Firefly Based Optimization in Web Page Recommendation System

Abstract

the personalized web page recommendation for individuals is evident these days. Web servers are loaded with recommendation systems that analyze and recommend web pages to the users. They use data that are implicitly obtained as a result of web browsing patterns of the uses for recommending web pages. The cluster based association rule mining based system collects the web logs and generates a cluster of similar users and recommends pages to the users by actively analyzing them in the online. However, the time for analyzing it in online is more. To optimize this and increase the accuracy of the recommendation systems, a method that applies Naïve Bayes technique for clustering and firefly algorithm based similarity measure for optimization is designed. Web logs are initially clustered in offline by Naïve Bayes clustering technique. To find the similarity between the active user queries with other users in the same cluster in online, firefly algorithm based similarity measure is used. Firefly algorithm meticulously searches the generated cluster of web logs of the active user and recommends top pages. Firefly algorithm utilizes time efficiently, thus it is used for processing in online. When web pages are obtained, they are ranked and the top pages that are more relevant to the query are recommended. Efficiency of the system is evaluated using the measures like precision, recall, f – score, Matthews correlation and fallout rate. Experimental evaluation with real data set shows that the proposed system produced better recommendations and uses less time for computation in online.