ASTRA: Adaptive Semantic Tree Reasoning Architecture for Complex Table Question Answering
2026-04-10 • Computation and Language
Computation and LanguageArtificial IntelligenceMachine Learning
AI summaryⓘ
The authors identify that current methods of turning tables into sequences for language models struggle with preserving the table’s structure and meaning. They propose ASTRA, which first reorganizes tables into semantic trees that show clear hierarchy and adapt based on table size. Then, their method combines a tree-based text search with code execution to better understand and verify answers. Experiments show this approach improves performance on difficult table questions.
Large Language ModelsTable SerializationSemantic TreesHierarchical DependenciesAdaptive MechanismTree SearchSymbolic Code ExecutionComplex Table Question AnsweringReasoning Framework
Authors
Xiaoke Guo, Songze Li, Zhiqiang Liu, Zhaoyan Gong, Yuanxiang Liu, Huajun Chen, Wen Zhang
Abstract
Table serialization remains a critical bottleneck for Large Language Models (LLMs) in complex table question answering, hindered by challenges such as structural neglect, representation gaps, and reasoning opacity. Existing serialization methods fail to capture explicit hierarchies and lack schema flexibility, while current tree-based approaches suffer from limited semantic adaptability. To address these limitations, we propose ASTRA (Adaptive Semantic Tree Reasoning Architecture) including two main modules, AdaSTR and DuTR. First, we introduce AdaSTR, which leverages the global semantic awareness of LLMs to reconstruct tables into Logical Semantic Trees. This serialization explicitly models hierarchical dependencies and employs an adaptive mechanism to optimize construction strategies based on table scale. Second, building on this structure, we present DuTR, a dual-mode reasoning framework that integrates tree-search-based textual navigation for linguistic alignment and symbolic code execution for precise verification. Experiments on complex table benchmarks demonstrate that our method achieves state-of-the-art (SOTA) performance.