Calculation method for user availability in collaborative filtering recommendation system

Abstract

The invention relates to a calculation method for user availability in a collaborative filtering recommendation system. The calculation method includes the following steps that a user-item rating matrix is established; the similarity among users is calculated through a Pearson similarity calculation method; availability coefficients of all the users for a target user are calculated, wherein the calculation method includes the following steps that when the similarity value between one certain user and the target user is smaller than 0 or items evaluated by the target user contain all the items evaluated by the user, the availability coefficient of the user for the target user is defined as 0, and when the similarity value between one certain user and the target user is not smaller than 0 or the items evaluated by the user contain items not evaluated by the target user, the availability coefficient of the user for the target user is defined as the ratio of the number of items, with grades larger than values in the rating interval, in common rating items of the two users to the number of the items rated by the target user; the availability of all the users for the target user is calculated; recommendation is carried out. The accuracy of recommendation can be improved.

Description

technical field [0001] The invention relates to a calculation method of user availability in a user-based collaborative filtering recommendation system. Background technique [0002] With the rapid development of the Internet, information and data resources are increasing day by day like a vast ocean, which has caused the Internet to fall into the problem of "information overload" when supporting information sharing, especially in the field of e-commerce. Faced with massive product information when purchasing online, it is often difficult for consumers to find their favorite or most suitable product in time. In order to obtain higher profits, attract more users and increase user stickiness, e-commerce websites must consider how to effectively present products or website content to users, improve service quality, and save users' time and energy. In this context, the personalized recommendation system came into being and has been widely used. The personalized recommendation ...

AI Technical Summary

Problems solved by technology

However, as the number of users and items on e-commerce platforms continues to increase, user ratings are sparse, resulting in large errors in the similarity between users, which in turn affects the selection of nearest neighbors.

Images