Coprocessing method and system capable of according with interest change of user

Abstract

The invention relates to a coprocessing method and system capable of according with the interest change of a user. The method comprises the following steps that: collecting user data; constructing a project explicit theme vector to obtain the adjacent list of a target project; for the constructed interest theme vector of the user, according to the behavior time information of a user for the project, scoring the project to obtain the neighbouring list of a target user; from the neighbouring list of the target user and the neighbouring list of the target project, obtaining the project list of the target user and the neighbouring list of the target user for the target project; for the target user and the neighbouring user of the target user, carrying out average value calculation on the project list of the target project to obtain the prediction score of the target user for the target project; and taking the projection score as a recommendation result. By use of the method, the influence of the time sequence of the user behavior on the interest of the user is considered, a concept drifting problem in the prior art is solved, the project is subjected to scoring prediction on the basis of the updated user interest vector and a user portrait, and better anti-sparse performance is carried out on data.

Description

technical field [0001] The present invention relates to the field of information processing, in particular to a collaborative processing method and system that meet changes in user interests. Background technique [0002] With the rapid development of computer networks, we are now in a world where information data is exploding, and huge amounts of data are exchanged and shared on the Internet every day. Emerging network information services such as online content services, instant messaging, social networking and cloud computing play an extremely important role in today's daily life. Faced with the massive amount of information generated by a wide range of applications, users need to invest a lot of energy and time in filtering and screening the information data provided by these network applications according to their own needs. [0003] With the passage of time, the recommendation system has gradually become a popular research direction in the fields of machine learning, ...

AI Technical Summary

Problems solved by technology

This brings many problems that reduce the accuracy of the recommendation system, such as: the data set represented by the matrix has a data sparsity problem caused by the sparsity, and the recommendation cannot reflect the user interest drift caused by the dynamic change of the user interest. Wait

Images