semantic Web, Proc. 21st National Conference on Artificial Intelligence (AAAI-06) , (2006.7) [PDF file(267KB)]
- Yutaka Matsuo, Junichiro Mori, Masahiro Hamasaki, Keisuke Ishida, Takuichi Nishimura, Hideaki Takeda, Koiti Hasida, and Mitsuru Ishizuka: POLYPHONET: An Advanced Social Network Extraction System, Proc. 15th International World Wide Web Conference (WWW2006) , (2006.5) [PDF file(438KB)]
POLYPHONET is a social network mining system that serves for promoting communication among researchers. It is operated in The Japanese Society for Artificial Intelligence (JSAI) annual conferences and a couple of international conferences.
- Service operated in JSAI2005 (Japanese)
- Service operated in UbiComp2006
- System operated at Yokohama Triennale 2005
User modeling from Location Information
This research, funded by Ministry of Internal Affairs and Communications from 2003-2005, aims at developing advanced representation of places and user modeling based on location information.
- Yutaka Matsuo, Naoaki Okazaki, Kiyoshi Izumi, Yoshiyuki Nakamura, Takuichi Nishimura, Koiti Hasida, and Hideyuki Nakashima: Inferring Long-term User Property based on Users' Location History, Proc. 20th International Joint Conference on Artificial Intelligence (IJCAI 2007), (2007.1) [PDF file(228KB)]
Analysis on weblogs, SNSs, and social tagging
We are provided weblogs data, social networking services (SNSs) data, and other user-generated data from various Japanese corporations that operate online social services. Especially, the analysis of mixi, which is the largest SNS in Japan, reveals some interesting characteristics found in Japanese communities.
By analysing social media, the prediction of stock market is possible. The innovative AI-based fund launched under the supervision of myself for technical aspects.