Pandora: Leveraging Code-driven Knowledge Transfer for Unified Structured Knowledge Reasoning
Year: 2025
Type: preprint
Abstract: Unified Structured Knowledge Reasoning (USKR) aims to answer natural language questions by using structured sources such as tables, databases, and knowledge graphs in a unified way. Existing USKR methods rely on task-specific strategies or bespoke representations, which hinder their ability to dismantle barriers between different SKR tasks, thereby constraining their overall performance in cross-task scenarios. In this paper, we introduce \textsc{Pandora}, a novel USKR framework that addresses the limitations of existing methods by leveraging two key innovations. First, we propose a code-based unified knowledge representation using \textsc{Python}'s \textsc{Pandas} API, which aligns seamlessly with the pre-training of LLMs. This representation facilitates a cohesive approach to handling different structured knowledge sources. Building on this foundation, we employ knowledge transfer to bolster the unified reasoning process of LLMs by automatically building cross-task memory. By adaptively correcting reasoning using feedback from code execution, \textsc{Pandora} showcases impressive unified reasoning capabilities. Extensive experiments on six widely used benchmarks across three SKR tasks demonstrate that \textsc{Pandora} outperforms existing unified reasoning frameworks and competes effectively with task-specific methods.
Source: arXiv (Cornell University)
Authors: Chen, Yongrui, He, Junhao, Fu, Linbo, Zhang, Shenyu, Jin, Rihui, Dai, Xinbang, Li, Jiaqi, Min, Dehai, Hu, Nan, Zhang, Yuxin, Qi, Guilin, Huang, Yi, Wu, Tongtong
Language: English
Topic: Topic Modeling
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
Award: G7659626447
OpenAlex ID: W7077974665
Cites:
Cited by:
Related to: