Social information based user attribute prediction method and system

Abstract

The invention provides a social information based user attribute prediction method and system, which can predict a real identity attribute of a user according to a social group in which the user joins. The social information based user attribute prediction method includes: data acquisition, text analysis, social group analysis and user attribute prediction; acquiring nationwide school names and major names, constructing a prior knowledge base, extracting major / school short name features, using a classifier to predict short names, and perfecting the prior knowledge base; analyzing social groups on the basis of keyword matching, and describing a user education background; calculating valuable information for prediction provided by each group according to experience and a 3 sigma principle; considering differences of each groups, screening a corresponding social group according to the fact if there is class and grade group information in the group in which the user joins, estimating information provided by each group, and predicting the user attribute through weighting calculation. The social information based user attribute prediction method and system can be used for detecting the authenticity of the user attribute, can be used for detecting user social information, and can perform message accurate pushing and friend recommendation according to the predicted user attribute.

Description

Technical field [0001] The invention belongs to the technical field of data mining, and particularly relates to a method and system for predicting user attributes based on social information. Background technique [0002] With the continuous development of Internet technology, the big data era of "data and resources" has come. In recent years, user information leakage incidents have emerged one after another. In 2011, the data of the CSDN website was dragged into the database. Then, in 2012, 40 million information leaks on the Tianya Forum, and in 2013, the personal information of 6 million users on the social networking site Facebook, such as Phone numbers, addresses, e-mail addresses, etc. were also leaked. In the same year, data from Adobe, a well-known American software company, was also leaked, including the user's name, credit card and credit card expiration time. In 2014, China's railway booking website 12306 also showed user data In the same year, Ctrip was also reported...

AI Technical Summary

Problems solved by technology

In order to infer the user's attributes, the traditional method needs to know all the user's friends, and then infer the user's attributes based on the friend information. However, the relationship between the user's friends is different, and the value of the information provided is also different.

Images