Recommendation method and system based on instant voice content detection

Abstract

The invention discloses a recommendation method and system based on instant voice content detection. The method comprises the following steps: collecting voice information of a user; converting the voice information into a text content and indexing to obtain an index database; retrieving a recommended article dictionary to obtain an instant voice keyword; calculating a user interest model; calculating the similarity between the user interest model and each preset article model; and generating a candidate article set in a recommended article information base, selecting a specified number of articles, and sending a recommendation list corresponding to the selected specified number of articles to a target user. The system comprises a voice collection module, a voice content analysis module, a voice keyword retrieval module, a user interest deduction module, an article model similarity calculation module and a list sending module, which are corresponding to the method. The recommendation method and system disclosed by the invention can make full use of the close correlation of daily voice communication chats, watched film and television drama lines and other voice information of the user with the current interest of the user, the pertinence of the recommended articles is strong, the accuracy is high, and the user experience is good.

Description

technical field [0001] The invention relates to an information recommendation system, in particular to a recommendation method and system based on real-time voice content detection. Background technique [0002] The recommendation system is a special form of information filtering system. The recommendation system will actively record the user's historical interest information during the interaction with the user, model the user's information needs, form a knowledge model about the user's interest and preference, and According to this model, the final information recommendation service is realized. With the trend of diversification of big data, in the process of users instant chatting and watching online streaming media, it is difficult to make the recommendation system relate to Voice-related information such as the user's real-time voice chat content, the lines of the film and television bureaus watched, and the voice information such as the user's chat content and the voi...

AI Technical Summary

Problems solved by technology

The patent document with Chinese patent application number 201310740533.9 discloses a method for detecting keywords in speech based on confusion network, but it does not combine this speech keyword detection method with a recommendation method

The patent document with Chinese patent application number 201410668170.7 discloses a recommendation system that can use user-input voice information to convert the voice information input by the target user into corresponding text information through voice recognition, and then perform subsequent query and recommendation processing, but the recommendation The method and system mainly treat voice input as a human-computer interaction interface, rather than performing relevant detection on voice content in specific application scenarios

Chinese patent application number 201310446815.8 discloses a personalized video content recommendation method and system. By extracting the voiceprint features of actors in videos watched by users, and then finding other interested actors with similar voiceprint matching, heuristically guide And recommending users to other videos they are interested in, this method is mainly by detecting the voiceprint features of the videos watched by the user, and does not detect the content of these voices

Images