VOICE ABSTRACT FORMING METHOD AND SYSTEM FOR CONFERENCE

摘要

The present invention relates to the field of voice recognition. Disclosed are a voice abstract forming method and system for a conference. According to the present invention, by analyzing a speaking position, identity information, personnel data, and other information of a speaker in a conference, a weight coefficient of the speaker is determined, so that candidate key speaking segments corresponding to different speakers can be obtained by using different preset strategies according to the weight coefficients of speakers. Furthermore, according to characteristics of speaking content, for example, an interval having a large probability of containing important content of a speech on a speaking timeline, or key adversatives and key conjunctions in the important content of the speech, a set of candidate key speaking segments is intercepted. The intercepted set of candidate key speaking segments is processed to obtain a set of audio/video segments for forming a voice abstract. More content is extracted from important statements, while less content is extracted from non-important statements. The content of a finally formed abstract is more reasonable, and the more effective help is provided for users.