FAIR: A Fuzzy ART Network Based Scheme for Retrieving Useful Information from Blogs

Abstract

Blogs could be viewed as the 4(th) crucial Internet application, after E-mail, Instant Message, and Bulletin Board System (BBS). The business world has experienced significant influence by the blogosphere. A hot topic in the blogosphere may affect a product's life period. Moreover, an exposure of an inside story in the blogosphere may influence a company's reputation. Nowadays, a lot of companies attempt to discover useful knowledge from that huge amount of blogs for business purposes. Therefore, the major objective of this study is to propose a Fuzzy Adaptive Resonance Theory (ART) network based Information Retrieval (FAIR) scheme by combining Fuzzy ART neural network, Latent Semantic Indexing (LSI), and association rules (AR) discovery to extract knowledge from blogs. In the proposed FAIR, Fuzzy ART network firstly has been employed to segment bloggers. Next, for each customer segment, we use LSI technique to retrieve important keywords. Then, in order to make the extracted keywords understandable, association rules mining is presented to organize these extracted keywords to form concepts. Finally, a real case of cosmetics products evaluation has been provided to demonstrate the effectiveness of the proposed FAIR scheme.