Interpretable Oracle Bone Script Decipherment through Radical and Pictographic Analysis with LVLMs

2025-08-13

AI Summary

Oracle bone script (OBS) decipherment has long been hindered by high glyph abstraction, semantic ambiguity, and severe scarcity of annotated data; existing deep learning approaches suffer from poor generalization, limited interpretability, and weak zero-shot recognition capability. To address these challenges, we propose the first Large Vision-Language Model (LVLM) framework explicitly designed for interpretable OBS decipherment. Our method integrates radical analysis with pictographic semantic modeling, introducing a dual radical–pictograph matching mechanism and a progressive training strategy to enable logical, stepwise reasoning from glyph to semantics. We construct a large-scale pictographic decipherment dataset containing 47,157 characters. On public benchmarks, our model achieves state-of-the-art Top-10 accuracy and significantly improves zero-shot performance. Crucially, it generates structured, human-readable explanation paths that support archaeological inference. This work pioneers the integration of interpretable reasoning systems into OBS research, establishing a novel paradigm for digital humanities.

Technology Category

• Computer Vision: Language and Vision
• Natural Language Processing: Lexical Semantics and Morphology
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning

Application Category

• Economics, Online Markets and Human Computation: Humans versus LLMs for data annotation and labeling
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

As the oldest mature writing system, Oracle Bone Script (OBS) has long posed significant challenges for archaeological decipherment due to its rarity, abstractness, and pictographic diversity. Current deep learning-based methods have made exciting progress on the OBS decipherment task, but existing approaches often ignore the intricate connections between glyphs and the semantics of OBS. This results in limited generalization and interpretability, especially when addressing zero-shot settings and undeciphered OBS. To this end, we propose an interpretable OBS decipherment method based on Large Vision-Language Models, which synergistically combines radical analysis and pictograph-semantic understanding to bridge the gap between glyphs and meanings of OBS. Specifically, we propose a progressive training strategy that guides the model from radical recognition and analysis to pictographic analysis and mutual analysis, thus enabling reasoning from glyph to meaning. We also design a Radical-Pictographic Dual Matching mechanism informed by the analysis results, significantly enhancing the model's zero-shot decipherment performance. To facilitate model training, we propose the Pictographic Decipherment OBS Dataset, which comprises 47,157 Chinese characters annotated with OBS images and pictographic analysis texts. Experimental results on public benchmarks demonstrate that our approach achieves state-of-the-art Top-10 accuracy and superior zero-shot decipherment capabilities. More importantly, our model delivers logical analysis processes, possibly providing archaeologically valuable reference results for undeciphered OBS, and thus has potential applications in digital humanities and historical research. The dataset and code will be released in https://github.com/PKXX1943/PD-OBS.

Problem

• Deciphering Oracle Bone Script using radical and pictographic analysis
• Improving interpretability and generalization in OBS decipherment
• Enhancing zero-shot performance for undeciphered OBS characters

Innovation

• Uses Large Vision-Language Models for OBS decipherment
• Combines radical and pictographic semantic analysis
• Introduces Radical-Pictographic Dual Matching mechanism