Damocles
The University of Geneva's "Little Contest"
Damocles Logo

The University of Geneva's "Little Contest"

The University of Geneva has a website dedicated to Fraud and Duty According to University Members, which sends letters every few months to its subscribers. The letter of the 16th of July 2007 proposed a little contest for the summer vacation: Two articles published in reputable journals in 2006 contained more than 50 identical lines. What had happened? You will find my response below.

David Squire.

Response using Damocles

I retrieved the articles:
[HEU2006] Sameer Hosany, Yuksel Ekinci and Muzaffer Uysal, "Destination image and destination personality: An application of branding theories to tourism places", Journal of Business Research, Vol. 59, No.5, 638-642, 2006
[MiC2006] Anastasios Michailidis and Fotis Chatzitheodoridis, "Scenarios Analysis of Tourism Destinations", Journal of Social Sciences, Vol. 2, No. 2, 41-47, 2006.
from the web for analysis with the Damocles plagiarism detection system. It was necessary to make two passes. The first pass identified various copies of these two articles on the web so that they could be excluded from further analysis. This pass also located http://responsable.unige.ch/plagiat.pdf, as well as some subsequent publications in which the authors of HEU2006 and MiC2006 had liberally reused some of the text from those articles.

Once the lists of URLs to exclude were provided to Damocles, a variety of reports was generated in the second pass. The main results can be seen from:

HosanyEkinciUysal_jbr2006_damocles_r10_t0.01_Sort2_pW.html

and

MichailidisChatzitheodoridis_jss2006_damocles_r10_t0.01_Sort2_pW.html

The first indicates that HEU2006 contains almost no text that overlaps with documents on the web (except for their own work and MiC2006, already excluded).

The second, however, shows that MiC2006 contains many passages from works by other authors - including even major parts of the introduction and conclusion! In particular, the introduction is adapted with minor changes from:

[HaP2003] D. A. Haralambopoulos and H. Polatidis, Renewable energy projects: structuring a multi-criteria group decision-making framework, Renewable energy, 28, 6, pp. 961-973, 2003.
paragraphs 49 and 50 are from:
[KBZ2005] Nikos Kalogeras, George Baourakis, Costantin Zopounidis and Gert van Dijk, Evaluating the financial performance of agri-food firms: a multicriteria decision-aid approach, Journal of Food Engineering, 70, 3, pp. 365-371, 2005.
and much of paragraphs 51-68 is adapted from:
[IBA2003] Lazaros Iliadis, Christos Batzios, Garyfallos Arabatzis and Vasilios Liakos, The APIDESSYS multicriteria analysis decision support system: a tool for financial evaluation & ranking of Greek beekeeping industry, In Proceedings of the EFITA 2003 Conference, Debrecen, Hungary, July 5-9 2003.
None of these works is cited (though another paper by Kalogeras and some co-authors of KBZ2005 is).

Perhaps most egregiously, MiC2006 have called their system "DESDESSYS", clearly adapted from "APIDESSYS" in IBA2003. Indeed some of the overlapping text is identical other than for the substitution of "DESDESSYS" for "APIDESSYS".

I was not able to identify a common source for the overlapping text in HEU2006 and MiC2006. Since HEU2006 is otherwise "clean" and MiC2006 is riddled with text from sources by other authors, circumstantial evidence suggests that the authors of MiC2006 obtained a copy of HEU2006 before submitting their paper. The possibility of a common source not available on the web cannot be excluded, but seems unlikely.

I note that HEU2006 was received by the Journal of Business Research on June 1st, 2005. I cannot establish when MiC2006 was submitted to the Journal of Social Sciences.

Other Damocles reports allow further details to be seen.

HosanyEkinciUysal_jbr2006_damocles_r10_t0.01_Sort2_pL.html

shows the overlap between HEU2006 and MiC2006.

HosanyEkinciUysal_jbr2006_damocles_r10_t0.01_Sort2_p_AW.html and
MichailidisChatzitheodoridis_jss2006_damocles_r10_t0.01_Sort2_p_AW.html

include a "Show all Matches" link for each paragraph that allows one to see matches other than the best one highlighted in the main report.

HosanyEkinciUysal_jbr2006_damocles_r5_t0.01_Sort2_pW.html and
MichailidisChatzitheodoridis_jss2006_damocles_r5_t0.01_Sort2_pW.html

show the matches with the web when a run length threshold of five is used, rather than 10. These reveal little new, but the one for MiC2006 further emphasizes the degree to which text from other authors has been minimally adapted for use in this paper.

Clearly those most at fault are the authors responsible for the plagiarism. Given that several of the papers used liberally in MiC2006 appeared in 2003, the editors of JSS and the reviewers of MiC2006 might have been able to detect the problem, but it is only likely if a tool such as Damocles is used systematically.