Model training method and device based on advertisement recall, equipment and storage medium

By obtaining multiple sample data and using knowledge system databases to enhance knowledge and training keyword matching models, the problem of not being able to dig deep into user needs and intentions in the existing technology is solved, and more accurate advertising recall and delivery is achieved.

CN119646189AActive Publication Date: 2025-03-18BEIJING BAIDU NETCOM SCI & TECH CO LTD
6 Cites 0 Cited by

Patent Information

Application Number
CN202411768071.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-12-03
Publication Date
2025-03-18
Estimated Expiration
2044-12-03

AI Technical Summary

Technical Problem

The existing technology cannot deeply explore the user needs and intentions behind the query words in the field of commercial search, resulting in limited recall capabilities of advertising systems and inability to achieve accurate targeted advertising delivery.

Method used

By obtaining the data of conventional click behavior and data other than click behavior, using the knowledge system database for knowledge enhancement and mixed samples for training, a keyword matching model is obtained, which is used to determine multiple keywords that match it based on query words and their knowledge enhancement information.

Benefits of technology

It improves the model's understanding of query terms, enhances the recall ability and accuracy of the advertising system, and improves the conversion rate of advertising delivery.

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

Abstract

The invention provides a model training method and device based on advertisement recall, equipment and a storage medium. The invention relates to the technical field of artificial intelligence, in particular to the technical fields of deep learning, large models, natural language processing, knowledge maps, data analysis and the like, and can be used for application scenes such as keyword recall, advertisement retrieval, advertisement optimization and the like. The specific scheme is as follows: acquiring a first type of samples; obtaining a second type of samples; obtaining corresponding multi-dimensional knowledge information through a knowledge system database interface, and performing knowledge enhancement on the query words in the first type of samples and the second type of samples; mixing the first class of samples and the second class of samples after knowledge enhancement to form a training data set; and training the pre-training model based on the training data set to obtain a keyword matching model. According to the scheme, the accuracy of matching the keyword for the query word by the model can be improved.
Need to check novelty before this filing date? Find Prior Art