Developing Hybrid-Based Recommender System with Naïve Bayes Optimization to Increase Prediction Efficiency
DOI:
https://doi.org/10.24203/ijcit.v10i2.75Keywords:
Collaborative filtering, Association rule mining, hybrid ensemble, recommender systemsAbstract
Commerce and entertainment world today have shifted to the digital platforms where customer preferences are suggested by recommender systems. Recommendations have been made using a variety of methods such as content-based, collaborative filtering-based or their hybrids. Collaborative systems are common recommenders, which use similar users’ preferences. They however have issues such as data sparsity, cold start problem and lack of scalability. When a small percentage of users express their preferences, data becomes highly sparse, thus affecting quality of recommendations. New users or items with no preferences, forms cold start issues affecting recommendations. High amount of sparse data affects how the user-item matrices are formed thus affecting the overall recommendation results. How to handle data input in the recommender engine while reducing data sparsity and increase its potential to scale up is proposed. This paper proposed development of hybrid model with data optimization using a Naïve Bayes classifier, with an aim of reducing data sparsity problem and a blend of collaborative filtering model and association rule mining-based ensembles, for recommending items with an aim of improving their predictions. Machine learning using python on Jupyter notebook was used to develop the hybrid. The models were tested using MovieLens 100k and 1M datasets. We demonstrate the final recommendations of the hybrid having new top ten highly rated movies with 68% approved recommendations. We confirm new items suggested to the active user(s) while less sparse data was input and an improved scaling up of collaborative filtering model, thus improving model efficacy and better predictions.
References
X. Guan, C. Li and Y. Guan, "Matrix Factorization with Rating Completion: An Enhanced SVD Model for Collaborative Filtering Recommender Systems," in IEEE Access, vol. 5, pp. 27668-27678, 2017, doi: 10.1109/ACCESS.2017.2772226.
Li, Y., Bradshaw, J., & Sharma, Y. (2019, May). Are generative classifiers more robust to adversarial attacks? In International Conference on Machine Learning (pp. 3804-3814). PMLR.
Lavecchia, A. (2015). Machine-learning approaches in drug discovery: methods and applications. Drugs discovery today, 20(3), 318-331.
Çano, E., & Morisio, M. (2017). Hybrid recommender systems: A systematic literature review. Intelligent Data Analysis, 21(6), 1487-1524.
Portugal, I., Alencar, P., & Cowan, D. (2018). The use of machine learning algorithms in recommender systems: A systematic review. Expert Systems with Applications, 97, 205-227.
Isinkaye, F. O., Folajimi, Y. O., & Ojokoh, B. A. (2015). Recommendation systems: Principles, methods and evaluation. Egyptian Informatics Journal, 16(3), 261-273.
Wei, J., He, J., Chen, K., Zhou, Y., & Tang, Z. (2017). Collaborative filtering and deep learning-based recommendation system for cold start items. Expert Systems with Applications, 69, 29-39.
Li, J., Jing, M., Lu, K., Zhu, L., Yang, Y., & Huang, Z. (2019, July). From zero-shot learning to cold-start recommendation. In Proceedings of the AAAI Conference on Artificial Intelligence (Vol. 33, pp. 4189-4196).
Hu, Y., Shi, W., Li, H., & Hu, X. (2017). Mitigating data sparsity using similarity reinforcement- enhanced collaborative filtering. ACM Transactions on Internet Technology (TOIT), 17(3), 1-20.
Yao, L., Xu, Z., Zhou, X., & Lev, B. (2019). Synergies Between Association Rules and Collaborative Filtering in Recommender System: An Application to Auto Industry. In Data Science and Digital Business (pp. 65-80). Springer, Cham.
Paradarami, T. K., Bastian, N. D., & Wightman, J. L. (2017). A hybrid recommender system using artificial neural networks. Expert Systems with Applications, 83, 300-313.
Seo, Y. D., Kim, Y. G., Lee, E., & Baik, D. K. (2017). Personalized recommender system based on friendship strength in social network services. Expert Systems with Applications, 69, 135-148.
Aggarwal, C. C. (2016). Recommender systems (Vol. 1). Cham: Springer International Publishing.
L. Zhang, W. Wang and Y. Zhang, “Privacy Preserving Association Rule Mining: Taxonomy, Techniques, and Metrics,” in IEEE Access, vol. 7, pp. 45032-45047, 2019, doi: 10.1109/ACCESS.2019.2908452.
Natarajan, S., Vairavasundaram, S., Natarajan, S., & Gandomi, A. H. (2020). Resolving data sparsity and cold start problem in collaborative filtering recommender system using linked open data. Expert Systems with Applications, 149, 113248.
Brownlee, J. (2019, November 27th). How to Choose a Feature Selection Method for Machine Learning. Retrieved on August 30, 2020 from: https://machinelearningmastery.com/feature-selection-with-real-and-categorical-data
Chen, V. X., & Tang, T. Y. (2019, August). Incorporating singular value decomposition in user-based collaborative filtering technique for a movie recommendation system: A comparative study. In Proceedings of the 2019 the International Conference on Pattern Recognition and Artificial Intelligence (pp. 12-15).
Falk, K. (2019). Practical recommender systems. Manning Publications.
Da Silva, J. F. G., de Moura Junior, N. N., & Caloba, L. P. (2018, July). Effects of data sparsity on recommender systems based on collaborative filtering. In 2018 International Joint Conference on Neural Networks (IJCNN) (pp. 1-8). IEEE.
Downloads
Published
Issue
Section
License
Copyright (c) 2021 Ndung’u Rachael Njeri, Rachael Njeri Ndung'u, Wambugu Geoffrey Mariga
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
The articles published in International Journal of Computer and Information Technology (IJCIT) is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
