Academic Journals Database
Disseminating quality controlled scientific knowledge

Collaborative Web Recommendation Systems based on Association Rule Mining

ADD TO MY LIST
 
Author(s): A. Kumar | P. Thambidurai

Journal: International Journal of Computer Science and Information Security
ISSN 1947-5500

Volume: 8;
Issue: 3;
Start page: 222;
Date: 2010;
Original page

Keywords: ijcsis | journal | computer science | doaj | information security | research | citation | june 2010 | volume 8

ABSTRACT
Massive development of internet in recent years necessitate the development of recommender systems that turn out to be user friendly in web applications. Recommender systems make an effort to outline user preferences over items, and model the relation between users and items. There are two elemental approaches that can be applied when generating recommendations systems. They are content based web recommender system and the other is collaborative web recommender system. This proposed paper presents a method of developing a collaborative web recommendation systems using association rule mining. The association rules were applied to personalization based on web usage data. The method utilize apriori algorithm to generate association rules. In general association rule mining is a technique common in data mining that attempts to discover patterns of products that are purchased together. The greater part of web page recommender systems that were proposed earlier utilized collaborative filtering. Web Content Recommendation has been an active application area for Information Filtering, Web Mining and Machine Learning research. The future work explains some of the modifications using other algorithms to generate the association rules that can be adopted on existing web recommendation system to make them functionally more effective. In order to explore the performance of the proposed web recommendation system experiments were conducted on available dataset. The performance of the proposed approach is best illustrated by comparing it with K-nearest neighboring algorithm.

Tango Jona
Tangokurs Rapperswil-Jona

     Save time & money - Smart Internet Solutions