Thursday, July 14, 2011

Research Visit to Gakushuin University

Located towards the northwest of central Tokyo, Gakushuin is a private academic corporation that comprises schools and colleges as well as a university that currently has 9000 students. It is famous in Japan because of its historical connections with the Imperial Court. When I arrived a couple of weeks ago, I found the Mejiro campus spacious and verdent, surprisingly so given its centrality; I was informed that among the universities inside the JR Yamanote loop, it's the second largest after the University of Tokyo.

The present campus, like much of Tokyo, has seen many changes, but among the modern blocks there still remain a few of the older buildings, a little over 100 years old:
University buildings old and new

The modern multi-storey block on the right (East building no. 1?) obviously offers more capacity, but in the distance are some well-maintained old classrooms, cherished by staff and students even today. To the left is the former library building, now housing the Museum of History, graced by a venerable old tree in front of a small pool and foundation:
Venerable tree at the Gakushuin University Museum of History.

However, I wasn't here to be a tourist (although during my week's stay I did have a chance to wander), but rather to explore aspects of e-learning. The opportunity had arisen following earlier exchanges of ideas in the UK: in 2005, Oxford University Computing Services (OUCS) received a visit from Professor Yukari Shirota of the Department of Management, Faculty Economics, Gakushuin University. I arranged for her to give a presentation on some interactive software she had developed that guided students through the study of some topics in mathematics. The system's architecture was based on solution plans to word problems and delivered using an intelligent agent (animated by the Microsoft wizard).

Prof. Shirota is a computer scientist of long-standing - for instance, she co-authored an introduction to UNIX in 1984. During the past decade, Prof. Shirota has been developing e-learning systems to aid in the teaching of mathematics to her Management students. Inspired by George Pólya, her research has been focused on problem solving, invoking techniques in A.I. and especially visualisation, to help make sense of how the formulae and equations are used in word problems in Economics. I was particularly struck by her idea that A.I. might be able replicate the rhythm of instruction from teacher to pupil, quite similar perhaps to the rhythm of communication between a mother and baby. I hadn't come across anything like this thinking in the UK.

More recently, Prof. Shirota's research has concerned the provision of integrated tools that enable academic staff to create a range of online materials to direct students step by step in tackling certain types of questions, particularly in the field of bond mathematics. These systems are typically Web applications with scripts that invoke computer algebra systems such as Maple and Maxima to deliver step by step instructions. Some recent examples of this work are evident in overview of activities for 2011.

With regard to the financial mathematics, Prof. Shirota and her colleagues have used a conceptual approach based on entity-relationship diagrams to relate variables to formulae and equations. So the tasks of understanding may be characterised by being able to visualise and understand this map and its relations. How may that process be best aided online? In its entirety, the complete diagram is too extensive and detailed to show all at once, so any online implementation will need navigation - to focus on particular formulae and relations, but also to 'zoom out' and see the model as a whole. So does this suggest navigating it like, say, Google Earth or some other way based more closely on the relationships?

It's in addressing these considerations that I have been offering some input with my background in mathematics, somewhat distant now, and more recently e-learning and Web development, gained mainly whilst in the Learning Technologies Group at OUCS. Determining effective solutions is a multi-faceted task: its design, especially in terms of user interface, should be soundly rooted in principles of cognition. As I am not trained in educational psychology, I have to deal more with the nuts and bolts of the learning context itself, primarily in terms of the current and emerging technologies. In this regard, I'd say that whilst the emphases in pedagogy vary from country to country, the predominance of personally own computing devices has presented new variables to factor into the education at all levels. It's particularly this phenomenon, which I tend to call mobile and ubiquitous computing, that I was encouraged to explore ahead of my visit in the form of a survey paper on e-learning systems for mathematics, with particular reference to business and economics.

To give some structure in this rather broad landscape, I initially made reference to some recent features of e-learning in the UK. There's a strong focus on learner-centred education; within that I'm particularly interested in processes of deep reflection, stemming from the RAMBLE project I led in mobile blogging and learning environments. I'm not sure about the merits of focusing so much on individual predispositions, but I felt that the work on open educational resources and podcasting as a delivery mechanism was a useful vehicle to illustrate how learning has emerged from being concentrated in a classroom or workshop and flowed out into less formal environments. It was with this image in mind that I developed a thread to show how computer assisted learning for mathematics has similarly emerged from the laboratory into the open and is now squarely aiming at handheld tablet and multi-touch devices.

The paper, whose full title is 'Ubiquitous e-Learning: Designing Web Systems for Economics and Business Mathematics', has been published in Gakushuin Economic papers and is now available online - in HTML and PDF formats.

No comments: