Information pushing method, server and system

Abstract

The disclosure relates to a method, a server and a system for information push. The method includes: obtaining the click information on an information push list to form an user click log; analyzing the distribution of user's click behavior based on said user click log, thus obtaining an analysis result; selecting the corresponding recommended items from said information push list according to said analysis result to issue. The disclosure forms a user click log through obtaining the click information on an information push list by client, and analyzes the distribution of user's click behavior based on said user click log, specifically analyze the distribution of click location, and combine the distribution in different times, so as to issue different items in different times. Thereby implements a flexible control of strategy for issuing information push results, which makes the push items both meet the demand for browsing of most users at a different time, and avoid a waste of traffic caused by redundant issue.

Description

technical field [0001] The invention relates to the technical field of the Internet, in particular to an information push method, server and system. Background technique [0002] At present, the explosive growth of information on the Internet has made the problem of information overload more and more serious. It is difficult for users to find the content they are interested in from a large number of information sources such as news websites, forums, and blogs. Thus, a personalized recommendation system is produced. Usually, a personalized recommendation system calculates the content that the user may be interested in based on the user's browsing history, and presents the content to the user preferentially. [0003] However, a very important problem facing a personalized recommendation system is how to determine the number of recommended results delivered each time a user requests. If the amount delivered is not enough, the user's browsing needs cannot be fully satisfied ea...

AI Technical Summary

Problems solved by technology

[0003] However, a very important problem facing a personalized recommendation system is how to determine the number of recommended results delivered each time a user requests

If the quantity delivered is not enough, the browsing needs of the user cannot be fully satisfied each time, and the user needs to request again after reading it; if too much is delivered, the user cannot fully consume it at one time, which will cause unnecessary traffic waste

[0004] The existing strategy for issuing recommendation results generally adopts a fixed artificial experience value combined with the business. Its defect is that it does not take into account that the scenarios of users using the product may be different at different times, so that the recommended solutions issued by them are not necessarily the same. Is the best delivery plan

Therefore, the existing strategy for delivering recommendation results cannot be flexibly controlled

Images