User geographical position positioning method and device, terminal equipment and storage medium

By constructing a multi-dimensional user relationship graph and adopting a multi-scale feature fusion model using graph attention networks, the problem of inaccurate user geolocation in existing technologies is solved, achieving more efficient and accurate user geolocation prediction.

CN120429750BActive Publication Date: 2026-06-26NANKAI UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202510540248.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-04-27
Publication Date
2026-06-26
Estimated Expiration
2045-04-27

AI Technical Summary

Technical Problem

Existing technologies fail to fully capture the geographic distribution characteristics of users in user geolocation, resulting in low positioning accuracy and efficiency, especially in social network graphs where high-order neighbors with close geographical proximity but weak social connections are ignored.

Method used

By acquiring target social data of social users, a multi-dimensional user relationship graph is constructed using a pre-established geolocation prediction model. A multi-scale feature fusion model based on graph attention networks, including a dual-channel graph encoder and a cross-gated feature fusion algorithm, is then used to process the social network graph and location association graph to obtain the overall representation and location label of the user.

Benefits of technology

It improves the accuracy and efficiency of user location prediction by constructing social network graphs and location association graphs through social relationships and latent location correlations, thus achieving more comprehensive user geolocation positioning.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120429750B_ABST
    Figure CN120429750B_ABST
Patent Text Reader

Abstract

The application discloses a positioning method and device for user geographic position, a terminal equipment and a storage medium, comprising: obtaining target social data of a social user, wherein the social data at least includes target longitude and latitude and target user text data; determining geographic position information of the social user corresponding to the target social data according to a previously established geographic position prediction model, wherein the previously established geographic position prediction model is obtained by processing sample data, determining a multi-dimensional user relationship graph corresponding to the sample data, and training a multi-scale feature fusion model based on a graph attention network by using the multi-dimensional user relationship graph, a social network graph and a position association graph are constructed through the social relationship and the hidden position correlation of the user; the multi-scale feature fusion model is processed to obtain the overall representation of the user; the position of the user is inferred by a user positioning classifier, and the accuracy and efficiency of the user position prediction are improved.
Need to check novelty before this filing date? Find Prior Art