Automatic Syllabus Creation of Computer Science Courses using Natural Language Processing

Abstract

Problem Statement: Curriculum and syllabus creation is an important part of education. The quality of computer science students depends upon the quality of computer science education and training given to them. Course experts consider various factors before creating a syllabus for a particular course. One of the major factors upon which the curriculum depends is industry feedback and syllabus repository. Research has been done on taking feedback from industries and including their suggestions in the syllabus. But they did not suggest a platform to take feedback, compare the feedback, and give suggestions about the courses and topics to be included in the curriculum. Research is done on creating syllabus repositories and extracting the contents from them. Research is done on creating a single platform to create, update and delete the syllabus. But it does not suggest a model for comparison of the contents of the syllabus and gives suggestions about the topic that can be included in the syllabus. newlineThere is a need for a systematic model for taking feedback from industries, comparing that feedback, and giving suggestions about topics that can be included in the curriculum. That model will also create the syllabus repository, extract the contents from the repository and suggest the topics after comparison of syllabuses in the repository that can be included in the syllabus. newlineMethod and Design: We proposed a model for syllabus creation. The model will provide the facility to create, update and save syllabi. This model considers two attributes industry feedback and open-source syllabus of various universities, after considering these attributes it will give suggestions to the syllabus creator about the contents. The syllabus creator will finalize the contents by considering the suggestions given by the model, their expertise, and other attributes upon which the syllabus depends. newlineTo implement this model, we used the stremlit python framework to design this framework and spacy for semantic comparison. newlineFindings: With this mo