Point-of-interest (POI) recommendation has attracted great attention in the field of recommender systems over the past decade. Various techniques, such as those based on matrix factorization and deep neural networks, have demonstrated outstanding performance. However, these methods are susceptible to the impact of data sparsity. Data sparsity is a significant characteristic of POI recommendation, where some POIs have limited interaction records and, in extreme cases, become cold-start POIs with no interaction history. To alleviate the influence of data sparsity on model performance, this paper introduced FAGRec, a POI-recommendation model based on the feature-aware graph. The key idea was to construct an interaction graph between POIs and their initial features. This allows the transformation of POI features into a weighted aggregation of initial features. Different POIs can share the learned representations of initial features, thereby mitigating the issue of data sparsity. Furthermore, we proposed attention-based graph neural networks and a user preference estimation method based on delayed time factors for learning representations of POIs and users, contributing to the generation of recommendations. Experimental results on two real-world datasets demonstrate the effectiveness of FAGRec in the task of POI recommendation.
Citation: Xia Liu, Liwan Wu. FAGRec: Alleviating data sparsity in POI recommendations via the feature-aware graph learning[J]. Electronic Research Archive, 2024, 32(4): 2728-2744. doi: 10.3934/era.2024123
Point-of-interest (POI) recommendation has attracted great attention in the field of recommender systems over the past decade. Various techniques, such as those based on matrix factorization and deep neural networks, have demonstrated outstanding performance. However, these methods are susceptible to the impact of data sparsity. Data sparsity is a significant characteristic of POI recommendation, where some POIs have limited interaction records and, in extreme cases, become cold-start POIs with no interaction history. To alleviate the influence of data sparsity on model performance, this paper introduced FAGRec, a POI-recommendation model based on the feature-aware graph. The key idea was to construct an interaction graph between POIs and their initial features. This allows the transformation of POI features into a weighted aggregation of initial features. Different POIs can share the learned representations of initial features, thereby mitigating the issue of data sparsity. Furthermore, we proposed attention-based graph neural networks and a user preference estimation method based on delayed time factors for learning representations of POIs and users, contributing to the generation of recommendations. Experimental results on two real-world datasets demonstrate the effectiveness of FAGRec in the task of POI recommendation.
| [1] |
Y. Hwangbo, K. J. Lee, B. Jeong, K. Y. Park, Recommendation system with minimized transaction data, Data Sci. Manage., 4 (2021), 40−45. https://doi.org/10.1016/j.dsm.2022.01.001 doi: 10.1016/j.dsm.2022.01.001
|
| [2] |
L. Shi, G. Song, G. Cheng, X. Liu, A user-based aggregation topic model for understanding user's preference and intention in social network, Neurocomputing, 413 (2020), 1−13. https://doi.org/10.1016/j.neucom.2020.06.099 doi: 10.1016/j.neucom.2020.06.099
|
| [3] |
W. Ji, X. Meng, Y. Zhang, STARec: Adaptive learning with spatiotemporal and activity influence for POI recommendation, ACM Trans. Inf. Syst., 40 (2021), 1−40. https://doi.org/10.1145/3485631 doi: 10.1145/3485631
|
| [4] |
L. Shi, J. Luo, C. Zhu, F. Kou, G. Cheng, X. Liu, A survey on cross-media search based on user intention understanding in social networks, Inf. Fusion, 91 (2023), 566−581. https://doi.org/10.1016/j.inffus.2022.11.017 doi: 10.1016/j.inffus.2022.11.017
|
| [5] |
