Beyond Linearization: Attributed Table Graphs for Table Reasoning

2026-01-13

AI Summary

This work addresses key limitations in existing large language model–based tabular reasoning approaches, which suffer from structural information loss due to table linearization, lack interpretable reasoning pathways, and are hindered by the “intermediate information loss” problem. To overcome these challenges, the authors propose Table Graph Reasoner (TABGR), a training-free method that introduces the Attributed Table Graph (ATG) to explicitly model row–column structures and employs a question-guided personalized PageRank (QG-PPR) to re-rank salient content for graph-based reasoning. By integrating graph neural networks with prompt engineering, TABGR achieves up to a 9.7% accuracy improvement over prior methods on two mainstream benchmarks, substantially mitigating information loss while enhancing both reasoning interpretability and robustness to noise.

Technology Category

• Knowledge Representation and Reasoning: Computational Complexity of Reasoning
• Reasoning under Uncertainty: Graphical Models
• Machine Learning: Graph-based Machine Learning

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Table reasoning, a task to answer questions by reasoning over data presented in tables, is an important topic due to the prevalence of knowledge stored in tabular formats. Recent solutions use Large Language Models (LLMs), exploiting the semantic understanding and reasoning capabilities of LLMs. A common paradigm of such solutions linearizes tables to form plain texts that are served as input to LLMs. This paradigm has critical issues. It loses table structures, lacks explicit reasoning paths for result explainability, and is subject to the"lost-in-the-middle"issue. To address these issues, we propose Table Graph Reasoner (TABGR), a training-free model that represents tables as an Attributed Table Graph (ATG). The ATG explicitly preserves row-column-cell structures while enabling graph-based reasoning for explainability. We further propose a Question-Guided Personalized PageRank (QG-PPR) mechanism to rerank tabular data and mitigate the lost-in-the-middle issue. Extensive experiments on two commonly used benchmarks show that TABGR consistently outperforms state-of-the-art models by up to 9.7% in accuracy. Our code will be made publicly available upon publication.

Problem

• table reasoning
• linearization
• structure preservation
• explainability
• lost-in-the-middle

Innovation

• Attributed Table Graph
• Graph-based Reasoning
• Question-Guided Personalized PageRank
• Table Reasoning
• Structure Preservation