Directly quoted from their FAQ:
Plagiarism.org has compiled a massive database of digital material by
continually cataloging and indexing online academic works with automated web
robots. Online paper mills are a major focus. Papers from participating
courses and other academic web sites are archived as well. To detect
plagiarism from the Internet at large we interface with the major search
engines. Students papers are then compared against this database by
iThenticate®, our advanced matching algorithm, and a report of originality
is generated. This report is sent to the instructor, and provides an
analysis of the structure and content of the paper. In the structure
analysis hyperlinks are provided to the closest five manuscripts with the
unoriginality of each being ranked by a numerical score in conjuction with a
color code to indicate the approximate level of unoriginality. In the
content analysis portion of the report unoriginal portions of the text are
highlighted and underlined using the same color as the hyperlink provided at
the top of the page.
- Jim