W. Ji, X. Meng, Y. Zhang, SPATM: A social period-aware topic model for personalized venue recommendation, IEEE Trans. Knowl. Data Eng., 34 (2020), 3997−4010. https://doi.org/10.1109/TKDE.2020.3029070 doi: 10.1109/TKDE.2020.3029070
|
| [6] |
L. Shi, J. P. Du, G. Cheng, X. Liu, Z. G. Xiong, J. Luo, Cross‐media search method based on complementary attention and generative adversarial network for social networks, Int. J. Intell. Syst., 37 (2021). https://doi.org/10.1002/int.22723 doi: 10.1002/int.22723
|
| [7] |
C. C. Chen, P. L. Lai, C. Y. Chen, ColdGAN: An effective cold-start recommendation system for new users based on generative adversarial networks, Appl. Intell., 53 (2023), 8302−8317. https://doi.org/10.1007/s10489-022-04005-1 doi: 10.1007/s10489-022-04005-1
|
| [8] |
Z. Zhang, M. Dong, K. Ota, Y. Zhang, Y. Kudo, Context-enhanced probabilistic diffusion for urban point-of-interest recommendation, IEEE Trans. Serv. Comput., 15 (2021), 3156−3169. https://doi.org/10.1109/TSC.2021.3085675 doi: 10.1109/TSC.2021.3085675
|
| [9] |
H. A. Rahmani, M. Aliannejadi, M. Baratchi, F. Crestani, A systematic analysis on the impact of contextual information on point-of-interest recommendation, ACM Trans. Inf. Syst., 40 (2022), 1−35. https://doi.org/10.48550/arXiv.2201.08150 doi: 10.48550/arXiv.2201.08150
|
| [10] |
W. Ji, X. Meng, Y. Zhang, STARec: Adaptive learning with spatiotemporal and activity influence for POI recommendation, ACM Trans. Inf. Syst., 40 (2021), 1−40. https://doi.org/10.1145/3485631 doi: 10.1145/3485631
|
| [11] |
Z. Sun, C. Li, Y. Lei, L. Zhang, J. Zhang, S. Liang, Point-of-interest recommendation for users-businesses with uncertain check-ins, IEEE Trans. Knowl. Data Eng., 34 (2021), 5925−5938. https://doi.org/10.1109/TKDE.2021.3060818 doi: 10.1109/TKDE.2021.3060818
|
| [12] |
Y. C. Chen, T. Thaipisutikul, T. K. Shih, A learning-based POI recommendation with spatiotemporal context awareness, IEEE Trans. Cybern., 52 (2020), 2453−2466. https://doi.org/10.1109/TCYB.2020.3000733 doi: 10.1109/TCYB.2020.3000733
|
| [13] |
K. Seyedhoseinzadeh, H. A. Rahmani, M. Afsharchi, M. Aliannejadi, Leveraging social influence based on users activity centers for point-of-interest recommendation, Inf. Process. Manage., 59 (2022), 102858. https://doi.org/10.48550/arXiv.2201.03450 doi: 10.48550/arXiv.2201.03450
|
| [14] |
Y. Liu, Z. Yang, T. Li, D. Wu, A novel POI recommendation model based on joint spatiotemporal effects and four-way interaction, Appl. Intell., 52 (2022), 5310−5324. https://doi.org/10.1007/s10489-021-02677-9 doi: 10.1007/s10489-021-02677-9
|
| [15] |
S. Liu, L. Yang, W. Zheng, Y. Xiao, L. Liu, An ensemble learning model for preference-geographical aware point-of interest recommendation, Appl. Intell., 52 (2022), 13763-13780. https://doi.org/10.1007/s10489-022-04035-9 doi: 10.1007/s10489-022-04035-9
|
| [16] |
C. Wang, M. Yuan, R. Zhang, K Peng, L Liu, Efficient point-of-interest recommendation services with heterogenous hypergraph embedding, IEEE Trans. Serv. Comput., 16 (2022), 1132−1143. https://doi.org/10.1109/TSC.2022.3187038 doi: 10.1109/TSC.2022.3187038
|
| [17] |
Y. Qin, C. Gao, Y. Wang, S. Wei, D. Jin, J. Yuan, et al., Disentangling geographical effect for point-of-interest recommendation, IEEE Trans. Knowl. Data Eng., (2022). https://doi.org/10.1109/TKDE.2022.3221873 doi: 10.1109/TKDE.2022.3221873
|
| [18] |
S. A. Puthiya Parambath, S. Chawla, Simple and effective neural-free soft-cluster embeddings for item cold-start recommendations, Data Min. Knowl. Discovery, 34 (2020), 1560−1588. https://doi.org/10.1007/s10618-020-00708-6 doi: 10.1007/s10618-020-00708-6
|
| [19] |
S. Li, W. Lei, Q. Wu, X. He, P. Jiang, T. S. Chua, Seamlessly unifying attributes and items: Conversational recommendation for cold-start users, ACM Trans. Inf. Syst., 39 (2021), 1−29. https://doi.org/10.1145/3446427 doi: 10.1145/3446427
|
| [20] |
