Test paper setting method and device based on test relevance

Abstract

The invention discloses a test paper setting method and device based on test relevance. The method comprises the steps that 1) the difficulty of each test and the relevance of all tests are calculatedaccording to previous answer records of a user in a test bank; 2) a relevance matrix is established; 3) according to the number of test knowledge points in the test bank and the total number of testsof test paper composition, the number of tests allocated for each knowledge point is calculated; 4) the number of tests allocated for each difficulty section is determined according to the difficultydistribution of test paper and the total number of tests of test paper composition; 5) a test map is established according to the relevance matrix; 6) the lowest relevance queue corresponding to eachknowledge point is acquired according to the test map; and 7) a rotation policy is used to select the vertex in the lowest relevance queue corresponding to each knowledge point in turn until the number of tests needed by test paper setting is reached. The device comprises a difficulty calculation module, a relevance calculation module, a matrix establishing module, an extraction module and a testpaper setting module. According to the invention, the method and device help to improve computer-assisted teaching; big data are used; and remote and automated teaching is realized.

Description

Technical field [0001] The present invention relates to the research field of automatic paper writing, in particular to a method and device for writing papers based on the correlation of test questions. Background technique [0002] With the rapid development of computer technology and its continuous deepening in the field of education informatization, the use of computer and automation technology to assist teaching has a huge role in promoting today's education and teaching reform. In daily teaching, examinations are still the main way to test students' knowledge mastery ability, and the main manifestation of examinations is still in the form of test papers. In the traditional test paper composition, the tester collects, selects and compiles test questions according to his own knowledge, experience, and style. Although the validity and credibility of the test questions are high, human factors will narrow the knowledge coverage of the test papers, the test sites are not comprehe...

AI Technical Summary

Problems solved by technology

However, the selection of test questions by these methods of making papers is carried out separately. In fact, there is a certain correlation between the test questions. Because the correlation between the test questions is not considered, there are too many test questions of the same knowledge point, and the test questions of other knowledge points are not. In addition, artificially distinguishing knowledge points through knowledge point chapters or learning points will make the classification not detailed enough, and it is difficult to find subtle differences between test questions, resulting in more similar test questions in the final test paper and fewer test points

Images