Search ranking method and apparatus, server, and storage medium
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- PING AN TECH (SHENZHEN) CO LTD
- Filing Date
- 2023-04-12
- Publication Date
- 2026-05-26
AI Technical Summary
Existing search ranking models require separate training for each content type when faced with search results of various content types. This results in long prediction times and is prone to error accumulation, leading to low accuracy in click probability.
By obtaining the popularity feature vector of online resources that match the search content, constructing the popularity feature vector using the same popularity influence factor, and running a unified click probability prediction model, we can handle online resources of different content types and avoid the accumulation of errors from multiple models.
It reduces the calculation time, improves the accuracy of click probability prediction, and greatly enhances the accuracy of search ranking results.
Smart Images

Figure CN116467519B_ABST