迈向更公平的健康建议：通过词义消歧寻找信息丰富的无偏样本
Towards Fairer Health Recommendations: finding informative unbiased samples via Word Sense Disambiguation
AI总结：
为解决医疗数据偏差导致健康推荐不公的问题，本研究借助词义消歧优化负样本数据集质量，实验发现微调BERT模型在医疗文本偏差检测上表现优于零/少样本LLM。
摘要：
人们对依赖有偏数据训练的模型的高风险应用日益担忧，这类模型会产生有偏预测，往往会伤害最脆弱的群体。具体而言，有偏的医疗数据可能导致健康相关应用和推荐系统生成的输出危及患者护理，并扩大健康结果的差异。近期名为Fairness via AI（借AI实现公平）的框架提出，研究人员不应试图修正模型偏差，而应聚焦偏差的根源，利用AI对数据去偏。受该框架启发，我们使用包括大语言模型（LLM）在内的NLP模型处理医疗课程中的偏差检测任务，并在一个金标准数据集上评估这些模型：该数据集包含从大型语料库中抽取、经医学专家标注偏差的4105条摘录。我们在合著者此前的研究基础上展开工作，该研究用包含社会身份标识术语的未标注文本扩充了负样本集。但这些术语中的一部分，尤其是与种族和族裔相关的术语，可能具有不同含义（例如“脊髓白质（white matter of spinal cord）”中的white）。为解决这一问题，我们提出使用词义消歧（Word Sense Disambiguation）模型，通过移除无关句子优化数据集质量。随后我们评估了微调后的各类BERT模型，以及采用零样本和少样本提示的GPT模型。我们发现，虽在许多NLP任务上被视为当前最优（SOTA），但LLM并不适用于偏差检测，而微调后的BERT模型在所有评估指标上普遍表现良好。
英文摘要：
There have been growing concerns around high-stake applications that rely on models trained with biased data, which consequently produce biased predictions, often harming the most vulnerable. In particular, biased medical data could cause health-related applications and recommender systems to create outputs that jeopardize patient care and widen disparities in health outcomes. A recent framework titled Fairness via AI posits that, instead of attempting to correct model biases, researchers must focus on their root causes by using AI to debias data. Inspired by this framework, we tackle bias detection in medical curricula using NLP models, including LLMs, and evaluate them on a gold standard dataset containing 4,105 excerpts annotated by medical experts for bias from a large corpus. We build on previous work by coauthors which augments the set of negative samples with non-annotated text containing social identifier terms. However, some of these terms, especially those related to race and ethnicity, can carry different meanings (e.g., "white matter of spinal cord"). To address this issue, we propose the use of Word Sense Disambiguation models to refine dataset quality by removing irrelevant sentences. We then evaluate fine-tuned variations of BERT models as well as GPT models with zero- and few-shot prompting. We found LLMs, considered SOTA on many NLP tasks, unsuitable for bias detection, while fine-tuned BERT models generally perform well across all evaluated metrics.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
