Self-Knowledge Retrieval Augmented Generation Framework for Patent Matching

2026-08-11

AI Summary

This work proposes a self-knowledge retrieval-augmented generation framework for patent matching, addressing the limitations of existing methods that either fail to capture subtle semantic differences or rely on costly annotated data while underutilizing the intrinsic knowledge of large language models. The proposed approach introduces, for the first time, a self-knowledge guidance mechanism that enables the model to autonomously extract key technical entities from queries and construct a hierarchical ontology structure to facilitate query expansion and precise retrieval. By integrating FAISS-based vector search with a generative matching strategy, the method leverages the model’s inherent semantic parsing capabilities without requiring external supervision. Evaluated on real-world patent datasets, it significantly improves matching accuracy and overcomes the conventional RAG paradigm’s insufficient exploitation of internal model knowledge.

Abstract

Patent retrieval and matching based on large language models (LLMs) play a vital role in intellectual property protection. However, due to the complex structure of patent documents, dense technical terminology, and multi-modal information, traditional methods struggle to accurately identify subtle differences between patents. Existing LLM-based patent matching approaches typically rely on domain-specific pretrained or instruction tuning, which often entail high manual labeling costs and catastrophic forgetting. While retrieval-augmented generation (RAG) methods introduce external knowledge they fail to fully leverage LLM's capability to automatically parse patents and mine deep semantic relationships. To address these limitations, this paper proposes a self-knowledge RAG framework that guides LLMs to autonomously extract key technical entities and construct hierarchical ontological structures from patent matching queries, thereby enabling query expansion and precise retrieval. The method integrates the FAISS retrieval with a generative matching mechanism, leveraging self-knowledge to enhance the model's understanding of patent innovations and significantly improve retrieval and matching accuracy. Experimental results demonstrate the outstanding performance of the proposed method on real-world patent datasets, validating its effectiveness and application potential.

Problem

• patent matching
• large language models
• retrieval-augmented generation
• technical terminology
• semantic relationships

Innovation

• Self-Knowledge RAG
• Patent Matching
• Hierarchical Ontology
• Query Expansion
• FAISS Retrieval