Summary: Design, Implementation and Deployment of PAIRwise
Allan Knight and Kevin Almeroth
Department of Computer Science
Department of Political Science
University of California, Santa Barbara
Increased access to the Internet has dramatically increased the sources from which students can
deliberately or accidentally copy information. This paper discusses our motivation to design,
implement, and deploy an Internet based plagiarism detection system, called PAIRwise, to
address this growing problem. We give details as to how we detect plagiarism and the various
development phases we went through before releasing PAIRwise as an open source software
tool. Our major conclusion is that it is possible to deploy an automated plagiarism system that is
an effective, customizable, and affordable tool for investigating how plagiarism affects academic
Plagiarism has long been a problem in academia. Compounding this problem more recently is
the creation of the Internet, which has greatly expanded the sources from which to plagiarize.
Furthermore, a high percentage of certain populations currently use the Internet. For example,
Nielsen estimates that 68.6% (Nielsen NetRatings, 2006) of the U.S. population uses the Internet.