A Structured Generation Framework for Transforming Scientific Papers into Patent

Researchers introduced FlowPlan-G2P, a graph-mediated framework that transforms scientific papers into patent descriptions through concept graph induction, section-level planning, and graph-conditioned generation. The system outperformed vanilla proprietary models on domain-specific evaluations, demonstrating that structured decomposition improves patent drafting quality more than model scale alone.

Computer Science Computation and Language Submitted on 5 Jan 2026 v1 https://arxiv.org/abs/2601.02589v1 , last revised 23 May 2026 this version, v4 Title:FlowPlan-G2P: A Structured Generation Framework for Transforming Scientific Papers into Patent Descriptions View PDF /pdf/2601.02589 HTML experimental https://arxiv.org/html/2601.02589v4 Abstract:Generating patent descriptions from scientific papers is challenging due to fundamental rhetorical and structural disparities between the two genres. Existing approaches treat this as surface-level rewriting, failing to capture the hierarchical reasoning and statutory constraints inherent in patent drafting. We propose FlowPlan-G2P, a graph-mediated generation framework that decomposes this transformation into three stages: 1 Concept Graph Induction, extracting technical entities and functional dependencies into a directed graph; 2 Section-level Planning, partitioning the graph into coherent subgraphs aligned with canonical patent sections; and 3 Graph-Conditioned Generation, synthesizing legally compliant paragraphs conditioned on section-specific subgraphs. Experiments on expert-validated benchmarks reveal that standard NLG metrics systematically favor legally non-compliant outputs over valid patent descriptions, motivating our domain-specific evaluation. Under this evaluation, FlowPlan-G2P with an open-weight backbone consistently outperforms vanilla proprietary models, demonstrating that structured decomposition is a stronger determinant of quality than model scale. Submission history From: Yoo Yongmin view email /show-email/f39b310c/2601.02589 Mon, 5 Jan 2026 22:40:15 UTC 10,029 KB v1 /abs/2601.02589v1 Tue, 14 Apr 2026 08:59:16 UTC 1,358 KB v2 /abs/2601.02589v2 Wed, 13 May 2026 12:15:30 UTC 1,358 KB v3 /abs/2601.02589v3 v4 Sat, 23 May 2026 02:45:09 UTC 7,092 KB References & Citations Loading... Bibliographic and Citation Tools Bibliographic Explorer What is the Explorer? https://info.arxiv.org/labs/showcase.html arxiv-bibliographic-explorer Connected Papers What is Connected Papers? https://www.connectedpapers.com/about Litmaps What is Litmaps? https://www.litmaps.co/ scite Smart Citations What are Smart Citations? https://www.scite.ai/ Code, Data and Media Associated with this Article alphaXiv What is alphaXiv? https://alphaxiv.org/ CatalyzeX Code Finder for Papers What is CatalyzeX? https://www.catalyzex.com DagsHub What is DagsHub? https://dagshub.com/ Gotit.pub What is GotitPub? http://gotit.pub/faq Hugging Face What is Huggingface? https://huggingface.co/huggingface ScienceCast What is ScienceCast? https://sciencecast.org/welcome Demos Recommenders and Search Tools Influence Flower What are Influence Flowers? https://influencemap.cmlab.dev/ CORE Recommender What is CORE? https://core.ac.uk/services/recommender arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs https://info.arxiv.org/labs/index.html .