可审计性原生：面向企业金融领域可信大语言模型分析的本体驱动框架
Auditable by Construction: An Ontology-Driven Framework for Trustworthy LLM Analytics in Enterprise Finance
浏览论文内容
中文总结 AI 辅助
本文针对企业金融大语言模型应用的信任问题，提出知识驱动分析框架KDAF，结合本体与CARP技术，在FinanceBench评估中，其可审计性优于基线方法，表明可审计性是该框架的核心优势。
中文摘要 AI 辅助
企业在金融领域采用大语言模型的制约因素较少在于流畅度，而更多在于信任：在财务规划与分析（FP&A）及其他受监管的工作流程中，只有当答案可追溯至权威来源且事后可审计时，才具备可用性。本文提出，企业金融领域的检索增强生成模型应在准确性之外，还需以可审计性作为评估标准，并提出知识驱动分析框架（KDAF），该框架通过六个迭代阶段构建本体驱动的知识系统，并通过上下文感知相关性传播（CARP）检索证据，确保每条检索到的事实都携带其关系类型、置信度及来源谱系。在FinanceBench（145个问题）上开展的评估，将KDAF与零上下文推理、BM25、概念加权词汇检索及无依据图遍历进行对比。结果显示：第一，检索是必要的：零上下文推理的正确率为4.1%，而检索增强条件下的正确率为10%-12%；第二，在答案正确性方面，各检索条件的结果无统计学差异（KDAF与BM25的差值为-0.007，95%置信区间为[-0.021, 0.000]），因此仅准确性不足以成为采用结构化检索的理由——本文明确报告了这一负面结果；第三，在可审计性方面，排序发生反转：KDAF的引用可追溯性F1值最高，为0.515，较无依据遍历高出0.027（置信区间[0.006, 0.050]），较BM25高出0.052（置信区间[0.024, 0.083]），区间不包含零。图结构化检索还不会引入问题主题实体之外的证据（426个条目里0个，而词汇基线方法的该比例为16.8%和20.2%），且每个选定条目都能形成完整的溯源链。本文认为，本体驱动检索的成本应基于可审计性而非准确性来衡量。
英文摘要
Enterprise adoption of large language models in finance is constrained less by fluency than by trust: in Financial Planning and Analysis (FP&A) and other regulated workflows, an answer is usable only if it is traceable to authoritative sources and auditable after the fact. This paper argues that retrieval-augmented generation for enterprise finance should be evaluated on auditability alongside accuracy, and presents the Knowledge-Driven Analytics Framework (KDAF), which builds ontology-driven knowledge systems through six iterative stages and retrieves evidence via Context-Aware Relevance Propagation (CARP), so that every retrieved fact carries its relationship type, confidence, and source lineage. An evaluation on FinanceBench (145 questions) compares KDAF against zero-context inference, BM25, concept-weighted lexical retrieval, and ungrounded graph traversal. First, retrieval is necessary: zero-context inference reaches 4.1% correctness against 10-12% for retrieval-augmented conditions. Second, on answer correctness the retrieval conditions are statistically indistinguishable (KDAF vs BM25: -0.007, 95% CI [-0.021, 0.000]), so accuracy alone does not justify structured retrieval here -- a negative result we report explicitly. Third, on auditability the ordering reverses: KDAF attains the highest citation traceability F1 (0.515), exceeding ungrounded traversal by +0.027 (CI [0.006, 0.050]) and BM25 by +0.052 (CI [0.024, 0.083]), intervals excluding zero. Graph-structured retrieval also admits no evidence from outside the question subject entity (0 of 426 items, against 16.8% and 20.2% for lexical baselines), and every selected item resolves to a complete provenance chain. We argue that auditability, not accuracy, is the axis on which ontology-grounded retrieval earns its cost.
