Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

FlowPlan-G2P: A Structured Generation Framework for Transforming Scientific Papers into Patent Descriptions

Kris W Pan, Yongmin Yoo

TL;DR

FlowPlan-G2P tackles transforming scientific papers into legally compliant patent descriptions by introducing a structured, three-stage pipeline that mirrors expert drafting: Concept Graph Induction, Paragraph/Section Planning, and Graph-Conditioned Generation. By grounding generation in a formal concept graph and section-level plans, the approach improves long-range coherence and enablement, addressing the core structural and legal constraints that end-to-end models struggle with. The authors demonstrate that traditional surface metrics poorly reflect patent quality and validate their method with Pat-DEVAL, showing superior legal compliance and structural coverage across multiple backbone models. The work highlights the Metric Paradox in patent evaluation and positions graph-guided, plan-based generation as a robust paradigm for domain-specific, legally constrained text generation with practical implications for scalable patent drafting.

Abstract

Over 3.5 million patents are filed annually, with drafting patent descriptions requiring deep technical and legal expertise. Transforming scientific papers into patent descriptions is particularly challenging due to their differing rhetorical styles and stringent legal requirements. Unlike black-box text-to-text approaches that struggle to model structural reasoning and legal constraints, we propose FlowPlan-G2P, a novel framework that mirrors the cognitive workflow of expert drafters by reformulating this task into three stages: (1) Concept Graph Induction, extracting technical entities and relationships into a directed graph via expert-like reasoning; (2) Paragraph and Section Planning, reorganizing the graph into coherent clusters aligned with canonical patent sections; and (3) Graph-Conditioned Generation, producing legally compliant paragraphs using section-specific subgraphs and tailored prompts. Experiments demonstrate that FlowPlan-G2P significantly improves logical coherence and legal compliance over end-to-end LLM baselines. Our framework establishes a new paradigm for paper-to-patent generation and advances structured text generation for specialized domains.

FlowPlan-G2P: A Structured Generation Framework for Transforming Scientific Papers into Patent Descriptions

TL;DR

FlowPlan-G2P tackles transforming scientific papers into legally compliant patent descriptions by introducing a structured, three-stage pipeline that mirrors expert drafting: Concept Graph Induction, Paragraph/Section Planning, and Graph-Conditioned Generation. By grounding generation in a formal concept graph and section-level plans, the approach improves long-range coherence and enablement, addressing the core structural and legal constraints that end-to-end models struggle with. The authors demonstrate that traditional surface metrics poorly reflect patent quality and validate their method with Pat-DEVAL, showing superior legal compliance and structural coverage across multiple backbone models. The work highlights the Metric Paradox in patent evaluation and positions graph-guided, plan-based generation as a robust paradigm for domain-specific, legally constrained text generation with practical implications for scalable patent drafting.

Abstract

Over 3.5 million patents are filed annually, with drafting patent descriptions requiring deep technical and legal expertise. Transforming scientific papers into patent descriptions is particularly challenging due to their differing rhetorical styles and stringent legal requirements. Unlike black-box text-to-text approaches that struggle to model structural reasoning and legal constraints, we propose FlowPlan-G2P, a novel framework that mirrors the cognitive workflow of expert drafters by reformulating this task into three stages: (1) Concept Graph Induction, extracting technical entities and relationships into a directed graph via expert-like reasoning; (2) Paragraph and Section Planning, reorganizing the graph into coherent clusters aligned with canonical patent sections; and (3) Graph-Conditioned Generation, producing legally compliant paragraphs using section-specific subgraphs and tailored prompts. Experiments demonstrate that FlowPlan-G2P significantly improves logical coherence and legal compliance over end-to-end LLM baselines. Our framework establishes a new paradigm for paper-to-patent generation and advances structured text generation for specialized domains.
Paper Structure (25 sections, 6 equations, 2 figures, 4 tables)

This paper contains 25 sections, 6 equations, 2 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Background: The Rhetorical and Legal Shift from Paper to Patent
  3. Rhetorical Disparity.
  4. The Enablement Requirement.
  5. Structural Complexity of Patent Specifications.
  6. Related Works
  7. Automated Patent Drafting and Generation
  8. Scientific Literature–Patent Linkage
  9. Graph-Guided and Plan-Based Text Generation
  10. Methodology
  11. Structured Reasoning-based Concept Graph Induction
  12. Paragraph and Section Planning
  13. Graph-Conditioned Generation
  14. Dataset
  15. Experiment & Result
  16. ...and 10 more sections

Figures (2)

  • Figure 1: Global patent application trends from 2009 to 2023, showing a steady increase to over 3.5 million filings in 2023, highlighting the need for automated drafting solutions wipo2023report.
  • Figure 2: Overview of FlowPlan-G2P, a three-stage framework for transforming scientific papers into legally compliant patent descriptions. Stage 1 induces a concept graph, Stage 2 organizes it into section-level plans, and Stage 3 generates graph-conditioned paragraphs.