Special Research I

Session 8: Study Case in IT Field (2) (Lecturer: Toshihiro Kita )

 Please carry out “This Session's Task” after reading and understanding the following contents.

[Miscellaneous Thoughts regarding My Research Themes]

 When I look back at the research themes I have covered so far, it seems that I can see a few characteristics or philosophies. I think that these are common to all my research subjects, although they changed over the time.

You should not forget the sense of fun when you do university research work.

 Although it is not wrong to do research on a straightforward theme which everyone in society would deem worthwhile, you will not get as much human resources or research funds as those available for the researchers working at private enterprises or certain famous universities. Therefore, my idea is that we should leave those research themes for such people. Instead, I would like to add some “twist” to the straightforward themes to make them unique and worthwhile. In this context, the most important thing here is our “innovative ideas”.

I like a poor research topic.

 I just “like” it. Of course I don't have a large research fund. Therefore what I am saying must sound like an excuse. However, the reason why my ideas are always related to poor topics comes from my personal character. To me it is fun and I like it.

 For example, the poor man's WebCT = Moodle. Also, I feel happy when I, by using what I have developed in my research, become able to do certain things that could not have been done before unless you used expensive software.

 Where there is a lot of money, there is a tendency for “fashionable” and “out of fashion” to go to extreme lengths. There I feel so self-conscious. That is why I don't like “lavish/grandiose” research themes.

When you try hard to achieve a small target in front of you, you often make unexpected discoveries.

 I have heard of a story about a world-renowned study group of a nonlinear system (which used to be my main area of research). Their motto was “Aim at the trivial things”. I understand that this means that you are likely to actually achieve a lot more when you concentrate on what you've got at hand rather than making a grand plan from the outset. You cannot put so called “discovery” on your timetable. I believe that the seeds of the discovery are hidden in any small research topic.

[Research Themes I Am Currently Conceptualizing]

 All of the following are no more than concepts.

Real-time estimation of the learner's state of mind and audio feedback

 I am interested in that which facilitates better learning and/or understanding during online learnnig, when you compare the silent environment and the environment where there is sound, for example music. However, it cannot be any sound, such as noise. Therefore, we monitor the feelings or attitude of the learner continuously in some way or another (for example, we can check the click frequency per second or rate of correct answers of a test) and provide an appropriate sound that facilitates the learning in accordance with the situation. One of the possibilities is to receive and play flash/mp3 streaming as a part of LMS functions so that the sound information can be controlled real-time from the server side. As there is a trial application to generate environmental music-like sound based on a chaotic time series, we can run that application at the client side and send the parameter values to select the kind of sound from the server side. This would be interesting.

A tool that can easily generate 3D animation

 This refers to the development of a tool with which you can generate animations by simply providing numerical data. This would be a tool to generate images to explain physical phenomena, or as a tool to generate animations of which characters and images move around three dimensionally so that things can be explained more easily.

 This tool has a function to create smoothly moving animations by linear interpolation, even if there are discrete data in terms of time. This tool can be operated through a Web interface, making it possible to operate from anywhere in the world. Web Plot3D or gnuplot engine is used.

Use of simple video, voice and animation for the feedback for a quiz (automatically marked test)

 It seems that it would be more effective if the feedback for the incorrect answers to a quiz (hint/explanation which guides the learner toward the correct answer) could be given in the form of video/animation or voice rather than in writing only. However, it is a lot of work to prepare such materials. Therefore, it would be nice if we could work out how to prepare them in as simple a way as possible. For example, when an incorrect answer is given, pattern recognition (template matching/clustering) is carried out to guide the learner to an appropriate video/animation. This should be integrated into the LMS so that it is readily available.

Automatic quiz generation tool for high educational effect and the creation of recipe library site

 Automatically marked quiz is a tool frequently used in e-Learning systems such as LMS. However, if the questions are prepared without careful consideration, learners just hit randomly to eventually get to the correct answer. I would like to develop, in relation to a variety of learning tasks, such a collection of quizzes that lets learners actually learn things when they try to answer them. I would like to develop a prototype of a quiz generation tool using artificial intelligence/expert-system like methods (to generate a quiz automatically by recording the way human specialists correct the answers and/or instruct the learners).

 It would be wonderful if we could create a site where you can post quiz generation tools and know-how, etc. in a similar way as Wikipedia so that community members have a chance to revise them.

I am looking for more incredible ideas than those mentioned above.

[Major Research Works I Have Carried Out So Far]

 Please visit and have a look at http://133.95.4.20/alt-www/research1.html (broken link). There is a list in chronological order starting from the most recent one.

 Reflecting my research career, the trend can be summarized in the following order: