Seventh International Workshop on SCIentific DOCument Analysis
(SCIDOCA 2023)
associated with JSAI-isAI 2023

Program

Day 1 (June 5, 2023)

• 09:50-10:00: Opening
• 10:00-11:00: Session 1 (SC: Prof. Le-Minh Nguyen)
• 10:00-10:30: Towards Enhancing Information Extraction via Public Discussions on Reddit about COVID-19 Research
  Vu Tran, Tomoko Matsui and Minh Le Nguyen
• 10:30-11:00: A Comparative Study of Language Models for Chemical Entity Recognition
  An Dao, Yuji Matsumoto and Akiko Aizawa
• 11:00-14:00: Lunch
• 14:00-15:30: Session 2 (SC: Dr. Vu Tran)
• 14:00-14:30: Reference classification using BERT models to support scientific-document writing
  Ryoma Hosokawa, Junji Yamato, Ryuichiro Higashinaka, Genichiro Kikui and Hiroaki Sugiyama
• 14:30-15:00: Directional Generative Networks
  Yasuaki Ito and Le-Minh Nguyen
• 15:00-15:30: Semantic Parsing for Questions and Answering Over DBLP Database
  Minh Nguyen, Khang Le, Anh Kieu and Y. Nagai
• 16:00-17:00: Invited talk 1 by Dr. Danilo Carvalho (SC: Dr. Vu Tran)
  Title: Learning Disentangled Representations for Natural Language Definitions
  Abstract: While large language models (LLMs) have been pushing the state-of-the-art in a variety of Natural Language Processing tasks, they have an important drawback, which is the difficulty in interpreting their internal representations and controlling them. This fact raises practical and ethical concerns about their use. To alleviate this problem, we propose a method for incorporating information about the role of each word in a dictionary definition towards the meaning of the defined term. This promotes disentanglement of said roles in the internal representations of the models, allowing better interpretation of their properties and a higher degree of control over the generated text.
  Bio: Danilo Carvalho is a Research Associate at the Department of Computer Science at the University of Manchester, working on Safe and Explainable Artificial Intelligence (XAI) architectures. He holds a Ph.D. in Information Science from Japan Advanced Institute of Science and Technology (JAIST), having worked as a systems analyst within Brazilian state oil company (Petrobras) on job safety analysis (JSA) and environmental licensing control systems, followed by several scientific and technological projects in the fields of Artificial Intelligence, Computational Linguistics and Information Systems, with applications to economics, legal informatics, media literacy and healthcare. His other research interests include Parallel and Distributed Computing and Software Engineering.

Day 2 (June 6, 2023)

• 10:00-12:00: Session 3 (SC: Prof. Le-Minh Nguyen)
• 10:00-11:00: Invited talk 2 by Prof. Yoshinobu Kano
  Title: Capabilities and limitations of LLMs: from viewpoint of our research in dialog system, medical, legal and political applications and in human-like NLP system
  Abstract: Recent advantages of the LLMs (Large Language Models, such as ChatGPT) affect not only to the NLP community but the global society of the oridinary people. A question is that the LLMs could solve what sort of NLP problems, though it is still under investigation. Another question is that to what extent the LLMs are "similar" to the humans -- they apparently differ in its architecture and the training data size. I introduce a couple of related NLP projects that I organize, such as automatic medical diagnosis, AIWolf project (dialog system in conversation games), Legal NLP (legal bar exam solver), and SNS analysis project for public opinions, discuss the ability and limitations of the LLMs with future directions.
  Bio: Yoshinobu Kano is an Associate Professor, a Faculty of Informatics in Shizuoka University, Japan. He received BS in physics (2001), MSc (2003) and PhD (2011) in information science and technology from the University of Tokyo, respectively. He was a research associate in University of Tokyo (2009), JST PRESTO researcher (2011), an associate professor (PI) in Shizuoka Univerity (2014-). He is interested in human-like natural language processing and its applications in medical, legal, political and conversational issues. See more here.
• 11:30-12:00: Constructing a Polymer Corpus for Named Entity Recognition and Relation Extraction
  Van-Thuy Phi and Yuji Matsumoto
• 12:00-13:30: Lunch
• 13:30-14:30: Session 4 (SC: Dr. Vu Tran)
• 13:30-14:00: Evaluating How Readable Paper Abstracts Are in Diverse Subcategories under Computer Science for Non-Native English Speakers
  Yo Ehara
• 14:00-14:30: Enhancing Legal Text Entailment with Prompt-Based ChatGPT: An Empirical Study
  Chau Nguyen and Le-Minh Nguyen
• 15:00-16:00: Invited talk 3 by Prof. Aiko Aizawa (SC: Prof. Yuji Matsumoto )
  Title: Natural Language Processing for Scientific Paper Analysis
  Abstract: The rapid development of recent language models has greatly improved the performance of various natural language processing tasks such as information extraction, summarization, or question answering. This talk will present some of the recent techniques in various areas of scientific text analysis to explore the challenges and future perspectives.
  Bio: Akiko AIZAWA is a professor at the National Institute of Informatics (NII). Aizawa is also an adjunct professor at the University of Tokyo as well as at the Graduate University of Advanced Studies. Aizawa's research interests include natural language understanding, dialogue systems, and scholarly document processing.
• 16:00-16:10: Closing