基于知识体系结构的组卷方法研究与应用

现有组卷算法未考虑组多份甚至数十份难度一致、知识点均衡试卷的情况,基于此提出了一种针对计算机类课程知识体系结构的改进随机组卷算法。该算法抽题前预先存取数据库中某课程试题,确定试题的抽取范围后,再从候选题集合中遵循未抽取题目的知识点优先原则,依次为多份试卷随机抽取试题。实验结果表明,基于知识体系结构的随机抽取算法与按章随机抽取算法、传统遗传算法相比,组卷效果有更好的提升。

Approach for Test Paper Generating Based on Knowledge Structure

The existing auto-generating test paper algorithm does not consider generating multiple or even dozens of test papers. Based on this, an improved random test paper algorithm for computer course knowledge structure is proposed. The algorithm pre-accesses a course test in the database before the question is drawn, and after determining the scope of the test, the knowledge point priority principle of the unextracted title is followed from the candidate question set, and the test questions are randomly selected for multiple test papers. The experimental results show that the random extraction algorithm based on knowledge architecture has better effect than the random extraction algorithm and the traditional genetic algorithm.