Similarity Texter

Additional Information

Softwaretest 2008

Results of the 2008 software test

The results of our investigations in 2008 should not be interpreted as a purchase recommendation.  We have only tried to assess how well the systems recognize plagiarism overall, and to evaluate the usability of those systems in daily work in universities, colleges,  and high schools. Institutions that plan to acquire software for detecting plagiarism should more closely investigate those systems which are seem to fit their situation best.

Overview of the test

The overview chart shows that the results of the test in a large chart. The test method and the evaluation criteria are described on separate pages. The test cases are also explained in more detail.

A summary of the test and the collusion recognition, as well as a press release and pictures, are also available.

Evaluation of  plagiarism detection systems

Very good systems

The year 2008 is no different from previous tests: there was no system in this category, although the criteria were a bit lower this time. Only humans can reach very good results here!

Good systems

Satisfying systems

Sufficient systems

Insufficient systems

Aborted tests

Other system

  • ETBlast – looks for plagiarism in the MEDLINE database
  • ProfNet – access denied
  • Eve2 – Money is accepted, but the software does not work

Evaluation of collusion recognition systems

Good systems

Satisfactory systems