T. Qian, Y. Liang, Q. Li, S. Wei, D. Jin, J. Yuan, et al., Attribute graph neural networks for strict cold start recommendation, IEEE Trans. Knowl. Data Eng., 34 (2020), 3597−3610. https://doi.org/10.1109/TKDE.2020.3038234 doi: 10.1109/TKDE.2020.3038234
|
| [21] |
J. Zhang, C. Ma, C. Zhong, P. Zhao, X. Mu, Combining feature importance and neighbor node interactions for cold start recommendation, Eng. Appl. Artif. Intell., 112 (2022), 104864. https://doi.org/10.1016/j.engappai.2022.104864 doi: 10.1016/j.engappai.2022.104864
|
| [22] |
H. Wu, C. W. Wong, J. Zhang, Y. Yan, D. Yu, J. Long, et al., Cold-start next-item recommendation by user-item matching and auto-encoders, IEEE Trans. Serv. Comput., (2023). https://doi.org/10.1109/TSC.2023.3237638 doi: 10.1109/TSC.2023.3237638
|
| [23] |
D. Cai, S. Qian, Q. Fang, J. Hu, C. Xu, User cold-start recommendation via inductive heterogeneous graph neural network, ACM Trans. Inf. Syst., 41 (2023), 1−27. https://doi.org/10.1145/3560487 doi: 10.1145/3560487
|
| [24] |
I. Rehman, W. Ali, Z. Jan, Z. Ali, H. Xu, J. Shao, CAML: Contextual augmented meta-learning for cold-start recommendation, Neurocomputing, 533 (2023), 178−190. https://doi.org/10.1016/j.neucom.2023.02.051 doi: 10.1016/j.neucom.2023.02.051
|
| [25] | S. Feng, G. Cong, B. An, Y. M. Chee, Poi2vec: Geographical latent representation for predicting future visitors, in 2017 Proceedings of the AAAI Conference on Artificial Intelligence (AAAI), (2017), 102−108. https://doi.org/10.1609/aaai.v31i1.10500 |
| [26] | R. Li, Y. Shen, Y. Zhu, Next point-of-interest recommendation with temporal and multi-level context attention, in Proceedings of the 2018 IEEE International Conference on Data Mining (ICDM), (2018), 1110−1115. https://doi.org/10.1109/ICDM.2018.00144 |
| [27] | Y. Luo, Q. Liu, Z. Liu, Stan: Spatio-temporal attention network for next location recommendation, in 2021 Proceedings of the web conference (WWW), (2021), 2177−2185. https://doi.org/10.48550/arXiv.2102.04095 |
| [28] | C. Ma, Y. Zhang, Q. Wang, X. Liu, Point-of-interest recommendation: Exploiting self-attentive autoencoders with neighbor-aware influence, in Proceedings of the 27th ACM International Conference on Information and Knowledge Management (CIKM), (2018), 697−706. https://doi.org/10.48550/arXiv.1809.10770 |
| [29] | M. Pontiki, D. Galanis, H. Papageorgiou, I. Androutsopoulos, S. Manandhar, M. AL-Smadi, et al., Semeval-2016 task 5: Aspect based sentiment analysis, in Proceedings of the 10th International Workshop on Semantic Evaluation (SemEval), (2016), 19−30. https://doi.org/10.18653/v1/S16-1002 |
| [30] | H. A. Rahmani, M. Aliannejadi, S. Ahmadian, M. Baratchi, M. Afsharchi, F. Crestani, LGLMF: local geographical based logistic matrix factorization model for POI recommendation, in Proceedings of the Information Retrieval Technology: 15th Asia Information Retrieval Societies Conference, (2020), 66−78. https://doi.org/10.48550/arXiv.1909.06667 |
| [31] |
P. Han, Z. Li, Y. Liu, P. Zhao, J. Li, H. Wang, et al., Contextualized point-of-interest recommendation, in Proceedings of the International Joint Conferences on Artificial Intelligence (IJCAI), 344 (2021), 2484−2490. https://doi.org/10.24963/ijcai.2020/344 doi: 10.24963/ijcai.2020/344
|
| [32] |
J. Fu, R. Gao, Y. Yu, J. Wu, J. Li, D. Liu, et al., Contrastive graph learning long and short-term interests for POI recommendation, Expert Syst. Appl., 238 (2024), 121931. https://doi.org/10.1016/j.eswa.2023.121931 doi: 10.1016/j.eswa.2023.121931
|
| [33] |
F. Mo, H. Yamana, EPT-GCN: Edge propagation-based time-aware graph convolution network for POI recommendation, Neurocomputing, 543 (2023), 126272. https://doi.org/10.1016/j.neucom.2023.126272 doi: 10.1016/j.neucom.2023.126272
|
Figures(6) / Tables(6)
Xia Liu, Liwan Wu. FAGRec: Alleviating data sparsity in POI recommendations via the feature-aware graph learning[J]. Electronic Research Archive, 2024, 32(4): 2728-2744. doi: 10.3934/era.2024123
DownLoad: