Tanja Mitrovic is a Professor at the Department of Computer Science and Software Engineering, and the Associate Dean of the College of Engineering, the University of Canterbury, New Zealand. She is the leader of the Intelligent Computer Tutoring Group, and the past President of the AIED society (2013-2015). She is an associate editor of the following journals: International Journal on Artificial Intelligence in Education, IEEE Transactions on Teaching and Learning Technologies, and Research and Practice in Technology Enhanced Learning (RPTEL). Tanja has co-authored more than 230 papers, most of which were published in key journals and conferences on intelligent educational systems. She was the local organizing chair of AIED 2011, the general chair of AIED 2015, and is currently the local chair for ICCE 2017.
From databases to prospective memory: the saga of CBM continued
Twelve years ago, I presented an invited talk at AIED 2005, which focused on the early days of the Intelligent Computer Tutoring Group, and the tutors we developed. Our early work focused on teaching design tasks, such as database querying and design. Since then, we have employed CBM successfully in many other domains. Some of those tutors also taught design tasks, such as Java programs and UML design, while other were procedural in nature. We also developed ASPIRE, an authoring system and deployment environment for constraint-based tutors. ASPIRE has served as the foundation for developing new tutors, ranging from teaching how to solve thermodynamics problems, manage oil palm plantations, diagnosing problems with X-ray images. ASPIRE allowed embedding constraint-based tutors into other software packages, such as accounting software and management information systems. It also allowed having sophisticated interfaces, such as the Augmented Reality interface of MAT. During these 12 years, we were successful in developing a constraint-based model of collaborative skills, modeling meta-cognitive skills and affect of our students. We also investigated feedback strategies, especially the effect of how feedback is phrased on learning, and the effect of positive feedback. The most recent studies focused on multiple teaching strategies: comparing learning from problem-solving, worked examples, and erroneous examples. And then we investigated whether we can model prospective memory using constraints; in a recently completed project, the prospective memory functioning of 15 stroke survivors increased significantly after 10 sessions of computer-based training on how to memorize prospective tasks and practicing in a Virtual Reality environment. In this talk, I will present highlights of our recent projects.
Riichiro Mizoguchi is a Research Professor at the Japan Advanced Institute of Science and Technology (JAIST) in the School of Knowledge Science in the Knowledge Management Area. His research focuses on the fundamental theory of ontology, upper ontology, domain ontologies such as medical, learning theories, sustainability science is conducted together with their application system development through Intelligent learning support systems. He has served as an Associate Editor for IEEE Transactions on Learning Technology; been an Editor -in-Chief for the Journal of Web Semantics and been a prior Conference Co-chair for the Joint International Semantic Technology Conference (JIST). He has numerous awards and peer-reviewed publications from his previous research including the Life Science Award for the Linked Open Data Challenge in Japan and numerous best paper awards.
An AI Methodology and a New Learning Paradigm
My talk consists of two topics: One is how ontological engineering (OE) as an AI methodology helps you modeling of AIED matters and the other is Negotiation-Driven Learning: NDL as a new learning paradigm. After reviewing several AI methodologies, I discuss OE to explain it is a promising methodology and it contributes to modeling rather than to metadata. I will try to convince you that it provides a powerful conceptual tool to tackle and handle complex objects/concepts/theories/systems/etc. It also enables you to design systems with clear separation between domain-dependent and domain-independent parts, which is exploited in the research on NDL. NDL is a new learning paradigm in OLM, in which I have been intensively involved with my former Ph.D. student, Raja Suleman recently. It is a framework built by integrating dialog-based tutoring, interest-based negotiation and affective computing in the negotiation process of OLM. I will discuss its role in AIED in terms of learning paradigm and methodology of system design.
Ronghuai Huang is a Professor in Faculty of Education and Dean of Smart Learning Institute in Beijing Normal University (BNU) in which educational technology is one of the National Key Subjects, and director of R&D Center for Knowledge Engineering, which is dedicated to syncretizing artificial intelligence and human learning.
Prof. Huang has being engaged in the research on educational technology as well as knowledge engineering since 1997. He has accomplished or is working on over 60 projects, including those of key science and technology projects to be tackled in the national “Ninth Five-year Plan”, “Tenth Five-year Plan” and “Eleventh Five-year Plan” and the projects in the national 863 plan as well as others financed by the government. His ideas have been widely spread, with more than 180 academic papers and over 20 books published in the domestic and overseas.
