Comparing Complex Concepts with Transformers: Matching Patent Claims Against Natural Language Text

TL;DR

This paper evaluates two new large language model-based methods for comparing patent claims with other texts, significantly improving performance in matching complex, domain-specific language to more general language, with potential applications beyond patents.

Contribution

Introduces and tests two novel LLM-based approaches for cross-domain claim matching, demonstrating substantial performance improvements over previous methods.

Findings

Both approaches outperform prior published results.

Enhanced ability to match dense patent claims with broader text.

Potential applications extend beyond patent analysis.

Abstract

A key capability in managing patent applications or a patent portfolio is comparing claims to other text, e.g. a patent specification. Because the language of claims is different from language used elsewhere in the patent application or in non-patent text, this has been challenging for computer based natural language processing. We test two new LLM-based approaches and find that both provide substantially better performance than previously published values. The ability to match dense information from one domain against much more distributed information expressed in a different vocabulary may also be useful beyond the intellectual property space.