A Study On Intelligent Examination System Based On Multi-feature Similarity And C4.5

At present, with the rapidly popularization and development of the information technology, the focus of each school is to develop information technology education. All kinds of examinations become hot spots of various industries with the gradual shift of the modern education to the lifelong education. It will be a trend of the examination ways that examination will realize the information, reduce the chance of miscarriage of justice, save the manpower and material resources, balance the degree of difficulty of the examination, improve the fairness of the examination and make the examination more scientific.This thesis analyzes the network intelligent examination system based on B/S structure. This system considers the information examination’s convenience, fairness and security, tries to build a multi-course examination system which involves the main courses of a number of professional and is better than the single-course examination system with these functions:the management of test questions, the creation of question paper, online examination, the judgment of the subjective and objective questions, the management and query of the scores and the assessment of the examination.To achieve this aim, this system combines ASP.NET and SQL SERVER technology and uses multi-feature fusion similarity calculation and C4.5algorithm to judge the subjective questions of difficult issues. For example, in the course of "Mechanical Design" and "Legal basis" in Hunan Chemical Vocational Technology College, all types of examination basis datas are entered into the system, It being used in daily teaching expands the teaching space, promotes the transformation and the development of the teaching methods. Finally, this thesis puts forward the research directions for the improvement of the network intelligent examination system.