The Implementation of a Learning Guidance Model Based on Attributes Correlation

Decision tree classification algorithm is one of the effective tools to realize ”intelligent” of Intelligent Guidance System. It can receive precise classification though the analyzing and mining of the data. It also has another positive characteristic that the decision tree and set of production rules are simple and efficient computing. This paper proposed the ”Intelligent Guidance System” and introduced the main modules of the system. This paper also proposed the new-C4.5r algorithm based the comparing of the ID3 algorithm and C4.5 algorithm and the requirements of the individual educations. The new algorithm based on the relative of the rules attribute, simultaneously, the model of evaluation system of Calculus has been established .Experiments show that new-C4.5 algorithm is better than C4.5 algorithm on the aspects of run-time, the size of rule sets and overhead of the production rules.