Monte Carlo Tree Search for Table-to-Multimodal Report Generation

基于蒙特卡洛树搜索的表格到多模态报告生成

Abstract: Automatically generating professional multimodal reports comprising both textual analysis and visual charts from structured tabular data is a critical challenge in data intelligence. Existing methods suffer from fixed linear pipelines and isolated subtask processing, which hinder joint optimization of factual accuracy, visual quality, and narrative coherence.

摘要： 从结构化表格数据中自动生成包含文本分析和可视化图表的专业多模态报告，是数据智能领域的一项关键挑战。现有的方法受限于固定的线性流水线和孤立的子任务处理，这阻碍了对事实准确性、视觉质量和叙事连贯性的联合优化。

To address these issues, this paper proposes MCTS-Report, a Monte Carlo Tree Search (MCTS)-driven framework that formulates multimodal table-to-report generation as a progressive construction process over a structured search space. The core idea is to decompose report generation into atomic actions, including chapter planning, visualization task identification, chart generation, insight organization, and narrative refinement, each executed by an LLM based on dynamic reasoning conditioned on the current report state.

为了解决这些问题，本文提出了 MCTS-Report，这是一个由蒙特卡洛树搜索（MCTS）驱动的框架，它将多模态表格到报告的生成过程建模为结构化搜索空间中的渐进式构建过程。其核心思想是将报告生成分解为原子操作，包括章节规划、可视化任务识别、图表生成、洞察组织和叙事润色，每一项操作都由大语言模型（LLM）根据当前报告状态的动态推理来执行。

We use an LLM to generate step-by-step reasoning and actions during MCTS, storing the reasoning trajectory in each node for context-aware, coherent report construction. To guide the search, we design a multi-dimensional reward function that jointly evaluates numerical fact consistency (via SQL), chart quality, chart-text alignment, and structural completeness, while incorporating a diversity penalty to suppress repeated charts and a precondition check to prune invalid actions.

我们利用大语言模型在 MCTS 过程中生成逐步推理和操作，并将推理轨迹存储在每个节点中，以实现上下文感知且连贯的报告构建。为了引导搜索，我们设计了一个多维奖励函数，联合评估数值事实一致性（通过 SQL）、图表质量、图文对齐度和结构完整性，同时引入多样性惩罚以抑制重复图表，并使用前提条件检查来剪枝无效操作。

We also construct MMRBench, a comprehensive benchmark comprising real-world tables from six domains, paired with expert-refined reference report structures and verifiable key insights. Experiments on MMRBench demonstrate that MCTS-Report significantly outperforms strong baselines across structural completeness, numerical accuracy, chart-text alignment, and insight novelty, achieving a 77.9 overall score.

我们还构建了 MMRBench，这是一个包含来自六个领域的真实世界表格的综合基准测试集，并配有经专家精炼的参考报告结构和可验证的关键洞察。在 MMRBench 上的实验表明，MCTS-Report 在结构完整性、数值准确性、图文对齐度和洞察新颖性方面显著优于强基线模型，总分达到 77.9 分。