International Journal of Computer Applications (0975 – 8887) Volume 109 – No. 5, January 2015 26 Intelligent Web Information Retrieval based on User Navigational Patterns Anupama Prasanth Research Scholar Karpagam University Coimbatore M. Hemalatha, Ph.D. Professor Department of Computer Sceince Karpagam University ABSTRACT This is a review conducted in the fields of web usage mining and its latest works for supporting the research on efficient information retrieval based on user access pattern. The foremost mission of any information retrieval algorithm is the efficient extraction of user interests. The rapid growth of web data, intense competition and user’s option to choose from several alternatives increase this issue. In this context Web usage mining can provide valuable contributions in terms of ideas and methods, as it fissures useful knowledge from the pattern of user interactions with the Web. The user interest can be identified by analyzing the access pattern of user browse, the web pages they save, collect, or print. These valuable items of information are available in server logs, which can be exploited to satisfy user needs by optimizing the document-retrieval task. This article is a review conducted in the field of web usage mining and its latest works for supporting the research on efficient information retrieval based on user access pattern. This survey analyzes 25 released information retrieval models to find out the major mining techniques applied in them and also to analyze the effect of diverse parameters like feedbacks, time, content, frequency etc in information retrieval. The goal of this survey is to find the best composition of features to be included in an efficient information retrieval model. Using those features a new retrieval model is then proposed. General Terms Information Retrieval, Web Usage Mining. Keywords Information retrieval, User navigational patterns, Web usage mining, Web personalization, Retrieval parameters. 1. INTRODUCTION World Wide Web has become a huge storeroom of web pages and links. The internet users can search out huge quantity of data from WWW. Millions of pages are added to the web day by day and its growing incredibly. Various surveys roughly estimated that around one million new pages are added and 600GB of data changes per day [1]. The manner of internet usage in business has been changed by the innovative application of e-services like e-commerce, e-learning, e – banking etc. In this scenario it is necessary to focus in providing more and more features and tailored products and services according to the specific individual needs to maintain loyal customers. The majority of web users are non experts, so they cannot cope with the rapid development of computer technologies. The rapid growth of the web data, intense competition and user’s option to choose from several alternatives forced to realize the necessity of intelligent web information retrieval. Information retrieval can be achieved by applying web mining techniques. Basically web mining is an unmitigated edition of data mining. Web mining techniques enables to work upon On- Line and helps in storing the data in server database and web log [2] [3]. Web mining is categorized into three phases, based on which part of the web is to mine: Web Content Mining, Web Structure Mining and Web Usage Mining. Out of these three web usage mining is mainly concern because which is purely based on user access pattern. The main purpose of using mining algorithms is to accept user query and retrieve more relevant information according to that, so compare with other web mining phases, the basic concept behind web usage mining is the hit elements in the result page is on the basis of user browsing behavior [4][5]. Also it provides a friendlier environment after reducing the problems of information overload. It is nothing but a task of providing web pages based on needs and interests of individual users by collecting information about their preferences. The research on web mining has been almost started in 1996, and large numbers of papers are published with the overview of what has happened in the area of Web Mining since 1996. Web mining, its categories, Web Structure Mining, Web Content Mining, Web Usage Mining, and a survey focuses on one of its categories, web content mining, is presented in the article [6]. A survey [7] presents the commercial solutions of web Usage mining focusing on WebSIFT[8] project. The applications of soft computing techniques like, neural networks, fuzzy logic, genetic algorithms, and rough sets, on used in Web Mining is presented in the article [9]. These are some major milestones in the history of Web Mining. This article is a review conducted in the field of web usage mining and its latest works for supporting the research in the field of information Retrieval. In contrast with [6], [7] and [9], here the focal point is on Web Usage Mining, specifically on the mining techniques applied various research results reported in the literatures. The paper is organized as follows. Initially, it discusses web usage mining and its processes. After that an overview on Mining techniques is cited. Next a cross reference among the typical techniques employed and the parameters considered for access behavior analysis is presented. In the next part the paper proposed a new IR model incorporating all major features. Then finally, the future research trends in this area are stated. 2. WEB USAGE MINING The main motivation of the survey on Web Usage Mining is its close relation in analyzing User Access Patterns. Web Usage Mining is the application of data mining to extract the user browsing pattern of web data and it is first raised by Cooley in 1997 [10]. It focuses mainly on web server and analyzes the user interactions with web server also it collects data from the web server access logs, proxy server logs, browser logs, user profiles, registration data, user sessions or transactions, cookies, user queries, bookmark data, mouse clicks and scrolls and any other data as the result of interactions [11] [12]. Its main intention is to discover general