Content recommendation method and device

Abstract

The invention provides a content recommendation method and device. The method comprises the steps that content classification recommendation information generated according to a user search behavior is received; the content classification recommendation information comprises at least one content classification; and the content classifications are subscribed for a current user. According to the method, by subscribing the content classifications recommended according to the user search behavior, a product can recommend information content, optimally recommended according to the user search behavior, to the user in a subscription page actively and continuously, recommendation frequency is guaranteed while recommendation accuracy is improved, and therefore the information distribution amount of the product is substantially increased.

Description

technical field [0001] The present application relates to the technical field of search, and in particular to a content recommendation method and device. Background technique [0002] Currently, the search business and information business in search products are usually completely separated. The search business is an independent user active behavior model, and the information service is an independent product active behavior model, and the two are independent of each other. [0003] The recommendation of information content only depends on the interest recommendation algorithm in the background. It takes a long time to establish user preferences, the recommended content is not accurate, and there is no user-related information obtained by using search services. Moreover, because the existing information classification method is based on the types of information content, the same classification method is applied to all users, and the classification granularity is very coarse,...

AI Technical Summary

Problems solved by technology

[0003] The recommendation of information content only depends on the interest recommendation algorithm in the background. It takes a long time to establish user preferences, the recommended content is not accurate, and there is no user-related information obtained by using search services.

Moreover, because the existing information classification method is based on the content of the information, the same classification method is applied to all users, and the granularity of the classification is very coarse, which cannot accurately cover the individual needs of users, and it is difficult to improve the user's acceptance of content.

[0004] For the above problems, one solution is to display the information content recommended according to this search while displaying the search results, but the defect of this type of solution is that it can only recommend information to the user every time the user conducts a search. content, and on the search result display page, users pay more attention to the search results than to the recommended content. Therefore, this type of solution has very limited improvement in the distribution of product information and cannot make full use of the benefits obtained from the search business. User related information

Images