The Acquisition of Common Sense Knowledge by Being Told: An Application of NLP to Itself

This paper shows how the knowledge of a semantic interpreter can be bootstrapped for other semantic interpretation tasks. Methods are described for automatically acquiring common sense knowledge and for applying this knowledge to noun sense disambiguation. Ordinary concepts are described by several plain English sentences that are parsed and semantically interpreted. The semantic interpreted sentences are stored under these concepts to be used for semantic interpretation tasks. This paper explains the description of the concepts, the interpretation of the sentences and two algorithms for noun sense disambiguation that use the acquired knowledge.