The question speech section extraction processing program of the voice data and the customer inquiry tendency presumption processing program which uses the question speech section of method and the device, and the voice data, method and the device

摘要

Question speech section extracting equipment 10 inputs the audio conversational data 3 where, sound of the operator and the customer is recorded to another channel calculates the power value of sound every of unit section with each channel, the preceding main leadership speech person and the te it decides the person where the time when you speak with fixed power in specified section is longer, at the same time preceding, speaks, from first the preceding main leadership speech person of speech continues and the period when you speak specifies. If and the particular preceding main leadership speech person is the customer, particular period the question speech section it specifies, extracts the said question speech section data 7. Speech input system 20, administering speech recognition processing to question speech section data 7, extracts keyword 9. Trend analysis device 25 analyzes the tendency of the inquiry of the customer making use of the keyword 9 which, is extracted inquires the result and outputs as tendency information 5.