Request rewriting method and device, equipment and computer storage medium

By identifying rewriting categories and training models using positive and negative samples, the problem of inaccurate search requests in e-commerce platforms was solved, improving the accuracy of information retrieval and product search, and increasing conversion rates.

CN122086964APending Publication Date: 2026-05-26ZHEJIANG TMALL TECH CO LTD
0 Cites 0 Cited by

Patent Information

Application Number
CN202610098316.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-01-23
Publication Date
2026-05-26

AI Technical Summary

Technical Problem

In e-commerce platforms, user search requests often contain spelling errors, colloquial expressions, and significant semantic gaps and ambiguities between standard product descriptions, leading to inaccurate search results and impacting product conversion rates.

Method used

By acquiring information query requests, determining their rewriting categories, and filtering positive and negative rewriting samples based on the rewriting categories for model training, a request rewriting model is obtained. This model is then used to rewrite the original search request, generating a rewritten request for information querying.

Benefits of technology

This improved the accuracy and reliability of information retrieval operations and the accuracy of product search results, thereby increasing product conversion rates.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The embodiment of the invention provides a request rewriting method and device, equipment and a computer storage medium. The method comprises the following steps: acquiring an information query request for realizing a model training operation; determining a rewriting category to which the information query request belongs; based on the rewriting category, determining a positive rewriting sample and a negative rewriting sample corresponding to the information query request; performing model training based on the information query request, the positive rewriting sample and the negative rewriting sample to obtain a request rewriting model; and carrying out request rewriting operation by utilizing the request rewriting model. In the embodiment, the sample screening operation is performed by adopting different sample screening strategies based on different rewriting categories to which the information query request belongs, and then the model training is performed based on the positive rewriting sample and the negative rewriting sample of the screening operation to obtain the request rewriting model, so that the quality and the effect of the model training operation are ensured; and then request rewriting operation is carried out based on the request rewriting model, so that the quality and efficiency of the request rewriting operation are effectively ensured.
Need to check novelty before this filing date? Find Prior Art