Pattern discovery in annotated dialogues using dynamic programming

We describe a scenario to discover patterns in annotated dialogues: first dialogues are transcripted and annotated; then two-dimensional patterns are extracted; finally the semantic pertinence of the extracted patterns can be evaluated by an expert. The dialogues are annotated by using a grid designed by the expert. Recurrent patterns are extracted in the annotated dialogues using dynamic programming with the help of a substitution matrix specifically designed for the task. In this article we focus on the method developed for extracting the patterns and we show some extracted patterns on annotation of dialogues between parents and four year old children during the narration of child stories whose context is mentalist.