Panel

Chair: Lun-Wei Ku, Academia Sinica

Title

“Challenges and Opportunities with SEA LLMs”

Theme

This panel will discuss the challenges and opportunities associated with Large Language Models (LLMs) in Southeast Asia (SEA). The panelists include:

  • Sarana Nutanong, VISTEC
  • Ayu Purwarianti, Bandung Institute of Technology (ITB), Indonesia
  • William Tjhi, AI Singapore

Each panelist will deliver a short statement (6-8 minutes) addressing key aspects of the panel’s main topic:

  • Challenges of Regional and/or Low-Resource Languages: Examining issues specific to languages such as Indonesian, Thai, and other SEA languages.
  • Existing Data and Future Needs: Discussing current datasets and identifying additional data required for training models in these languages.
  • Opportunities for LLMs in SEA Languages: Exploring how LLMs can benefit SEA languages and the opportunities they present.
  • Progress of SEA LLMs and Application Scenarios: Discussing current status of SEA LLMs and their practical applications.
  • Development and Maintenance of LLMs for Regional Languages: Debating the necessity and strategies for developing and maintaining LLMs for regional languages.

The presentations will be followed by a discussion between the panelists and the audience.

Panelists

Dr. Sarana Nutanong

Researcher, VISTEC

Sarana Nutanong is a prominent researcher in Thai Natural Language Processing, leading the Natural Language Processing and Representation Learning Lab (NRL) at VISTEC. His research interests include benchmarking, low-resource natural language processing, and robust representation learning methods in multi- and cross-lingual environments. Sarana has numerous publications in top-tier conferences such as ACL, EMNLP, and TACL. He received the Best Short Paper award at EMNLP 2022 for his work titled “Topic-Regularized Authorship Representation Learning.” His work particularly focuses on advancing NLP and AI capabilities in Southeast Asia, aiming to bridge the gap between research and practical applications in the region and closing the low-resource gap.

Dr. Ayu Purwarianti

Bandung Institute of Technology (ITB), Indonesia

Ayu Purwarianti received her Ph.D. degree from the Toyohashi University of Technology in December 2007, with a dissertation titled “Cross Lingual Question Answering System (Indonesian Monolingual QA, Indonesian-English CLQA, Indonesian-Japanese CLQA).” Since then, she has been working at the Bandung Institute of Technology (ITB), Indonesia. In addition to teaching and conducting research, she has been actively involved in the Indonesian Association for Computational Linguistics, where she served as the Chair from 2016 to 2018. She was also the Chair of the IEEE Education Chapter of the Indonesian Section from 2017 to 2019. She has been with IABEE since 2015. She founded a start-up named Prosa.ai in 2018. She has been the Chair of the Artificial Intelligence Center at ITB since August 2019, continuing in this role until early 2024.

Dr. William Tjhi

AI Singapore, Singapore

William Tjhi has been practicing machine learning to solve industry problems for more than a decade. He received his PhD from NTU in 2008, with a thesis on unsupervised learning for text data. He spent a few years with A*STAR where he learned how to scale up ML with distributed systems. When in GovTech, he contributed to the initial data science efforts in the organisation. For a while, William was also the lead NLP at Traveloka. There, he appreciated the challenges of doing NLP in the then low-resource Bahasa Indonesia. This became an inspiration for him when he later initiated a programme that focuses on building NLP resources for Southeast Asia languages in AI Singapore. In the early days of AI Singapore, he was one of the first engineers that set the foundation for the execution of the 100Experiments and AI Apprenticeship Programme. Today, William leads the applied research work on Regional LLMs in the AI Products division of AI Singapore. He also holds a technical advisor position with ASEAN Applied Research Centre, and a part-time ML consultant to an early-stage tech startup, Sembly. In his spare time, William contributes regularly to tech communities like the Data Science SG and Data Science Indonesia (DSI).