Curriculum Vitae

Thanh-Nghia Truong · Assistant Professor / Applied Scientist / Data Scientist / Machine Learning Researcher

Tokyo, Japan · thanhnghiadk@gmail.com · Website · Google Scholar · researchmap · LinkedIn · GitHub · ResearchGate

Summary

I am an Assistant Professor and applied AI researcher with a Ph.D. in Computer Science, specializing in machine learning, computer vision, pattern recognition, and data-driven AI systems. My work focuses on practical AI models for user-generated data, including handwriting recognition, automatic scoring, image/text understanding, and recommendation-oriented applications.

I have both academic research experience and industry experience as a Data Scientist / Team Leader, where I worked on social media data platforms, brand-influencer recommendation, and content classification for business analytics. I bring a strong publication record, experience in model evaluation and benchmarking, and hands-on ability to translate research prototypes into deployable AI services.

Core Strengths

Research and Communication

Ph.D.-level research, peer-reviewed publications, international conference presentations, cross-functional collaboration, and English professional communication.

Applied Machine Learning and Data Mining

ML models for business applications, user-generated data, social media analytics, classification, ranking, and recommendation tasks.

Recommendation and Personalization

Brand-influencer recommendation, user/content profiling, campaign intelligence, data-driven matching, and business-oriented model evaluation.

Computer Vision and Multimodal AI

Object detection, classification, tracking, segmentation, face recognition, document analysis, handwriting recognition, and image-text classification.

AI Evaluation and Model Reliability

Offline evaluation protocols, benchmark comparison, answer similarity, confidence/rejection mechanisms, human-in-the-loop verification, and LLM-assisted semantic scoring.

Deployment and Engineering

PyTorch, ONNX, REST API, Linux, Git, AWS, FFmpeg, AI servers, backend integration, data pipelines, and model optimization.

Experience

Oct 2022 – Present

Assistant Professor / Applied AI Researcher

WACOM-TUAT Joint Research Lab, Tokyo University of Agriculture and Technology · Tokyo, Japan
  • Research and develop AI systems for handwriting recognition, automatic scoring, digital ink processing, and educational assessment.
  • Build end-to-end deep learning and graph-based models for handwritten mathematical expression recognition and structured relation prediction, with emphasis on model evaluation and system reliability.
  • Develop recognition servers and practical prototypes connecting AI models with REST APIs, LMS platforms, and user-facing educational applications.
  • Contribute to automatic scoring pipelines for short-answer and long-form handwritten answers, including recognition outputs, answer similarity, confidence-based rejection, and human-in-the-loop verification.
  • Collaborate with academic and industry partners to design, evaluate, and deploy applied AI systems for document understanding, learning analytics, and educational assessment.
Oct 2019 – Sep 2022

Researcher / Applied AI Researcher

Nakagawa Lab, Tokyo University of Agriculture and Technology · Tokyo, Japan
  • Developed online/offline handwritten mathematical expression recognition models using deep learning and end-to-end learning architectures.
  • Built AI models and recognition services for mathematical expression, English, and Japanese handwriting recognition.
  • Published multiple first-author papers and achieved competitive performance in international HME recognition benchmarks.
  • Translated research prototypes into deployable recognition systems and applied AI services.
Oct 2017 – Jun 2019

Data Scientist / Team Leader

HIIP Asia · Ho Chi Minh City, Vietnam
  • Led and developed data science projects focused on social media data mining, viral marketing analytics, campaign intelligence, and business-oriented ML applications.
  • Built a company data platform with crawlers for collecting data from multiple social networks and storing/analyzing large-scale crawled data.
  • Developed a brand-influencer recommendation system based on social media data, user/content profiles, and campaign requirements.
  • Developed image and text classification models to analyze crawled social media content and extract useful signals for downstream analytics.
Jan 2016 – Jun 2019

Research Assistant / Computer Vision Engineer

GV Lab, Ho Chi Minh University of Technology · Ho Chi Minh City, Vietnam
  • Worked on machine learning and computer vision applications including liver segmentation, 3D reconstruction, face detection/recognition, vehicle detection/tracking, license plate recognition, and video streaming.
  • Developed traffic video analytics systems for estimating and visualizing traffic parameters using real-world traffic videos.
  • Built an online video streaming system for an e-learning application using FFmpeg and real-time video processing techniques.

