Understanding Transparent and Complicated Users as Instances of Preference Learning for Recommender Systems

  • P. VojtasEmail author
  • M. Kopecky
  • M. Vomlelova
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 9548)


In this paper we are concerned with user understanding in content based recommendation. We assume having explicit ratings with time-stamps from each user. We integrate three different movie data sets, trying to avoid features specific for single data and try to be more generic. We use several metrics which were not used so far in the recommender systems domain. Besides classical rating approximation with RMSE and ratio of order agreement we study new metrics for predicting Next-k and (at least) 1-hit at Next-k. Using these Next-k and 1-hit we try to model display of our recommendation – we can display k objects and hope to achieve at least one hit.

We trace performance of our methods and metrics also as a distribution along each single user. We define transparent and complicated users with respect to number of methods which achieved at least one hit.

We provide results of experiments with several combinations of methods, data sets and metrics along these three axes.


Content based recommendation Explicit ratings Integration of three movie datasets User understanding User preference learning RMSE Next-k 1-hit Order agreement metrics 



This work was supported by Czech grants P103-15-19877S and P46.


  1. 1.
    Amatriain, X., Pujol, J.M., Oliver, N.: I like it… i like it not: evaluating user ratings noise in recommender systems. In: Houben, G.-J., McCalla, G., Pianesi, F., Zancanaro, M. (eds.) UMAP 2009. LNCS, vol. 5535, pp. 247–258. Springer, Heidelberg (2009)CrossRefGoogle Scholar
  2. 2.
    Chapman, P.: The CRISP-DM user guide. In: 4th CRISP-DM SIG Workshop in Brussels in March 1999 (1999).
  3. 3.
    Guillou, F., Gaudel, R., Mary, J., Preux, P.: User engagement as evaluation: a ranking or a regression problem? In: ACM 2014, pp. 7–12 (2014)Google Scholar
  4. 4.
    Horváth, T., Vojtáš, P.: Induction of fuzzy and annotated logic programs. In: Muggleton, S.H., Otero, R., Tamaddoni-Nezhad, A. (eds.) ILP 2006. LNCS (LNAI), vol. 4455, pp. 260–274. Springer, Heidelberg (2007)CrossRefGoogle Scholar
  5. 5.
    Kifer, M., Subrahmanian, V.S.: Theory of generalized annotated logic programming and its applications. J. Logic Programm. 12(4), 335–367 (1992)MathSciNetCrossRefGoogle Scholar
  6. 6.
    Kopecky, M., Peska, L., Vojtas, P., Vomlelova, M.: Monotonization of user preferences. In: Andreasen, T., et al. (eds.) FQAS 2015. AISC, vol. 400, pp. 29–40. Springer, Heidelberg (2015)CrossRefGoogle Scholar
  7. 7.
    Peska, L., Vojtas, P.: Hybrid recommending exploiting multiple DBPedia language editions. In: ESWC 2014 Linked Open Data-enabled Recommender Systems Challenge (2014)Google Scholar
  8. 8.
    Peska, L., Vojtas, P.: Hybrid biased k-NN to predict movie tweets popularity, poster.
  9. 9.
  10. 10.
    Smith-Miles, K.: Understanding strengths and weaknesses of optimization algorithms with new visualization tools and methodologies. In: IFORS Conference Plenary Session, Barcelona, July 2014.
  11. 11.
    Vomlelova, M., Kopecky, M., Vojtas, P.: Transformation and aggregation preprocessing for top-k recommendation GAP rules induction. In: CEUR Proceedings Vol-1417 Rule Challenge and Doctoral Consortium @ RuleML 2015, Track 2: Rule-based Recommender Systems for the Web of Data.
  12. 12.
    Kuchar, J.: Augmenting a feature set of movies using linked open data. In: CEUR Proceedings Vol-1417 Rule Challenge and Doctoral Consortium @ RuleML 2015, Track 2: Rule-based Recommender Systems for the Web of Data., Dataset

Copyright information

© Springer International Publishing Switzerland 2016

Authors and Affiliations

  1. 1.Faculty of Mathematics and PhysicsCharles University in PraguePragueCzech Republic

Personalised recommendations