Authorship identification and author fuzzy "fingerprints"

• Published in: 2011 Annual Meeting of the North American Fuzzy Information Processing Society pp. 1 - 6
• Main Authors: Homem, Nuno, Carvalho, Joao Paulo
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.03.2011
• Subjects: Algorithm design and analysis; approximate algorithms; Approximation algorithms; data streams; Feature extraction; Fingerprint recognition; Frequency selective surfaces; frequent elements; fuzzy fingerprints; Monitoring; Radiation detectors; similarity; vector valued fuzzy sets
• DOI: 10.1109/NAFIPS.2011.5751998

Fingerprint identification is a well-known technique in forensic sciences. The basic idea of identifying a subject based on a set of features left by the subject actions or behavior can be applied to other domains. Identifying text authorship based on an author "fingerprint" is one such application. This paper considers the problem of extracting "fingerprints" from texts and matching them with those obtained from a set of known authors. It presents an innovative fuzzy fingerprint algorithm based on vector valued fuzzy sets. Words and other stylometric features are used to create the fingerprint. The implementation is based on an approximated fast and compact algorithm that allows the method to be used on near real time, even for a large number of authors and texts.