Social network analysis for weblog behavior

• Main Authors: Bing-Shian Lin, 林炳憲
• Other Authors: 黃夙賢
• Format: Others
• Language: zh-TW
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/49452854640793703375

碩士 === 明新科技大學 === 資訊管理研究所 === 98 === Rapid growth of Web2.0 has made Weblogs become the most important application of social network services. Weblogs provide customized web pages and service integration of Web2.0 and become the popular platform for knowledge-sharing and product promotion for publics, companies and organizations. Compared to conventional social network, people may own more than one weblog and hence produce different relationship of social network. This study proposed a four-layer architecture of weblog social network analysis to study and demonstrate the relationships of the weblog and personal layer. In this architecture, behavior features and implicit relationships among bloggers are studied. This study uses social network theory and natural language processing to discover the implicit relationship in the account, weblog title and article similarity. Through this information, blogger have multiple weblogs can be detected. This study illustrates the social network between bloggers and weblogs with the implicit relationship appended. The result showed that the behavior feature of those bloggers with multiple weblogs can be found from keywords search. Combine both explicit and implicit relationships can derive more precise blogger social network. Accordingly, the architecture proposed herein can express more accurate blogosphere social network and provide a basis for future research about behavior analysis of blogger social network.