Summary of Claimbrush: a Novel Framework For Automated Patent Claim Refinement Based on Large Language Models, by Seiya Kawano et al.
ClaimBrush: A Novel Framework for Automated Patent Claim Refinement Based on Large Language Models
by Seiya Kawano, Hirofumi Nonaka, Koichiro Yoshino
First submitted to arxiv on: 8 Oct 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper proposes ClaimBrush, a novel framework for automated patent claim refinement, which includes a dataset and rewriting model. The dataset was constructed by collecting actual patent claim rewriting cases from the patent examination process. A large language model was fine-tuned using this dataset to build an automatic patent claim rewriting model. To enhance performance, preference optimization based on a prediction model of patent examiners’ Office Actions was applied. Experimental results show that ClaimBrush outperformed heuristic baselines and zero-shot learning in state-of-the-art large language models. Preference optimization further boosted the performance of patent claim refinement. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary ClaimBrush is a new way to improve patent claims using artificial intelligence. Right now, patent examiners have to rewrite patent claims by hand, which can be time-consuming and lead to errors. The authors of this paper created ClaimBrush to automate this process. They made a dataset with examples of rewritten patent claims from the patent examination process. Then, they used a big language model to learn how to write good patent claims. To make it even better, they added a preference optimization step based on what patent examiners look for in Office Actions. The results show that ClaimBrush is more accurate than other methods and can be very useful. |
Keywords
» Artificial intelligence » Language model » Large language model » Optimization » Zero shot
