Fugee Tsung
Professor Fugee Tsung is a globally recognized expert in industrial analytics and quality engineering, listed among the top 2% of most influential scientists worldwide by Stanford-Elsevier Mendeley Data 2023. As a Chair Professor at HKUST and HKUST(GZ), he directs the Industrial and Intelligence Institute (Triple-I Institute) and the Quality and Data Analytics Lab (QLab). He has held prominent positions such as Editor-in-Chief for the Journal of Quality Technology (JQT), Head of the Department of Industrial Engineering and Decision Analytics, and founding acting Dean of the Information Hub at HKUST(GZ). A fellow of esteemed organizations like ASA, ASQ, IISE, IAQ, and HKIE, Prof. Tsung has an extensive publication record and has mentored many successful doctoral students. He holds a Ph.D. and an MSc from the University of Michigan, Ann Arbor.
Keynote Title
“Harnessing Industrial Informatics and Intelligence in Service Science: A Pathway to Future Innovations”
Keynote Abstract
This keynote presentation delves into the transformative role of industrial informatics and intelligence in driving digital transformation across service industries within the framework of Industry 4.0. It highlights how the integration of artificial intelligence, particularly through AI-generated content (AIGC) and large language models (LLMs), is reshaping interactions between technology and human expertise, emphasizing the cultivation of innovation as a critical skill. Insights will be drawn from recent advancements at the Industrial Informatics and Intelligence Institute (Triple-I Institute) and the Quality and Data Analytics Lab. Additionally, the session will feature the HKUST 2.0 initiative, illustrating the integration of technology, arts, and education in innovating service science. The discussion aims to spark dialogue on leveraging interdisciplinary approaches to enhance service sectors, showcasing technology’s transformative potential in education and industry.
Robin Qiu
Dr. Qiu is a worldwide pioneer leading the development of Service Science (an interdisciplinary field of AI, Data Analytics, Big Data, Management Science, Social Science, and Computing Science). Dr. Qiu has actively promoted service science education and research internationally, aimed at developing the needed knowledge and skills required in today and the future’s service-led global economy. In addition to initiating international conferences in the area of Service Science, Dr. Qiu worked with many international scholars to have founded the Service Science Section of INFORMS (in 2006) and the Logistics and Services Technical Committee in the IEEE Intelligent Transportation Systems Society (in 2005). Service Science, a fully refereed journal, was launched in 2008 and became an official INFORMS journal in 2011 under his leadership and vision. He has been working diligently with many other pioneers in this emerging research, education, and application field to make Service Science an INFORMS flagship journal, facilitating the development of Service Science to better serve academics and practitioners in this field worldwide. Dr. Qiu served as the editor-in-chief of INFORMS Service Science. He was an associate editor of IEEE Transactions on Systems, Man, and Cybernetics and an associate editor of IEEE Transactions on Industrial Informatics. Dr. Qiu is the Editor-in-chief of Digital Transformation and Society published Emerald Publishing and SpringerBriefs in Service Science by Springer. He has had more than 180 publications, including 3 books.
Keynote Title
“Feeding Right Data to Large Language Models”
Keynote Abstract
The quick advances and powerful capabilities of large language model (LLM) have recently stimulated the emergence of a lot of LLM applications worldwide. Due to the flocking-in effects of capitalism or marketing propaganda, many of those LLM applications were developed and deployed for the purpose of entertaining or show-off. In the field of machine learning, “garbage in, garbage out” is a well-known norm. Because of limited or sometime bad data, AI policymakers and educators are deeply concerned with the negativity of the outcomes derived from LLM applications, including but not limited to serious nonsense, disinformation, bias, and ethical implications. This talk presents an effective and efficient approach by focusing on collecting massive trusted data across networked platforms to train an LLM and aggregating quality data to finetune the LLM, aimed at deploying problem-solving LLM applications in the field, in particular in the field, such as healthcare and manufacturing, that current, accurate, and precise knowledge is the key to all stakeholders. The explored framework for developing LLM applications can help avoid the potential of generating a serious nonsense as an answer to an end user’s question, ensuring that developed LLM applications are of high-quality and readily acceptable in practice.
Two essential data-oriented technologies will be explored. First, blockchains enabling distributed data will be applied for collecting massive trusted data across distributed and networked platforms. Secondly, graph-based knowledge bases aggregating quality data in a logic manner will be utilized to generate the input to finetune LLMs, resulting in high-quality and readily acceptable domain-specific LLM applications. As an example, an LLM application in an automated automotive body welding line will be presented.