A real-time plagiarism detection tool for computer-based assessments

Abstract

Aim/Purpose The aim of this article is to develop a tool to detect plagiarism in real time amongst students being evaluated for learning in a computer-based assessment setting. Background Cheating or copying all or part of source code of a program is a serious con-cern to academic institutions. Many academic institutions apply a combination of policy driven and plagiarism detection approaches. These mechanisms are either proactive or reactive and focus on identifying, catching, and punishing those found to have cheated or plagiarized. To be more effective against plagia-rism, mechanisms that detect cheating or colluding in real-time are desirable. Methodology In the development of a tool for real-time plagiarism prevention, literature re-view and prototyping was used. The prototype was implemented in Delphi pro-gramming language using Indy components. Contribution A real-time plagiarism detection tool suitable for use in a computer-based as-sessment setting is developed. This tool can be used to complement other exist-ing mechanisms. Findings The developed tool was tested in an environment with 55 personal computers and found to be effective in detecting unauthorized access to internet, intranet, and USB ports on the personal computers. Recommendations for Practitioners The developed tool is suitable for use in any environment where computer-based evaluation may be conducted. Recommendation for Researchers This work provides a set of criteria for developing a real-time plagiarism pre-vention tool for use in a computer-based assessment. Impact on Society The developed tool prevents academic dishonesty during an assessment pro-cess, consequently, inculcating confidence in the assessment processes and re-spectability of the education system in the society. Future Research As future work, we propose a comparison between our tool and other such tools for its performance and its features. In addition, we want to extend our work to include testing for scalability of the tool to larger settings.