Remote Examination System Design And Implementation

With the rapid development of the Internet and increasing improvement in informatization, the computer network technology has been applied more and more widely in education realm. The traditional examination method has fallen behind the times and has to be reformed. It is an inevitable trend to replace the traditional method by network test. Therefore,the market has an urgent demand for a fully functioning network test system.The thesis analyzes the characteristics of C/S and B/S network structures and uses web database technology as a theoretic basis to design and implement the network test system in combination with my teaching requirements and working practice. By taking B/S three-tier architecture as the overall structure and the web database technology as its basis, this system uses Microsoft ASP.NET and ADO.NET to realize the access and query to web database. The functions provided include examination data maintenance, question selection, online testing, automatic grading, results inquiry and etc.The thesis mainly studies test theories and the realization of intellectual test sheet-composition algorithms. The objective of intellectual test sheet composition is to form a set of test paper that could best satisfy the user's requirements. Therefore, test sheet composition is a solving process for classical constraint satisfaction problem. A newly developed overall optimization algorithm -- Genetic Algorithm (GA) could solve this kind of problem perfectly due to its ability in global search, its randomness, its appropriate encoding mode and high parallelism. Thus, Genetic Algorithm is used in this thesis to solve the problem of test sheet composition and the views of biogenetics are borrowed to improve the adaptability of individuals through natural selection, heredity, variation and other mechanisms. Five elements are involved in Genetic Algorithm: initial group setting, parametric coding, genetic manipulation design, fitness function design and the setting of control parameters, among which the genetic manipulation includes crossover, selection and variation. The thesis transforms the constraint conditions such as the difficulties of composition, sections, question types and values into objective function. In genetic manipulation, to ensure the new generation could satisfy the constraints as well, overall crossover of question types would be adopted in the process of one-point crossover. Test sheet composition strategy could be well realized through the above-mentioned algorithm.