发明名称 PROCESSING PROGRAM AND PROCESSING APPARATUS FOR SPECIFYING PART OF VOICE DATA DESIRABLY TO BE HEARD
摘要 PROBLEM TO BE SOLVED: To extract a "part desirably to be heard" from voice data in which interactive contents between a customer and an operator are recorded. SOLUTION: An utterance length calculation section 11 calculates an utterance length (utterance length information 24) of each utterance in utterance voice information from interactive voice information 20 where interactive contents between the customer and the operator are separately recorded. An utterance characteristic change point of time specifying section 13 estimates an interactive structure of the interactive voice information 20 on the basis of the utterance length information 24 to specify an utterance length characteristic change point of time (utterance length characteristic change point of time information 26) indicating a change in talkers for making utterance for a long period of time. A part desirably to be heard specifying section 15 extracts a time of the operator for searching information (information searching point of time) on the basis of operation log information 22 as to operations which the operator makes during the reception with the customer to determine an end point of time of the part desirably to be heard and determines a start point of time of the part desirably to be heard on the basis of the utterance length characteristic change point of time. COPYRIGHT: (C)2007,JPO&INPIT
申请公布号 JP2007184699(A) 申请公布日期 2007.07.19
申请号 JP20060000559 申请日期 2006.01.05
申请人 FUJITSU LTD 发明人 ONODERA SACHIKO;NANBA ISAO
分类号 H04M3/53 主分类号 H04M3/53
代理机构 代理人
主权项
地址