Abstract:
Group recommendations often include two processes, dividing users into groups and aggregating group members’ preferences for recommendation. Because of the increasing num...Show MoreMetadata
Abstract:
Group recommendations often include two processes, dividing users into groups and aggregating group members’ preferences for recommendation. Because of the increasing number of users and the fact that it is more cost-effective to recommend to homogeneous groups than to heterogeneous groups, the group recommendation prefers to use the automatic identification group method to divide users into groups. However, with the continuous increase of the number of items and users, the time cost required for the process of dividing users into groups also increases sharply. Therefore, in order to effectively deal with massive high-dimensional data, this paper proposes an LSH-based automatic identification group approach called GRLSH. Extensive experiments on the movielens 100k dataset prove that GRLSH can greatly reduce the time cost of the process while ensuring the accuracy.
Date of Conference: 12-15 September 2022
Date Added to IEEE Xplore: 13 December 2022
ISBN Information: