Third-party website promotion method and system based on user behavior characteristic portrait

By constructing a cross-site user behavior feature system and dynamic interest tags, combined with a dual-channel strategy generation model, the problem of insufficient cross-domain data integration and dynamic user profile in existing third-party website promotion methods is solved, thus achieving precise promotion strategy generation and resource optimization.

CN122240920APending Publication Date: 2026-06-19TIANJIN XIAJIANG TECHNOLOGY CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
TIANJIN XIAJIANG TECHNOLOGY CO LTD
Filing Date
2026-03-17
Publication Date
2026-06-19

AI Technical Summary

Technical Problem

Existing third-party website promotion methods rely on coarse-grained metrics, lack cross-domain behavioral data integration, have weak user profile dynamism and portability, and lack a closed-loop promotion mechanism centered on behavioral characteristics, resulting in limited adaptability and foresight of promotion strategies.

Method used

By collecting heterogeneous data from multiple sources, performing timestamp alignment, device fingerprint matching, and session reconstruction, multi-dimensional behavioral features are extracted to generate user behavior feature vectors. Unsupervised clustering is used to divide user groups, dynamic interest tags are constructed, gradient boosting tree model is combined to evaluate conversion contribution, a dual-channel strategy is introduced to generate a model for promotion priority ranking, and cross-site data collaboration is achieved through federated learning.

Benefits of technology

It enables detailed profiling of user behavior across different sites, dynamically captures interest evolution, improves the relevance and timeliness of promotional content, enhances user reach efficiency and conversion quality, and optimizes the allocation of promotional resources while protecting data privacy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122240920A_ABST
    Figure CN122240920A_ABST
Patent Text Reader

Abstract

This invention relates to the field of internet information analysis and promotion technology, and discloses a method and system for promoting third-party websites based on user behavior feature profiling. The method includes cross-site user behavior sequence construction, multi-dimensional behavior feature vectorization, dynamic user segmentation, generation of interest tags and conversion potential scores, interest shift detection, and identification of dominant factors. Based on this, a weighted promotion strategy model is constructed to achieve targeted content distribution based on priority scores and preferred channel combinations. This application can improve the accuracy of user profiling and the real-time nature of promotion, and enhance cross-domain behavior integration capabilities and closed-loop optimization effects.
Need to check novelty before this filing date? Find Prior Art