Advertisement

GeoInformatica

, Volume 23, Issue 3, pp 329–352 | Cite as

Regularized topic-aware latent influence propagation in dynamic relational networks

  • Shuhui Wang
  • Liang LiEmail author
  • Chenxue Yang
  • Qingming Huang
Article
  • 70 Downloads

Abstract

On social networks, investigating how the influence is propagated is crucial in understanding the network evolution and the social impact of different topics. In previous study, the influence propagation is either modeled based on the static network structure, or the infection between two connected users is recovered from some given event cascades. Unfortunately, existing solutions are incapable of identifying the user susceptibility delivered by user generated content. In this paper, we propose RegInfoIbp, a general regularized learning framework for modeling topic-aware influence propagation in dynamic network structures. Specifically, the observed time-sequential user topic preference and user adjacency information are factorized by the prior information reflected by a user-influential bipartite relation graph. The influence propagation is approximated with a nonparametric regularized Bayesian matrix factorization model with tractable polynomial complexity. and the influential users are identified by several sampling algorithms with slightly different approximation qualities. To further model dynamic temporal evolution, we construct Markov conditional probabilistic model on the compact latent feature representation. By integrating both topic and structure information into the regularized non-parametric probabilistic learning process, RegInfoIbp is more efficient and accurate in discovering the key factors in the content and influential users in dynamic network structure. Extensive experiments demonstrate that RegInfoIbp better adapts to real data, and achieves better approximation in influence propagation over existing approaches.

Keywords

Bayesian nonparametric matrix factorization Influence propagation Dynamic relational networks 

Notes

Acknowledgments

This work was supported in part by National Natural Science Foundation of China: 61672497, 61620106009, 61771457, 61732007, U1636214 and 61836002, in part by National Basic Research Program of China (973 Program): 2015CB351800, and in part by Key Research Program of Frontier Sciences of CAS: QYZDJ-SSW-SYS013.

