{"ID":2889247,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2507.21884","arxiv_id":"2507.21884","title":"Exploration on Demand: From Algorithmic Control to User Empowerment","abstract":"Recommender systems often struggle with over-specialization, which severely limits users' exposure to diverse content and creates filter bubbles that reduce serendipitous discovery. To address this fundamental limitation, this paper introduces an adaptive clustering framework with user-controlled exploration that effectively balances personalization and diversity in movie recommendations. Our approach leverages sentence-transformer embeddings to group items into semantically coherent clusters through an online algorithm with dynamic thresholding, thereby creating a structured representation of the content space. Building upon this clustering foundation, we propose a novel exploration mechanism that empowers users to control recommendation diversity by strategically sampling from less-engaged clusters, thus expanding their content horizons while explicitly exposing the relevance-diversity trade-off. Experiments on the MovieLens dataset demonstrate the system's effectiveness, showing that exploration significantly reduces intra-list similarity from 0.34 to 0.26 while simultaneously increasing unexpectedness to 0.73. Furthermore, our Large Language Model-based A/B testing methodology, conducted with 300 simulated users, reveals that 72.7% of long-term users prefer exploratory recommendations over purely exploitative ones. Additional relevance metrics, including NDCG@k, Recall@k, and HitRate@k, reveal the expected relevance-diversity trade-off against CF and MMR baselines, positioning the method as a controllable exploration layer for promoting meaningful content discovery.","short_abstract":"Recommender systems often struggle with over-specialization, which severely limits users' exposure to diverse content and creates filter bubbles that reduce serendipitous discovery. To address this fundamental limitation, this paper introduces an adaptive clustering framework with user-controlled exploration that effec...","url_abs":"https://arxiv.org/abs/2507.21884","url_pdf":"https://arxiv.org/pdf/2507.21884v2","authors":"[\"Edoardo Bianchi\"]","published":"2025-07-29T14:57:26Z","proceeding":"cs.IR","tasks":"[\"cs.IR\"]","methods":"[\"Transformer\",\"Language Model\",\"LoRA\"]","has_code":false}