Invited Keynote Speakers
Title: Design of Moderately Ill-Structured Task and Intelligent Support with Information Structure Open Approach
Abstract: Problem-solving tasks are often categorized into two types: ill-structured one and well-structured one, in the context of education/learning, cognitive science and artificial intelligence. Then, from an educational viewpoint, ill-structured tasks are further more important because they are useful to promote a learner to think about target learning contents deeply and to master computational or logical thinking skills, including metacognition and self-regulation. This talk presents an additional characterization of problem-solving tasks by using two factors, (1) well/ill-structured domain model and (2) well/ill-structured task setting. Here, a well-structured domain model provides a problem space as a set of states and a set of operators linking one state with the next. Then, a well-structured task setting provides a specific problem space with an initial sate and a goal state. Based on this characterization, “moderately ill-structured tasks” can be defined as a category of tasks specified by “well-structured domain model” and “ill-structured task setting”. If a task is set in well-structured domain model, it is possible to realize computer-based monitoring and diagnosis of learner’s activities for the task. If the task setting is ill-structured, for example, open-ended, a learner is required to engage in the task as ill-structured one. Therefore, moderately ill-structured tasks are promising to realize computer-based intelligent support for solving the tasks, while keeping educational advantages of ill-structured tasks. In this paper, a definition of moderately ill-structured tasks is described. As a method to realize scaffolding and intelligent support, information structure open approach where the domain model is represented as information structure and open for learners to direct manipulation are proposed. In this talk, using arithmetic word problems as an example of learning contents, (1) well-structured domain model of arithmetic word problems, and (2) design of moderately ill-structured task as “problem-posing assignment” based on the information structure model are introduced. Moreover, (3) implementation and practical uses of several intelligent learning environments that support learners to solve the moderately ill-structured tasks as problem-posing are reported.
Bio: Tsukasa Hirashima received his B.E., M.E. and PhD from Osaka University in 1986, 1988, and 1991 respectively. He worked at The Institute of Scientific and Industrial Research, Osaka University as a research associate and lecturer from 1991 to 1997. During 1997-2003, he worked in Graduate School of Information Engineering at Kyushu Institute of Technology as an associate professor. He has been a professor of Graduate School, Department of Information Engineering, Hiroshima University since 2004. Dr. Hirashima's contributions in Computers in Education, especially, in artificial intelligence in education include modeling of problem-solving process, error-visualization for error-awareness, information filtering, question/problem generation, learning by problem posing and design method of learning game. His research activities are rich in originality, and have impacted the flow of researches on artificial intelligence in education. Because of the originality and impact, he has received several awards from major international conferences about computer and education (World Conference on Educational Multimedia and Hypermedia (ED-MEDIA1995, Outstanding Paper Award), International Conference on Computer in Education (Best Paper Awards at ICCE 2001&2002, Best Technical Design Paper Award at 2015), and Artificial Intelligence and Education (AIED2009, Honorable Mention Award)). In domestic in Japan, he received Best Journal Paper Award from Japanese Society for Information and Systems in Education in 2008 and 2009 respectively. He served as PC co-chair of host country and city in ICCE2007. He also served as President of APSCE from 2012 to 2013. Now he is a member of executive committee of APSCE.
Title: From OLMs + LAK + PI to PIL, Personal Informatics for Learners
Abstract: This talk presents a vision for Personal Informatics for Learners (PIL). This builds on three foundations. The first, and oldest of these, comes from the decades of research on Open Learner Models (OLMs) in Artificial Intelligence in Education. These fields strive to create high quality teaching systems with personalisation driven by a fine-grained, carefully crafted model of the learner. The second is the far newer, but very fast growing LAK, Learning Analytics and Knowledge. This arose with wide deployment of learning technology that captures huge quantities of learning data and the tantalising possibilities such data offers. The third, Personal Informatics (PI) comes from Ubicomp research, aiming to harness personal sensor data about diverse aspects of life, particularly for health and wellness. PIL is grounded on the view that the learner should be empowered to take responsibility for their own learning. To achieve this, PIL needs to make progress on many fronts. We need mechanisms for collecting the right learning data and managing it effectively. We also need new classes of the interfaces and systems that enable learners to control their data, do personal data mining and engage in meta-cognitive processes of reflection, self-monitoring, planning.
The talk begins by reviewing the broad scope and vision described above. Taking that lens, I then present a two sets of case studies. The first relates to group-work, with interfaces onto data harvested as groups use an online collaboration tool or collaborate around an interactive tabletop or wall display, with diverse models, from the simple to richer ones built by data mining. Key insights come from the series of studies, both in the lab and in authentic classrooms. The second set of case studies is for individual learning, ranging from mastering computer software and curriculum-wide learner modelling to personal informatics, for health and wellness, harnessing data from activity trackers, virtual reality and mobile food logging. The talk concludes with key lessons learnt and a research agenda for PIL.
