maechannelerResearch Scientist

Bridging patent examination and AI research. Specializing in NLP applications for patent analysis and fluid dynamics simulation technologies.

About Me

A research scientist working at the intersection of artificial intelligence and patent analysis. My work focuses on developing innovative NLP solutions for intellectual property research and exploring computational approaches to complex engineering problems.

With a background spanning aerospace engineering and machine learning, I bridge traditional engineering disciplines with cutting-edge AI technologies. My research contributes to both academic knowledge and practical applications in patent examination.

I believe in the power of interdisciplinary research to solve complex problems and am passionate about developing technologies that can make a meaningful impact on how we analyze and understand technical information.

Research Interests

Natural Language Processing

Specializing in patent domain-specific language models, I developtransformer-based solutions for patent analysis and examination support. My work focuses on applying state-of-the-art NLP techniques to intellectual property research and knowledge extraction from technical documents.

Key Achievements

• Developed first patent-domain BERT in organizational context
• Implemented transformer-based automatic reading estimators
• Created NLP solutions for SDGs patent landscape analysis

Patent-domain BERTTransformer ModelsPatent Analysis

Fluid Dynamics

With expertise in aerospace engineering and numerical analysis, I focus onadvanced propulsion systems and computational fluid dynamics. My research bridges theoretical fluid mechanics with practical aerospace applications.

Research Focus

• MHD generator technology for scramjet engines
• Numerical simulation of complex flow phenomena
• Aerospace propulsion system optimization

MHD GeneratorScramjet EngineAerospace CFD

Web Applications

Patent Analysis Dashboard

Interactive dashboard for visualizing and analyzing patent data with advanced NLP insights

In Development

ReactD3.jsPython

NLP Research Tool

Web-based toolkit supporting natural language processing research and experimentation

Planning

Next.jsTensorFlow.jsFastAPI

Flow Simulation Viewer

Browser-based visualization tool for computational fluid dynamics simulation results

Planning

Three.jsWebGLReact

These applications aim to transform research findings into practical, accessible tools

Recent Publications

Multi-stage fine-tuning of patent domain-specific DeBERTa for advanced patent landscape on SDGs/Decarbonization

World Patent Information, Vol. 812025

Macro analysis of decarbonization-related patent technologies by patent domain-specific BERT

World Patent Information, Vol. 692022

"Visualization" of SDGs-related Patent Technologies by Natural Language Processing Technology

Business and Management Studies, Vol. 7, No. 32021

Exploring patent examination support using transformer

Tokugikon, Vol. 2972020

Numerical Analysis of Application of MHD Generator Technology to Scramjet Engine

Proceedings of 41st Fluid Dynamics Conference2009

Get In Touch

Interested in research collaboration on patent analysis, AI applications, or fluid dynamics? Let's explore innovative possibilities together.