This paper presents an interactive interface to create visually structured summaries of human conversations via ontology mapping. We have built highly accurate classifiers for mapping the sentences of a conversation in an ontology, which includes nodes for the Dialog Acts (DA) properties such as decision and subjective, along with nodes for the conversation participants. In contrast with previous work, our classifiers do not rely on features specific to any particular conversational modality.

more »

... are currently developing an interactive interface that allows the user to generate visual structured summaries by searching a conversation for sentences according to the ontology mapping. Our first prototype comprises two panels. The right panel displays the ontology, while the left panel of the our prototype displays the whole conversation, where sentences are temporally ordered. Given the information displayed in the two panels, the user can generate visual, structured summaries by selecting nodes in the ontology. As a result, the sentences that were mapped in the selected nodes will be highlighted. Our initial prototype builds on a component of the GATE system, which was originally developed as a tool for text annotation.