Data query method, device, equipment, storage medium and program product

By constructing a target text library and querying it at different granularities, the problem of mismatched query granularity is solved, achieving high efficiency and accuracy in data querying and meeting users' multi-level query needs.

CN122173635APending Publication Date: 2026-06-09BEIJING KECHENG TECH DEV CO LTD
View PDF 6 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
BEIJING KECHENG TECH DEV CO LTD
Filing Date
2026-05-11
Publication Date
2026-06-09

AI Technical Summary

Technical Problem

Existing technologies, when querying text data, use a fixed splitting method, which leads to a mismatch between the granularity of the content the user wants to query and the actual text blocks retrieved, thus reducing the accuracy of data querying.

Method used

By constructing a target text library containing text blocks of the original text at different candidate granularities, the target granularity is determined based on the user's query request, and a preliminary query is performed. When the query confidence is low, a supplementary granularity is selected for supplementary query. Feature extraction and fusion processing are performed by combining the target text block and the candidate text block to generate the final query result.

Benefits of technology

It improves the accuracy and efficiency of data querying, ensures the reliability of initial queries, and enhances the accuracy of results through supplementary queries when necessary, thus meeting users' query needs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122173635A_ABST
    Figure CN122173635A_ABST
Patent Text Reader

Abstract

This application relates to a data query method, apparatus, device, storage medium, and program product. The method includes: responding to a user query request, querying a target text database to obtain candidate text blocks and query confidence scores at a target granularity; if, based on the query confidence scores, it is determined that supplementary queries are needed, at least one candidate granularity other than the target granularity is used as a supplementary granularity; a target text block is determined based on the candidate text blocks and the size relationship between the supplementary granularity and the target granularity; and a target query result is determined based on the target text block and the candidate text blocks. The target granularity is the granularity of the text to be queried as indicated by the user query request, and the target text database contains text blocks obtained by dividing the original text at different candidate granularities. This method can improve the accuracy of data queries.
Need to check novelty before this filing date? Find Prior Art