计算机课程相似度计算方法及其改进

Calculation methods and improvement of similarity in computer courses

在传统描述方法上,计算机科学课程一般使用教学计划和教学大纲来描述专业知识和课程知识点的整体结构,由于计算机科学课程结构复杂,各个课程间联系紧密,依靠传统方法不足以把握计算机科学课程的总体结构和专业课之间的普遍联系。文章针对传统方法的不足,提出构建基于计算课程知识图谱的方法,具体包括:①课程知识图谱的构建,量化课程知识点关系,构建知识点关系矩阵,量化各知识点间的关系;②课程关系的构建,量化所有课程间的关系;③国外计算机名校关系的构建分析。研究发现,计算课程知识图谱的构建和分析将对目前计算机科学专业课程的改革研究具有一定的创新意义,这种计算机科学专业课程知识图谱的构建方法,可弥补传统课程描述方法上的不足,为计算机科学专业课程改革分析提供一定的数据支持,同时,对相似度的衡量提供了一种新的方法,提高了相似度衡量速度。该方法也可移植到其他学科。

In traditional methods of describing computer science courses,instructional plans and course outlines are commonly used to represent the overall structure of computer science s specialized knowledge and course content.However,due to the intricate nature of computer science courses and their interconnectedness,relying on traditional methods makes it challenging to grasp the overall structure of computer science courses and the connections between specialized courses.This article addresses the limitations of traditional methods in describing computer science courses and proposes a method for constructing a computer course knowledge graph to address these shortcomings.Specifically,this includes the construction of a course knowledge graph,quantifying the relationships between courses and knowledge points,building a knowledge point relationship matrix to quantify relationships between various knowledge points,constructing course relationships to quantify relationships between all courses,and analyzing relationships between top computer science schools internationally.The construction and analysis methods of the computer course knowledge graph are expected to have innovative significance for current reforms in computer science education.They provide a new approach to building a knowledge graph for computer science education and can be easily adapted to other disciplines.This new computer course knowledge graph fills in the gaps of traditional course description methods,and offers data support for the reform and analysis of computer science education.Additionally,it introduces a novel method for measuring similarity,thereby enhancing similarity calculation speed.