Assessing the Viability of the Urban Dictionary as a Resource for Slang

The use of slang is ubiquitous, especially in internet communities. This paper evaluates the success of conventional dictionary and thesaurus-based semantic similarity assessments on The Urban Dictionary, an online, user-contributed dictionary for contemporary slang. Conventional methods are shown to perform poorly, and problematic aspects of the corpus are examined. Language use on the internet is found to be very unconventional, and techniques designed for conventional, well-formed language are likely to perform poorly. Future work is suggested in order to understand unconventional language use and the development of neologisms.