We are pleased to announce the 4th joint Workshop on NLP/AI R&D which will take place on November 5, 2022 at Chiang Mai, Thailand. It will be colocated with iSAI-NLP-AIoT 2022. The workshop will bring together researchers, professors, and students in NLP and AI. The aim of the proposed workshop is to bring together the community of Southeast Asian Countries’ researchers interested in these areas.

We invite submissions on topics that include but are not limited to the following:


  • POS-tagging
  • Tokenization
  • Named entity recognition (NER)
  • Morphological analysis
  • Semantic analysis
  • Chunking
  • Disambiguation
  • Machine Translation (MT)
  • Automatic Speech Recognition (ASR)
  • Text to Speech (TTS)


  • Statistical Learning
  • Machine Learning
  • Deep Learning
  • AI in image and speech processing
  • Bayesian Machine Learning
  • Graph Neural Networks
  • Meta Learning
  • Reinforcement Learning
  • Adversarial Machine Learning

Invited Speakers

Prof. Taro Watanabe

Nara Institute of Science and Technology (NAIST)

Multilingual Machine Translation by Disentangling Language Dependency


Multilingual machine translation leverages the capacity of neural networks to encode multiple language representations into a single model. This leads to a success in zero-shot translation in which the model is capable of translating language pairs which are unseen during training. However, the zero-shot translation is unstable in that it is largely influenced by the dominant language, e.g., English, in training data, and decoupling the paired representations has been a challenging topic. In this work, we propose a simple, lightweight yet effective language-specific modeling by adapting to non-centered languages, i.e., non-English, and combining the shared information and the language-specific information to counteract the instability of zero-shot translation.


Taro Watanabe received his B.E. and M.E. degrees in information science from Kyoto University in 1994 and 1997, respectively, and obtained an M.S. degree in Language and Information Technologies from the School of Computer Science, Carnegie Mellon University in 2000. In 2004, he received a Ph.D. in informatics from Kyoto University. After working as a researcher at ATR, NTT and NICT, and as a software engineer at Google, he is a professor at the Nara Institute of Science and Technology starting in 2020. His research interests include natural language processing, machine learning and machine translation.

Dr. Ohnmar Htun

Rakuten Institute of Technology, Rakuten Group Inc.

A Journey of NLP Researcher from Academia to Industry


Natural Language Processing (NLP) holds the ground for advances in current and future Artificial Intelligence (AI). I started my research in the sphere of NLP in 2008 as a fresh graduate student at Nagaoka University of Technology. My initial research focused on measuring and extracting phonetically similar words between two languages as an information extraction task, which is useful for various linguistic research and machine learning applications. One of my research papers proposed a technique for integrating human knowledge into stochastic models of string similarity. This approach has been utilized to extract Japanese synonyms from e-commerce item data to improve our e-commerce search engine and translation systems. As NLP is one of the core technologies behind AI applications, various deep learning NLP techniques are utilized and developed for many business solutions. In addition, Neural Machine Translation (NMT) is a major part of the research, and many experiments are carried out to improve the quality of the translation of R-Translate models. Augmenting paraphrase generation is the useful approach to improve translation quality of NMT; the current SOTA models (e.g., GPT-2, mT5, etc.) are used to generate multilingual synthetic data. Beyond that, the complexity of Japanese sentence levels from general conversation to business use cases has been considered in the translation from NMT system as well. Recently, my colleagues and I did some research on controlling the complexity of target Japanese text based on the vocabulary and kanji of JLPT levels. We have investigated some limitations of this approach in our workshop submission at TSAR2022.


Natural Language Processing; Neural Machine Translation


Ohnmar Htun is a research scientist at Rakuten Institute of Technology at Rakuten Group Inc., Tokyo-Japan, and Rakuten Asia Pte.Ltd., Singapore. She received her Doctorate in Information Science and Control Engineering from the Nagaoka University of Technology in 2013. She also holds a Master in e-Business Management from the International University of Japan in 2008 and a Bachelor of Honus in Computing and Information Systems from the London Guild Hill University (UK) in 2001. Her research engages in Natural Language Processing (NLP), Cross-Language Information Retrieval (CLIR), Machine Translation (MT), and Deep Learning AI. She has authored several papers in international journals, conferences, and workshops within these domains. She received the IEEE Shin-Etsu Young Researcher Paper Awards (Japan) in 2010. She joined Rakuten Inc., Tokyo, Japan as a research engineer in 2015 and was involved in several research and development projects. She is passionate about practical research innovation on real-world business for a multilingual society.

Paper Submission Instructions

Submission Deadline: September 25, 2022
Notification of Acceptance: October 10, 2022
Camera-ready: November 1, 2022

Paper length:
6-10 pages including figures, tables references.

Submission Format

Papers must be prepared under JIIST format as shown in the link below: https://jiist.aiat.or.th/page/authorGuide

Submission System

The papers must be submitted via the email address: yekyaw.thu@nectec.or.th

Review Process

For papers that are accepted for review, program committee (PC) members will be matched to submissions based on research expertise and interest. Authors will have a limited opportunity to respond to initial reviews. This author feedback will then be taken into account in the final recommendations and reviews may be changed accordingly.

Journal Opportunities

A select set of rated papers may be nominated for fast track reviewing at the Journal of Intelligent Informatics and Smart Technology (JIIST).

Workshop Date

November 5, 2022

Conference Registration and Attendance

At least one author is required to register for the conference to present the paper, and we encourage all authors to attend if possible.

For the workshop presenters, please select “Full Registration for Participant” (2400 Baht). That will cover the whole iSAI-NLP-AIoT conference.

Program Chairs

Ye Kyaw Thu (NECTEC, Thailand)
Thepchai Supnithi (NECTEC, Thailand)
Nongnuch Ketui (RMUTL, Thailand)