Subject generation method, device, equipment and storage medium

By processing document data through a topic generation model and generating document topics using a pre-trained model and encoder, the problem of low generation efficiency and susceptibility to human intervention in existing technologies is solved, achieving efficient and reliable document topic generation.

CN119066191BActive Publication Date: 2026-07-21PING AN TECH (SHENZHEN) CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202411095139.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-09
Publication Date
2026-07-21
Estimated Expiration
2044-08-09

AI Technical Summary

Technical Problem

In existing technologies, the process of generating document topics for documents to be processed is cumbersome, resulting in low generation efficiency and susceptibility to human intervention.

Method used

A topic generation model is adopted. By acquiring preset document data and preset topics from the sample set, the model is used to generate predicted topics, calculate the loss value, and adjust the model in multiple rounds of training to generate document topics. The encoder and average pooling layer are combined to process semantic units and generate document topics.

Benefits of technology

It improves the efficiency of document topic generation, reduces generation time, enhances the reliability of generated results, and avoids the impact of human intervention.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119066191B_ABST
    Figure CN119066191B_ABST
Patent Text Reader

Abstract

The application relates to the fields of artificial intelligence and financial technology, and discloses a topic generation method, device, equipment and storage medium. The method comprises the following steps: acquiring a loss value between a preset topic and a predicted topic; determining a trained topic generation model according to a predefined training mode and the loss value; acquiring a to-be-processed document corresponding to the topic generation model, decomposing the to-be-processed document into a plurality of semantic units; inputting a first embedding vector corresponding to the semantic unit into an encoder, and generating a second embedding vector corresponding to the semantic unit through the encoder; determining a third embedding vector according to a predefined merging mode and the second embedding vector corresponding to the semantic unit; inputting the third embedding vector into the trained topic generation model; and acquiring a document topic generated by the trained topic generation model based on the third embedding vector. The application is beneficial to improving the generation efficiency of the document topic.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Multi-document automatic abstract generation method

    CN113468854A

  • Training method and device of data enhancement model

    CN117892134A