Pandora: Leveraging Code-driven Knowledge Transfer for Unified Structured Knowledge Reasoning

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: 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
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866908501890039808
author 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
author_facet 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
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2508_17905
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Pandora: Leveraging Code-driven Knowledge Transfer for Unified Structured Knowledge Reasoning
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
Computation and Language
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.
title Pandora: Leveraging Code-driven Knowledge Transfer for Unified Structured Knowledge Reasoning
topic Computation and Language
url https://arxiv.org/abs/2508.17905