Developing Web-based tutoring agents using CORBA

We introduce an agent-based approach for the design and implementation of a Web-based learning environment in the general chemistry pre-laboratory domain by taking advantage of the Common Object Request Broker Architecture (CORBA). Four kinds of agents are investigated: the student modeler that models students´ learning progress; expert agents that simulate domain expertise; the intelligent tutoring tool (ITT) that scaffolds students´ problem-solving processes; and adaptive lecture guidance (ALG) that presents students with an adaptive navigation plan in a Web-based environment. The advantages of this approach are two-fold. First, CORBA middleware provides the distributed communication infrastructure for agents to cooperate and to produce intelligent tutoring behavior. This in turn is important for achieving learning efficiency in a WWW environment. Second, the modularity and platform independent character of these agents has great potential in supporting cost-effective reuse and cooperative development of educational content