Question and answer method, table question and answer model training method and device, equipment and medium

By converting table documents into coding formats and generating table questions, and entering a pre-trained table question and answer model, the problem of low utilization of complex table data by natural language models is solved, and more efficient table information utilization and question-answer accuracy are achieved.

CN120337939APending Publication Date: 2025-07-18HANGZHOU ALICLOUD FEITIAN INFORMATION TECH CO LTD
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202410039472.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-01-10
Publication Date
2025-07-18

Smart Images

  • Figure CN120337939A_ABST
    Figure CN120337939A_ABST
Patent Text Reader

Abstract

The invention provides a question and answer method, a table question and answer model training method and device, equipment and a medium, and relates to the technical field of artificial intelligence. The method comprises the following steps: selecting a target table of which table information is matched with a natural language question from a plurality of tables to be selected, wherein the tables to be selected are in a coding format; generating a corresponding table question based on the natural language question and table information of the target table; and inputting the generated table question into a pre-trained table question and answer model to obtain a question and answer result corresponding to the natural language question. According to the technical scheme, the utilization rate of the table information in the table question and answer model can be increased, and the complex table and multi-table question and answer ability is improved.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Cited By

  • Document question and answer method and device and storage medium

    CN120910203A

  • Question and answer method and device based on Excel multilevel header and medium

    CN122065989A