PatentScore: Multi-dimensional Evaluation of LLM-Generated Patent Claims

TL;DR

PatentScore is a comprehensive evaluation framework tailored for assessing the quality of LLM-generated patent claims, addressing the limitations of traditional metrics by incorporating structural, semantic, and legal dimensions.

Contribution

This paper introduces PatentScore, a novel multi-dimensional evaluation method specifically designed for complex patent claims, outperforming existing metrics in correlating with expert judgments.

Findings

PatentScore achieved a correlation of 0.819 with expert annotations.

It significantly outperformed traditional NLG metrics.

The framework effectively captures structural, semantic, and legal aspects of patent claims.

Abstract

High-stakes texts such as patent claims, medical records, and technical reports are structurally complex and demand a high degree of reliability and precision. While large language models (LLMs) have recently been applied to automate their generation in high-stakes domains, reliably evaluating such outputs remains a major challenge. Conventional natural language generation (NLG) metrics are effective for generic documents but fail to capture the structural and legal characteristics essential to evaluating complex high-stakes documents. To address this gap, we propose PatentScore, a multi-dimensional evaluation framework specifically designed for one of the most intricate and rigorous domains, patent claims. PatentScore integrates hierarchical decomposition of claim elements, validation patterns grounded in legal and technical standards, and scoring across structural, semantic, and legal dimensions. In experiments on our dataset which consists of 400 Claim1, PatentScore achieved the highest correlation with expert annotations ($r = 0.819$), significantly outperforming widely used NLG metrics. This work establishes a new standard for evaluating LLM-generated patent claims, providing a solid foundation for research on patent generation and validation.