Abstract
How to capture user interest accurately to enhance the user experience is a great practical challenge in recommender systems. Through preliminary investigation, we find that each user has his personalized interest model which may contain multiple kinds of interests, and the strength of each user interest usually has a dynamic evolution process which can be divided into two stages: rising stage and declining stage. The evolution rate of the user interests also differ from each other. Based on this finding, a recommendation framework called SimIUC is proposed, which can identify multiple user interests and adapt the inverted-U-curve to model the dynamic evolution process of user interests. Specifically, SimIUC differs from the traditional user preference based methods which use monotonously decreasing function to model user interest. It can predict the evolutionary trends of interests and make recommendations by inverted-U-interest-based collaborative filtering. We studied a large subset of data from MovieLens and netflix.com respectively. The experimental results show that our method can significantly improve the accuracy in 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
Silvia, P.J.: Exploring the Psychology of Interest. Oxford University Press, New York (2006)
Stewart, B., Mark, M.: The U-curve adjustment hypothesis revisited: a review and theoretical framework. J. Int. Bus. Stud. 22, 225–247 (1991)
Zajonc, R.B.: Attitudinal effects of mere exposure. J. Pers. Soc. Psychol. 19(2), 77–78 (1968)
Stigler, G.J.: The adoption of the marginal utility theory. Hist. Polit. Econ. 4(2), 571–586 (1972)
Karypis, G.: Evaluation of item-based top-n recommendation algorithms. In: CIKM 2001, pp. 247–254 (2001)
Ding, Y., Li, X.: Time weight collaborative filtering. In: CIKM 2005, pp. 485–492 (2005)
Quan, Y., Gao, C., Aixin, S.: Graph-based point-of-interest recommendation with geographical and temporal influences. In: CIKM 2014, pp. 659–668 (2014)
Chen, W., Hsu, W., Lee, M.L.: Modeling user’s receptiveness over time for recommendation. In: SIGIR 2013, pp. 373–382 (2013)
Deshpande, M., Karypis, G.: Item-based top- N recommendation algorithms. ACM TOIS 22(1), 143–177 (2004)
Newman, M.: Power laws, pareto distributions and Zipf’s law. Contemp. Phys. 46(5), 323–351 (2005)
Sun, Y., Han, J., Yan, X., Wu, T.: Pathsim: meta path-based top-k similarity search in heterogeneous information networks. In: VLDB 2011 (2011)
Chen, J., Wang, C., Wang, J.: Modeling the interest-forgetting curve for music recommendation. In: MM 2014, ACM (2014)
Toscher, A., Jahrer, M., Bell, R.M.: The bigchaos solution to the Netflix Grand prize (2008)
Koren, Y.: Collaborative filtering with temporal dynamics. In: KDD 2009, pp. 447–456 (2009)
Koychev, I., Schwab, I.: Adaptation to drifting user’s interests. In: ECML 2000 Workshop: Machine Learning in New Information Age (2000)
Lathia, N., Hailes, S., Capra, L., Amatriain, X.: Temporal diversity in recommender systems. In: SIGIR 2010, pp. 210–217 (2010)
Wang, H., Fan, W., Yu, P.S., Han, J.: Mining concept-drifting data streams using ensemble classifiers. In: KDD 2003, pp. 226–235 (2003)
Senzhang, W., Xia, H., Philip, S.Y., Zhoujun, L.: MMRate: inferring multi-aspect diffusion networks with multi-pattern cascades. In: KDD 2014, pp. 1246–1255 (2014)
Liang, X., Quan, Y., Zhao, S., Chen, L., Zhang, X.: Temporal recommendation on graphs via long-and short-term preference fusion. In: KDD 2010, pp. 723–732 (2010)
Senzhang, W., Sihong, X., Xiaoming, Z., Zhoujun, L., Philip, S.Y., Xinyu, S.: Future influence ranking of scientific literature. In: SDM 2014, pp. 749–757 (2014)
Kruskal, J.B.: On the shortest spanning subtree of a graph and the traveling salesman problem. Am. Math. Soc. 7, 48–50 (1956)
MovieLens: http://grouplens.org/datasets/movielens
Netflix: http://www.netflixprize.com
Acknowledgements
This work is supported by NSF of China (No. 61303005), 973 Program (No. 2015CB352500), NSF of Shandong, China (No. ZR2013FQ009), the Science and Technology Development Plan of Shandong, China (No. 2014GGX101047, No. 2014GGX101019). This work is also supported by US NSF grants III-1526499, and CNS-1115234.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2016 Springer International Publishing Switzerland
About this paper
Cite this paper
Xu, Y., Hong, X., Peng, Z., Yang, G., Yu, P.S. (2016). Temporal Recommendation via Modeling Dynamic Interests with Inverted-U-Curves. In: Navathe, S., Wu, W., Shekhar, S., Du, X., Wang, X., Xiong, H. (eds) Database Systems for Advanced Applications. DASFAA 2016. Lecture Notes in Computer Science(), vol 9642. Springer, Cham. https://doi.org/10.1007/978-3-319-32025-0_20
Download citation
DOI: https://doi.org/10.1007/978-3-319-32025-0_20
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-32024-3
Online ISBN: 978-3-319-32025-0
eBook Packages: Computer ScienceComputer Science (R0)