Bio: Judy Kay is Professor of Computer Science. She leads the Human Centred Technology Research Cluster, in the Faculty of Engineering and IT at the University of Sydney and the CHAI, Computer Human Adapted Interaction, Research Group. Her research areas are in human computer interaction (HCI), ubiquitous computing (Ubicomp) and Artificial Intelligence in Education (AIED). A core focus of her research has been to create infrastructures, tools and interfaces for personalised lifelong life-wide learning. Central to this has been in the design of Open Learner Model interfaces that enable people to scrutinise and control the system's model of them and personalisation processes based on it. Her interface work has created the Cruiser Natural User Interaction (NIU) software which provides new ways for people to interact with interactive large surfaces, on tabletops and walls. Her research has been commercialised and deployed and she has extensive publications in leading venues for research in AIED, human computer interaction and personalisation. She is co-Editor-in-Chief of the International Journal of Artificial Intelligence in Education.
Title: From Augmented to Virtual Learning: Design Affordances of Different Mixes of Reality for Learning
Abstract: Mixed realities that combine digital and real experiences are now becoming a true reality. These experiences are being delivered over smartphones as well as increasingly accessible and practical head mounted displays. This ubiquity of devices is in turn making mixed reality the next digital frontier in entertainment, education and the workplace. But what do we know about where these technologies have value? Where do they add to the learning experience? And what theories and evidence can we generate and build upon to provide a foundation for using these technologies productively for learning? We have been working on mixed realities in education for over a decade and have started to learn about where, when and for whom they can add value. Part of this understanding stems from differentiating the wide variety of mixed realities and focusing on affordances. Landscape based Augmented Realities, popularized by Pokemon Go, have fundamentally different affordances than smartphone based Virtual Realities like Google Cardboard, which in turn are different than immersive experiences delivered by headsets like the Oculus Rift and HTC Vive. The core of our work has been doing research and development to identify these affordances that match with key learning challenges, particular in Science, Technology, Engineering and Mathematics (STEM). In this talk, I will draw upon our work in location-based Augmented Reality games, as well as work in Virtual Reality. In the realm of Augmented Reality, I will discuss a long series of design experiments through which we have learned about where these technologies play an important role in learning, primarily around socio-scientific issues. In the space of Virtual Reality our newest designs and experiments focus on the concept of scale, and how we can use Virtual Realities to teach about STEM systems at radically different scales. This talk will provide a history and overview of these experiences, including iterations of design research experiments.
Bio: Eric Klopfer is Professor and Director of the Scheller Teacher Education Program and The Education Arcade at MIT. He is also a co-faculty director for MIT’s J-WEL World Education Lab. His work uses a Design Based Research methodology to span the educational technology ecosystem, from design and development of new technologies to professional development and implementation. Much of Klopfer's research has focused on computer games and simulations for building understanding of science, technology, engineering and mathematics. He is the co-author of the books, "Adventures in Modeling", "The More We Know", and the upcoming “Resonant Games”, as well as author of "Augmented Learning,” His lab has produced used by millions of people, as well as online courses that have reached hundreds of thousands. His work has been funded by federal agencies including NIH, NSF and the Department of Education, as well as the Gates Foundation, the Hewlett Foundation, and the Tata Trusts. Klopfer is also the co-founder and past President of the non-profit Learning Games Network (www.learninggamesnetwork.org).
Title: New directions in personalized learning: Open, informal, social
Abstract: A goal of educational technology since the 1930s has been to adapt teaching to the personal needs of each student. Significant developments have included programmed instruction, branched instruction, intelligent tutoring systems, and adaptive courseware. Personalized learning is coming back into prominence with the development of new techniques for linking learning analytics to adaptive teaching. Research challenges include how to enable personalization of informal and inquiry-led learning, and how to link personalization with learning through conversation and social networking.
Personalized open learning must offer opportunities for students from widely differing backgrounds to learn in ways that match their needs and abilities. This requires new designs for flexibility of timing, pace, facilitation and assessment. For informal learning, personalization must align with changes in context, learning materials co-created by students, and self-directed study. In social networked learning, students need support to merge their individual pathways through the curriculum into shared goals, positive interdependence and productive conversation.
I shall discuss recent work at The Open University on predictive analytics and flexible pathways for learners, as part of a strategic university initiative in personalized open learning. Our iSpot and nQuire-it platforms combine informal science learning with personalization through reputation management and student authoring. Adaptive crowdsourcing may offer mechanisms for personalized social networked learning.
Bio: Mike Sharples is Professor of Educational Technology in the Institute of Educational Technology at The Open University, UK. He leads the Minerva project to transform the University’s approach to course development. He also has a post as Academic Lead for the FutureLearn company. His research involves human-centred design of new technologies and environments for learning. He inaugurated the mLearn conference series and was Founding President of the International Association for Mobile Learning. He is Associate Editor in Chief of IEEE Transactions on Learning Technologies. He is author of over 300 papers in the areas of educational technology, science education, human-centred design of personal technologies, artificial intelligence and cognitive science.