Source code and dataset for our paper "Beyond Scaling: Predicting Patent Approval with Domain-specific Fine-grained Claim Dependency Graph"
Our code consists of the following two parts: (1) Scaling up SOTA with LLMs and (2) FLAN Graph.
The PatentAP dataset we use in the paper is on 🤗 Huggingface [link].
The code base for embedding-based training and inference can be found here.
The prompt templates we used are provided here.
The construction process and the saved results of the FLAN Graph can be found here.
If these codes and data help you, please consider citing us as follows.
@misc{gao2024scaling,
title={Beyond Scaling: Predicting Patent Approval with Domain-specific Fine-grained Claim Dependency Graph},
author={Xiaochen Kev Gao and Feng Yao and Kewen Zhao and Beilei He and Animesh Kumar and Vish Krishnan and Jingbo Shang},
year={2024},
eprint={2404.14372},
archivePrefix={arXiv},
primaryClass={cs.CL}
}