Web Page Recommendation Using Domain Knowledge and Improved Frequent Sequential Pattern Mining Algorithm

Abstract

Web page recommendation is the technique of web site customization to fulfil the needs of every particular user or group of users. The web has become largest world of knowledge. So it is more crucial task of the webmasters to manage the contents of the particular websites to gather the requirements of the web users. The web page recommendation systems most part based on the exploitation of the patterns of the site's visitors. Domain ontology’s provide shared and regular understanding of a particular domain. Existing system uses pre-order linked WAP-tree mining (PLWAP Mine) algorithm that helps web recommendation system to recommend the interested pages but it has some drawbacks, it require more execution time and memory. To overcome the drawbacks of existing system paper utilizes PREWAP algorithm. The PREWAP algorithm recommends the interested results to web user within less time and with less memory and improves the efficiency of web page recommendation system. In work, various models are presented; the first model is Web Usage Mining which uses the web logs. The second model also utilizes web logs to represent the domain knowledge, here the domain ontology is used to solve the new page problem. Likewise the prediction model, which is a network of domain terms, which is based on the frequently viewed web-pages and represents the integrated web usage. The recommendation results have been successfully verified based on the results which are acquired from a proposed and existing web usage mining (WUM) technique.