Problem
Research questions and friction points this paper is trying to address.
Table-to-Report Generation
Exploration Bias
Verifiable Composite Insights
Evidence Chains
Automated Data Science
Structured Composition of Verifiable Atomic Insights for Table-to-Report Generation
🤖 AI Summary
This study addresses the issues of missing cross-table evidence and unverifiable facts in table-to-report generation caused by exploration bias. To this end, we propose ComInsight, a novel framework that introduces the concept of "atomic insights" to reformulate complex insight discovery as the composition of atomic evidence. By constructing multi-relational graphs and incorporating logical combination operators to synthesize higher-order conclusions, ComInsight integrates SQL execution-based verification to enable fully traceable and verifiable report generation. Extensive experiments across three benchmarks demonstrate that the proposed method significantly outperforms existing baselines in factual correctness, novelty, and structural completeness.
📝 Abstract
Table-to-report generation refers to the task of automatically generating article-level analyt- ical reports from relational tables and is an essential capability for automated data science and decision support. Its central challenge lies in systematically discovering verifiable com- posite insights across tables, attributes, and analytical perspectives, and organizing them into coherent, complete, and traceable evidence chains. Existing methods primarily rely on sequential, reactive data agents or direct Large Language Model(LLM) generation. They suffer from exploration bias: early local observations constrain subsequent actions, causing models to focus prematurely on local analyzes and miss cross-table or cross-dimensional evidence. We propose ComInsight, which reformulates insight discovery as the composition of atomic evidences. We first define an atomic insight as the smallest executable analytical unit conforming to a predefined analysis pattern and enumerate all valid atomic insights from database schema and content. These atoms are then organized into a multi-relational insight graph, where nodes represent verified data facts and edges encode logical, temporal, or hierarchical relations. Finally, a set of composition operators systematically fuses atomic nodes into higher-order composite conclusions. Every composite output is accompanied by executable SQL and fine-grained provenance, ensuring full verifiability. Across three benchmarks InsightBench, DDR-Bench, and T2R-Bench, ComInsight consistently outperforms strong baselines in factual correctness, novelty, and structural completeness. We believe ComInsight offers a reliable, efficient, and explainable path toward table-to-report generation.
Innovation
Methods, ideas, or system contributions that make the work stand out.
Table-to-Report Generation
Atomic Insight Composition
Multi-relational Insight Graph
Verifiable Evidence Chain
Exploration Bias Mitigation
🔎 Similar Papers
No similar papers found.
