Class model adaptation for speech summarisation

Abstract

The performance of automatic speech summarisation has been improved in previous experiments by using linguistic model adaptation. We extend such adaptation to the use of class models, whose robustness further improves summarisation performance on a wider variety of objective evaluation metrics such as ROUGE-2 and ROUGE-SU4 used in the text summarisation literature. Summaries made from automatic speech recogniser transcriptions benefit from relative improvements ranging from 6.0% to 22.2% on all investigated metrics.

A Cascaded Broadcast News Highlighter

“…Other works on broadcast news data include [4]- [6]. Related tasks such as summarisation of multiparty meetings, voicemails and lectures/talks have also attracted a growing amount of interest [7]- [11]. Broadcast news highlighting is closely related to works on speech gisting and headline generation, where short sentences are generated as part of a speech understanding system [12]- [14].…”