L. K. Joshila Grace ¹, V. Maheswari ¹

Affiliations
1 Sathyabama University, Jeppiaar Nagar, Chennai, IN

Abstract

The proposed work does a web utility mining process for identifying the useful web navigation patterns. An optimal prefix tree is generated from the log file details and the mining is performed. From the set of mined web navigational patterns the efficiency is found by considering certain parameters. The parameters like frequency, utility, downloads, book mark, selection are considered for each web page and efficiency for the web navigation pattern is found. The work is done by using real data set extracted from an e-commerce web site and with synthetic data set. This provides a better analysis of the web site. The result provided by this proposed work can be used or various application in developing the web site contents.

Keywords

Non Sequential Pattern, Pattern Weight, Sequential Pattern, Utility, Web Path Traversal

Full Text

   

V. Vijeya Kaveri ¹, V. Maheswari ¹

Affiliations
1 Sathyabama University, Chennai - 600119, Tamil Nadu, IN

Abstract

Web (2.0) is the place where people can upload, share and access various sources of information. Web (2.0) has given rise to information overloading problem and knowledge starvation. Recommender Systems (RS) helps in alleviating this overloading problem and gaining the exact information what we need. RS suggest user items or products based on their browsing or purchasing history. RS suggest list of items by identifying similar users with explicit user-item rating. But, in real time applications most users do not rate items. In current web (2.0) social tagging applications help us to find useritem ratings implicitly based on the user’s interest and preferences they give for the list of items. In this paper we have proposed a model based resource recommendation on social tagging information which has improved the performance of the RS. In the proposed system the topic is identified from the tagged data, based on the topic user profile is constructed by semantic approach and the recommendation is done for the user.

Keywords

Explicit Rating, Resource Recommendation, Recommender System, Social Tagging, User Profile.

Full Text