References

  1. 1.
    Aggarwal CC, Lin S, Yu PS (2012) On influential node discovery in dynamic social networks. SDM ’12, pp 636–647Google Scholar
  2. 2.
    Barbieri N, Bonchi F, Manco G (2012) Topic-aware social influence propagation models. ICDM ’12, pp 81–90Google Scholar
  3. 3.
    Bi B, Tian Y, Sismanis Y, Balmin A, Cho J (2014) Scalable topic-specific influence analysis on microblogs. WSDM ’14, pp 513–522Google Scholar
  4. 4.
    Borgs C, Brautbar M, Chayes J, Lucier B (2014) Maximizing social influence in nearly optimal time. SODA ’14, pp 946–957Google Scholar
  5. 5.
    Brin S, Page L (1998) The anatomy of a large-scale hypertextual web search engine. Comput Netw ISDN Syst 30:107–117CrossRefGoogle Scholar
  6. 6.
    Caron F (2012) Bayesian nonparametric models for bipartite graphs. NIPS ’12, pp 2051–2059Google Scholar
  7. 7.
    Chen S, Fan J, Li G, Feng J, lee Tan K, Tang J (2015) Online topic aware influence maximization. In: Proceedings of VLDB Endowment, vol 8Google Scholar
  8. 8.
    Chen W, Wang C, Wang Y (2010) Scalable influence maximization for prevalent viral marketing in large-scale social networks. KDD ’10, pp 1029–1038Google Scholar
  9. 9.
    Chen W, Wang Y, Yang S (2009) Efficient influence maximization in social networks. KDD ’09, pp 199–208Google Scholar
  10. 10.
    Du N, Song L, Gomez-Rodriguez M, Zha H (2013) Scalable influence estimation in continuous-time diffusion networks. NIPS ’13, pp 3147–3155Google Scholar
  11. 11.
    Foulds JR, Dubois C, Asuncion AU, Butts CT, Smyth P (2011) A dynamic relational infinite feature model for longitudinal social networks. AISTATS ’11, pp 287–295Google Scholar
  12. 12.
    Gael JV, Teh YW, Ghahramani Z (2008) The infinite factorial hidden markov model. NIPS ’08, pp 1697–1704Google Scholar
  13. 13.
    Gopalan P, Ruiz FJ, Ranganath R, Blei D (2014) Bayesian nonparametric poisson factorization for recommendation systems. AISTATS ’14, pp 275–283Google Scholar
  14. 14.
    Goyal A, Lu W, Lakshmanan LV (2011) Celf++ optimizing the greedy algorithm for influence maximization in social networks. WWW ’11, pp 47–48Google Scholar
  15. 15.
    Griffiths TL, Ghahramani Z (2005) Infinite latent feature models and the indian buffet process. In: NIPS, pp 475–482Google Scholar
  16. 16.
    Heaukulani C, Ghahramani Z (2013) Dynamic probabilistic models for latent feature propagation in social networks. ICML ’13, pp 275–283Google Scholar
  17. 17.
    Huo Z, Huang X, Hu X (2018) Link prediction with personalized social influence. In: AAAIGoogle Scholar
  18. 18.
    Kempe D, Kleinberg J, Tardos E (2003) Maximizing the spread of influence through a social network. KDD ’03, pp 137–146Google Scholar
  19. 19.
    Kulesza A, Taskar B (2012) Determinantal point processes for machine learning. arXiv:1207.6083
  20. 20.
    Le Cam L (1960) An approximation theorem for the poisson binomial distribution. Pac J Math 10:1181–1197CrossRefGoogle Scholar
  21. 21.
    Lei S, Maniu S, Mo L, Cheng R, Senellart P (2015) Online influence maximization. In: ACM SIGKDD international conference on knowledge discovery and data mining, pp 645–654Google Scholar
  22. 22.
    Liu L, Tang J, Han J, Jiang M, Yang S (2010) Mining topic-level influence in heterogeneous networks. CIKM ’10, pp 199–208Google Scholar
  23. 23.
    Liu S, Qu Q, Wang S (2018) Heterogeneous anomaly detection in social diffusion with discriminative feature discovery. Inf Sci 439-440:1–18CrossRefGoogle Scholar
  24. 24.
    Liu S, Wang S (2017) Trajectory community discovery and recommendation by multi-source diffusion modeling. IEEE Trans Knowl Data Eng 29(4):898–911CrossRefGoogle Scholar
  25. 25.
    Liu S, Wang S, Zhu F (2015) Structured learning from heterogeneous behavior for social identity linkage. IEEE Trans Knowl Data Eng 27(7):2005–2019CrossRefGoogle Scholar
  26. 26.
    Miller K, Jordan MI, Griffiths TL (2009) Nonparametric latent feature models for link prediction. NIPS ’09, pp 1276–1284Google Scholar
  27. 27.
    Nguyen HT, Thai MT, Dinh TN (2016) Stop-and-stare:optimal sampling algorithms for viral marketing in billion-scale networks. In: ACM international conference on management of data (SIGMOD), pp 695–710Google Scholar
  28. 28.
    Pan T, Kuhnle A, Li X, Thai MT (2017) Popular topics spread faster: New dimension for influence propagation in online social networks. arXiv:1702.01844
  29. 29.
    Phan N, Ebrahimi J, Dou D, Kil D, Piniewski B (2015) Topic-aware physical activity propagation with temporal dynamics in a health social network. ACM transactions on intelligent systems and technologyGoogle Scholar
  30. 30.
    Qu Q, Liu S, Yang B, Jensen CS (2014) Efficient top-k spatial locality search for co-located spatial web objects. In: IEEE MDM, pp 269–278Google Scholar
  31. 31.
    Qu Q, Liu S, Zhu F, Jensen CS (2016) Efficient online summarization of large-scale dynamic networks. IEEE Trans Knowl Data Eng 28(12):3231–3245CrossRefGoogle Scholar
  32. 32.
    Richardson M, Domingos P (2002) Mining knowledge-sharing sites for viral marketing. KDD ’02, pp 61–70Google Scholar
  33. 33.
    Rodriguez MG, Schölkopf B (2012) Influence maximization in continuous time diffusion networks. ICML ’12, pp 313–320Google Scholar
  34. 34.
    Scott SL (2002) Bayesian methods for hidden markov models: Recursive computing in the 21st century. J Am Stat Assoc 97:337–351CrossRefGoogle Scholar
  35. 35.
    Song D, Meyer DA, Tao D (2015) Efficient latent link recommendation in signed networks. In KDD, pp 1105–1114Google Scholar
  36. 36.
    Tang J, Sun J, Wang C, Yang Z (2009) Social influence analysis in large-scale networks. KDD ’09, pp 807–816Google Scholar
  37. 37.
    Tang Y, Xiao X, maximization Y. S. h. i. (2014) Influence Near-optimal time complexity meets practical efficiency. SIGMOD ’14, pp 75–86Google Scholar
  38. 38.
    Tong G, Wu W, Tang S, Du DZ (2017) Adaptive influence maximization in dynamic social networks. IEEE/ACM Trans Networking 25(1):112–125CrossRefGoogle Scholar
  39. 39.
    Wang B, Chen G, Fu L, Song L, Wang X (2017) DRIMUX dynamic rumor influence minimization with user experience in social networks. IEEE Trans Knowl Data Eng PP(99):1–1Google Scholar
  40. 40.
    Wang C, Tang J, Sun J, Han J (2011) Dynamic social influence analysis through time-dependent factor graphs. ASONAM ’11, pp 239–246Google Scholar
  41. 41.
    Weng J, Lim E-P, Jiang J, He Q (2010) Twitterrank: Finding topic-sensitive influential twitterers. WSDM ’10, pp 261–270Google Scholar
  42. 42.
    Wood F, Griffiths TL, Ghahramani Z (2006) A non-parametric Bayesian method for inferring hidden causes. UAI ’06, pp 536–543Google Scholar
  43. 43.
    Zhan Q, Zhang J, Wang S, Yu P, Xie J (2015) Influence maximization across partially aligned heterogenous social networks. In: PAKDD, pp 58–69Google Scholar
  44. 44.
    Zhang J, Yu PS (2014) Link prediction across heterogeneous social networks: A surveyGoogle Scholar
  45. 45.
    Zhang J, Yu PS (2015) Integrated anchor and social link predictions across partially aligned social networks. In: IJCAIGoogle Scholar
  46. 46.
    Zheng W, Kveton B, Valko M, Vaswani S (2017) Online influence maximization under independent cascade model with semi-bandit feedback. In: NIPSGoogle Scholar
  47. 47.
    Zhuang H, Sun Y, Tang J, Zhang J, Sun X (2013) Influence maximization in dynamic social networks. ICDM ’13, pp 1313–1318Google Scholar

Copyright information

© Springer Science+Business Media, LLC, part of Springer Nature 2019

Authors and Affiliations

  1. 1.Key Laboratory of Intelligent Information Processing (CAS)Institute of Computing Technology (CAS)BeijingChina
  2. 2.National Key Laboratory of Pattern Recognition (NLPR)Institute of Automation (CAS)BeijingChina
  3. 3.School of Computer Science and TechnologyUniversity of Chinese Academy of SciencesBeijingChina

Personalised recommendations