Information Technology Reference
In-Depth Information
actions, and landmarks [22]. Landmark is essential and it is more important
than orientation and actions in guidance [10]. Sorrows and Hirtle [18] have
proposed three landmark categories: visual, cognitive, and structure [18].
Each landmark category has to use a different navigation method.
16.3 Multi-Agent Architecture
Each learner is running the front-end of a learner agent on his or her own
mobile phone or PDA supporting J2ME with the Mobile Information Device
Profile (MIDP) 1.0 or greater APIs and Bluetooth JSR-82 or Wi-Fi APIs. A
back-end mediator in the JADE-LEAP architecture, running on a host with a
known, fixed, and visible address can automatically manage the back-ends
of all front-ends active on the learner's device. Some types of agents located
in a JADE agent platform are as follows:
•
A main container, which stores default JADE agents.
•
One or more containers, which execute system agents.
•
Location-aware learner agents, one for each learner, which present each
learner logged in within the area. The learner agent has capabilities,
using the expressed preferences of learners or their previous behavior
when navigating the system to provide information for the student
model.
Connection agents manage the connection between the mobile
•
devices and agent platform.
Service agents are available for each service such as location-aware
•
service.
Resource agents can manage resources, such as problem-based
•
learning objects.
Figure 16.2 shows the Multi-Agent Architecture for ubiquitous learning.
16.4 Location-Based optimal Grouping
The location-based optimal grouping service is to group, geographically,
nearby students together to create ad hoc online learning groups. In this
service, the positioning accuracy is not critical. Proposed conceptual system
architecture for the grouping service is shown in Figure 16.3.
