Abstract
Relationships among items, especially complementarity, have shown great potential to empower the performance and explainability of recommender systems. However, there are two key limitations: 1) Most previous methods use co-occurrence to quantify item complementary relationship, which lacks theoretical support and overlooks the fact that co-occurrence is only a necessary but not sufficient condition to identify item complementarity. 2) Most studies do not consider the time-sensitive nature of item complementarity, which does exist in real scenarios.
In this study, we propose a Graph Neural Network (DCGNN) to model the dynamic item complementarity for the recommendation. First, to improve the reliability of item relationships, complementary item pairs are mined according to the ‘cross elasticity of demand’ concept in economic theory and the mined relationships are applied to enrich the user-item graph. Second, considering the time-sensitive nature of item complementarity, we design a time-transfer mechanism to distillate historical knowledge of item complementarity by using graph neural networks. Finally, extensive experiments and analysis were conducted on two real-world data sets, which demonstrate the effectiveness of DCGNN in capturing dynamic item complementarity and recommendation.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Covington, P., Adams, J., Sargin, E.: Deep neural networks for YouTube recommendations. In: Recsys 2016, pp. 191–198 (2016)
Dong, Y., Chawla, N.V., Swami, A.: metapath2vec: scalable representation learning for heterogeneous networks. In: SIGKDD 2017, pp. 135–144 (2017)
Fan, W., et al.: Graph neural networks for social recommendation. In: The Web Conference, pp. 417–426 (2019)
He, X., Deng, K., Wang, X., Li, Y., Zhang, Y., Wang, M.: LightGCN: simplifying and powering graph convolution network for recommendation. arXiv preprint arXiv:2002.02126 (2020)
He, X., Liao, L., Zhang, H., Nie, L., Hu, X., Chua, T.S.: Neural collaborative filtering. In: The Web Conference, pp. 173–182 (2017)
Hidasi, B., Karatzoglou, A., Baltrunas, L., Tikk, D.: Session-based recommendations with recurrent neural networks. arXiv preprint arXiv:1511.06939 (2015)
Kang, W.C., Wan, M., McAuley, J.: Recommendation through mixtures of heterogeneous item relationships. In: CIKM, pp. 1143–1152 (2018)
Ma, W., et al.: Jointly learning explainable rules for recommendation with knowledge graph. In: The World Wide Web Conference, pp. 1210–1221 (2019)
Marshall, A.: Elements of Economics of Industry: Being the First Volume of Elements of Economics, vol. 1. Macmillan (1898)
McAuley, J., Pandey, R., Leskovec, J.: Inferring networks of substitutable and complementary products. In: SIGKDD, pp. 785–794 (2015)
Rendle, S., Freudenthaler, C., Gantner, Z., Schmidt-Thieme, L.: BPR: Bayesian personalized ranking from implicit feedback. arXiv preprint arXiv:1205.2618 (2012)
Shao, Y., Zhang, M., Ma, W., Wang, C., Liu, Y., Ma, S.: Integrating latent item-item complementarity with personalized recommendation systems. J. Softw. 31(4), 1090–1100 (2020)
Shocker, A.D., Bayus, B.L., Kim, N.: Product complements and substitutes in the real world: the relevance of “other products’’. J. Mark. 68(1), 28–40 (2004)
Sun, P., Wu, L., Wang, M.: Attentive recurrent social recommendation. In: SIGIR, pp. 185–194 (2018)
Vasile, F., Smirnova, E., Conneau, A.: Meta-prod2vec: product embeddings using side-information for recommendation. In: RecSys, pp. 225–232 (2016)
Wan, M., Wang, D., Liu, J., Bennett, P., McAuley, J.: Representing and recommending shopping baskets with complementarity, compatibility and loyalty. In: CIKM, pp. 1133–1142 (2018)
Wang, C., Zhang, M., Ma, W., Liu, Y., Ma, S.: Modeling item-specific temporal dynamics of repeat consumption for recommender systems. In: The World Wide Web Conference, pp. 1977–1987 (2019)
Wang, J., Ding, K., Zhu, Z., Zhang, Y., Caverlee, J.: Key opinion leaders in recommendation systems: opinion elicitation and diffusion. In: WSDM, pp. 636–644 (2020)
Wang, Z., Jiang, Z., Ren, Z., Tang, J., Yin, D.: A path-constrained framework for discriminating substitutable and complementary products in e-commerce. In: WSDM, pp. 619–627 (2018)
Wu, L., Sun, P., Fu, Y., Hong, R., Wang, X., Wang, M.: A neural influence diffusion model for social recommendation. In: SIGIR, pp. 235–244 (2019)
Xiang, L., et al.: Temporal recommendation on graphs via long-and short-term preference fusion. In: SIGKDD, pp. 723–732 (2010)
Ying, R., He, R., Chen, K., Eksombatchai, P., Hamilton, W.L., Leskovec, J.: Graph convolutional neural networks for web-scale recommender systems. In: SIGKDD, pp. 974–983 (2018)
Acknowledgements
This work is supported by the National Key Research and Development Program of China (2018YFC0831900), Natural Science Foundation of China (Grant No. 62002191, 61672311, 61532011) and Tsinghua University Guoqiang Research Institute.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 Springer Nature Switzerland AG
About this paper
Cite this paper
Shiu, Y., Ma, W., Zhang, M., Liu, Y., Ma, S. (2021). Modelling Dynamic Item Complementarity with Graph Neural Network for Recommendation. In: Lin, H., Zhang, M., Pang, L. (eds) Information Retrieval. CCIR 2021. Lecture Notes in Computer Science(), vol 13026. Springer, Cham. https://doi.org/10.1007/978-3-030-88189-4_4
Download citation
DOI: https://doi.org/10.1007/978-3-030-88189-4_4
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-88188-7
Online ISBN: 978-3-030-88189-4
eBook Packages: Computer ScienceComputer Science (R0)