System and method for digging key information of telephony nature conversation

Abstract

The invention provides a system which can excavate the key information in a telephone natural dialogue, as well as a method thereof. The system comprises a telephone voice detection module, a voice keyword extraction module, a voice keyword database, a data excavation module and a user preference database which are connected sequentially, wherein, the voice keyword extraction module, the voice keyword database, the data excavation module and the user preference database are arranged in a computer; the telephone voice detection module is connected with a voice output device through a public switching telecommunication network (PSTN) or Internet. The invention can excavate the user's interests or data information such as the consumption demand, the consumption habit, the consumption field, the consumption category, etc. from the telephone voice, and create the user preference database, thereby can provide more reliable and more comprehensive and effective supports for an enterprise to make a marketing decision; the invention is applicable to a fixed telephone, a mobile telephone, a voice chat, etc., and has a wide business application prospect.

Description

technical field [0001] The communication network voice data mining technology of the present invention specifically refers to a system and method for mining key information in natural telephone conversations. Background technique [0002] In recent years, various text data mining (Data Mining) technologies (such as latent semantic analysis technology LSA, etc.) have been widely used in product recommendation in e-commerce, similar document recommendation in document search, etc. It can find out what kind of other products the user who buys this kind of product usually buys, and what kind of other useful information the user who searches for the document usually consults, so that the user can enjoy the information service more comfortably. Information providers collect statistics and predictive analysis of data and information based on data mining technology, providing strong support for enterprises to analyze and discover business opportunities, and conduct business developm...

AI Technical Summary

Problems solved by technology

However, the behavioral characteristics selected by it only belong to the data information mining analysis at the "face" level, and cannot obtain and analyze customers' hobbies or consumption needs, habits, scope, categories, etc., which are more specific and belong to the "point" level data information. , so that information providers cannot provide more specific and targeted data information for enterprises, and the commercial value of the provided data information is limited

Images