An Accountability of Various Clustering Techniques for Improvement of Page Searching Process and Analysis.

Abstract

In internet era we have coming across tons of data and information as the usage of internet has been growing at a great speed. These data had been stored somewhere either with our concern or by policies of the providers of the website, internet etc. along with this these are used for good work like suggesting the pages, surfing made easy by tracking the surfing pattern, giving customer personalization facility for product buying and alternative product suggestions. It is also helping us to understand what one need today without being told to anyone. Web takes care of the same. These things are created analyzing interaction of users with the web structure in any form through social medias, shopping sites, blogs, e learning etc. In this paper we discuss in details about the web log mining process along with the mining methods and then clustering the data based on the requirement to study the logs etc. web log has also been growing tremendously due to increase data on WWW. But we need to keep in mind that these logs not only contain useful data but some incomplete data, noisy data, duplicate and ambiguous data. So we here focus on techniques to filter these data, convert it into required format, select specific fields needed and then classify them to understand the pattern which may be of our interest. Here the main focus is on discussing various techniques used by other authors in past and try to compare them.