A spam speech recognition method and system based on social relationship graph

Abstract

The invention discloses a spam voice recognition method based on a social relationship graph. The method comprises the following steps: establishing a spam voice library for a user, extracting a historical call record of the user, constructing a social relationship graph of the user based on the historical call record, judging whether a newly-accessed calling number of the user is in the spam voice library, if not, judging whether the newly-accessed calling number of the user can be connected with a called number through an intermediate node through the established social relationship graph, if the connection fails, calculating the global reliability of the newly-accessed calling number of the user according to the established social relationship graph, and judging whether the obtained global reliability of the calling number is greater than a global reliability threshold. According to the spam voice recognition method disclosed by the invention, the spam voice recognition is carried out based on the global reliability and local reliability of the user, and thus the technical problem that the accuracy rate of the manually-initiated spam voice call recognition is not high in the existing spam voice recognition method can be solved.

Description

technical field [0001] The invention belongs to the technical field of wireless communication, and more specifically relates to a method and system for recognizing garbage speech based on a social relationship graph. Background technique [0002] With the material life of modern society becoming richer and richer, people's living standards are getting higher and higher, and the demand for mobile communication equipment is also increasing, and the number of mobile phone users continues to increase. However, with the number of mobile phone users The increase of spam voice and harassing calls is also showing a blowout trend, and the growth rate is obvious. In order to achieve their economic interests or political goals, more and more companies, groups, and individuals are harassing their target groups through voice spam, mainly in the form of telemarketing, dissemination of reactionary information, and fraud. For ordinary users, voice spam not only disrupts people's normal lif...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the above defects or improvement needs of the prior art, the present invention provides a spam speech recognition method and system based on a social relationship graph. Garbage speech recognition is performed based on the user's global trust degree and local trust degree, thereby solving the problem that the accuracy of artificially initiated spam voice call recognition is not high in the existing spam speech recognition method, which will violate user privacy in a certain sense, The technical problem of poor recognition efficiency for spam voice originating numbers whose communication behavior is not fixed

Images