Abstract
Understanding the web user browsing behaviour in order to adapt a web site to the needs of a particular user represents a key issue for many commercial companies that do their business over the Internet. This paper presents the implementation of a Knowledge Base (KB) for building web-based computerized recommender systems. The Knowledge Base consists of a Pattern Repository that contains patterns extracted from web logs and web pages, by applying various web mining tools, and a Rule Repository containing rules that describe the use of discovered patterns for building navigation or web site modification recommendations. The paper also focuses on testing the effectiveness of the proposed online and offline recommendations. An ample real-world experiment is carried out on a web site of a bank.
Original language | English |
---|---|
Pages (from-to) | 793-828 |
Number of pages | 36 |
Journal | International Journal on Artificial Intelligence Tools |
Volume | 16 |
Issue number | 5 |
DOIs | |
Publication status | Published - Oct 2007 |
Externally published | Yes |
Keywords
- Adaptive web sites
- Computerized recommender systems
- Knowledge bases
- Web usage mining
- Web-based systems
ASJC Scopus subject areas
- Artificial Intelligence