PRISM-BN：面向文本到参数化贝叶斯网络抽取的可控语料库与基准

中文摘要 AI 辅助

概率图模型（PGMs），尤其是贝叶斯网络（BNs），揭示了有向结构和概率参数，使其成为神经符号AI的自然符号目标。然而，训练文本到参数化贝叶斯网络系统需要成对的文本到BN资源，而这些资源无法大规模获得。我们引入了PRISM-BN，一个包含5054个基于BN的描述的可控语料库，这些描述与包含变量、状态、有向边、根先验和跨五个领域的完整多父条件概率表（CPDs）的离散参考BN配对。这些实例源自50个以维基百科为种子的骨干网络，其概率是内部构建的基准目标，而非外部验证的因果估计。PRISM-BN使用PRISM构建，这是一种边际优先的流水线，它引出边际和局部联合分布，解析地恢复归一化CPDs，并构建局部重新参数化的子图。我们定义了一个基准，包括语义节点和状态对齐、条件结构评分以及严格的完整CPD评估。在六个LLM提取器中，节点F1分数范围从0.56到0.83，条件边F1分数从0.90到0.97，CPD-KL从1.11到3.14。条件状态和边恢复始终表现强劲，而严格的完整CPD一致性仍然具有挑战性。这些趋势在独立生成的GPT-5.5参考中持续存在，人工试点研究证实了结构的可恢复性和相似的概率解释。PRISM-BN支持对结构恢复和概率参数估计进行独立评估。

英文摘要

Probabilistic Graphical Models (PGMs), especially Bayesian Networks (BNs), expose directed structure and probabilistic parameters, making them natural symbolic targets for neurosymbolic AI. Yet training text-to-parameterized-BN systems requires paired text-to-BN resources unavailable at scale. We introduce PRISM-BN, a controlled corpus of 5054 BN-grounded descriptions paired with discrete reference BNs containing variables, states, directed edges, root priors, and full multi-parent CPDs across five domains. The instances are derived from 50 Wikipedia-seeded backbones, and their probabilities are internally constructed benchmark targets rather than externally validated causal estimates. PRISM-BN is built with PRISM, a marginal-first pipeline that elicits marginal and local joint distributions, analytically recovers normalized CPDs, and constructs locally reparameterized subgraphs. We define a benchmark with semantic node and state alignment, conditional structural scoring, and strict full-CPD evaluation. Across six LLM extractors, Node F1 ranges from 0.56 to 0.83, conditional Edge F1 from 0.90 to 0.97, and CPD-KL from 1.11 to 3.14. Conditional state and edge recovery remain consistently strong, whereas strict full-CPD agreement remains challenging. These trends persist with independently generated GPT-5.5 references, and a human pilot corroborates structural recoverability and similar probabilistic interpretations. PRISM-BN supports separate evaluation of structural recovery and probabilistic parameter estimation.

发表机构

• Dartmouth College（达特茅斯学院）
• RAND Corporation（兰德公司）

机构由 AI 辅助整理，请以论文原文为准。