A hybrid GA-CS based Meta Heuristic Approach for Web Personalization

Internet data is increasing exponentially every second leading to increase in data on World Wide Web. As a result, it becomes a challenge to extract needed and useful information from it which could be presented to individual users. This is possible by learning the user's behavior of navigating the web. For this user generally uses any search engine to find the required information. There is need to record the keywords used by the user. These keywords will become the basis for the proposed approach. The proposed approach finds the relevant pages based on the keywords used by the user as well as the relevance of the documents which is found by important web pages. Important web pages are decided based on unique visitors, keywords frequency, active time spent by the user on a webpage, hubs and authority values. Finding most relevant content from huge source of the web is a NP complete problem. Metaheuristic Optimization algorithms have been successfully applied for this purpose. In this paper, a hybrid approach based on Genetic Algorithm (GA) followed by Cuckoo Search Algorithm (CSA) has been proposed to find the relevant web pages. Experimentally it has been shown that the proposed GA-CS approach select utmost relevant web pages in comparison to basic algorithms proposed in the literature.