Research On Adaptive Test Model For Supporting Learners' Knowledge Evaluation

With the popularization of emerging science and technology,which is highlighted by networking,digitization,and intelligence,student's personal appeal has been responded.Recommendation technology is gradually being applied to the education field,which has led to the rise of "adaptive testing systems." In education,testing is an important part,the emergence of the computerized adaptive testing provides new technical and tool for more accurate and efficient evaluation of participants.Through the information provided by the participants,we not only obtained test scores,but also began to consider the knowledge structure,cognitive defects and other content.And it has been become the focus of current research.At present,there are two kinds of adaptive test theory that are popular,—Item response theory and Knowledge space theory,which is also the two main research directions in the CAT field.The former one measures the student's knowledge level and knowledge breakpoints,and the latter one measures the level of measurement ability.Based on these two theories,this paper designs and develops two measurement modules: Knowledge state diagnosis module and Capability level test module.This study combines the advantages of the two theories,and it mainly studies the testing functions in the adaptive testing system,improves the relevant core components,and build a feasible adaptive test model in order to position learners and reflect the individualized theory.Specific research work includes:At first,by reviewing relevant literature,the research status of the theory and practice of adaptive test systems,Including domestic and foreign has been sorted out.We understand the current development trend and know the research direction of this article.At the same time,we summarize the core ideas of knowledge space theory and Item response theory and define the specific concept of “knowledge evaluation”.Secondly,the main content first analyzes the typical model of the adaptive learning system,and studies its design concept,hierarchical structure,the logical relationship between the various parts and the functional service positioning,etc.,which provides the basis for the analogying computerized adaptive testing model.Then the research of the test model mainly emphasizes the system structure,and designs the interaction relationship between each part.Based on the above research,we try to build an adaptive test model and analyze its core components in detail.Finally,using the “WeChat applet + Tomcat + Java + Mysql” as the development tools and the "C language programming" as the specific course,an adaptive test system based on the B/S architecture was developed.We conducted a small-scale test of the test results,and the results showed that the design goals were basically achieved.