Developing a Web Page Recommendation System based on Hierarchical Genetic Algorithm

• Main Authors: Tien-Hua Jiang, 蔣天華
• Other Authors: Chih-Chin Lai
• Format: Others
• Language: zh-TW
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/65797658634327691896

碩士 === 樹德科技大學 === 資訊管理研究所 === 92 === With development of Internet, the problem of information overload has deteriorated when the number of web servers increases rapidly. In this paper, we build a personal web information space (PWIS) from web pages that attracting personal interest of user. The method for mining such user interests is then presented. In this way, each user is associated with a set of interests, which is stored in the PWIS. This paper proposes a method based on hierarchical genetic algorithm (HGA) combined with vector space model (VSM) in order to solve Google’s problems on searching by single keyword and recommend web pages to the user. For experimental results and performance evaluation, we suppose and implement a Webpage Recommendation System that will facilitate to save web searching time and ameliorate the knowledge gap between professional and amateur. In performance evaluation, we compare with genetic algorithm (GA), metagenetic algorithm (MGA) and Google’s PageRank™ on similarity (fitness), accuracy rate, stability and time-cost. Last, the results show the average accuracy rate of amateur is up to 92% from 15%, and professional is up to 92% from 28%.