Plagiarism Detection

This page contains information about the plagiarism detection process. We discuss the four stage process for detecting plagiarism described in Culwin and Lancaster (2001a), clues found in students work that could cause suspicion, and various plagiarism detection resources.

A Four Stage Process for Plagiarism Detection

Culwin and Lancaster (2001a) define a four stage process for detecting plagiarism.

(Picture courtesy of Culwin and Lancaster)

The collection stage may be defined as the process of electronically collecting and pre-processing student submissions into a suitable format (Culwin and Lancaster 2001b). Analysis is defined as “where the submissions are compared with each other and with documents obtained from the Web and the list of those submissions, or pairs, that require further investigation is produced.” Verification (confirmation) is required to ensure that those pairs reported as being suspicious are worth investigating with a view to possible disciplinary action (this is a task normally undertaken by humans, since value judgements may be involved). The final stage, investigation, will determine the extent of the alleged misconduct and will “also involve the process of deciding culpability and possible penalties.”

A useful reference for the process of investigation and punishment is the book "Crisis on Campus: Confronting Academic Misconduct " (Decoo 2002). This book mainly deals with the problem of plagiarism by post-graduates or academics but is also an authoritative source for anyone having to prosecute suspected plagiarism and the issues that have to be dealt with in such a prosecution.

Clues for Detecting Plagiarism

In the event when the academic suspects that the student may have inappropriately used material taken from another source s/he can use electronic resources (such as search engines and electronic databases) for investigating the suspicious work. Also, in the event where the academic suspects plagiarism between students, software tools can be used for comparing the works in question. There are clues within work that can cause suspicion. Such clues include:

• Unusual formatting such as changes in margins, strange spacing, inconsistent font sizes and colours, greyed out words, strange page numbers.
• Changes in writing style and grammar. For example mixed paragraph styles could be a sign that the students has copied and pasted material.
• Changes in spelling style, i.e. from English to American.
• Sections or sentences that do not relate to each other.
• Use of jargon or advanced vocabulary.
• Mixed citation styles.
• Missing references or quotes. Use of quotes that are not referenced in the bibliography.
• Using anachronisms, failure to mention current events, and using old references are signs that the work was written at some past date.
• Changes in the quality of the student’s work in different parts of the assignment or from one assignment to another.
• Clues that the work was printed of the internet.

Detection Using Software Tools

Various tools exist for detecting plagiarism between students work, and can be divided into three categories.

• Text analysis tools work with plain text, such as essays, and relate to the “analysis” phase.
• Source code tools work with (high-level) computer programs, and we compare and evaluate the most popular based on their efficiency in detecting plagiarism, and relate to the “analysis” phase.
• Other tools assist in the process, for example course management tools which support the collection of data which can then be input to the detection tools. These relate to the “collection” phase.

Comparisons of these tools and demonstrations of the major source code tools, are available.

Detection Using Web Sources

Detection using web sources involves using Internet search engines for detecting plagiarism. This is a free, easy and fast way of detecting plagiarism. The user can copy and paste or type in suspicious phrases taken from students work into a search engine in an attempt to find on-line material containing the suspicious phrases.

In addition, many Internet search engines allow the user to conduct exact phrase searching for finding phrases exactly as they were typed in the search box. Examples of major search engines include Google, Yahoo, Altavista, Lycos, Microsoft Live, Ask.com, Ujiko, Nutch and the commercial source-code search engine Krugle

In addition to conducting searches using individual search engines, one can use meta-search engines for transmitting single searches across various search engines simultaneously.  The results are then displayed together all in one page. Some examples of metasearch engines include Dogpile, Vivisimo, Kartoo, Mamma and SurfWax.

Detection Using Full-Text Library Databases

Full-text electronic library databases can also be used when searching for sources containing suspicious phrases that may have been taken from papers, articles, etc. Academics can use various databases they have subscriptions to as part of their personal or academic organisation subscriptions.

References

• Fintan Culwin and Thomas Lancaster (2001a). “Plagiarism Issues for Higher Education”. VINE 31(2) pp. 36-41.
• Fintan Culwin and Thomas Lancaster (2001b). Plagiarism, Prevention, Deterrence and Detection. South Bank University, UK.
• Wilfried Decoo, Crisis on Campus: Confronting Academic Misconduct, Cambridge, MA: MIT Press, 2002.