Article recommending method and device based on ant colony collaborative filtering algorithm

Abstract

The invention discloses an article recommending method and a device based on an ant colony collaborative filtering algorithm, which are used for recommending proper articles for a user according to the favor of the user. The method comprises the following steps: acquiring the information element of user ui and the information element of any article vj based on the rating factor of the article vj and the decay factor of the information element given by the user ui, and calculating the similarity between the user ui and the article vj according to the information element of the user ui and the information element of the article vj; sequencing the similarity obtained by calculating, and recommending articles with high similarity for the user ui. According to the method provided by the embodiment of the invention, modelling is performed according to the fact that the favor of user can be changed along with time by means of a mechanism that the information element is volatilized along with time, so that articles the user really likes can be recommended for the user, and the recommending function of a recommending system is optimized.

Description

technical field [0001] The invention relates to the field of artificial intelligence, in particular to an item recommendation method and device based on an ant colony collaborative filtering algorithm. Background technique [0002] With the continuous development of the Internet and communication technology, people are more and more inclined to shop online, e-commerce is a successful example. If the system can make a certain judgment based on the user's shopping situation, that is, to judge which items the user still likes, it can recommend the items he likes to the user, which not only saves the user's selection time, but also greatly improves the e-commerce business. profitability. Of course, this method of recommending other items that users like based on what they have purchased in the e-commerce field can also play a role in physical stores. If the "items" here are further abstracted, for example, abstracted into virtual items such as knowledge and information, the me...

AI Technical Summary

Problems solved by technology

[0004] However, existing recommender systems are less efficient when dealing with large-scale sparse data; on the other hand, existing recommender systems do not take into account the role played by the time factor, which is crucial to the effectiveness of recommendations

For example, as time goes by, the user who originally liked the item A no longer likes the item A. If the recommendation system still recommends the item A to the user, the recommendation effect will obviously be greatly reduced.

Images