Cross-social network user identity matching method

Abstract

The invention provides a cross-social network user identity matching method, which is characterized by comprising the following steps of: S1, matching known user name data to obtain positive sample pairs, disrupting a corresponding user name sequence, randomly extracting data and obtaining negative sample pairs to form a user name sample pair data set; and S2, performing vectorization representation on each user name sample pair in the user name sample pair data set by using an alphabet. And S3, constructing a user identity matching network model. And S4, inputting the sample pair training data set into the user identity matching network model, performing supervised training by using a cross entropy loss function, and obtaining a trained user identity matching network model when the number of times of training meets an end condition. And S5, pairing the two pieces of to-be-tested user name data to obtain to-be-tested sample pair data. And S6, inputting to-be-tested sample pair data into the trained user identity matching network model to obtain a user name matching result.

Description

technical field [0001] The invention relates to a method for user identity matching across social networks. Background technique [0002] In recent years, with the development of information technology, people have gradually entered the Internet age. As a form of online network service, online social network provides people with new social channels besides real life, and also represents a new way of information dissemination and interaction. The diversification of social networks makes each Internet user leave clues of user behavior information on different social platforms, and there are huge value of user behavior data hidden in major social networking platforms. But at present, even with the support of big data platform data analysis, most social network analysis researches are carried out for a single social platform. The data under the same platform has limitations in the range of user groups and behavioral content limitations. There is no information sharing mechanis...

AI Technical Summary

Problems solved by technology

However, as the data size increases, the repetition rate of user names will increase, which increases the difficulty of matching algorithm research based only on user name information.

In addition, the premise of username-based user identity matching is that users tend to use the same or similar usernames in different social networks. If this assumption is not true, the recall rate of the algorithm will be reduced.

Images