ПРИМЕНА ФАКТОРИЗАЦИЈЕ МАТРИЦА У СИСТЕМИМА ПРЕПОРУКА

Leopoldina Đanić

doi:10.24867/34BE34Djanic

Elektrotehničko i računarsko inženjerstvo

God. 41 Br. 02 (2026): Зборник радова Факултета техничких наука

ПРИМЕНА ФАКТОРИЗАЦИЈЕ МАТРИЦА У СИСТЕМИМА ПРЕПОРУКА

Leopoldina Đanić

34BE35Djanic.pdf

DOI:: https://doi.org/10.24867/34BE34Djanic
Predato: February 17, 2026
Objavljeno: 2026-03-09

Apstrakt

У овом раду се разматра и изучава примена медода факторизације матрица у системима препорука. Проучавају се различите методе факторизације матрица. Циљ је да се побољшају тачности и персонализација препорука. Проблем који се решава се односи на што прецизније предвиђање оцена које би корисник дао филмовима. Технике коришћене у експерименту су СВД, СВД++ и ТимеСВД++.

Reference

[1] X. Zhou, J. He, G.Huang, Y. Yhang, “SVD based incremental approacher for recommender systems”, Joiurnal of Computer and System Sciences, Vol. 81, pp. 717-733, June 2015.
[2] R. Bell, C. Volinosky, “Matrix factorization techniques for recommender systems”, Computer, pp. 30-37, August 2009.
[3] Y. Koren, “Collaborative filtering with temporal dynamics”, Proceeding of the 15th ACM SIGKDD international conference on Knowledge discovery and data mining, pp. 447-456, June 2009.
[4] F. O. Isinkaye, “Matrix Factorization in Recommender Systems: Algorithms, Applications and Peciuliar Challenges”, IETE Journal of research, pp. 6087-6100, November 2021.
[5] R. Mehta, K. Rana, “A Review on Matrix Factorization Techniques in Recommender Systems”, CSCITA, pp. 269-274, April 2017.
[6] M. Jallouli, S. Lajmi, I. Amous, “When contextual information meets recommender systems: extended SVD++ models”, International Journal of Computers and Applications, pp. 349-356, May 2020.
[7] B. Zhang, X. Zhou, J. Li, L. Li, “Recommendation Algorithm Based on Matrix SVD with Exponential Correction”, CIPAE 2020, pp. 71-75, October 2020.
[8]https://www.kaggle.com/datasets/parasharmanas/movie-recommendation-system, (pristupljeno u oktobru 2025.)
[9] S. Jiang, J. Li, W. Zhou, “An Application of SVD++ Method in Collaborative Filtering”, IEEE, pp. 192-197, January 2021.
[10] https://www.datacamp.com/tutorial/rmse, (pristupljeno u oktobru 2025.)

[1] [1] X. Zhou, J. He, G.Huang, Y. Yhang, “SVD based incremental approacher for recommender systems”, Joiurnal of Computer and System Sciences, Vol. 81, pp. 717-733, June 2015.

[2] [2] R. Bell, C. Volinosky, “Matrix factorization techniques for recommender systems”, Computer, pp. 30-37, August 2009.

[3] [3] Y. Koren, “Collaborative filtering with temporal dynamics”, Proceeding of the 15th ACM SIGKDD international conference on Knowledge discovery and data mining, pp. 447-456, June 2009.

[4] [4] F. O. Isinkaye, “Matrix Factorization in Recommender Systems: Algorithms, Applications and Peciuliar Challenges”, IETE Journal of research, pp. 6087-6100, November 2021.

[5] [5] R. Mehta, K. Rana, “A Review on Matrix Factorization Techniques in Recommender Systems”, CSCITA, pp. 269-274, April 2017.

[6] [6] M. Jallouli, S. Lajmi, I. Amous, “When contextual information meets recommender systems: extended SVD++ models”, International Journal of Computers and Applications, pp. 349-356, May 2020.

[7] [7] B. Zhang, X. Zhou, J. Li, L. Li, “Recommendation Algorithm Based on Matrix SVD with Exponential Correction”, CIPAE 2020, pp. 71-75, October 2020.

[8] [8]https://www.kaggle.com/datasets/parasharmanas/movie-recommendation-system, (pristupljeno u oktobru 2025.)

[9] [9] S. Jiang, J. Li, W. Zhou, “An Application of SVD++ Method in Collaborative Filtering”, IEEE, pp. 192-197, January 2021.

[10] [10] https://www.datacamp.com/tutorial/rmse, (pristupljeno u oktobru 2025.)