Teaching System Based On Student Score Prediction Using C4.5 Decision Tree

The rapid development of university informatization and software engineering makes a big step forward on educational informatization.There are a great deal of students data accumulated by the teachers.However,the data never have a role to play but only remain in the system.By data mining,knowledge and patterns are discovered from the data,which will provide evidences for the optimization of the teaching process.Therefore,decision tree is adopted in this research.By inproving the C4.5 arithmetic according to the reality,analyzation is performed using the students data of a certain major in the school of information science and technology.The source of database is dean's office of Southwest Jiaotong University.The purpose of the arithmetic is finding out which of the basic courses has the greatest effect on different specialized courses,meanwhile,generating classification rules for performance prediction.To put the arithmetic into pratice,a teaching system based on student score prediction is built.Before the project,a detailed requirement analysis is made to make clear the goal of design and module partition of the system.In this thesis,architecture of the system,module of function,database as well as technology selection are discussed.As for system architecture,layered architecture is used,including client layer,presentation layer,business layer,integration layer and resource layer.Modules of function is divided into data maintenance,examination management,intelligent test paper composition,after-class studying,performance prediction,communication management.The system adopts B/S structure,Structs2 and Hibernate is used in the development of the system.In this thesis,The implementation of the system and application effect is demonstrated.After deploying the system,all the functions of the system are tested,the result shows that the system is under normal operation.