用案例互辩机制提升案件推理准确率
Supported Abstract Argumentation for Case-Based Reasoning
- 案例间通过支持或反驳进行二元辩论，动态决定分类标签
- 新模型消除旧版中无关案例干扰，保持推理纯净性
- 适合需要可解释性推理的法律与医疗决策场景
我们提出支持型抽象论证的案例推理（sAA-CBR），一种二分类模型，其中过往案例通过支持自身标签、攻击或支持对立标签来展开辩论。引入支持机制后，sAA-CBR克服了其前身AA-CBR存在的冗余案例（即‘尖峰’）问题，这些尖峰在原模型中不参与辩论却影响结果。我们证明，sAA-CBR在不牺牲关键模型性质的前提下，完全消除了此类尖峰现象。
原文摘要 · Abstract (English)
We introduce Supported Abstract Argumentation for Case-Based Reasoning (sAA-CBR), a binary classification model in which past cases engage in debates by arguing in favour of their labelling and attacking or supporting those with opposing or agreeing labels. With supports, sAA-CBR overcomes the limitation of its precursor AA-CBR, which can contain extraneous cases (or spikes) that are not included in the debates. We prove that sAA-CBR contains no spikes, without trading off key model properties
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。