about

About me

Hi, I’m Ting-Chun Lin (林庭均), a master’s student in Computer Science at National Taipei University of Education. My research focuses on multimodal emotion forecasting across social media platforms using temporal transformer models, integrating insights from artificial intelligence, psychology, and behavioral science.

In 2025, my paper titled
“Weather Effects on Social Media: A Cross-disciplinary Analysis of Digital Behavior and Emotional Change”
was accepted at the 2nd Taiwan Digital Generation Research Conference, hosted by the Taiwan i-Generation Panel Study, co-organized by the National Academy for Education Research, and supervised by the National Science and Technology Council.

This study investigates how weather conditions affect emotional expressions and user behavior by analyzing over 5,000 Threads posts combined with official meteorological data. We applied a multimodal framework combining NLP, sentiment classification, and temporal modeling.

Currently, I lead the development of Serenity Forecast (馨晴預報), a collaborative project involving three teammates from AI engineering, counseling psychology, and digital media. We aim to create a forecasting and support system that links meteorological changes with emotional health patterns, providing personalized suggestions and early warnings to promote mental well-being.

I am passionate about applying AI to real-world challenges, especially in the domains of emotional resilience, psychological care, and public mental health.

關於我

我是林庭均，目前就讀於國立臺北教育大學資訊科學研究所碩士班。我的研究聚焦於跨社群平台之多模態情緒預測模型，運用時間序列 Transformer 架構結合社群語意與氣象條件，探索 AI 在心理健康推動上的應用潛力。

我的論文〈社群媒體中的天氣效應：數位行為與情緒的跨領域分析〉已獲選於2025 年臺灣數位世代研究第二屆國際學術研討會發表。該研討會由臺灣數位世代青少年的成長歷程追蹤研究主辦，國家教育研究院共同主辦，並由國家科學及技術委員會指導。

本研究分析 5731 筆 Threads 平台的貼文與中央氣象署氣象資料，結合 NLP 模型（mDeBERTa + GPT-4o）、時間序列預測（LSTM），探討氣象條件對群體情緒與數位行為的潛在影響。

我目前帶領「馨晴預報 Serenity Forecast」團隊進行系統開發，成員涵蓋 AI 工程、心理諮商、插畫設計等專業背景。系統結合氣象資料、社群語意與情緒趨勢，期望提供日常化的心理健康提醒與個人化支持，協助使用者建立自我照護機制。

我相信 AI 不只是一種技術，更是促進理解、療癒與連結的可能性。