Method and system for identifying vest account numbers of forum users

Abstract

The invention provides a method for identifying vest account numbers of forum users. The method comprises the steps of training a classification model depending on various user account numbers in a training set and characteristic vector of each text, determining the user account number of the training set to which each text in a test set is classified through a trained classification model, and identifying vest account numbers depending on the classification result. The method disclosed by the invention selects characteristics from speech text data of forum user account numbers, and judges relation between various account numbers of a same person through mining similarity of language styles of the account numbers, thus improving the probability of identifying vest account numbers. Aiming at the characteristics of network language which is less rigorous in grammar and has many cyber words, the method extracts and analyzes effective characteristics in the speech text, so as to avoid a problem that lexicon updating content and speed are unable to keep pace with trend of the network language, thus reducing complex operation of maintaining a segmentation dictionary and improving accuracy of identifying the vest account numbers.

Description

technical field [0001] The invention belongs to the field of network security, in particular to a method for identifying vest accounts of forum users. Background technique [0002] Today, more than 1.5 billion people around the world use social networks, and the number of monthly active users of global social networks has already exceeded 2 billion, generating a large amount of speech every day. It is difficult to implement real-name registration in the background of social networking sites, and it has not yet been fully popularized; even if the background of the website is based on the real-name system, most of the online speeches in the front-end of the website are anonymous, and it is difficult to know the real identity of the network users who are the speeches on the Internet. In a social network, it is very common for a person to have multiple accounts (ID), or have several accounts on the same website or register accounts on different websites. When a person registers...

AI Technical Summary

Problems solved by technology

However, in these methods, the user information that can be used is limited to the IP address used by the account, the registration information of the account, or the operation behavior of the account, so the identification range is small and the identification accuracy is low.

Images