Development of Interactive Clustering based on Higher Order Statistics

2014 Fiscal Year Final Research Report

• Project/Area Number: 23300063
• Research Category: Grant-in-Aid for Scientific Research (B)
• Allocation Type: Single-year Grants
• Section: 一般
• Research Field: Intelligent informatics
• Research Institution: Central Research Institute of Electric Power Industry
• Principal Investigator: ONODA Takashi 一般財団法人電力中央研究所, システム技術研究所, 副研究参事 (40371661)
• Co-Investigator(Kenkyū-buntansha): YAMADA Seiji 国立情報学研究所, 大学共同機関等の部局等, 教授 (50220380)
• Co-Investigator(Renkei-kenkyūsha): OKABE Masayuki 豊橋技術科学大学, 情報メディア基盤センター, 助教 (50362330)
• Research Collaborator: NISHIGAKI Takahiro 東京工業大学, 大学院知能システム科学専攻
• Project Period (FY): 2011-04-01 – 2015-03-31
• Keywords: クラスタリング / 高次統計 / 知識発見 / データマイニング / HAI

Outline of Final Research Achievements

Recently, quite many documents have overflowed on the Internet by utilization of Web news, twitter, Social Network System, and so on. In stored documents on the Internet, many documents describe about the same “topic” and this thing is a common thing. Our research proposed the Independent Semantic Analysis (ISA) which extracts the “topic” from the quite many documents. The proposed ISA can extract independent “topics” from a high order statistical point of view. It is easy to generate an abstract by using the extracted independent “topics” and group the quite many documents by using the extracted independent “topics”.

Free Research Field

統計的機械学習