摘要:
An implementation for detecting interruptions in audio conversations between a plurality of participants comprising continuously recording the most recent predetermined time segment of the conversation stream of one participant in the conversation, monitoring the conversation stream of the participant for interruptions in said conversation stream and storing the most recent segment prior to the interruption in response to a detection of an interruption. The segment is represented by a conversation marker. The user may then select the conversation marker to playback this most recent segment, so that he may then decide whether he wishes to resume or pursue the interrupted conversation fragment.