Abstract:

To improve the quality and effect of CAI (Computer Assisted Instruct) courseware system, compared with the traditional CAI courseware system,the network-based CAI courseware system has its advantages and disadvantages.A hierarchical level-generate algorithm is proposed, based on the data mining theory, to cope with its disadvantages by building a CAI courseware tree and organizing the whole courseware objects into a reasonable structure. By doing this, the system can provide the student with personalized service. But this algorithm overlooked the students. Thus, mining on students were considered. Using Hamming distance theory, students were clustered into different groups based on the courseware they accessed. The performance is greatly improved.

Key words: CAI courseware, personalized study, hierarchical level-generate algorithm, Hamming distance