CREDENCE: 面向分解与增强可信度的声明缩减——语义度量与收敛性分析
CREDENCE: Claim Reduction for Decomposition & Enhanced Credibility -- Semantic Metrics and Convergence Analysis
AI总结：
提出CREDENCE框架，通过语义F1度量解决Jaccard度量对释义声明的低估问题，并形式化分析修复管道的收敛性，实验表明语义F1比Jaccard F1提升15-32个百分点，规则修复将原子性违反率降低47-100%。
摘要：
将复合句分解为原子化的、可验证的声明是可靠自动化事实核查的前提。先前工作依赖基于词重叠（Jaccard）的度量，系统性地低估了释义声明的分解质量，并且缺乏对修复循环的形式化终止分析。我们提出CREDENCE，一个改进的声明分解与评估框架，解决了这两个缺陷。我们的贡献包括：(1) 语义F1：我们使用BGE-large余弦相似度保真度度量，解决了Jaccard的惩罚问题，并提高了下游事实核查的准确性；(2) 收敛定理：我们形式化地表征了修复管道的四个性质，确立了在预言解析器假设下基于规则的修复是单调且有限终止的；基于LLM的自修复被证明是非单调的，需要早期退出保护；(3) 三个评估基准，涵盖社交媒体、百科全书和新闻领域，用于跨领域泛化度量；(4) 跨四个分解器模型（3.8B-12B）和一个封闭API模型的多模型基准测试。在SocialClaimSplit、WikiSplitBench和ClaimDecompBench上的实验表明，语义F1比Jaccard F1提升15-32个百分点。在SocialClaimSplit和WikiSplitBench上，EPR范围为0.94至1.00，而ClaimDecompBench由于更难的新闻领域构造，包含较低的基线EPR情况（低至0.824），规则修复相对于基线模型将原子性违反率（AVR）降低了47-100%，且不降低保真度。
英文摘要：
Decomposing compound sentences into atomic, verifiable claims is a prerequisite for reliable automated fact-checking. Prior work has relied on token-overlap (Jaccard) metrics that systematically underestimate decomposition quality for paraphrastic claims, and has lacked formal termination analysis for the repair loop. We present Credence, a revised claim decomposition and evaluation framework addressing both shortcomings. Our contributions are: (1) Semantic-F1: we use BGE-large cosine similarity fidelity metric that resolves Jaccard's penalisation and improves downstream fact-checking accuracy; (2) Convergence theorems: we formally characterise four properties of the repair pipeline, establishing that rule-based repair is monotone and finitely terminating under an oracle parser assumption; LLM-based self-repair is provably non-monotone and requires an early-exit guard; (3) Three evaluation benchmarks spanning social-media, encyclopaedic, and news domains for cross-domain generalisation measurement; (4) Multi-model benchmarking across four decomposer models (3.8B-12B) and a closed API model. Experiments on SocialClaimSplit, WikiSplitBench, and ClaimDecompBench show that Semantic-F1 outperforms Jaccard-F1 by +15-32pp. EPR ranges from 0.94 to 1.00 on SocialClaimSplit and WikiSplitBench, while ClaimDecompBench includes lower base EPR cases (down to 0.824) due to harder news-domain constructions, and rule-repair reduces the Atomicity Violation Rate (AVR) by 47-100% relative to the base model without degrading fidelity.
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。
