| Distance Learning via Web-based media is becoming an efficient way to convey education and training to a continuously growing and diverse community of people. This rich multi-media environment appears as a promising alternative to traditional classroom instruction, and helps to overcome problems like geographical distribution, accessibility limitations, and scarce human resources. Unfortunately, at the same time it can leave students isolated from peer support. Identifying student needs and providing adequate remedial help can be an awkward and challenging process.; This research focuses in the development of a general framework called Partner-Finder that supports the study and analysis of the peer-to-peer interaction problem in Web-based education. The Partner-Finder framework creates and maintains a representation of the overall state of knowledge of an entire class of students taking a particular distance learning course via the World Wide Web. The framework's core components are a Web-based instructional management system, a model of the domain's knowledge, a model for the human resource inventory including personal traits and individual styles, a set of instruments to assess the learner's mastery level, and a peer-matching system.; As a proof of concept, the framework was prototyped and tested in a real-life distance education environment at the University of Central Florida. The prototype can be tuned so that alternative remedies for participants' knowledge and skill deficits are provided through the paradigm of peer collaboration. A three-phase overall experiment (pilot experiment, principal experiment, and follow-up experiment) was conducted during three consecutive semesters. The results of this experimental evaluation are not yet conclusive. However, partial results suggest that indeed the Partner-Finder framework can be useful for improving peer collaboration in Web-based education. |
