Electronic Detection of Plagiarism in Student Work

1
Electronic Detection of Plagiarism in Student Work

• Judith Barlow, LuAnn Bean, David Hott
• School of ManagementFlorida Institute of
  Technology

http//my.fit.edu/jbarlow/papers/sci2003.ppt
2
Plagiarism abounds!

• In 1950, 20 of college students admitted to
  having cheated in high school.
• Davis, Stephen, Keeping plagiarism at bay in the
  Internet age, Monitor on Psychology, Vol. 33.
  No. 2. February 2002. lthttp//www.apa.org/monitor/
  feb02/plagiarism.htmlgt
• Recent studies show between 75 and 98 admit to
  having cheated in high school.
• The cheating game, US News and World Report,
  November 22, 1999. lthttp//www.usnews.com/usnews/e
  du/college/articles/brief/cocheata.htmgt.

3
Why do students cheat?

• Students are natural economizers. 
• Students are faced with too many choices, so they
  put off low priorities. 
• Many students have poor time management and
  planning skills. 
• Some students fear that their writing ability is
  inadequate. 
• A few students like the thrill of rule breaking. 

4
Common Sources of Plagiarized Material

• Downloading a free research paper. 
• Buying a paper from a commercial paper mill. 
• Copying an article from the Web or an online or
  electronic database. 
• Copying a paper from a local source. 
• Cutting and pasting to create a paper from
  several sources -- "assembly-kit" papers

5
Some Sources of Term Papers

• Free Papers!
• School Sucks
• Other People's Papers
• Absolutely Free Online Essays
• Free Termpapers International
• Dorian's Paper Archive

• Papers for Sale!
• Evil House of Cheat
• Research Papers Online A
• A1 Termpaper
• Genius Papers
• TermPapers.com

6
Plagiarism Detection Tools

• TurnItIn http//www.turnitin.org
• deterring plagiarism for nearly 5 million
  students and educators worldwide
• growing at a rate of one new user every 20
  seconds
• widely used in over 50 countries, at some of the
  world's most respected institutions

7
TurnItIn Faculty Interface
TurnItIn Faculty Interface
8
Other Plagiarism Detection Software

• Plagiarism.org - http//plagiarism.org/ Online
  service for preventing plagiarism, deterring
  scholarly dishonesty and cheating, and raising
  academic ethics. Detects plagiarized papers by
  comparing with submitted papers and papers from
  online term paper mills. 
• WCopyfind - http//plagiarism.phys.virginia.edu/Ws
  oftware.html Software for comparing specific
  documents for plagiarism. Free.
• CopyCatch - http//www.copycatch.freeserve.co.uk/
  Compares student documents with those available
  on the user's computer system or from other
  electronic sources. Free. 
• MOSS - http//www.cs.berkeley.edu/aiken/moss.html
  Free Internet service for detecting plagiarism
  in software code from Berkeley. 

9
Other Plagiarism Detection Software

• WordCHECK KeyWord Software - http//www.wordchecks
  ystems.com/ Compares on-line documents based on
  percentage of key words and phrases. Displays
  result in graphical form. Demo version available 
  
• EVE The Essay Verification Engine -http//www.can
  exus.com/eve/ Software to detect plagiarism.
  Free 15 day trial.
• JPlag - http//wwwipd.ira.uka.de2222/ An
  on-line plagiarism detection program for Java, C,
  C and Scheme programming code. 
• Glatt Plagiarism Program - http//www.plagiarism.c
  om/index.htm Software to detect plagiarism.
• Comparing Plagiarism Detection Software

10
Some Indicators of Plagiarism

• Mixed citation styles. 
• Lack of references or quotations. 
• Unusual formatting. 
• Off topic. 
• Signs of datedness. 
• Anachronisms .
• Anomalies of style formats, margins, spacing. 
• Smoking guns.  -- "blunders of the clueless"

11
A Structural Approach to Plagiarism Detection

• Regardless of the source of plagiarized materials
  and their availability, there are some structural
  document characteristics that may indicate the
  increased likelihood of plagiarism including
  variations in writing style, font properties,
  margin sizes, and paragraph formats

12
A Structural Approach to Plagiarism Detection
part 2

• The plagiarizer may leave behind some telltale
  signs of electronic plagiarism.
• When cutting and pasting plagiarized materials,
  the plagiarizer may not take the time to change
  the font color, size, and style to match their
  normal document template.
• Paragraph formats (e.g., indents, line spacing,
  paragraph spacing, and alignment) may also vary
  for parts of plagiarized documents from different
  sources.

13
A Structural Approach to Plagiarism Detection
part 3

• Readability scores, which are calculated based on
  the average number of syllables per word and
  words per sentence, can be used to identify parts
  of a student document that may be plagiarized.
• Documents whose individual paragraphs or
  sections have significantly different readability
  indices may indicate plagiarism.
• While none of these structural characteristics
  of is proof of plagiarism, they are easily
  automated and can be used as leading indicators
  of plagiarism.

14
A Structural Approach to Plagiarism Detection
part 4

• Visual Basic for Applications (VBA) is a
  programming language that can be run from within
  Microsoft Office applications.
• VBA gives the programmer access to the same tools
  that the Microsoft developers used to create
  MS-Office.
• VBA can automatically detect the existence of
  leading indicators of plagiarism including number
  of different font sizes, styles, and colors
  paragraph format parameters such as indents and
  line spacing.

15
A Structural Approach to Plagiarism Detection
part 5

• Sets of readability index scores are calculated,
  not only for each source document, but also for
  each of the individual paragraphs and/or
  sentences in a document.
• Each VBA procedure identifies instances of
  different plagiarism indicators.
• Some look only at a single document for
  characteristics that are often present in
  plagiarized work.
• Different types of source document databases can
  be used in concert with VBA procedures to
  identify potential cases of plagiarism.

16
Future Directions

• Run studies to see how effective VBA procedures
  are in identifying plagiarized documents.
• Develop user-interface for faculty.
• Identify custom Document Properties (MS-Word
  DocumentProperties Object)