2015 Fiscal Year Final Research Report
Social Media User Search/Recommendation Using External Knowledge Bases
Project/Area Number |
25330362
|
Research Category |
Grant-in-Aid for Scientific Research (C)
|
Allocation Type | Multi-year Fund |
Section | 一般 |
Research Field |
Web informatics, Service informatics
|
Research Institution | Tokyo Institute of Technology |
Principal Investigator |
NORO TOMOYA 東京工業大学, 情報理工学(系)研究科, 助教 (80401553)
|
Co-Investigator(Kenkyū-buntansha) |
TOKUDA TAKEHIRO 東京工業大学, 大学院情報理工学研究科, 教授 (30111644)
|
Project Period (FY) |
2013-04-01 – 2016-03-31
|
Keywords | ソーシャルメディア / マイクロブログ / 推薦 / 検索 / コミュニティ検出 / クラスタリング / オントロジー / ニュース記事 |
Outline of Final Research Achievements |
Finding users on social media such as Twitter and Facebook who usually post valuable messages related to a topic of interest will help us get information on the topic efficiently. However, it is not easy for us to find such users by general text mining methods, which is based on only the message text, since each message text is short. In this research, we developed methods for finding/recommending topic-related Twitter users by using news articles and ontologies (e.g. DBpedia) as external knowledge bases. News articles are used for detecting news-topic-related hashtag communities and finding two types of influential users: content-based influential users and authority-based influential users. Ontologies are used for checking topic consistency of each user's messages and finding users who usually post messages on the same topic. In addition, we applied the methods to a method for finding topic-related messages and a method for finding hierarchical user clusters.
|
Free Research Field |
自然言語処理
|