Selected Applied AI and Research Projects

Automatic Scoring of Handwritten Answers

AI-based scoring systems for mathematics, English, and Japanese handwritten answers using handwriting recognition, answer similarity, confidence-based rejection, and human-in-the-loop verification. Current collaborative work also uses recognized answer outputs as inputs to LLM-based semantic scoring for long-form answers.

Handwritten Mathematical Expression Recognition

Multimodal deep learning methods for online/offline HME recognition, including encoder-decoder models and graph-based methods using GNNs for symbol-relation structure analysis.

E-compass Automatic Scoring for Geometric Construction

Process-aware automatic scoring of geometric construction answers using digital ink from an electronic drawing compass, including arc/line extraction, stroke-level clustering, step-level segmentation, geometric reasoning, and step-wise evaluation.

Brand-Influencer Recommendation System

Recommendation models to match brands with suitable influencers for marketing campaigns using social media data, user/content profiles, and campaign requirements.

Social Media Data Platform and Analytics

Crawlers and data systems for collecting, storing, and analyzing large-scale social media data to support viral marketing analytics and campaign insights.

Computer Vision Recognition Projects

Practical recognition models for sushi recognition, Japanese wine classification with more than 200 classes, Asian face recognition, traffic video analytics, and medical image segmentation.

Education

Oct 2019 – Sep 2022

Ph.D. in Computer Science

Tokyo University of Agriculture and Technology · Tokyo, Japan

Handwritten mathematical expression recognition by deep neural networks.

Oct 2016 – Jun 2018

M.Sc. in Computer Science

Ho Chi Minh University of Technology · Ho Chi Minh City, Vietnam

Deep learning for medical image diagnosis, organ segmentation, and 3D reconstruction.

Sep 2011 – Mar 2016

B.Sc. in Computer Science, Honor Program

Ho Chi Minh University of Technology · Ho Chi Minh City, Vietnam

Built an online video streaming system for a real e-learning application using FFmpeg and streaming protocols.

Skills

Programming

Python, C/C++, C#, Java

Machine Learning / Deep Learning

PyTorch, PyTorch Geometric, TensorFlow, NumPy, ONNX, OpenCV

Data Science

SQL, MySQL, MongoDB, NoSQL, data crawling, recommendation systems, social media analytics, data mining

Computer Vision

Detection, classification, tracking, segmentation, face recognition, document analysis, handwriting recognition, image-text classification

Deployment

Linux, Git, REST API, AWS, FFmpeg, AI servers, model optimization and deployment

Data Formats

InkML, LG, JSON, digital ink traces, stroke-based data

Selected Publications and Research Outputs

Hierarchical Stroke-Level Clustering and Step-Level Segmentation for Automatic Scoring of Geometric Construction Answers with an Electronic Drawing Compass
ICDAR, 2026 · Accepted for presentation
Automatic Scoring System for Digitized Handwritten Answers
IEEE Access, 2026
Automated Recognition and Scoring of Handwritten Short Answer: Insights from Japanese Elementary and Junior High Schools
ICDAR, 2025
Content-Based Similarity for Automatic Scoring of Handwritten Descriptive Answers
ICDAR, 2024
A Survey on Handwritten Mathematical Expression Recognition: The Rise of Encoder-Decoder and GNN Models
Pattern Recognition, 2024
ICDAR 2023 CROHME: Competition on Recognition of Handwritten Mathematical Expressions
ICDAR, 2023
Syntactic Data Generation for Handwritten Mathematical Expression Recognition
Pattern Recognition Letters, 2022

Full publication list: website publications, Google Scholar, and researchmap.

Languages

Languages: Vietnamese native; English professional communication and research presentations; Japanese intermediate level, around JLPT N3.

Interests

Football Badminton Running Photography Cycling