Publication Date

Fall 2015

Document Type

Project Summary

Degree Name

Master of Science


Computer Science

First Advisor

Soon-Ok Park, Ph.D.

Second Advisor

(Clare) Xueqing Tang, Ph.D.

Third Advisor

Neng-Shin Chen, M.S.


The drastic development in the information accessible on the World Wide Web has made the employment of automated tools to locate the information resources of interest, and for tracking and analyzing the same a certainty. Web Mining is the branch of data mining that deals with the analysis of World Wide Web. The concepts from various areas such as Data Mining, Internet technology and World Wide Web, and recently, Semantic Web can be said as the origin of web mining. Web mining can be defined as the procedure of determining hidden yet potentially beneficial knowledge from the data accessible in the web. Web mining comprise the sub areas: web content mining, web structure mining, and web usage mining. Web content mining is the process of mining knowledge from the web pages besides other web objects. The process of mining knowledge about the link structure linking web pages and some other web objects is defined as Web structure mining. Web usage mining is defined as the process of mining the usage patterns created by the users accessing the web pages.

The search engine technology has led to the development of World Wide. The search engines are the chief gateways for access of information in the web. The ability to locate contents of particular interest amidst a huge heap has turned businesses beneficial and productive. The search engines respond to the queries by employing the process of web crawling that populates an indexed repository of web pages. The programs construct a confined repository of the segment of the web that they visit by navigating the web graph and retrieving pages.

There are two main types of crawling, namely, Generic and Focused crawling. Generic crawlers crawls documents and links of diverse topics. Focused crawlers limit the number of pages with the aid of some prior obtained specialized knowledge. The systems that index, mine, and otherwise analyze pages (such as, the search engines) are provided with inputs from the repositories of web pages built by the web crawlers. The drastic development of the Internet and the growing necessity to incorporate heterogeneous data is accompanied by the issue of the existence of near duplicate data. Even if the near duplicate data don’t exhibit bit wise identical nature they are remarkably similar. The duplicate and near duplicate web pages either increase the index storage space or slow down or increase the serving costs which annoy the users, thus causing huge problems for the web search engines. Hence it is inevitable to design algorithms to detect such pages.


Co-authored capstone with authors listed in alphabetical order by OPUS staff.