Making Wishlists by Personal Interests and Favorites Extracted from Blogs

• Main Authors: Wei-Yu Chen, 陳威宇
• Other Authors: Chuan-Jie Lin
• Format: Others
• Language: zh-TW
• Online Access: http://ndltd.ncl.edu.tw/handle/58958576873283588675

碩士 === 國立臺灣海洋大學 === 資訊工程學系 === 96 === The blogsphere grows up rapidly in recent years. Many people begin to write articles in their blog space and share them to everyone. Blogs have become a rich source of information. Many researches are focused on blog mining recently. Personal blog provides various kinds of information, like age, gender, interest and favorite lists of the blogger. By extracting interest information from a blogger's profile, we can recommend products that he may be interested in as his wishlist. In this paper, we first established a list of interests and a list of products, and measured relationships among them by similarity, probability, bookmarks, web page counts, etc. An interest in the interest list will be aligned to some related products. Then we mining interest from blogger's info page in blog and detect their favorite. After all, we generated and recommend blogger a list. We call this list is wishlist. Our work will be apply in many domains in the future, such as advertiser: what offers should be made to a blogger, what advertisements should be shown to their readers. After we evaluated our work, the result showed that the log-likelihood is powerful method.