Font Size: a A A

Research And Design Of Intelligent System Composing Test Paper

Posted on:2011-12-26Degree:MasterType:Thesis
Country:ChinaCandidate:C S LangFull Text:PDF
GTID:2178360308974647Subject:Signal and Information Processing
Abstract/Summary:PDF Full Text Request
Online examination system is a hotspot of current research and it is studied by many educators that how we can generate the high-quality papers efficiently and fastly on line. In this paper the popular and traditional genetic algorithm composing examination paper is improved. The test paper composed by the improved Memetic algorithm overcomes the weaknesses of the traditional genetic algorithm and it avoids premature phenomenon which primary algorithm is easy to become.In order to meet all kinds of test needs, the system can compose different types of test paper according to users'requirements. It can set not only difficulty, distinctiveness and knowledge point etc in the light of test types such as the final exam, the unit test,the passing test and the selecting examination but also corresponding parameter for the different types of test needs. The second chapter mainly describes the system's overall framework and main functional blocks. A good paper is determined by the algorithm.The genetic algorithm is popular to compose the paper now but it has weakness such as prematurity. In the third chapter we use the better Memetic algorithm to compose paper and the improved algorithm is obviously superior to the traditional genetic algorithm in fitness through experimental analysis. In fitness function design, the main parameters are taken into account such as difficulty, distinctiveness and knowledge point as far as possible. In order to expedite the speed of test paper composed, the total score and the total amount of examination questions are pre-designed. The fourth chapter introduces the concrete realization of online examination system. The traditional database can not be updated dynamically and Manually updating the database has workload and error-prone so that the data of the database has not updated for years and it does not reflect the actual condition truly. For this shortcoming in the design process, we should update the database after each test, so as to express real data. Of course, in the database updating process, we also update the difficulty of the dadabase questions so that the difficulty in the database is to reflect the true level of the database better. This article also introduces auto-marking subjective paper. The general procedure aims at correcting objective paper. This article mainly describes how to correct the examination questions of office on line auomatically. Finally, the safety of the online test is designed at the end of this chapter so that candidates can not cheat in the test as far as possible in order to reflect a fair and impartial nature of the examination.This article is a renovation according to the traditional online examination system not only in the test paper method but also in designing of the database questions and both of them have carried out an effective update. The system owns better popularity, if the item bank meets its requirements, it can make the appropriate online testing in order to achieve effective combination of real learning and online exams.
Keywords/Search Tags:genetic algorithms, Memetic algorithms, automatic grading, intelligent test paper
PDF Full Text Request
Related items