Yazar "Bilge, Alper" seçeneğine göre listele

Listeleniyor 1 - 4 / 4
  • [ X ]
    Öğe
    A Novel Pre-Processing Technique to Combat Popularity Bias in Personality-Aware Recommender Systems
    (Ieee-Inst Electrical Electronics Engineers Inc, 2024) Waris, Madiha; Zaman Fakhar, Muhammad; Gulsoy, Mert; Yalcin, Emre; Bilge, Alper
    Recommender systems aid users in discovering items of interest across various domains. However, these systems often suffer from popularity bias, disproportionately recommending popular items and neglecting less popular ones that may still appeal to users. We propose an efficient pre-processing technique to mitigate popularity bias in personality-aware recommender systems, which leverage users' personality traits to enhance personalization and deliver higher-quality recommendations. Our method infuses synthetic ratings into less popular items, increasing their visibility in the recommendation process. We evaluate this technique using both accuracy and beyond-accuracy metrics but recognize that these metrics alone do not fully capture a recommender system's performance. To provide a holistic evaluation, we introduce the General Performance Indicator-a comprehensive metric combining accuracy and beyond-accuracy measures. Experimental results on two publicly available real-world datasets demonstrate that while our approach may cause minor decreases in accuracy, it significantly improves the beyond-accuracy aspects of recommendation quality. These enhancements underscore the effectiveness of our method in delivering a more balanced and diverse set of recommendations, addressing the limitations of traditional accuracy-focused approaches.
  • [ X ]
    Öğe
    DUoR: Dynamic User-oriented re-Ranking calibration strategy for popularity bias treatment of recommendation
    (Academic Press Ltd- Elsevier Science Ltd, 2025) Gulsoy, Mert; Yalcin, Emre; Tacli, Yucel; Bilge, Alper
    Recommender systems are widely used to provide personalized recommendations to users to help them navigate the vast amount of available content. They have become pervasive in various online applications. However, they often suffer from popularity bias, where popular items receive more recommendations, leading to potential issues such as limited diversity, homogenized user experience, perpetuating existing inequalities, and filter bubble effects. In this paper, we propose a novel approach to mitigate popularity bias by incorporating users' inclination towards item popularity. The proposed method incorporates a practical popularity inclination measuring strategy considering the dynamic preference tendencies of individuals to capture their unique propensities towards item popularity better and to provide more calibrated referrals regarding expectations of individuals on item popularity. Experimental results on benchmark datasets demonstrate that our proposed method effectively mitigates popularity bias by generating more diverse and balanced recommendations compared to several benchmark post-processing methods and outperforming them in diversity and fairness metrics according to the Borda count system. Overall, the proposed method presents a promising approach to addressing popularity bias in recommender systems by incorporating users' inclination towards item popularity and opens up potential directions for further research in the field.
  • [ X ]
    Öğe
    EquiRate: balanced rating injection approach for popularity bias mitigation in recommender systems
    (Peerj Inc, 2025) Gulsoy, Mert; Yalcin, Emre; Bilge, Alper
    Recommender systems often suffer from popularity bias problem, favoring popular items and overshadowing less known or niche content, which limits recommendation diversity and content exposure. The root reason for this issue is the imbalances in the rating distribution; a few popular items receive a disproportionately large share of interactions, while the vast majority garner relatively few. In this study, we propose the EquiRate method as a pre-processing approach, addressing this problem by injecting synthetic ratings into less popular items to make the dataset regarding rating distribution more balanced. More specifically, this method utilizes several synthetic rating injection and synthetic rating generation strategies: (i) the first ones focus on determining which items to inject synthetic ratings into and calculating the total number of these ratings, while (ii) the second ones concentrate on computing the concrete values of the ratings to be included. We also introduce a holistic and highly efficient evaluation metric, i.e., the FusionIndex, concurrently measuring accuracy and several beyond-accuracy aspects of recommendations. The experiments realized on three benchmark datasets conclude that several EquiRate's variants, with proper parameter-tuning, effectively reduce popularity bias and enhance recommendation diversity. We also observe that some prominent popularity-debiasing methods, when assessed using the FusionIndex, often fail to balance the referrals' accuracy and beyond-accuracy factors. On the other hand, our best-performing EquiRate variants significantly outperform the existing methods regarding the FusionIndex, and their superiority is more apparent for the high-dimension data collections, which are more realistic for real-world scenarios.
  • [ X ]
    Öğe
    Robustness of privacy-preserving collaborative recommenders against popularity bias problem
    (Peerj Inc, 2023) Gulsoy, Mert; Yalcin, Emre; Bilge, Alper
    Recommender systems have become increasingly important in today's digital age, but they are not without their challenges. One of the most significant challenges is that users are not always willing to share their preferences due to privacy concerns, yet they still require decent recommendations. Privacy-preserving collaborative recommenders remedy such concerns by letting users set their privacy preferences before submitting to the recommendation provider. Another recently discussed challenge is the problem of popularity bias, where the system tends to recommend popular items more often than less popular ones, limiting the diversity of recommendations and preventing users from discovering new and interesting items. In this article, we comprehensively analyze the randomized perturbation-based data disguising procedure of privacy-preserving collaborative recommender algorithms against the popularity bias problem. For this purpose, we construct user personas of varying privacy protection levels and scrutinize the performance of ten recommendation algorithms on these user personas regarding the accuracy and beyond-accuracy perspectives. We also investigate how well-known popularity-debiasing strategies combat the issue in privacy-preserving environments. In experiments, we employ three well-known real-world datasets. The key findings of our analysis reveal that privacy-sensitive users receive unbiased and fairer recommendations that are qualified in diversity, novelty, and catalogue coverage perspectives in exchange for tolerable sacrifice from accuracy. Also, prominent popularity-debiasing strategies fall considerably short as provided privacy level improves.