arXiv cs.CL· Furkan Yasir Goksu, Yaren Yilmaz, Sule Gunduz Oguducu·· 12 小时前AI 评分23
微调数据预处理选择影响 LLM 推荐效果：基于 MovieLens-1M 的效用与人口敏感性研究
Preprocessing Choices in Fine-Tuning Data Matter: Utility and Demographic Sensitivity in LLM-Based Recommendation
AI 导读
研究在 MovieLens-1M 上考察微调数据预处理对生成式 LLM 推荐系统的影响，采用监督微调、目录约束解码与提示词增强构建受控流程，目录约束解码可消除目录外条目并保证完整输出十条结果。
正文 · 原文
Abstract:Large language models (LLMs) have introduced a generative approach to recommendation, producing item lists from textual representations of user histories and item information. However, this generative setting raises challenges concerning output validity, prompt consistency, and sensitivity to demographic descriptors. In this work, we investigate how preprocessing choices in fine-tuning data affect recommendation utility and fairness-related behavior in a generative LLM-based recommender system. We first establish a controlled recommendation pipeline that combines supervised fine-tuning, catalog-constrained decoding, and prompt augmentation. We then study three preprocessing choices: item-title representation, demographic balancing, and auxiliary item-knowledge enrichment. Experiments on MovieLens-1M show that the catalog-constrained pipeline eliminates out-of-catalog items and ensures complete ten-item outputs, while prompt augmentation improves prompt consistency and utility. Original MovieLens titles increase HR@10 from 10.67% to 12.47% and Coverage@10 from 41.36% to 47.88%. Fairness-related effects are attribute-dependent: age balancing and knowledge enrichment reduce selected group-level utility gaps while largely preserving utility, whereas gender balancing reduces utility without improving the evaluated gender-related indicators. Overall, preprocessing choices in fine-tuning data should be evaluated jointly in terms of utility, output behavior, and demographic sensitivity.
| Comments: | 11 pages, 2 figures, 5 tables. Accepted at the 3rd Workshop on Evaluating and Applying Recommender Systems with Large Language Models (EARL) at CIKM 2026, November 8, 2026, Rome, Italy |
| Subjects: | Information Retrieval (cs.IR); Computation and Language (cs.CL); Machine Learning (cs.LG) |
| ACM classes: | H.3.3; I.2.7 |
| Cite as: | arXiv:2611.03113 [cs.IR] |
| (or arXiv:2611.03113v1 [cs.IR] for this version) | |
| https://doi.org/10.48550/arXiv.2611.03113 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Yaren Yilmaz [view email]
[v1]
Mon, 2 Nov 2026 13:32:28 UTC (54 KB)
来源：arXiv cs.CL · arxiv.org