Prof. Huang is very active in academic organizations both at home and abroad. He was Co-chairs of the Program Committee of the 6th and 8th Global Chinese Conference on Computers in Education, and chairs of the Organizing Committee of the International Conference on Computers in Education (ICCE2006), the 10th IASTED International Conference on Computers and Advanced Technology in Education (CATE2007) and the 5th International Conference on Wireless, Mobile and Ubiquitous Technologies in Education (WMUTE2008). He was the General Co-Chairs of the 5th International Conference on Advanced Data Mining and Applications (ADMA2009) and the International Conference on Hybrid Learning (ICHL2010). He is editor-in-chief of Global Chinese Journal for Computers in Education (GCJCE). In addition, he is currently serving on the Executive Committee of Asia-Pacific Society for Computers in Education (2008-2011), Chairman of CSCL2011 post-Conference, Chair of the 13th 14th, 15th and 16th International Conference on Advanced Learning Technology (ICALT 2013, ICALT 2014, ICALT 2015, ICALT 2016) and General Chair of 1st, 2nd, and 3rd International Conference on Smart Learning Environment (ICSLE 2014, ICSLE 2015, ICSLE 2016). He is also the president of International Association of Smart Learning Environments (IASLE) and Editor-in-Chief of Springer’s Journal of Smart Learning Environment and Journal of Computers in Education.
A Conceptual Framework for Smart Learning Engine
In a life-long learning society, learning scenarios can be categorized into five types, which are “classroom learning”, “self-learning”, “inquiry learning”, “learning in doing” and “learning in working”. From a life-wide learning perspective, all these scenarios play vital roles for personal development. How to recognize these learning scenarios (including learning time, learning place, learning peers, learning activities, etc.) and provide the matched learning ways (including learning path, resources, peers, teachers, etc.) are the basis for smart learning environments, however, few research could be found to address this problem.
In order to solve this problem, we propose a conceptual framework of smart learning engine that is the core of integrated, interactive and intelligent (i3) learning environments. The smart learning engine consists of three main functions.
The first function is to identify data from student, teacher, subject area, and the environment using wireless sensors, the established learning resources and scenarios, and a learner modeling technology. The acquired data includes prior knowledge, theme-based context, learner/teacher profile, physical environments, etc.
The second function is to compute the best ways of learning based on the learning scenario and learning preference. In detail, this function includes modeling learner’s affective data, building knowledge structure, optimizing knowledge module, and connecting learners.
The third function is to deploy personalized and adaptive strategy, resources and tools for students and teachers based on the computed results in the second function. Deploy interactive strategies, learning paces, learning resources, and delivery approaches are the core elements for this function.
Dr. Sannyuya Liu is the deputy director and a professor of National Engineering Research Center for E-Learning. He received a bachelor’s degree in Environmental Engineering in 1996 from Northeastern University and a master’s degree in Artificial Intelligence from Hohai University in 1999. In 2003, he graduated from Huazhong University of Science & Technology (HUST) and received his Ph.D. degree in Systems Engineering. After two-year experience in Xiamen University as a post-doctoral researcher, he joined National Engineering Research Center for E-Learning in 2006. His research interests include data analytics and education, artificial intelligence, and ICT in education. He published more than 40 SCI/SSCI/EI academic articles and 6 books in the past decade. He’s the copyright holder of more than 20 national patents and has been awarded First and Secondary Prizes for Scientific and Technological Progress of Hubei Province.
As the project leader, Dr. Liu participates in many national scientific and technological projects and makes remarkable attributions. In 2011, he was awarded the “New Century Excellent Researcher” Award by Ministry of Education of China. In 2012, he was enlisted as the “Elite Talent in the New Century” by Human Resources Department of Hubei Province. For his remarkable contribution, Dr. Liu was awarded the “Innovative and Entrepreneurial Talent” Award by Hubei Association of Science and Technology in 2016. His professional affiliations include ICT in Education Committee of Chinese Society of Educational Development Strategy, ICT in Education Expert Panel of Ministry of Education of China and Hubei System Engineering Society.
Emerging technologies, including internet of things and big data, are leading to educational revolutions in the learning environment, learning applications, and learning approaches. Recent advancement in data collection and data analysis offers opportunities for accurate description and quantification of learning activities. Quantified Learning, refers to the process of utilizing appropriate approaches and methods to gain insights from students’ explicit and implicit behavioral features, and offering analysis and intervention services to accommodate students’ personalized learning needs. With “learner-centered” philosophy, Quantified Learning will develop data-oriented perception and effectively facilitate knowledge construction and personal development. With data, learners, stakeholders, and connected learning services, Quantified Learning is a closed-loop with adaptive feedbacks. The four stages of quantified learning, including quantification, data collection, integration and analysis, and intelligent services will enhance research and practices of teaching and learning with more accuracy and intelligence.