About

Academic profile

Thanh Le is a lecturer in the Computer Science Department, Faculty of Information Technology, University of Science, VNU-HCM. His research focuses on data mining, knowledge graphs, big data, machine learning, and graph-based reasoning.

Experience

Working experience

Vice Dean, Faculty of Information Technology

Faculty of Information Technology, University of Science, VNU-HCM

Vice Head, Department of Computer Science

Department of Computer Science, Faculty of Information Technology, University of Science, VNU-HCM

Lecturer

Computer Science Department, Faculty of Information Technology, University of Science, VNU-HCM

Teaching Assistant

Computer Science Department, Faculty of Information Technology, University of Science, VNU-HCM

Software Engineer

DTV Team, Renesas Design Vietnam Company, Ho Chi Minh City, Vietnam

Research

Research interests

Knowledge Graphs

Embedding models, link prediction, completion, and temporal knowledge graph reasoning.

Graph Learning

Graph neural networks, relational reasoning, tensor decomposition, and quaternion representations.

Machine Learning

Data mining, deep learning, representation learning, and recommender systems.

Big Data Analytics

Scalable data-driven methods for intelligent systems and educational applications.

Research timeline

Research development

A Google Scholar-inspired timeline combined with an academic CV layout, highlighting the evolution of research directions without citation counts.

Google Scholar →
2025–2026
Current focus

Temporal and multimodal knowledge graph reasoning

Recent work focuses on temporal knowledge graph reasoning, multimodal knowledge graph completion, bicomplex/quaternion representations, tensor decomposition, and graph neural reasoning.

Selected works

  • CCGCN - Complex composition graph convolutional network for temporal knowledge graph reasoning.
  • MESN - Multimodal KG embedding with expert fusion and relational attention.
  • HGCT - Temporal KG reasoning through extrapolated historical fact extraction.
  • FTPComplEx / TCrossE / TBicomR - Temporal KG completion with flexible time perspectives and bicomplex representations.

Research keywords

Temporal KGMultimodal KGBicomplex embeddingsQuaternion modelsTensor decomposition
2023–2024
Expansion

Advanced KG embedding, temporal completion, and explainable graph learning

This period emphasizes interaction-rich embedding models, Fourier-enhanced representations, graph transformers, contrastive learning, and interpretable graph neural networks.

Selected works

  • Knowledge graph embedding by relational rotation and complex convolution, Expert Systems with Applications.
  • Knowledge graph embedding with the special orthogonal group in quaternion space, Knowledge-Based Systems.
  • MixER, FouriER, TouriER, and temporal graph transformer work.
  • FleX and nearest-neighbor interpretations for explainable GNNs.

Publication venues

Knowledge-Based SystemsESWAApplied IntelligenceCoopISESANN
2021–2022
Foundation

Knowledge graph completion, graph neural networks, and applied AI systems

The research agenda expands from embedding methods to hyperplanes, rotations, complex/quaternion spaces, graph neural networks, and applied AI systems in education and healthcare.

Selected works

  • RotatHS and rotation embedding on hyperplanes for knowledge graph link prediction.
  • Negative sampling and hypernetwork-based embeddings for KG completion.
  • Prescription recognition, student monitoring, attendance tracking, and facial acne severity assessment.

Research keywords

KG completionGraph neural networksHyperplanesMedical AIEducation AI
2020 & earlier
Early direction

Data mining, recommender systems, sparse networks, and semantic relation discovery

Earlier work includes MOOC recommendation, chatbot response filtering, sparse network representation learning, next-purchase prediction, and detecting hidden semantic relations in geographic data.

Selected works

  • Deep matrix factorization and hybrid loss functions for MOOC recommendation.
  • Feature learning for representing sparse networks based on random walks.
  • Detecting hidden relations in geographic data.

Research keywords

Data miningRecommendationSparse networksSemantic processingRandom walks

Teaching

Courses

Publications

Publications by year

Full list on Google Scholar →

    Contact

    Get in touch

    Department of Computer Science, Faculty of Information Technology, University of Science, VNU-HCM.

    lnthanh@fit.hcmus.edu.vn Tel: (+84 28) 62884499 (Ext: 4000) fit.hcmus.edu.vn/~lnthanh