DeepInnovation AI: A Global Dataset Mapping the AI innovation from Academic Research to Industrial Patents

2025-03-12

AI Summary

This study addresses the opacity of academic-to-industrial technology transfer in AI and the fragmentation of innovation-related data. To this end, we construct the first global AI innovation mapping dataset, featuring a novel tripartite architecture integrating patents, scholarly papers, and cross-modal semantic similarity. We propose a hypergraph-based innovation quantification framework and DeepCosineAI—a scalable paper-patent matching method leveraging multilingual large language models, hierarchical BERT classification, and cosine similarity computation—aligned across one million semantically grounded pairs. The dataset encompasses 2.35 million patents and 3.51 million papers, yielding approximately 100 million high-quality semantic matches. Our approach enables the first dynamic, cross-national, and longitudinal quantitative analysis of AI technology transfer, offering an extensible infrastructure for evidence-based innovation policy design and industry–academia–research collaboration.

Technology Category

• Humans and AI: Other Foundations of Human Computation & AI
• Philosophy and Ethics of AI: AI & Epistemology
• Application Domains: Humanities & Computational Social Science

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Security and Privacy: Data transparency and provenance

Abstract

In the rapidly evolving field of artificial intelligence (AI), mapping innovation patterns and understanding effective technology transfer from research to applications are essential for economic growth. However, existing data infrastructures suffer from fragmentation, incomplete coverage, and insufficient evaluative capacity. Here, we present DeepInnovationAI, a comprehensive global dataset containing three structured files. DeepPatentAI.csv: Contains 2,356,204 patent records with 8 field-specific attributes. DeepDiveAI.csv: Encompasses 3,511,929 academic publications with 13 metadata fields. These two datasets leverage large language models, multilingual text analysis and dual-layer BERT classifiers to accurately identify AI-related content, while utilizing hypergraph analysis to create robust innovation metrics. Additionally, DeepCosineAI.csv: By applying semantic vector proximity analysis, this file presents approximately one hundred million calculated paper-patent similarity pairs to enhance understanding of how theoretical advancements translate into commercial technologies. DeepInnovationAI enables researchers, policymakers, and industry leaders to anticipate trends and identify collaboration opportunities. With extensive temporal and geographical scope, it supports detailed analysis of technological development patterns and international competition dynamics, establishing a foundation for modeling AI innovation and technology transfer processes.

Problem

• Mapping AI innovation patterns from academia to industry
• Bridging the gap between academic papers and industrial patents
• Enhancing understanding of technology transfer with comprehensive data

Innovation

• Uses large language models for AI content identification
• Applies hypergraph analysis for robust innovation metrics
• Employs semantic vector proximity for paper-patent similarity