Advertisement

An Evidential Semi-supervised Label Aggregation Approach

  • Lina AbassiEmail author
  • Imen Boukhris
Conference paper
  • 841 Downloads
Part of the Lecture Notes in Computer Science book series (LNCS, volume 11775)

Abstract

Crowdsourcing is a powerful concept that typically takes advantage of human intelligence to deal with problems in many fields most importantly in machine learning. Indeed, it enables to collect training labels in a fast and cheap way for supervised algorithms. The only major challenge is that the quality of the contributions is not always guaranteed because of the expertise heterogeneity of the participants. One of the basic strategies to overcome this problem is to assign each task to multiple workers and then combine their answers in order to obtain a single reliable one. This paper provides a new iterative approach that aggregates imperfect labels using the supervision of few gold labels under the evidence theory. Besides of inferring the consensus answers, the workers’ accuracies and the questions difficulties are as well estimated. A comparative evaluation on synthetic and real datasets confirms the effectiveness of our semi-supervised approach over the baselines.

Keywords

Crowdsourcing Answer aggregation Evidence theory Gold labels 

References

  1. 1.
    Zheng, Y., Wang, J., Li, G., Feng, J.: QASCA: a quality-aware task assignment system for crowdsourcing applications. In: International Conference on Management of Data, pp. 1031–1046 (2015)Google Scholar
  2. 2.
    Yan, T., Kumar, V., Ganesan, D.: Designing games with a purpose. Commun. ACM 51(8), 58–67 (2008)Google Scholar
  3. 3.
    Snow, R., O’Connor, B., Jurafsky, D., Ng, A.Y.: Cheap and fast but is it good? Evaluation non-expert annotations for natural language tasks. In: The Conference on Empirical Methods in Natural Languages Processing, pp. 254–263 (2008)Google Scholar
  4. 4.
    Kuncheva, L., Whitaker, C., Shipp, C., Duin, R.: Limits on the majority vote accuracy in classifier fusion. Pattern Anal. Appl. 6, 22–31 (2003)MathSciNetCrossRefGoogle Scholar
  5. 5.
    Sheng, V.S., Provost, F., Ipeirotis, P.G.: Get another label? Improving data quality and data mining using multiple, noisy labellers. In: International Conference on Knowledge Discovery and Data Mining, pp. 614–622 (2008)Google Scholar
  6. 6.
    Shafer, G.: A Mathematical Theory of Evidence, vol. 1. Princeton University Press, Princeton (1976)zbMATHGoogle Scholar
  7. 7.
    Dempster, A.P.: Upper and lower probabilities induced by a multivalued mapping. Ann. Math. Stat. 38, 325–339 (1967)MathSciNetCrossRefGoogle Scholar
  8. 8.
    Jousselme, A.-L., Grenier, D., Bossé, É.: A new distance between two bodies of evidence. Inf. Fusion 2, 91–101 (2001)CrossRefGoogle Scholar
  9. 9.
    Lefèvre, E., Elouedi, Z.: How to preserve the confict as an alarm in the combination of belief functions? Decis. Support Syst. 56, 326–333 (2013)CrossRefGoogle Scholar
  10. 10.
    Smets, P.: The combination of evidence in the transferable belief model. IEEE Trans. Pattern Anal. Mach. Intell. 12(5), 447–458 (1990)CrossRefGoogle Scholar
  11. 11.
    Raykar, V.C., Yu, S.: Eliminating spammers and ranking annotators for crowdsourced labelling tasks. J. Mach. Learn. Res. 13, 491–518 (2012)MathSciNetzbMATHGoogle Scholar
  12. 12.
    Dawid, A.P., Skene, A.M.: Maximum likelihood estimation of observer error-rates using the EM algorithm. Appl. Stat. 28, 20–28 (2010)CrossRefGoogle Scholar
  13. 13.
    Karger, D.R., Oh, S., Shah, D.: Budget-optimal task allocation for reliable crowdsourcing systems. Oper. Res. 62, 1–24 (2014)CrossRefGoogle Scholar
  14. 14.
    Raykar, V.C., et al.: Supervised learning from multiple experts: whom to trust when everyone lies a bit. In: Proceedings of the 26th Annual International Conference on Machine Learning, pp. 889–896 (2009)Google Scholar
