|
With the development of the Internet , the world is moving towards the information economy ; information resources are not scarce , scarce resources is a means of discovery information . The search engine is therefore came into being , but now the popular search engines generally only text-based information search , even for voice information, such as news record , transcript , etc., it is just only retrieve some meta information : author, title, description, file size, etc., but not content -based audio retrieval. With voice resources on the Internet , more and more people will increasingly want to be able to search the content of speech . For to this demand, the paper presents key information based voice search engine , the so-called critical information refers to the speaker's voice resources or special keywords. For example : We want to search the Internet with \In this thesis, a key figure speaker recognition technology to search the conversation , lectures and other recordings ; recognition technology to search by keywords contain the keyword voice resources . The main contents of this paper : First, put forward a framework for voice search engine . Second , keyword recognition technology , speech recognition technology with search engine technology integration . Third, a key figure in the search engine Voice Search application of resources . Fourth, the design of the search engine optimization techniques local paper get the following funding : National Development and Reform Commission \( 60273059 ) , the National outstanding Youth Science Foundation ( 60525202 ) , the Ministry of Education \
|