RECOMMENDATION GENERATION BY INTEGRATING SEQUENTIAL PATTERN MINING AND SEMANTICS

Geethapriya Uvaraja

Abstract: As the Internet usage keeps increasing, the number of web sites and hence the number of web pages also keeps increasing. A recommendation system can be used to provide personalized web service by suggesting the pages that are likely to be accessed in future. Most of the recommendation systems are based on association rule mining or based on keywords. Using the association rule mining the prediction rate is less as it doesn’t take into account the order of access of the web pages by the users. The recommendation systems that are key-word based provides lesser relevant results. This paper proposes a recommendation system that uses the advantages of sequential pattern mining and semantics over the association rule mining and keyword based systems respectively.

Keywords: Sequential Pattern Mining, Taxonomy, Apriori-All, CS-Mine, Semantic, Clustering

DOI: https://doi.org/10.15623/ijret.2014.0301032