基于知识图谱的学科单选题考点提取研究

在人工智能快速发展的今天，智能教育逐渐成为一大研究热点。在自然语言处理方面对智能教育中智慧学习的探究，提出根据知识图谱和学科规则确定单选题考点，主要介绍知识图谱的构建和单选题考点的提取这两方面内容。通过建立一个开放性的知识图谱，不断实现学科知识的扩充。为提取考点，首先将单选题分类、分词以及替换相似词，然后通过检索图谱得到单选题的候选考点集，最后通过学科规则定位知识点及其所属章节，便于学生有针对性地复习教材知识。在所收集的C++试题集上的实验结果表明，通过知识图谱和规则可较为准确地提取出试题考点。

Research on extraction of subject single knowledge points based on knowledge map

At present, artificial intelligence develops rapidly, intelligent education has gradually become a hot research. This paper is a study of intelligent learning in intelligent education in the aspect of natural language processing. This paper proposed to determine the single choice knowledge according to the knowledge map and discipline rules. It mainly introduced the construction of knowledge map and the extraction of the single choice knowledge, through the establishment of an open knowledge map, expand the subject knowledge. In order to extract the test knowledge, first, classification of single choice, word segmentation and replacement of similar words, then, the candidate knowledge set of single choice can be obtained by searching the knowledge map, finally, through the discipline rules to locate the knowledge points and their chapters, help students to targeted review the knowledge of teaching materials. The experimental results on the collected C++ test set show that the test points can be accurately extracted by knowledge map and rules.