IdeaForge: Multi-Agent Framework for Innovation Analysis and Patent Generation
IdeaForge is a multi-agent framework rooted in a knowledge graph, aimed at analyzing innovation across various methodologies and generating patent claims. It combines several innovation approaches—TRIZ, Design Thinking, and SCAMPER—utilizing specialized agents that operate on a continuous FalkorDB knowledge graph. These agents provide structured entities and relationships that encompass contradictions, inventive principles, user needs, transformations, analogies, and potential claims. A key feature is the cross-methodology convergence mechanism, which employs graph-based claim linking. This framework tackles the issue of fragmented insights by maintaining an intermediate reasoning structure, facilitating traceability, synthesis, and systematic novelty evaluation. The research is available on arXiv under ID 2605.13311.
Key facts
- IdeaForge is a multi-agent framework for innovation analysis and patent claim generation.
- It integrates TRIZ, Design Thinking, and SCAMPER methodologies.
- Specialist agents operate over a FalkorDB knowledge graph.
- The framework preserves intermediate reasoning structure for traceability.
- It uses a cross-methodology convergence mechanism via graph-based claim linking.
- The paper is available on arXiv under ID 2605.13311.
- The framework aims to overcome fragmentation in current AI-assisted innovation systems.
- Each agent contributes structured entities and relationships.
Entities
Institutions
- arXiv