基于多特征融合的主观题智能阅卷算法研究

为了实现在线考试主观题智能阅卷，本文提出了一个基于多特征融合的主观题智能阅卷算法，首先，分别设计两种算法：关键词相似度算法和语义相似度算法，计算学生答案和标准答案之间的关键词相似度和语义相似度；然后，利用题目的题型和科目作为特征值，分别计算出关键词相似度的权值和语义相似度的权值，并利用两权值计算出学生答案的最终得分；最后，以某在线学习平台历年考试数据作为数据集，对该主观题智能阅卷算法进行训练和实验。实验表明，本文提出的基于多特征融合的主观题智能阅卷算法，能有效主观题智能阅卷算法的准确性，具有较好的应用前景。

Research on Subjective Intelligent Marking Algorithm based on Multi-feature Fusion

in order to realize intelligent marking of subjective questions in online exams, this paper proposes an intelligent mark-ing algorithm for subjective questions based on multi-feature fusion. First, two algorithms are designed respectively: keyword similarity algorithm and semantic similarity algorithm, to calculate the keyword similarity and semantic similarity between students'' answers and standard answers.Then, using the question type and subject as the eigen-value, the weights of keyword similarity and semantic similarity were calculated respectively, and the final scores of students'' answers were calculated by using the two weights.Finally, taking the test data of a middle school online learning platform over the years as the data set, this paper conducts training and experiment on the subjective intelli-gent marking algorithm.The experiment shows that the intelligent marking algorithm based on multi-feature fusion proposed in this paper is effective and accurate, and has a good application prospect.