Article Preview
TopIntroduction
Lots of research has been going on in the field of information retrieval through Web to improve its performance while accessing and sharing information across the globe. Among all the services available today on the internet, the World Wide Web (WWW) plays a substantial role in information distribution and management. However, delay in accessing a Webpage may reduce or lessen end-users interest (especially in an e-learning websites) from accessing Web. Also, it causes frustration among online learners and decreases the Website popularity. Therefore, in this present internet era, the speed of information sharing and access plays a crucial role in sustaining end-users. To address these problems over the Web, Web pre-fetching and caching are most commonly used where it predicts and load Webpages which are accessed very often. Based on the location, it can be implemented either on the client or on the proxy or on the server. Among these three, the proxy-based pre-fetching and caching is more popular because it sits in-between the client and the Web server (Arshi & Pushpraj, 2018). It predicts and loads Webpages into the cache storage (Pallis, Vakali & Pokorny, 2008) then serves when a request arrives. Here, Web mining plays a predominant role in improving the performance of WWW through proxy server since it acts as an intelligent system to predict webpages that are to be preloaded. This is an application of data mining techniques over Webpages to improve its performance in terms of response time, throughput and latency. Based on the sources, web mining is categorized into three major areas as follows (Kumar & Meenu, 2017; Li, 2017):
In this research work, the Web usage mining is used to optimize the existing proxy-based Web caching system for better performance especially for an e-learning system (Baskaran & Kalaiarasan, 2016).
The rest of this paper is organized as follows: Sections 2 gives the importance of web mining research in the field of information retrieval and Section 3 examines some of the existing works related to Web pre-fetching and Web caching. It summarizes various prefetching techniques based on its location with merits and demerits. Also, gives the application of various data mining techniques over a Web based system and identifies the problem in the existing system then proposes a novel technique called a clustering-based pre-fetching technique to overcome the problem. Section 4 presents performance metrics related to Web caching and prefetching. The specification of the proposed system is provided in Section 5. Section 6 gives the empirical analysis of the proposed system and then concludes the research with future research directions for researchers.
TopBackground
At present Web-based information management and retrieval system, there are various factors that are affecting the performance of a Webpage such as high cost of bandwidth, broken bandwidth and latency, ever-increasing network distance and high bandwidth demands from end-users (Sathiyamoorthi, 2016; Mitali, Garg & Mishra, 2018). Hence, Web mining plays a major role in civilizing the performance of web-based information retrieval. Therefore, the objective of this research is to improve the performance of a Web-based information retrieval system through a clustering-based pre-fetching technique called Modified Adaptive Resonance Theory 1(MART1). It is mainly useful in e-learning environments. This integrated system framework is arrived at after attaining the following objectives, namely: