Recommendation model training method and training apparatus

Abstract

The present invention provides a recommendation model training method and training apparatus, which can take the behavior of the user after the recommendation result is displayed as recommendation feedback, and can effectively alleviate the sparseness of the acceptation amount in the training data compared with the recommended amount and the imbalance of the proportion of the positive and negative samples. The recommendation model training method comprises: obtaining training data of a recommendation model, wherein the training data is the data generated in the latest time period, and the latest time period has a predefined time length; marking the training data according to the predetermined training data marking rule to obtain the latest marked data, wherein the latest marked data comprises the user feedback after the recommendation result of the latest time period is displayed, and the user feedback is determined according to the user behavior after the recommendation result is displayed; and training the latest marked data to obtain the recommendation model.

Description

technical field [0001] The invention relates to the field of computer technology, in particular to a training method and training device for a recommendation model. Background technique [0002] Recommendation is an important way for information promotion to enable people to obtain information, and it is applied in all aspects of life, such as: search engine recommendation, input word recommendation, website information recommendation, and friend circle information recommendation, commodity Recommend and more. By analyzing and processing the collected user information to generate a recommendation model, user recommendations can be made more conveniently, quickly and accurately. [0003] Take a search engine as an example. When a user enters a keyword, the search engine will use the recommendation model trained and produced based on the search records of other users recorded on the website to make recommendations to the user. For example: when the user enters the keyword "JA...

AI Technical Summary

Problems solved by technology

[0007] 1. At present, most recommendation systems affect the recommendation results by inputting user behavior data as a sample feature into the recommendation model, but there is no specific reference standard for whether the recommendation results are recognized by customers and whether the recommendation model can be optimized ;

[0008] 2. The current recommendation system has problems such as sparse data labeling and unbalanced distribution. For example, among the recommended results, only a small number of results are adopted by users. Compared with the amount of presentation, the amount of adoption is very sparse

Images