Coping with aboutness complexity in information extraction from spoken dialogues

We report on the strategies in automatic summariza-tion of spontaneous spoken dialogues. Our dialogue sum-marization system, MIMI, recognizes key linguisitc patterns and merges information to construct a summary of conference room scheduling dialogues. Dialogues about single reservations can be accurately summarized with a simple merging scheme, but we need several extensions and changes to cope with the aboutness complexity in unrestricted dialogues | knowing exactly how many reservations are being discussed and which reservation values must be updated by subsequent utterances. A side eeect of these extensions is a new problem of overrecognition caused by spoken language dissuencies.