Digital creative marketing scene intelligent management method based on multi-modal feature fusion

By using a self-attention mechanism and a bidirectional GRU network to model user interaction sequences in parallel, and combining a gating mechanism and dynamic edge weight updates, the problem of single user interest representation dimension and limited recommendation personalization is solved, achieving higher accuracy personalized content recommendation and improving marketing conversion efficiency.

CN122335341APending Publication Date: 2026-07-03DEEPIN MEDIA GROUP CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
DEEPIN MEDIA GROUP CO LTD
Filing Date
2026-04-08
Publication Date
2026-07-03

AI Technical Summary

Technical Problem

Existing technologies struggle to accurately capture the temporal evolution of user interests. The lack of coordinated embedding of location and time information in multimodal feature fusion results in a single dimension of user interest representation and insufficient dynamic adaptability, making it difficult to meet the needs of rapidly changing user intentions in creative marketing scenarios. Furthermore, the static edge weights during graph neural network propagation limit personalized recommendations.

Method used

The user interaction sequence is modeled in parallel using a self-attention mechanism and a bidirectional GRU network. A gating mechanism is introduced to achieve adaptive fusion of short-term and long-term interest vectors. The edge weights are dynamically updated in the graph neural network to generate personalized creative content recommendations.

Benefits of technology

It significantly improves the accuracy and scenario adaptability of user dynamic interest representation, and enhances the accuracy of personalized content recommendation and marketing conversion efficiency in creative marketing scenarios.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122335341A_ABST
    Figure CN122335341A_ABST
Patent Text Reader

Abstract

This invention discloses an intelligent management method for digital creative marketing scenarios based on multimodal feature fusion, relating to the field of multimodal feature fusion technology. The method includes the following steps: processing raw data to obtain short-term and long-term interaction sequences; performing fixed-length processing on the short-term interaction sequences to obtain fixed-length short-term interaction sequences; performing fusion processing, first embedding processing, and second embedding processing on the multimodal feature set of creative content to obtain creative content embedding matrices, location embedding matrices, and time embedding matrices; processing the fixed-length short-term and long-term interaction sequences to obtain user short-term and user long-term interest vectors; adaptively fusing the user short-term and user long-term interest vectors to obtain a user dynamic interest representation; calculating a balanced interest score matrix based on the user dynamic interest representation; and performing propagation processing through an improved graph neural network model to generate a personalized creative content recommendation list.
Need to check novelty before this filing date? Find Prior Art