Abstract
Hawkes processes have become very popular in modeling multiple recurrent user–item interaction events that exhibit mutual-excitation properties in various domains. Generally, modeling the interaction sequence of each user–item pair as an independent Hawkes process is ineffective since the prediction accuracy of future event occurrences for users and items with few observed interactions is low. On the other hand, multivariate Hawkes processes (MHPs) can be used to handle multi-dimensional random processes where different dimensions are correlated with each other. However, an MHP either fails to describe the correct mutual influence between dimensions or become computational inhibitive in most real-world events involving a large collection of users and items. To tackle this challenge, we propose local low-rank Hawkes processes to model large-scale user–item interactions, which efficiently captures the correlations of Hawkes processes in different dimensions. In addition, we design an efficient convex optimization algorithm to estimate model parameters and present a parallel algorithm to further increase the computation efficiency. Extensive experiments on real-world datasets demonstrate the performance improvements of our model in comparison with the state of the art.
Similar content being viewed by others
Notes
References
Aalen O, Borgan O, Gjessing H (2008) Survival and event history analysis: a process point of view. Springer, New York
Bell R, Koren Y, Volinsky C (2007) Modeling relationships at multiple scales to improve accuracy of large recommender systems. In: Proceedings of the ACM SIGKDD international conference on knowledge discovery and data mining (KDD), pp 95–104
Chi EC, Kolda TG (2012) On tensors, sparsity, and nonnegative factorizations. SIAM J Matrix Anal Appl 33(4):1272–1299
Cox DR, Lewis PAW (1972) Multivariate point processes. In: Proceedings of the 6th Berkeley symposium on mathematical statistics and probability, vol 3, pp 401–448
Du N, Wang Y, He N, Sun J, Song L (2015) Time-sensitive recommendation from recurrent user activities. In: Proceedings of the annual conference on neural information processing systems (NIPS), pp 3492–3500
Eichler M, Dahlhaus R, Dueck J (2017) Graphical modeling for multivariate Hawkes processes with nonparametric link functions. J Time Ser Anal 38(2):225–242
Etesami J, Kiyavash N, Zhang K, Singhal K (2016) Learning network of multivariate Hawkes processes: a time series approach. In: Proceedings of the conference on uncertainty in artificial intelligence (UAI), pp 162–171
Farajtabar M, Du N, Rodriguez MG, Valera I, Zha H, Song L (2014) Shaping social activity by incentivizing users. In: Proceedings of the annual conference on neural information processing systems (NIPS), pp 2474–2482
Hall EC, Willett RM (2016) Tracking dynamic point processes on networks. IEEE Trans Inf Theory 62(7):4327–4346
Hawkes AG (1971) Spectra of some self-exciting and mutually exciting point processes. Biometrika 58(1):83–90
Krumin M, Reutsky I, Shoham S (2010) Correlation-based analysis and generation of multiple spike trains using Hawkes models with an exogenous input. Front Comput Neurosci 4:147
Lan G (2012) An optimal method for stochastic composite optimization. Math Program 133(1):365–397
Lan G (2013) The complexity of large-scale convex programming under a linear optimization oracle. arXiv:1309.5550
Lee J, Kim S, Lebanon G, Singer Y (2013) Local low-rank matrix approximation. In: Proceedings of the international conference on machine learning (ICML), pp 82–90
Lee J, Sun M, Kim S, Lebanon G (2012a) Automatic feature induction for stagewise collaborative filtering. In: Proceedings of the annual conference on neural information processing systems (NIPS), pp 314–322
Lee J, Sun M, Lebanon G (2012b) PREA: personalized recommendation algorithms toolkit. J Mach Learn Res 13(1):2699–2703
Lemonnier R, Scaman K, Kalogeratos A (2017) Multivariate Hawkes processes for large-scale inference. In: Proceedings of the AAAI conference on artificial intelligence, pp 2168–2174
Liniger TJ (2009) Multivariate Hawkes processes. PhD thesis, ETH Zurich
Nesterov Y (2013) Gradient methods for minimizing composite functions. Math Program 140(1):125–161
Sastry S (1990) Some NP-complete problems in linear algebra. Honors projects
Shang J, Sun M (2018) Local low-rank Hawkes processes for temporal user–item interactions. In: Proceedings of the IEEE international conference on data mining (ICDM), IEEE
Sun M, Lebanon G, Kidwell P (2011) Estimating probabilities in recommendation systems. In: Proceedings of the international conference on artificial intelligence and statistics (AISTATS), pp 734–742
Wand MP, Jones MC (1995) Kernel smoothing. Chapman and Hall/CRC, Boca Raton
Wang Y, Du N, Trivedi R, Song L (2016) Coevolutionary latent feature processes for continuous-time user–item interactions. In: Proceedings of the annual conference on neural information processing systems (NIPS), pp 4547–4555
Xiao W, Xu X, Liang K, Mao J, Wang J (2016) Job recommendation with Hawkes process: an effective solution for RecSys challenge 2016. In: Proceedings of the recommender systems challenge
Xu H, Farajtabar M, Zha H (2016) Learning granger causality for Hawkes processes. In: Proceedings of the international conference on machine learning (ICML), pp 1717–1726
Xu H, Wu W, Nemati S, Zha H (2017) Patient flow prediction via discriminative learning of mutually-correcting processes. IEEE Trans Knowl Data Eng 29(1):157–171
Yu AW, Ma W, Yu Y, Carbonell J, Sra S (2014) Efficient structured matrix rank minimization. In: Proceedings of the annual conference on neural information processing systems (NIPS), pp 1350–1358
Zhou K, Zha H, Song L (2013) Learning social infectivity in sparse low-rank networks using multi-dimensional Hawkes processes. In: Proceedings of the international conference on artificial intelligence and statistics (AISTATS), pp 641–649
Acknowledgements
This work was supported in part by the Louisiana Board of Regents under Grant LEQSF(2017-20)-RD-A-29. The authors would also like to thank Yichen Wang and Le Song from Georgia Tech for their helpful discussions.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This paper is an extended version of the full paper published in the proceedings of the IEEE International Conference on Data Mining (ICDM) 2018 [21].
Rights and permissions
About this article
Cite this article
Shang, J., Sun, M. Local low-rank Hawkes processes for modeling temporal user–item interactions. Knowl Inf Syst 62, 1089–1112 (2020). https://doi.org/10.1007/s10115-019-01379-6
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10115-019-01379-6