Advertisement

GLaSS: Semi-supervised Graph Labelling with Markov Random Walks to Absorption

  • Max Glonek
  • Jonathan Tuke
  • Lewis Mitchell
  • Nigel Bean
Conference paper
Part of the Studies in Computational Intelligence book series (SCI, volume 812)

Abstract

Graph labelling is a key activity of network science, with broad practical applications, and close relations to other network science tasks, such as community detection and clustering. While a large body of work exists on both unsupervised and supervised labelling algorithms, the class of random walk-based supervised algorithms requires further exploration, particularly given their relevance to social and political networks. This work proposes a new semi-supervised graph labelling method, the GLaSS method, that exactly calculates absorption probabilities for random walks on connected graphs, whereas previous methods rely on simulation and approximation. The proposed method models graphs exactly as a discrete time Markov chain, treating labelled nodes as absorbing states. The method is applied to a series of undirected graphs of roll call voting data from the United States House of Representatives. The GLaSS method is compared to existing supervised and unsupervised methods, demonstrating strong and consistent performance when estimating the labels of unlabelled nodes in graphs.

Keywords

Community detection Graph labelling Random walk Markov chain Political networks 

Notes

Acknowledgements

The authors thank Data to Decisions CRC and the ARC Centre of Excellence for Mathematical and Statistical Frontiers for their financial support.

References

  1. 1.
    Andris, C., Lee, D., Hamilton, M.J., Martino, M., Gunning, C.E., Selden, J.A.: The rise of partisanship and super-cooperators in the US house of representatives. PLoS One 10(4), e0123507 (2015)Google Scholar
  2. 2.
    Azran, A.: The rendezvous algorithm: multiclass semi-supervised learning with markov random walks. In: Proceedings of the 24th International Conference on Machine Learning (ICML), pp. 49–56 (2007)Google Scholar
  3. 3.
    Fish, B., Huang, Y., Reyzin, L.: Recovering social networks by observing votes. In: Proceedings of the 2016 International Conference on Autonomous Agents & Multiagent Systems, pp. 376–384 (2016)Google Scholar
  4. 4.
    Fortunato, S.: Community detection in graphs. Phys. Rep. 486(3–5), 75–174 (2010)Google Scholar
  5. 5.
    Garimella. K., Weber, I.: A long-term analysis of polarization on Twitter. arXiv:1703.02769 (2017)
  6. 6.
    Grinstead, C.M., Snell, J.L.: Introduction to probability. Amer. Math. Soc. (2012)Google Scholar
  7. 7.
    Hassan, A., Radev, D.: Identifying text polarity using random walks. In: Proceedings of the 48th Annual Meeting of the Association for Computational Linguistics (ACL), pp. 395–403 (2010)Google Scholar
  8. 8.
    Csardi, G., Nepusz, T.: The igraph software package for complex network research. InterJ. Complex Syst. 1695 (2006). https://igraph.org. Accessed 28 Aug 2018
  9. 9.
    Lewis, J.B., Poole, K., Rosenthal, H., Boche, A., Rudkin, A., Sonnet, L.: Voteview: congressional roll-call votes database (2018). https://voteview.com/data. Accessed 21 Aug 2018
  10. 10.
    Pons, P., Latapy, M.: Computing communities in large networks using random walks. In: International Symposium on Computer and Information Sciences, pp. 284–293 (2005)Google Scholar
  11. 11.
    Rizoiu, M.A., Graham, T., Zhang, R., Zhang, Y., Ackland, R., Xie, L.: #debatenight: the role and influence of socialbots on twitter during the 1st us presidential debate. arXiv:1802.09808 (2018)
  12. 12.
    Shai, S., Stanley, N., Granell, C., Taylor, D., Mucha, P.J.: Case studies in network community detection. arXiv:1705.02305 (2017)
  13. 13.
    Talukdar, P.P., Reisinger, J., Paşca, M., Ravichandran, D., Bhagat, R., Pereira, F.: Weakly-supervised acquisition of labelled class instances using graph random walks. In: Proceedings of the Conference on Empirical Methods in Natural Language Processing, pp. 582–590 (2008)Google Scholar
  14. 14.
    Zhou, H., Lipkowsky, R.: Network brownian motion: A new method to measure vertex-vertex proximity and to identify communities and subcommunities. In: International Conference on Computational Science (ICCS), pp. 1062–1069 (2004)Google Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  • Max Glonek
    • 1
  • Jonathan Tuke
    • 1
  • Lewis Mitchell
    • 1
  • Nigel Bean
    • 1
    • 2
  1. 1.School of Mathematical SciencesUniversity of AdelaideAdelaideAustralia
  2. 2.ARC Centre of Excellence for Mathematical and Statistical Frontiers, University of AdelaideAdelaideAustralia

Personalised recommendations