Abstract
This paper overviews 15 plagiarism detectors that have been evaluated
within the fourth international competition on plagiarism detection at PAN12.
We report on their performances for two sub-tasks of external plagiarism detection:
candidate document retrieval and detailed document comparison. Furthermore,
we introduce the PAN plagiarism corpus 2012, the TIRA experimentation
platform, and the ChatNoir search engine for the ClueWeb. They add scale and
realism to the evaluation as well as new means of measuring performance.