An Automated System for Plagiarism Detection Using the Internet


Allan Knight, Kevin Almeroth            Bruce Bimber
Dept of Computer Science                Dept of Political Science
University of California                University of California
Santa Barbara, CA 93106-5110            Santa Barbara, CA 93106-5110
{aknight, almeroth}@cs.ucsb.edu         bimber@polsci.ucsb.edu

In this paper we present a framework for plagiarism detection. 
Rather than exhaustively or randomly searching sentences in a 
student paper on the Internet for possible sources of borrowed 
ideas in student’s text, we take a middle ground. With intelligent 
selection of sentences from papers we will show that examples of 
plagiarism can be found just as frequently and with fewer queries 
to Internet search engines.  One of our goals is to use this 
technology to develop freeware that any instructor or teaching 
assistant, even modestly non-technical ones, can use to detect 
plagiarism in their classes.