  15. 15.
    Khattak, F.K., Salleb, A.: Quality control of crowd labelling through expert evaluation. In: The Neural Information Processing Systems 2nd Workshop on Computational Social Science and the Wisdom of Crowds, pp. 27–29 (2011)Google Scholar
  16. 16.
    Lee, K., Caverlee, J., Webb, S.: The social honeypot project: protecting online communities from spammers. In: International World Wide Web Conference, pp. 1139–1140 (2010)Google Scholar
  17. 17.
    Smets, P., Mamdani, A., Dubois, D., Prade, H.: Non Standard Logics for Automated Reasoning, pp. 253–286. Academic Press, London (1988)zbMATHGoogle Scholar
  18. 18.
    Ben Rjab, A., Kharoune, M., Miklos, Z., Martin, A.: Characterization of experts in crowdsourcing platforms. In: Vejnarová, J., Kratochvíl, V. (eds.) BELIEF 2016. LNCS (LNAI), vol. 9861, pp. 97–104. Springer, Cham (2016).  https://doi.org/10.1007/978-3-319-45559-4_10CrossRefGoogle Scholar
  19. 19.
    Watanabe, M., Yamaguchi, K.: The EM Algorithm and Related Statistical Models, p. 250. CRC Press, Boca Raton (2003)CrossRefGoogle Scholar
  20. 20.
    Li, J., Li, X., Yang, B., Sun, X.: Segmentation-based image copy-move forgery detection scheme. IEEE Trans. Inf. Forensics Secur. 10, 507–518 (2015)CrossRefGoogle Scholar
  21. 21.
    Liu, K., Cheung, W.K., Liu, J.: Detecting multiple stochastic network motifs in network data. Knowl. Inf. Syst. 42, 49–74 (2015)CrossRefGoogle Scholar
  22. 22.
    Whitehill, J., Wu, T., Bergsma, J., Movellan, J.R., Ruvolo, P.L.: Whose vote should count more: optimal integration of labels from labellers of unknown expertise. In: Neural Information Processing Systems, pp. 2035–2043 (2009)Google Scholar
  23. 23.
    Abassi, L., Boukhris, I.: Crowd label aggregation under a belief function framework. In: Lehner, F., Fteimi, N. (eds.) KSEM 2016. LNCS (LNAI), vol. 9983, pp. 185–196. Springer, Cham (2016).  https://doi.org/10.1007/978-3-319-47650-6_15CrossRefGoogle Scholar
  24. 24.
    Abassi, L., Boukhris, I.: A gold standards-based crowd label aggregation within the belief function theory. In: Benferhat, S., Tabia, K., Ali, M. (eds.) IEA/AIE 2017. LNCS (LNAI), vol. 10351, pp. 97–106. Springer, Cham (2017).  https://doi.org/10.1007/978-3-319-60045-1_12CrossRefzbMATHGoogle Scholar
  25. 25.
    Abassi, L., Boukhris, I.: Iterative aggregation of crowdsourced tasks within the belief function theory. In: Antonucci, A., Cholvy, L., Papini, O. (eds.) ECSQARU 2017. LNCS (LNAI), vol. 10369, pp. 159–168. Springer, Cham (2017).  https://doi.org/10.1007/978-3-319-61581-3_15CrossRefGoogle Scholar
  26. 26.
    Abassi, L., Boukhris, I.: A worker clustering-based approach of label aggregation under the belief function theory. Appl. Intell. 49, 53–62 (2018)CrossRefGoogle Scholar
  27. 27.
    Abassi, L., Boukhris, I.: Imprecise label aggregation approach under the belief function theory. In: Abraham, A., Cherukuri, A.K., Melin, P., Gandhi, N. (eds.) Intelligent Systems Design and Applications, vol. 941, pp. 607–616. Springer, Cham (2018)CrossRefGoogle Scholar
  28. 28.
    Koulougli, D., HadjAli, A., Rassoul, I.: Handling query answering in crowdsourcing systems: a belief function-based approach. In: Annual Conference of the North American Fuzzy Information Processing Society (NAFIPS), pp. 1–6 (2016)Google Scholar
  29. 29.
    Welinder, P., Branson, S., Perona, P., Belongie, S.J.: The multidimensional wisdom of crowds. In: Neural Information Processing Systems, pp. 2424–2432 (2010)Google Scholar
  30. 30.
    Frank, A.: UCI machine learning repository (1987). http://archive.ics.uci.edu/ml

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  1. 1.LARODEC Laboratory, Institut Supérieur de Gestion de TunisUniversity of TunisTunisTunisia

Personalised recommendations