CV

Mavin Sao

mavinsao11@gmail.com
+82 1030139109
Gwangju, South Korea, KR

Summary

Cambodian Master's student in Data Science at Chonnam National University with expertise in NLP, deep learning, and LLMs. Focused on classification models, topic modeling, and empathic computing.

Education

  • Master’s in Data Science
    Chonnam National University
  • Korean Language Program
    2023-07
    Chonnam National University
  • Computer Science and Engineering
    2020
    Royal University of Phnom Penh
  • Software Development Expert (iOS and Web)
    2020-02
    Korea Software HRD Center
  • English Diploma
    2018
    Paññāsāstra University of Cambodia

Work Experience

  • Research Member
    2023-08 -
    Advanced research in Natural Language Processing (NLP), specializing in classification, topic modeling, model interpretability, and transformer architectures.
  • IT Instructor
    2020-01 - 2022-07
    Taught web and iOS mobile app development, including advanced courses in React Native, UI/UX, Linux, and Docker.

Skills

Programming

  • HTML/CSS/JS
  • Python
  • Java
  • Swift
  • C/C++/C#

AI & NLP

  • LLMs
  • Text Classification
  • Topic Modeling
  • Sentiment Analysis
  • LangChain
  • BERTopic
  • Transformers

Tools

  • Docker
  • React Native
  • Linux
  • Streamlit

Publications

  • MIRoBERTa: Mental Illness Text Classification With Transfer Learning on Subreddits
    2024
    IEEE Access
    This paper introduces MIRoBERTa, a RoBERTa-based model adapted for mental health domain via transfer learning on Reddit data. Achieves state-of-the-art performance in multiclass classification.
  • Personalized E-Learning Course Recommendations: A Chatbot Approach Using LangChain
    2024
    2024 International Conference on Digital Contents
    A LangChain-powered chatbot for personalized e-learning course recommendations using private datasets and RAG architecture.

Presentations

  • Personalized E-Learning Course Recommendations: A Chatbot Approach Using LangChain
    2024
    2024 International Conference on Digital Contents
    South Korea
    Presented an AI-powered chatbot system using LangChain and RAG to recommend personalized e-learning courses through a Streamlit interface.

Teaching

  • Web & iOS App Development
    2020
    Korea Software HRD Center
    Role: IT Instructor
    Delivered basic and advanced courses in web design and mobile application development with tools like React Native and Docker.

Portfolio

  • Mental Health Text Classification
    2024
    Nlp project
    Used traditional ML, deep learning, and transformers (MIRoBERTa, MIBERT) to classify Reddit mental health posts.
  • Academic Ally: Course Recommendation Chatbot
    2024
    Chatbot
    RAG-based chatbot built using LangChain to deliver personalized course recommendations via Streamlit.
  • Topic Refinement with BERTopic and Mistral
    2025
    Topic modeling
    Used BERTopic with LLM integration to extract and interpret user perception from smartphone forum discussions.

Languages

  • Khmer
    Native
  • English
    Professional
  • Korean
    Conversational (TOPIK Level 3)

Interests

  • Empathic Computing
  • Multimodal NLP
  • Educational Technology

References

  • Prof. Hoi-Jeong Lim
    Professor, Chonnam National University — hjlim@jnu.ac.kr / +82 62-530-5790
  • Mr. Chen Phirum
    Deputy Director, Korea Software HRD Center — phirum.gm@gmail.com / +855 12-998-919