A dynamic trust model between users in social networks

2015 Fiscal Year Final Research Report

• Project/Area Number: 25330365
• 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: University of Shizuoka
• Principal Investigator: Mutoh Nobuaki 静岡県立大学, 経営情報学部, 准教授 (40275102)
• Co-Investigator(Kenkyū-buntansha): SAITO KAZUMI 静岡県立大学, 経営情報学部, 教授 (80379544) ; IKEDA TETSUO 静岡県立大学, 経営情報学部, 教授 (60363727) ; OKUBO SEIYA 静岡県立大学, 経営情報学部, 助教 (90422576)
• Project Period (FY): 2013-04-01 – 2016-03-31
• Keywords: 動的信頼モデル / オンラインレビューサイト / レビューランキング / 時間的イベント / 時間的・空間的減衰関数 / レビューカテゴリ / 時間的区間分割 / ネットワーク構造特性

Outline of Final Research Achievements

The word of mouth information of online review sites are affecting various activities from person to person. In review sites, it may occur that the future review tendency of an item changes a lot by one review. In this study, we propose a dynamic trust model between users who review the items on the online review site, in order to allow for high-accuracy prediction of information diffusion and share of opinions between such users. Based on the dynamic trust model, we established a method for change-point detection in time series data of online review sites, a ranking method equipped with some temporal or special dynamics of either the exponential decay or the power-law decay, an extraction method of categories containing significantly large numbers of highly ranked objects, and so on.

Free Research Field

組合せ論、グラフ論、観光情報学