EvidFuse: Writing-Time Evidence Learning for Consistent Text-Chart Data Reporting

2026-01-09

arXiv.org

AI Summary

This work addresses the limitations of existing large language models in generating data reports, which typically rely on staged pipelines that result in misaligned text and visuals and inflexible analytical perspectives, hindering dynamic evidence construction. To overcome these issues, we propose EvidFuse—a training-free multi-agent framework that enables an analyst agent and a writer agent to collaboratively interweave text and visualization generation during the writing process, dynamically constructing evidence on demand to support narrative coherence. EvidFuse introduces the novel concept of “evidence learning during writing,” breaking free from fixed evidence spaces to achieve fine-grained, dynamic text-visual co-generation. By integrating exploratory data analysis knowledge, direct access to raw tabular data, and real-time evidence synthesis, EvidFuse decouples long-form drafting from visual analytics, significantly improving chart quality, text-visual alignment, and report utility in both automatic and human evaluations.

Technology Category

• Data Mining & Knowledge Management: Mining of Visual, Multimedia & Multimodal Data
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Web Mining and Content Analysis: Web data visualization
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Data-driven reports communicate decision-relevant insights by tightly interleaving narrative text with charts grounded in underlying tables. However, current LLM-based systems typically generate narratives and visualizations in staged pipelines, following either a text-first-graph-second or a graph-first-text-second paradigm. These designs often lead to chart-text inconsistency and insight freezing, where the intermediate evidence space becomes fixed and the model can no longer retrieve or construct new visual evidence as the narrative evolves, resulting in shallow and predefined analysis. To address the limitations, we propose \textbf{EvidFuse}, a training-free multi-agent framework that enables writing-time text-chart interleaved generation for data-driven reports. EvidFuse decouples visualization analysis from long-form drafting via two collaborating components: a \textbf{Data-Augmented Analysis Agent}, equipped with Exploratory Data Analysis (EDA)-derived knowledge and access to raw tables, and a \textbf{Real-Time Evidence Construction Writer} that plans an outline and drafts the report while intermittently issuing fine-grained analysis requests. This design allows visual evidence to be constructed and incorporated exactly when the narrative requires it, directly constraining subsequent claims and enabling on-demand expansion of the evidence space. Experiments demonstrate that EvidFuse attains the top rank in both LLM-as-a-judge and human evaluations on chart quality, chart-text alignment, and report-level usefulness.

Problem

• text-chart inconsistency
• insight freezing
• data-driven reporting
• evidence space
• LLM-based generation

Innovation

• EvidFuse
• text-chart alignment
• real-time evidence construction
• multi-agent framework
• data-driven reporting