Figment Software Limited

A renowned University academic specialising in algorithmics requested that we develop a plagiarism detection system. The system employed the Smith-Waterman string matching algorithm (used in DNA matching) to identify similarities and dissimilarities between two or more ASCII text documents.

The similarity detection mechanism differs from a simple "diff" program by being capable of ignoring/overlooking reordering of sections of documents or use of synonyms to hide plagiarism.

The system was developed using Java and provided a full graphical user interface, including a estimate of how long the time-consuming comparison process might take based on a sampling technique.

The system was a great success and was featured in several white papers, proving that the Smith-Waterman algorithm, normally used on a string length of just four symbols, can be used to great effect on larger string lengths - here, full documents such as University dissertations - within acceptable time-requirements and in a variety of domains.

For further information, Contact Us.