Mizuo Kansen

A personalized ubiquitous education support environment by comparing learning instructional requirement with learner's behavior

With the progress of ubiquitous techniques and devices, such as RFID and sensors, etc., invisible computers are embedded in ones life and working environment, such as applicants, furniture, stationaries etc. Research on ubiquitous environment becomes very important. On the other hand, E-learning using WWW has been studying prosperously. Merging creatively the techniques in the two fields to construct a support environment of ubiquitous learning is demanded. Specially, the ubiquitous learning environment is suitable for outdoor learning to increase the students in middle school and high school for increasing their experiences and knowledge from real world, not only from text books. In this paper, we propose a new personalized ubiquitous learning support environment. In the ubiquitous learning environment, every learners behavior or learning habit can be sensed and observed precisely. By comparing them with learning instructional requirement set by an instructor, the learners situation can be grasped. Personalized support suitable for the learners situation can be provided to the learner. The proposed environment will be very useful for promoting ubiquitous learning.

A proposal on a learner's context-aware personalized education support method based on principles of behavior science

With the progress of ubiquitous techniques and devices, such as RF-ID and sensors, etc., invisible computers are embedded in ones life and working environment, such as applicants, furniture, stationeries etc. A ubiquitous network environment provides personalized context-aware services to a user by making good use of ubiquitous techniques such as RF-IDs, sensors, actuators. The services can be considered as an extension of Internet services in a cyber world into the real physical world. As an application of ubiquitous network environment, an ubiquitous learning support environment can provide personalized learning support to a learner according to the learners learning situations, in not only Web-based learning environment but also in real world, through the books, text, pens, tables and other objects surrounding him/her. In this paper, we propose a ubiquitous learning method, which is based on principles of behavior science to provide suitable learning support by collecting the learners data regarding the learners learning behaviors, and provide guidance messages based on analysis of the data.

A New Security Testing Method for Detecting Flash Vulnerabilities by Generating Test Patterns

Flash has a number of security defects even though Flash Player is installed on most of world’s PC. Protection using sandbox has limitation to protect a user from vulnerabilities of Flash application because an attacker can attack a vulnerable Flash application when a sandbox can’t work if an engineer or a web administrator set sandbox permission wrongly. Another way to solve it is testing. As a testing, penetration testing is useful for detecting vulnerability of Flash Application. Existing penetration testing performs penetration test through UI manually, which is inefficient and time consuming. In this paper, to overcome a problem of existing penetration test, we design a new penetration testing, which enables to generate as many test patterns as possible from VM inputs, inputting test patterns into VM, and checks the existence of vulnerabilities from VM outputs automatically. We demonstrate our testing method using an example, which can detect Flash Parameter Injection that is a one kind of vulnerability of Flash application.

Troublesome detectable context-aware services based on game-theory

With rapid progress on ubiquitous technologies, various services become available at anytime and anywhere. One of hot topics in ubiquitous computing is context-aware services, which provide the services at proper time and proper place, in a proper way personalized to a user. However, there may be conflicts of services among users whose contexts are overlapped. Moreover, a context-aware service convenient for a user may cause another users unpleasantness even troublesome. In this paper, we consider the relations between contexts of different users. A mechanism consisting of multi-agents, to find and reduce the unpleasantness of users due to context-aware services is constructed based on game-theory.

A courseware platform for safety education through dual experience based learning using an integrated environment of a miniaturized world and a computerized world

Ubiquitous learning is one of fields utilizing widespread ubiquitous technologies and gives learners more effective and personalized education than the traditional one. Safety education is necessary for children since accidents and crimes against them happen increasingly in recent years. The children might need to protect them from dangers by themselves. Though it is important for the children to learn safety knowledge through real experiences on danger, it is not always realizable because of lack of opportunities and possibilities of hurting the children. In contrast, a virtual experience is possible and popular to educate the children in some ways such as using computers. However, only virtual experiences might be insufficient for children to practice what they learnt in the real world. Therefore we propose a novel approach by fusing a real world and virtual space to let a learner have dual experiences in both real and virtual worlds. We design a support system which employs the following techniques: (1) a platform of dual experiencing learning, (2) dual experiencing learning scenario design policies, and (3) personalized dual experiencing learning course.

A composition method for manipulating detectable situations to provide situation-aware services

One of the hot topics in ubiquitous computing is context-aware services, which automaticlly adapt user based on users context information, e.g., location, time, favorite etc. Situation-aware services are advanced context-aware services, such that more complex context information and relations among users, objects, and environment are taken into consideration. In order to provide the situation-aware services, it is necessary to describe the situations, to define and link services which will be provided in the situations, and to detect the happening of the situations in the real world. Though many researches have worked on context-aware services, there are not so many on situation aware services, especially on describing and detecting the situations uniformly. In this paper, we take a composition approach to solve the problems. We first define primitive situations which are based on J. Barwises situation theory and can be detected by current ubiquitous techniques. And then, we present several manipulation rules, by which more complex situations can be composed. Our methods can be used as a building block in construction situation-aware services, such as warning to a user to prevent him/her from dangerous situations, reminding a user not forgetting to bring belongings with him/her, navigation to a handicapped user, etc.

M-LT/R·W method based Mobile-Learning supporting systems

In recent years, e-Learning has been used very widely because of rapid spread of computers and network. Furthermore the mobile-phone is so well spread that almost everyone owns a mobile-phone because of its simplicity. Many people carry their mobile phone anywhere with them. We can say that a mobile-phone becomes a commodity for everyone. The purpose of this paper is to propose a mobile - learning support system using LT/R ·W method. LT/R ·W method is one of techniques for getting learning diagnosis information from each answer in multiple-choice test. Using LT/R ·W method we are able to grasp the learners understanding state not only from judging if the answer is right or wrong but also from a relationship between the level of difficulty of each question and the level of understanding of each learner. And also we are able to grasp from a relationship between the amount of time that a particular learner takes to solve each question and the amount of time to do so in general. We design the learning support system using LT/R ·W method and implement it as an “i-application (for NTT docomos mobile phone)”. Moreover we carry out the evaluation of the system through a questionnaire to users.