Information pushing method and device

Abstract

The invention discloses an information pushing method. The information pushing method comprises the following steps: obtaining the behavior information of the behavior objects according to the behavior of each user; confirming the weight value of each behavior of the behavior objects according to the behavior information; sorting the weight value of each behavior and generating a first sorting result; confirming the recommendation priority value and the collaborative filtering recommendation priority value of the behavior objects according to the behavior information of each behavior of the behavior objects and the weight value of each behavior while the first sorting result is the same as the preset sort result; sorting all of the behavior objects according to sum of the recommendation priority value and the the collaborative filtering recommendation priority value and generating a second sorting result; obtaining the behavior object matched with the behavior of the user in the second sorting result according to the behavior information of the user behavior and outputing the matched behavior object to the user as the information to be recommended. The invention also discloses an information pushing device.

Description

technical field [0001] The invention relates to information push technology, in particular to an information push method and device. Background technique [0002] Since the currently widely used information push system usually adopts content-based recommendation algorithm and collaborative filtering recommendation algorithm; among them, the content-based recommendation algorithm refers to recommending products purchased by users through the characteristics of products purchased by users. For products with similar products, the content-based recommendation algorithm can deal with the cold start problem and the newly added product problem, because it does not depend on the user's rating of the product, so it will not suffer from the problem of scoring sparsity. The collaborative filtering algorithm uses the matrix of user ratings on products to calculate the similarity between users or products, uses neighboring users with high similarity to predict the ratings of other produc...

AI Technical Summary

Problems solved by technology

[0003] However, the defect of content-based recommendation algorithm is that it cannot deal with commodities such as graphics, video and music, which are difficult to analyze and extract content features.

The disadvantage of using the collaborative filtering algorithm is that since the number of products purchased by each user is usually less than 1% of the total number of products, it is easy to cause the user's rating matrix for the product to be very sparse, resulting in poor recommendation results.

Images