Hilke Reckman, Jeff Orkin, Deb Roy

Abstract

We discuss the use of data from a virtual world game for automated learning of words and grammatical constructions and their meanings. The language data is an integral part of the social interaction in the game and consists of chat dialogue, which is only constrained by the cultural context, as set by the nature of the provided virtual environment. This paper presents a preliminary exploration of syntactic and semantic aspects of the dialogue in the corpus. We show how simple association metrics can be used to extract words, phrases and more abstract syntactic patterns with targeted meanings or speech-act functions, by making use of the nonlinguistic context.