Fusing Bidirectional Chains of Thought and Reward Mechanisms A Method for Enhancing Question-Answering Capabilities of Large Language Models for Chinese Intangible Cultural Heritage

2025-05-13

AI Summary

To address erroneous knowledge inheritance, bias amplification, and catastrophic forgetting in fine-tuning large language models (LLMs) for Chinese intangible cultural heritage (ICH), this paper proposes a bidirectional chain-of-thought (CoT) collaborative training framework based on ICH-Qwen. It innovatively introduces dual pathways—forward reasoning and backward questioning–reasoning—to activate latent domain knowledge. A structured, content-aware multi-dimensional weighted reward function is designed to guide decision-making optimization during generation. Integrated with domain-adaptive fine-tuning and systematic ablation analysis, the framework achieves significant improvements over zero-shot, standard CoT, knowledge distillation, and question-augmentation baselines on ICH question answering (superior across Accuracy, BLEU-4, and ROUGE-L). Furthermore, it demonstrates strong cross-domain generalization on financial QA, Wikidata fact verification, and StrategyQA tasks.

Technology Category

• Natural Language Processing: Question Answering
• Data Mining & Knowledge Management: Intelligent Query Processing
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work

Abstract

The rapid development of large language models (LLMs) has provided significant support and opportunities for the advancement of domain-specific LLMs. However, fine-tuning these large models using Intangible Cultural Heritage (ICH) data inevitably faces challenges such as bias, incorrect knowledge inheritance, and catastrophic forgetting. To address these issues, we propose a novel training method that integrates a bidirectional chains of thought and a reward mechanism. This method is built upon ICH-Qwen, a large language model specifically designed for the field of intangible cultural heritage. The proposed method enables the model to not only perform forward reasoning but also enhances the accuracy of the generated answers by utilizing reverse questioning and reverse reasoning to activate the model's latent knowledge. Additionally, a reward mechanism is introduced during training to optimize the decision-making process. This mechanism improves the quality of the model's outputs through structural and content evaluations with different weighting schemes. We conduct comparative experiments on ICH-Qwen, with results demonstrating that our method outperforms 0-shot, step-by-step reasoning, knowledge distillation, and question augmentation methods in terms of accuracy, Bleu-4, and Rouge-L scores on the question-answering task. Furthermore, the paper highlights the effectiveness of combining the bidirectional chains of thought and reward mechanism through ablation experiments. In addition, a series of generalizability experiments are conducted, with results showing that the proposed method yields improvements on various domain-specific datasets and advanced models in areas such as Finance, Wikidata, and StrategyQA. This demonstrates that the method is adaptable to multiple domains and provides a valuable approach for model training in future applications across diverse fields.

Problem

• Enhancing QA accuracy for Chinese ICH using bidirectional reasoning
• Addressing bias and knowledge errors in domain-specific LLMs
• Improving model outputs via reward-based decision optimization

Innovation

• Bidirectional chains of thought enhance reasoning
• Reward mechanism optimizes decision-making process
• Combined method improves accuracy and adaptability