On explainable fuzzy recommenders and their performance evaluation

Rutkowski, Tomasz ; Łapa, Krystian ; Nielek, Radosław

International Journal of Applied Mathematics and Computer Science, Tome 29 (2019) no. 3, pp. 595-610.

Résumé

This paper presents a novel approach to the design of explainable recommender systems. It is based on the Wang–Mendel algorithm of fuzzy rule generation. A method for the learning and reduction of the fuzzy recommender is proposed along with feature encoding. Three criteria, including the Akaike information criterion, are used for evaluating an optimal balance between recommender accuracy and interpretability. Simulation results verify the effectiveness of the presented recommender system and illustrate its performance on the MovieLens 10M dataset.

Détail

Keywords: recommender system, explainable recommendation, fuzzy system, Akaike information criterion
Mots-clés : system rekomendacyjny, system rozmyty, kryterium informacyjne Akaike

Bibliographie

[1] Alvarez-Estevez, D., Moret-Bonillo, V. (2018). Revisiting the Wang–Mendel algorithm for fuzzy classification, Expert Systems 35(4): 35:312268.

[2] Bologna, G. and Hayashi, Y. (2017). Characterization of symbolic rules embedded in deep DIMLP networks: A challenge to transparency of deep learning, Journal of Artificial Intelligence and Soft Computing Research 7(4): 265–286.

[3] Cpalka, K. (2017). Design of Interpretable Fuzzy Systems, Studies in Computational Intelligence 684, Springer Verlag, Cham.

[4] Harper, F.M. and Konstan, J.A. (2015). The MovieLens datasets: History and context, ACM Transactions on Interactive Intelligent Systems 5(4):19:1–19:19.

[5] Ishibuchi H. and T. Nakashima (2001). Effect of rule weights in fuzzy rule-based classification systems, IEEE Transactions on Fuzzy Systems 9(4): 506–515.

[6] Ishibuchi H. and T. Yamamoto (2005). Rule weight specification in fuzzy rule-based classification systems, IEEE Transactions on Fuzzy Systems 13(4): 428–435.

[7] Jin, Y. (2000). Fuzzy modeling of high-dimensional systems: Complexity reduction and interpretability improvement, IEEE Transactions on Fuzzy Systems 8(2): 212–221.

[8] Kuncheva, L. (2000). Fuzzy Classifier Design, Studies in Fuzziness and Soft Computing, Vol. 49, Springer Verlag, New York, NY.

[9] Liu, H., Gegov, A. and Cocea, M. (2017). Rule based networks: An efficient and interpretable representation of computational models, Journal of Artificial Intelligence and Soft Computing Research 7(2): 111–123.

[10] Lops, P., Gemmis, M. and Semeraro, G. (2011). Content-based recommender systems: State of the art and trends, in F. Ricci et al. (Eds), Recommender Systems Handbook, Springer, New York, NY, pp.73–105.

[11] Nauck, D. and R. Kruse (1998). How the learning of rule weights affects the interpretability of fuzzy systems, IEEE International Conference on Fuzzy Systems 1998 (FUZZ-IEEE’98), Ancorage, AK, USA, pp.1235–1240.

[12] Portugal, I., Paulo S.C. Alencar and Cowan, D.D. (2018). The use of machine learning algorithms in recommender systems: A systematic review, Expert System Applications 97: 205–227.

[13] Prasad, M., Liu, Y.-T., Li, D.-L., Lin, C.-T., Shah, R.R. and Kaiwartya, O.P. (2017). A new mechanism for data visualization with TSK-type preprocessed collaborative fuzzy rule based system, Journal of Artificial Intelligence and Soft Computing Research 7(1): 33–46.

[14] Riid, A. and Preden, J.-S. (2017). Design of fuzzy rule-based classifiers through granulation and consolidation, Journal of Artificial Intelligence and Soft Computing Research 7(2): 137–147.

[15] Rutkowska, D. (2002). Neuro-Fuzzy Architectures and Hybrid Learning, Studies in Fuzziness and Soft Computing, Springer Verlag, New York, NY.

[16] Rutkowski, L. (2004). Flexible Neuro-Fuzzy Systems: Structures, Learning and Performance Evaluation, Kluwer Academic Publishers, Boston, MA/Dordrecht/London.

[17] Rutkowski, L. (2008). Computational Intelligence: Methods and Techniques, Springer, Berlin.

[18] Rutkowski, T., Romanowski, J., Woldan, P., Staszewski, P. and Nielek, R. (2018). Towards interpretability of the movie recommender based on a neuro-fuzzy approach, 17th International Conference on Artificial Intelligence and Soft Computing, ICAISC 2018, Zakopane, Poland, pp. 752–762.

[19] Rutkowski, T., Romanowski, J., Woldan, P., Staszewski, P., Nielek, R. and Rutkowski, L. (2018). A content-based recommendation system using neuro-fuzzy approach, 2018 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE 2018), Piscataway, NJ, USA, pp. 1–8, DOI:10.1109/FUZZ-IEEE.2018.8491543.

[20] Simiński, K. (2010). Rule weights in a neuro-fuzzy system with a hierarchical domain partition, International Journal of Applied Mathematics and Computer Science 20(2): 337–347, DOI: 10.2478/v10006-010-0025-3.

[21] Söderström, T. and Stoica, P. (1989). System Identification, Prentice Hall International, Upper Saddle River, NJ.

[22] Wang, L.-X. and Mendel J.M. (1992). Generating fuzzy rules by learning from examples, IEEE Transactions on Systems, Man, and Cybernetics: Systems 22(6): 1414–1427.

[23] Wei, J., He, J., Chen, K., Zhou, Y. and Tang, Z. (2017). Collaborative filtering and deep learning based recommendation system for cold start items, Expert Systems with Applications 69: 29–39.

[24] Zhang, S., Yao L., Sun A. and Y. Tay (2018). Deep learning based recommender system: A survey and new perspectives, ACM Computing Surveys 52(1), Article No. 5.