Information recommendation method, information recommendation device and computer equipment
By combining the recommendation information and the multimodal characteristics of the recommended object, an initial recommendation strategy is generated and the strategy is optimized based on the dynamically changing recommendation object characteristics, the problem of poor recommendation results in the existing technology is solved, and higher recommendation accuracy and personalization are achieved.
CN119691270APending Publication Date: 2025-03-25MASHANG CONSUMER FINANCE CO LTD
0 Cites 0 Cited by
Patent Information
- Application Number
- CN202411755931.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-02
- Publication Date
- 2025-03-25
AI Technical Summary
Technical Problem
Existing information recommendation technologies are difficult to accurately capture the personalized interests and needs of the recommended subjects, resulting in poor recommendation results.
Method used
By combining the recommendation information and the multimodal characteristics of the recommended object, an initial recommendation strategy is generated and the strategy is optimized based on the dynamically changing recommendation object characteristics to improve the accuracy of recommendations.
Benefits of technology
It has achieved a deeper understanding of the needs of the recommended object, thereby improving the accuracy and personalization of information recommendations and improving the recommendation effect.
✦ Generated by Eureka AI based on patent content.
Abstract
The invention discloses an information recommendation method, an information recommendation device and computer equipment. First, a first feature of first recommendation information and a second feature of a recommendation object are determined. A first recommendation policy is then determined based on the first feature and the second feature. And when the second feature of the recommendation object changes, adjusting the first recommendation strategy according to the changed second feature to obtain a second recommendation strategy. And finally, based on a second recommendation strategy, recommending second recommendation information in the first recommendation information to the recommendation object. Thus, the first recommendation strategy is generated based on the first recommendation information and the multi-modal features formed by the first features and the second features of the recommendation objects, the features of the recommendation information and the features of the recommendation objects can be extracted more comprehensively, and then the first recommendation strategy is optimized based on the dynamically changing second features of the recommendation objects. The demand of the recommended object can be understood more deeply, so that the accuracy of information recommendation is improved.
Need to check novelty before this filing date? Find Prior Art