Pap2Pat: Benchmarking Outline-Guided Long-Text Patent Generation with Patent-Paper Pairs
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909652084588544 |
|---|---|
| author | Knappich, Valentin Razniewski, Simon Hätty, Anna Friedrich, Annemarie |
| author_facet | Knappich, Valentin Razniewski, Simon Hätty, Anna Friedrich, Annemarie |
| contents | Dealing with long and highly complex technical text is a challenge for Large Language Models (LLMs), which still have to unfold their potential in supporting expensive and timeintensive processes like patent drafting. Within patents, the description constitutes more than 90% of the document on average. Yet, its automatic generation remains understudied. When drafting patent applications, patent attorneys typically receive invention reports (IRs), which are usually confidential, hindering research on LLM-supported patent drafting. Often, prepublication research papers serve as IRs. We leverage this duality to build PAP2PAT, an open and realistic benchmark for patent drafting consisting of 1.8k patent-paper pairs describing the same inventions. To address the complex longdocument patent generation task, we propose chunk-based outline-guided generation using the research paper as invention specification. Our extensive evaluation using PAP2PAT and a human case study show that LLMs can effectively leverage information from the paper, but still struggle to provide the necessary level of detail. Fine-tuning leads to more patent-style language, but also to more hallucination. We release our data and code https://github.com/boschresearch/Pap2Pat. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2410_07009 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Pap2Pat: Benchmarking Outline-Guided Long-Text Patent Generation with Patent-Paper Pairs Knappich, Valentin Razniewski, Simon Hätty, Anna Friedrich, Annemarie Computation and Language Artificial Intelligence Dealing with long and highly complex technical text is a challenge for Large Language Models (LLMs), which still have to unfold their potential in supporting expensive and timeintensive processes like patent drafting. Within patents, the description constitutes more than 90% of the document on average. Yet, its automatic generation remains understudied. When drafting patent applications, patent attorneys typically receive invention reports (IRs), which are usually confidential, hindering research on LLM-supported patent drafting. Often, prepublication research papers serve as IRs. We leverage this duality to build PAP2PAT, an open and realistic benchmark for patent drafting consisting of 1.8k patent-paper pairs describing the same inventions. To address the complex longdocument patent generation task, we propose chunk-based outline-guided generation using the research paper as invention specification. Our extensive evaluation using PAP2PAT and a human case study show that LLMs can effectively leverage information from the paper, but still struggle to provide the necessary level of detail. Fine-tuning leads to more patent-style language, but also to more hallucination. We release our data and code https://github.com/boschresearch/Pap2Pat. |
| title | Pap2Pat: Benchmarking Outline-Guided Long-Text Patent Generation with Patent-Paper Pairs |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2410.07009 |