Hyperparameter Optimization for Predicting the Tolerance Level of Religious Discourse

  • Donald E. BrownEmail author
  • Hope McIntyre
  • Peter J. Grazaitis
  • Riannon M. Hazell
  • Nicholas Venuti
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 10354)


To address the rising tide of religious violence as it affects U.S. Military deployments, the Army requires analytic methods that can be generalized to predict religious group violence around the globe and scalable to the number of potential groups in an area of operation. Current computational methods based on semantics and topics are lacking in predictive performance for the generalized problem and topic modeling performs poorly in predicting the tolerance level of new groups towards U.S. Military presence. The research in this paper aims to discover the association between religious speech and behaviors and provide a foundation for proactive engagement with these groups. The approach builds on the work from ethnolinquistics to model how things are said (performative analysis) rather than word meanings (semantic analysis). Recent research has developed computational approaches to streamline the manually intensive performative analysis of religious text. While producing promising results, these computational methods lack systematic optimization of the hyperparamters in the learning algorithms. Hence, we do not know the sensitivity of the results to parameter settings. This paper reports on results for predicting religious tolerance by optimizing the parameters in the signal processing algorithms and shows that the predictive power of performative approach is robust to parameter settings.


Behavior analysis Military Computational linguistics 


  1. 1.
    Abu-Nimer, M.: Conflict resolution, culture, and religion: toward a training model of interreligious peacebuilding. J. Peace Res. 38(6), 685–704 (2001)CrossRefGoogle Scholar
  2. 2.
    Barsalou, L.: Flexibility, structure, and linguistic vagary in concepts: manifestations of a compositional system of perceptual symbols. Theor. Mem. 1, 29–31 (1993)Google Scholar
  3. 3.
    Blatter, B., Patel, V.: Exploring dangerous neighborhoods: latent semantic analysis and computing beyond the bounds of the familiar. In: AMIA 2005 Symposium Proceedings, pp. 151–155 (2005)Google Scholar
  4. 4.
    Boussidan, A., Ploux, S.: Using topic salience and connotational drifts to detect candidates to semantic change. In: Proceedings of the Ninth International Conference on Computational Semantics, pp. 315–319. Association for Computational Linguistics (2011)Google Scholar
  5. 5.
    Bullinaria, J.A., Levy, J.: Extracting semantic representations from word co-occurrence statistics: stop-lists, stemming, and SVD. Behav. Res. Methods 44(3), 890–907 (2012)CrossRefGoogle Scholar
  6. 6.
    Cook, P., Stevenson, S.: Automatically identifying changes in the semantic orientation of words. In: LREC (2010)Google Scholar
  7. 7.
    Csardi, G., Nepusz, T.: The igraph software package for complex network research. InterJ. Complex Syst. 1695(5), 1–9 (2006)Google Scholar
  8. 8.
    Estrada, E., Rodriguez-Velazquez, J.A.: Subgraph centrality in complex networks. Phys. Rev. E 71(5), 056103 (2005)MathSciNetCrossRefGoogle Scholar
  9. 9.
    Hassner, R.E.: War on Sacred Grounds. Cornell University Press, Ithica (2009)Google Scholar
  10. 10.
    Loper, E., Bird, S.: NLTK: the natural language toolkit. In: Proceedings of the ACL-02 Workshop on Effective tools and methodologies for teaching natural language processing and computational linguistics, vol. 1, pp. 63–70. Association for Computational Linguistics (2002)Google Scholar
  11. 11.
    Marcus, B.: Global covenant of religion. (2015).
  12. 12.
    Ochs, P.: The possibilities and limits of inter-religious dialogue. In: Omer, A., Appleby, R.S., Little, D. (eds.) Religion, Conflict, and Peacebuilding, pp. 488–534. Oxford University Press, New York (2015)Google Scholar
  13. 13.
    Sagi, E., Kaufmann, S., Clark, B.: Semantic density analysis: Comparing word meaning across time and phonetic space. In: Proceedings of the Workshop on Geometrical Models of Natural Language Semantics, pp. 104–111. Association for Computational Linguistics (2009)Google Scholar
  14. 14.
    Tompkins, P.J.: Human factors considerations of undergrounds in insurgencies. U.S. Army Special Operations Command, Ft. Bragg, NC, January 2013Google Scholar
  15. 15.
    United States Special Operations Command: Gray zone, September 2015.
  16. 16.
    Venuti, N., Sachtjen, B., McIntyre, H., Mishra, C., Hays, M., Brown, D.E.: Predicting the tolerance level of religious discourse through computational linguistics. In: 2016 IEEE Systems and Information Engineering Design Symposium (SIEDS), pp. 309–314. IEEE Press (2016)Google Scholar
  17. 17.
    Yang, M., Wong, S., Coid, J.: The efficacy of violence prediction: a meta-analytic comparison of nine risk assessment tools. Psychol. Bull. 136(5), 740 (2010)CrossRefGoogle Scholar

Copyright information

© Springer International Publishing AG 2017

Authors and Affiliations

  • Donald E. Brown
    • 1
    Email author
  • Hope McIntyre
    • 1
  • Peter J. Grazaitis
    • 2
  • Riannon M. Hazell
    • 2
  • Nicholas Venuti
    • 1
  1. 1.Data Science InstituteUniversity of VirginiaCharlottesvilleUSA
  2. 2.Human Research and Engineering Directorate, U.S. Army Research LaboratoryAberdeen Proving GroundUSA

Personalised recommendations