Mathematical Similarity Models: Do We Need Incomparability to Be Precise?

Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 11530)


The expanded version of Tversky’s feature model by Geist, Lengnink and Wille (GLW) provides a compelling representation of similarity by adding an incomparability option. However, previous research has found the GLW model lacking when compared with participants’ responses. We present several new models based on the GLW model and validate them on data from previous similarity experiments. The results indicate that human similarity perception is not well described by a partial order.


Similarity Knowledge structure Order theory 



This work was supported by the DFG SFB/TRR 135 “Cardinal Mechanisms of Perception”, project C6.


  1. 1.
    Endres, D., Adam, R., Giese, M.A., Noppeney, U.: Understanding the semantic structure of human fMRI brain recordings with formal concept analysis. In: Domenach, F., Ignatov, D.I., Poelmans, J. (eds.) ICFCA 2012. LNCS (LNAI), vol. 7278, pp. 96–111. Springer, Heidelberg (2012). Scholar
  2. 2.
    Geist, S., Lengnink, K., Wille, R.: An order-theoretic foundation for similarity measures. In: Lengnink, K. (ed.) Formalisierungen von Ähnlichkeit aus Sicht der Formalen Begriffsanalyse, pp. 75–87. Shaker Verlag (1996)Google Scholar
  3. 3.
    Schubert, M., Endres, D.: Empirically evaluating the similarity model of Geist, Lengnink and Wille. In: Chapman, P., Endres, D., Pernelle, N. (eds.) ICCS 2018. LNCS (LNAI), vol. 10872, pp. 88–95. Springer, Cham (2018). Scholar
  4. 4.
    Tversky, A.: Features of similarity. Psychol. Rev. 84(4), 327–352 (1977). Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  1. 1.Philipps University MarburgMarburgGermany

Personalised recommendations