Keynote Speaker I
Prof. Janusz Kacprzyk, The Polish Academy of Sciences, Poland
Keynote Speaker II
Assoc. Prof. Dr. Anirach Mingkhwan, King Mongkut's University of Technology North Bangkok, Thailand
Assoc. Prof. Dr. Anirach Mingkhwan is a researcher and lecturer in the Department of Information Technology, Faculty of Industrial Technology and Management, King Mongkut's University of Technology North Bangkok (KMUTNB). He specializes in AI Engineering, with research interests spanning Artificial Intelligence, Big Data Analytics, Cloud Computing, Knowledge Management, and Nature-Inspired Computing. From 2012 to 2020, he served as Dean of the faculty, driving innovation in curriculum design and research development.
Dr. Mingkhwan has led multiple interdisciplinary research projects focused on intelligent system design and data-driven technologies. His work has been widely published and is accessible through platforms such as Google Scholar. As a mentor and academic leader, he continues to shape Thailand’s technological education landscape and contribute to global AI research.
Speech: From Hypothesis to Prototype: Accelerating NLP Research with Vibe Coding
This talk provides a practical introduction to Vibe Coding, an agile, iterative development approach designed to assist researchers in translating NLP hypotheses into functional prototypes. Instead of offering a one-size-fits-all solution, Vibe Coding is presented as a flexible framework that can help streamline the processes of hypothesis formulation, model implementation, and empirical evaluation—particularly in applied research settings.
To illustrate its potential, the talk draws on a case study exploring the hypothesis: “Transformer-based language models can significantly improve the detection of misinformation in short-form social media posts compared to traditional keyword-based classifiers.” The session walks through how Vibe Coding structures the workflow into manageable development sprints, enables integration with pre-trained models such as BERT and RoBERTa, and supports comparative evaluation using real datasets and explainability tools.
The emphasis is not on proving a particular research outcome, but on demonstrating how Vibe Coding can make the research process more approachable, reproducible, and adaptable. Attendees will take away ideas on how to use this method to explore their own NLP questions with greater clarity and efficiency. This talk may be of interest to researchers, students, and practitioners who are looking for lightweight, practical strategies to bridge theory and experimentation in NLP work.