Data recommendation method and device

Abstract

The embodiment of the invention discloses a data recommendation method and device. The method comprises the steps that all user clusters and all item clusters are acquired; in the user cluster which atarget us belongs to, a neighbor user corresponding to the target user is acquired; an item evaluated by the target user and an item evaluated by the neighbor user are both determined as target items, and in the item cluster which each target item belongs to, a neighbor item corresponding to the target item is acquired; in the target item and the neighbor item corresponding to the target item, the items evaluated by the target user are removed to obtain a to-be-recommended item set, forecast scores between the target user and each item in the to-be-recommended items are respectively calculated, and according to the forecast scores, a recommendation item corresponding to the target user is selected in the to-be-recommended item set. By adopting the data recommendation method and device, the system power consumption can be reduced, and the accuracy of the recommendation is improved.

Description

technical field [0001] The present invention relates to the technical field of the Internet, in particular to a data recommendation method and device. Background technique [0002] With the rapid development of Internet technology, the amount of data on network platforms has increased exponentially, and information technology has profoundly affected and changed human life. Since there is a huge and complicated information ocean in the Internet, how to recommend personalized information services for massive users in the information ocean is particularly important. [0003] In the current information recommendation method, it is usually based on the ratings of users on items (such as movies, songs, articles, etc. Internet information), find similar users among a large number of users for the users to be recommended, and then compare the ratings of these similar users. Select the recommended item for the user to be recommended from the past items. Since it is necessary to fin...

AI Technical Summary

Problems solved by technology

Since it is necessary to find users similar to the user to be recommended among a large number of users, it will bring a huge amount of calculation due to the large search range, which will greatly increase the power consumption of the system; and the items evaluated by these similar users will be used as The recommendation candidate set will cause the number of items in the recommendation candidate set to be too small due to the small number of items evaluated by the user, which will affect the accuracy of the recommendation

Images