Recommender systems aim to provide personal and satisfaction-oriented recommendations based on users’ preferences. However, although efforts are made to act in line with the goal, the problem of popularity bias is encountered in recommendations. This study addresses the use of recommendation systems in the field of personality prediction and aims to examine the popularity bias based on the personality traits of individuals (Openness, Agreeableness, Emotionality, Conscientiousness and Extraversion). In addition, the effectiveness of a model developed to evaluate the accuracy of predictions made according to these personality dimensions was analyzed in the study. Ranking-based (nDCG, GAP) and coverage-based (APLT, LTC) metrics were used in the evaluations. The obtained results show that the model has the highest ranking success in the Openness dimension, while the negative GAP values for all dimensions indicate significant deficiencies in ranking accuracy. 

Keywords - Recommender systems, Popularity bias, User tendency

Öneri sistemleri, kullanıcıların tercihlerine dayalı olarak kişisel ve memnuniyetin ön planda olduğu öneriler sunmayı amaçlamaktadır. Ancak her ne kadar amaç doğrultusunda hareket edilmeye çalışılsa da, önerilerde popüler yanlılığı problemi ile karşılaşılmaktadır. Bu çalışma, öneri sistemlerinin kişilik tahmini alanındaki kullanımını ele almakta ve bireylerin kişilik özellikleri (Açıklık, Uyumluluk, Duygusallık, Sorumluluk ve Dışadönüklük) temelinde popülerlik yanlılığını incelemeyi amaçlamaktadır. Çalışmada ayrıca, bu kişilik boyutlarına göre yapılan tahminlerin doğruluğunu değerlendirmek amacıyla geliştirilen bir modelin etkinliği analiz edilmiştir. Değerlendirmelerde sıralama tabanlı (nDCG, GAP) ve kapsama tabanlı (APLT, LTC) metrikler kullanılmıştır. Elde edilen sonuçlar, Açıklık boyutunda modelin sıralama başarısının en yüksek olduğunu gösterirken, GAP değerlerinin tüm boyutlar için negatif olması sıralama doğruluğunda önemli eksikliklere işaret etmektedir. 

Keywords^Turkish - Öneri sistemleri, Popüler yanlılığı, Kullanıcı eğilimi

[1] G. Adomavicius and A. Tuzhilin, Toward the next generation of recommender systems: A survey of the state-of-the-art and possible extensions, IEEE Transactions on Knowledge and Data Engineering, pp. 734-749, 2005.

[2] J. S. Breese, D. Heckerman, and C. Kadie, "Empirical analysis of predictive algorithms for collaborative filtering," arXiv preprint arXiv:1301.7363, 2013.

[3] H. Abdollahpouri, "Popularity Bias in Ranking and Recommendation," in Proc. AAAI/ACM Conf. on AI, Ethics, and Society (AIES’19), Honolulu, HI, USA, Jan. 2019.

[4] E. Brynjolfsson, Y. J. Hu, and M. D. Smith, "From niches to riches: Anatomy of the long tail," Sloan Management Review, vol. 47, no. 4, pp. 67–71, 2006.

[5] T. T. Nguyen, P.-M. Hui, F. M. Harper, L. Terveen, and J. A. Konstan, "Exploring the filter bubble: the effect of using recommender systems on content diversity," in Proc. 23rd Int. Conf. on World Wide Web, ACM, pp. 677–686, 2014.

[6] P. Resnick, R. K. Garrett, T. Kriplean, S. A. Munson, and N. J. Stroud, "Bursting your (filter) bubble: strategies for promoting diverse exposure," in Proc. 2013 Conf. on Computer Supported Cooperative Work Companion, ACM, pp. 95–100, 2013.

[7] Ò. Celma and P. Cano, "From hits to niches?: or how popular artists can bias music recommendation and discovery," in Proc. 2nd KDD Workshop on Large-Scale Recommender Systems and the Netflix Prize Competition, ACM, p. 5, 2008.

[8] S. M. Lundberg and S. Lee, "A unified approach to interpreting model predictions," in Proc. 31st Int. Conf. on Neural Information Processing Systems, pp. 4768–4777, 2017.

[9] “L. S. Shapley, "A value for n-person games," RAND Corporation, Paper P-295, 1953.

[10] R. Sanders, "The Pareto principle: its use and abuse," Journal of Services Marketing, vol. 1, no. 2, pp. 37–40, 1987.

[11] T. Chamorro-Premuzic, Personality and Individual Differences, John Wiley & Sons, 2016.

[12] E. Yalcin and A. Bilge, "Evaluating unfairness of popularity bias in recommender systems: A comprehensive user-centric analysis," Information Processing & Management, vol. 59, no. 6, 2022, doi: https://doi.org/10.1016/j.ipm.2022.103100.

[13] H. Abdollahpouri, M. Mansoury, R. Burke, and B. Mobasher, "The unfairness of popularity bias in recommendation," arXiv preprint arXiv:1907.13286, 2019.

[14] H. Abdollahpouri, "Popularity Bias in Recommendation: A Multi-stakeholder Perspective," arXiv preprint arXiv:2008.08551, 2020.

[15] H. Abdollahpouri, M. Mansoury, R. Burke, and B. Mobasher, "The connection between popularity bias, calibration, and fairness in recommendation," in Proc. 14th ACM Conf. on Recommender Systems (RecSys'20), pp. 726–731, 2020.