摘要：This paper constructs an authorship-linked collection or corpus of anonymous, sex-seeking ads found on the classified ad website Craigslist. This corpus is then used to validate an authorship attribution approach based on identifying near duplicate text within ad clusters, providing insight into how often anonymous individuals post sex-seeking ads and where they meet for encounters. We find that while a near duplicate detection approach fails to identify all ads written by a single author, it does identify subsets of a single author's ads with high precision, meaning the system is capable of measuring the posting behavior of real Craigslist users.
Loading...
