Using CNL for knowledge elicitation and exchange across story generation systems

Abstract

Story generation is a long standing goal of Artificial Intelligence. At first glance, there is a noticeable lack of homogeneity in the way in which existing story generation systems represent their knowledge, but there is a common need: their basic knowledge must be expressed unambiguously to avoid inconsistencies. A suitable solution could be the use of a controlled natural language (CNL), acting both as an intermediate step between human expertise and system knowledge and as a generic format in which to express knowledge for one system in a way that can be easily mined to obtain knowledge for another system - which might use a different formal language. This paper analyses the suitability of using CNLs for representing knowledge for story generation systems.