Font Size: a A A

Design And Implementation Of Intelligent Generating Test Paper System Based On Genetic Algorithm

Posted on:2016-03-29Degree:MasterType:Thesis
Country:ChinaCandidate:D M WeiFull Text:PDF
GTID:2308330482464375Subject:Computer technology
Abstract/Summary:PDF Full Text Request
With the development of computer-aided instruction and artificial intelligence technology, the study on the intelligent generating test paper system is an important research area. The establishment of the test database based on computer and realization of auto composing test paper is the important measure to make the examination more scientific and standard, then it is a key way to separate the examinatiom from teaching.It is very helpful for the improvement of teaching quality. Genetic algorithm is a kind of stochastic search and optimization method,which solves complex problems by simulating natural selection and genetic variability mechanism of biological world. Auto-generating test problem is a multi-objective parameters optimization issue under certain constraints. The efficiency and quality of auto-generating test paper largely depend on the design of test bank and question selection algorithm. Traditional test paper algorithm has the shortcomings of generating test paper slow, low success rate, poor test paper quality and so on.As for these shortcomings, on the basis of full analysis and research on the intelligent generating test paper algorithm, this paper discusses the auto-generating test paper strategy based on genetic algorithm. Auto-generating test paper system under certain constraints analyzes the evaluation indicators of paper, the roles of indicators and their relationship, quantifies them as a set of index score distribution describing the characteristics of the target papers, and then generating paper distributed model was builded by adapting various evaluation indicators. On the basis of generating paper model, the multi-objective constrained optimization model which is closest to the consumer requirements was established. The genetic algorithm was utilized to solve the auto-generating test problem. This paper constructs an auto-generating test paper system, which is based on B/S structure, relays on ASP.NET technology. The system include the functions of customer management, test bank management, test paper management, auto-generating test paper, paper type management, and examitatiom management. This system significantly improves the success rate of test paper and convergence speed.This mathod is not only suitable for the larger test bank system, but also can meet complex test paper conditions. What’s more, several suits of different papers that meet the requirements can be obtained only by one generating prosess. The proposed system can greatly improve the efficiency and quality of the exam, exactly reflect the examinationers’ degrees of knowledge possessed.
Keywords/Search Tags:Auto-generating test paper, Genetic algorithm(GA), Mathematical mode, Multi-objective optimization
PDF Full Text Request
Related items