一种大语言模型智慧搜索分组方法、系统、设备及介质
By constructing a search operation matrix and calculating saliency parameters, the problem of existing search systems being difficult to adjust multiple times was solved, enabling synchronous feedback from multiple search objects and improving the efficiency and accuracy of search interaction.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- JIMENG COMPUTER (BEIJING) CO LTD
- Filing Date
- 2026-01-29
- Publication Date
- 2026-07-17
AI Technical Summary
Existing search systems struggle to simultaneously characterize the differences between multiple potential search objects during a single search, requiring users to adjust their search criteria multiple times to obtain search feedback from different directions or emphases, resulting in low search interaction efficiency.
By acquiring users' search requests and historical behavior data, we extract search status features, construct a search operation matrix, calculate saliency parameters and update group weights, and call a large language model to generate differentiated search results.
This allows for the simultaneous output of feedback results for multiple search objects within a single search request, improving the coverage and accuracy of search results.
Smart Images
Figure CN122019569B_ABST