Commodity recommendation method and system based on commodity popularity and user dynamic interest

Abstract

The invention discloses a commodity recommendation method and system based on commodity popularity and user dynamic interests. The system comprises a popularity calculation module, a popularity punishment weight calculation module, a commodity similarity calculation module, a user interest attenuation value calculation module, a commodity score prediction value calculation module and a commodity recommendation module. According to the method and the system, the problem of high similarity of hot commodities is solved, and meanwhile, the time function is added into the prediction formula according to the behavior characteristics of each user, so that the contribution of historical data to prediction preferences is weakened. The recommendation accuracy is improved, unpopular commodities in a data set are effectively mined and recommended, the coverage rate of recommendation is improved, the long tail effect problem of a recommendation system is relieved, and the commodity recommendation quality is improved.

Description

technical field [0001] The invention belongs to the technical field of commodity recommendation systems in electronic commerce, and in particular relates to a recommendation method and system based on commodity popularity and user dynamic interests. Background technique [0002] Collaborative filtering is considered to be one of the most promising recommendation algorithms. It is used to help users find products they may like. Due to its simplicity and ease of implementation, it is widely used in e-commerce websites and various music, short video and other apps. Item-based collaborative filtering identifies the similarity between items by comparing user ratings on items, and then predicts users' ratings on unrated items based on the similarity between items. Therefore, whether the two key steps of calculating the similarity of products and predicting the ratings of products not rated by users are reasonable and effective is directly related to the quality of product recommen...

AI Technical Summary

Problems solved by technology

The "long tail effect" in the recommendation system means that due to the large number of ratings of popular products, they account for a large proportion of the recommended list, and the unpopular products that can better reflect the user's personalization only account for a small part, resulting in single recommendation results. lack of novelty

Images