Analysis and Reuse of Plots Using Similarity and Analogy

Abstract

A plot is a partially ordered set of events. Plot analysis is a relevant source of knowledge about the agents' behavior when accessing data stored in the database. It relies on logical logs which register the actions of individual agents. This paper proposes techniques to analyze and reuse plots based on the concepts of similarity and analogy, borrowed from cognitive science and linguistics. The concept of similarity is applied to organize plots as a library, and to explore the reuse of plots in the same domain. By contrast, the concept of analogy helps reuse plots across different domains. The techniques proposed in this paper find applications in areas such as computer games and emergency response information systems, as well as some traditional business applications.