A search result ranking method and device, a computer device and a storage medium
By acquiring and analyzing the text attributes, feature dimensions, and source information of search results, and combining them with click-through rate and intent prediction models, the problem of unreasonable ranking of multi-genre search results has been solved, achieving more accurate and reasonable ranking and improving user experience and usage rate.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- BEIJING ZITIAO NETWORK TECH CO LTD
- Filing Date
- 2023-05-15
- Publication Date
- 2026-05-29
AI Technical Summary
In existing technologies, the ranking methods for search results have poor rationality, which affects the usage rate of search results and user experience. In particular, in search results of various genres, the neglect of the correlation between result documents leads to inaccurate determination of click-through rates.
By acquiring text attribute information of search results, dimensional information under multiple preset feature dimensions, and source information of search terms, and combining click-through rate prediction model and intent prediction model, the document features and association features of the result documents are determined, and these features are used for multi-angle ranking optimization.
It improved the rationality and accuracy of search result ranking, enhanced user experience, and increased the usage rate of search results.
Smart Images

Figure CN116578725B_ABST