/ Qun Jin / Associate Professor
/ Timothy K. Shih / Visiting Professor
/ Jianhua Ma / Assistant Professor
We seek to exploit the rich interdependence between theory and practice in our works. We conduct research works extensively related to information systems, including internet computing, multimedia computing, and agent-based cooperative computing.
For the fiscal year of 1999, we have been engaged in research works on virtual universities and distance education, computer supported collaborative work/learning, collaborative virtual environemts and communication support, intelligent user interface, agent-based cooperative framework for knowledge creation and exchange, behavioral modeling of autonomous agents in distributed environments, and others.
Major research projects are described as follows.
The primary goal of this project is to propose and build an innovative educational system for the coming new century. We propose a conceptual framework for Community-Based Collaborative Virtual University based on collaborative learning theories, in whcih computation is effectively used to actively support human-human communication, collaboration and interaction in addition to human-computer interaction.
Community-Based Collaborative Virtual University widely opens to large and diverse group of people who have the will to learn and to share their knowledge with others across the networks. It is participants-driven. That is, participants or learners share a common interest in a topic or area, share a way of knowing and a set of practices. Community-Based Collaborative Virtual University supports different ways for novices and experts to work in the same environment to accomplish similar goals. It may be large, the task general, and the communication open. Alternatively, it can be small, the task specific, and the communication close.
We propose and develop a collaborative virtual environment to flexibly support community learning, which has been designed so that a learner can navigate through it, select relevant information, respond to questions using computer input devices such as a keyboard, mouse, or voice command system, solve problems, complete challenging tasks, create knowledge representations, collaborate with others, or otherwise engage in meaningful learning activities.
In recent years, the Internet based online communities, also known as virtual communities, are becoming an increasing part of how we work, play, and learn. They cover a very broad field. Due to the diversity of community members (users of the system), it becomes important and critical to provide a real user-freindly interface to access the virtual environments, e.g., Community-Based Collaborative Virtual Universities.
In this study, we propose a new model of every-citizen user interface, and design and develop a prototype software that is highly integrated with multimedia and flexibly adapts to user's behavior, to support easy access to the community-based virtual universities. Key technologies for a user interface to be every-citizen easily accessible include: Mechanism to facilitate the electronic communication between large and diverse groups of people, depending on an underlying social conduct of co-presence and information sharing across the networks; Support for social interaction between users and between human users and software agents that co-exist in the virtual community, by applying a cooperative multi agent paradigm; Predictive and demonstrational interface paradigm with user-centered awareness support; Multimedia (such as MPEG1, MPEG2, and/or MP3 data) integration to the virtual enviroment and seemless integration with the Internet resources.
During the past several years, the spread of concurrent and distributed computing and deeper integration of computing into the lives of people have led to increasing collaborations among large collections of interacting people and large groups of interacting machines. Agent-based computing has been proposed to support such collaborations flexibly and reliably, in which an agent supports a user, represents the user to the system, and handles complex interactions with other cooperating agents and system resources. On the other hand, in this information age, people need to do more than just find, collect and/or transmit information; they need to know how to separate the fluff from the substance. Creating or deriving knowledge from recent flooding information requires more and more effort. There is a growing desire to aid such effort using computer systems.
This study focuses on effective use of information on the Internet, or Internet resources. We propose a model of knowledge creation from ill-formed, diverse and dynamically changing information with reorganizing process using internally existing knowledge and external knowledge. We further propose a distributed cooperative framework between users and the system and between users, by incorporating intelligent agents both on the interface and within the system.
Software agents have been invented to facilitate the creation of software programs able to interoperate - to exchange information and services with other programs and thereby solve problems that cannot be solved alone. When interacting and acting in a dynamic multiagent environment, an agent must ask itself what is presently its best course of action given what it now knows about what the environment will be like when it intends to act. It requires some ability to make decision by computing the probability that relevant propositions will hold at a specified point of time. It has been a critical issue to represent and model coordination, interaction, and communication among agents. Although a variety of research work on agent theories, architectures and languages have been done, little work has been conducted so far in the aspect of modeling and analyzing for agents' behavior. Several particular representation techniques have been developed, but it still lacks of effective representation and modeling method.
Petri nets are well known as a graphical and mathematical modeling tools applicable to various systems. Petri nets and their extensions (e.g., stochastic Petri nets, high level Petri nets) are a promising tool for describing and studying systems that are characterized as being concurrent, asynchronous, distributed, parallel, nondeterministic, and/or stochastic. Petri nets could be effective in representing and modeling coordination, interaction, and communication for software agents. In this study, we define a new class of stochastic Petri net: non-regenerative stochastic Petri net (NRSPN), which automatically generates a bounded reachability graph that is equivalent to a generalization of the Markov renewal process. We use NRSPN to model and analyze autonomous agents' probabilistic behavior.
Refereed Journal Papers