Research On Testing System In Mobile Learning

With the successful development of WAP technology, Bluetooth technology, 3G technology and the GPRS system, as well as the rapid popularization of mobile phones and mobile computing technology, it causes a new learning form-mobile learning. At home and abroad there are a series of studies, including technical support, system architecture, platform implementation, and study materials, which have provided technical and material conditions for the study of testing system in the mobile environment.Testing system in mobile learning is a adaptive testing system which is based on item response theory in the mobile environment by mobile terminal, this testing system can test which meet the needs of candidates anytime and anywhere, it will take the initiative to adapt to the actual level of candidates, select the questions which similar with the level of candidates from the item database, and in accordance with the results to assess their abilities until get a more precise the ability level of candidates , it has scientific, efficiency and convenience features.In this paper, mobile learning theory and item response theory have been done a systematic research, on the basis of them design and realize testing system in mobile learning based on item response theory. Focus on topics of item-selecting and the ability estimation in the testing system. Item-selecting strategy is the core issue of the testing system, how to avoid too much exposure, how in the mobile environment to reduce the system calculating. The maximum information amount algorithm in item-selecting has been optimized, which solve this problem; ability estimation is the another key issue in the testing system, how to evaluate when the initial item parameters is not inaccurate, how to solve the problem in the special circumstances which can not be assessed, as well as apply what kind of ability estimation algorithms at different stage, in this paper, maximum likelihood estimation algorithm and the joint maximum likelihood estimation algorithm have been used together and optimized which better meet the ability estimation requirements. After the test, at various stages there are given different ability estimation algorithms.This paper introduces content of mobile learning and adaptive testing, development and research significance, discusses the technologies of item response theory and mobile learning, and then analyzes the system design and the modular design, focusing on the achievement of important functions such as item-selecting strategy and ability estimation, at the end of the article points out the problems and future prospects of the study.