A Recommendation Method for Items with Classified Information

Abstract

In many network applications, it is often necessary to recommend to users, and contextual information needs to be used to improve recommendation accuracy and enhance user experience. However, existing context-aware recommendation methods still face the challenge of data sparsity. In order to further alleviate the problem of data sparsity, this patent proposes a new recommendation method, which combines user rating data and user category preferences to recommend items, so as to solve the problem of low rating prediction accuracy when user rating data is sparse. This method is suitable for large-scale data. Experimental results show that this method has a better recommendation effect than the current mainstream methods.

Description

technical field [0001] The invention belongs to the field of recommendation systems and relates to a recommendation method for an application with an item classification function. Background technique [0002] The existing context-based recommendation systems are all recommendation methods that directly use the user's historical data. Although it is convenient, easy to use widely, and easy to get a wide range of evaluation benefits, but because the user's historical behavior data is usually very sparse , so these methods all face serious data sparsity problems. It is difficult to model user preferences based on sparse historical user behavior data, resulting in low accuracy of the recommendation system and affecting user experience. [0003] We want to recommend items for an application system. Generally, we need to analyze the composition of the recommendation system. Here we will discuss some of the main bodies that make up the recommendation system. The following is a b...

AI Technical Summary

Problems solved by technology

Most of the recommendation methods are aimed at specific application systems, which loses the universality of the recommendation methods

Images