POI Recommendation for Random Groups Based on Cooperative Graph Neural Networks
Date
Authors
Liu, Zhizhong
Meng, Lingqiang
Sheng, Quan Z
Chu, Dianhui
Yu, Jian
Song, Xiaoyu
Supervisor
Item type
Journal Article
Degree name
Journal Title
Journal ISSN
Volume Title
Publisher
Elsevier BV
Abstract
Group Point-of-Interests (POI) recommendation devotes to find the optimal POIs for groups, which has extracted extensive attention. This work first brings forward a novel POI recommendation model for random groups based on Cooperative Graph Neural Networks (named as CGNN-PRRG). We have done three innovative work. (1) We propose a new fitted presentation learning method for generating the fitted representations of random groups. (2) To conquer the cold start issues in recommending POI for a new random group, we propose to take similar users’ (which have the similar representations with that of the random group) POI interaction data as the learning data. (3) We propose an Edge-learning enhanced Bipartite Graph Neural Network (EBGNN) to learn similar users’ POI comprehensive interaction preferences. Specially, EBGNN can learn the information on the edges of the graph. Meanwhile, we propose to learn similar users’ POI transfer preferences with the Session-based Graph Neural Networks (SRGNN). We verify our proposed model on the three public benchmark datasets (Foursquare, Gowalla and Yelp), which contain 124,933 to 860,888 POI check-in records. The comparison between our proposed model and ten representative baseline models demonstrates the outstanding performance of CGNN-PRRG. In terms of Precision@K and NDCG@K, our model achieves about 24.9% and 62.5% improvement compared with the best baseline models on the three benchmark datasets averagely. Adequate ablation experiments prove the effectiveness of the fitted representation generation method, similar users’ POI comprehensive interaction preferences learning method and the method for overcoming the cold start problem. The source code of the CGNN-PRRG model is available on github1.Description
Keywords
46 Information and Computing Sciences, 4611 Machine Learning, Clinical Research, 0804 Data Format, 0806 Information Systems, 0807 Library and Information Studies, Information & Library Sciences, 4609 Information systems, 4610 Library and information studies
Source
Information Processing & Management, ISSN: 0306-4573 (Print), Elsevier BV, 61(3), 103676-103676. doi: 10.1016/j.ipm.2024.103676
Publisher's version
Rights statement
This is the Preprint version of an article published in Information Processing & Management © 2024 Elsevier Ltd. All rights reserved. The Version of Record is available at DOI: 10.1016/j.ipm.2024.103676