Analysis of CS Curriculum Using Automatic Syllabus Crawling Methodology
IEEE Access · 2026 · 14 · 78231-78241
DOI: 10.1109/access.2026.3694893 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Access, 2026. Verfasst von Takayuki Sekiya und 2 weiteren Autor:innen. Die Arbeit nennt 34 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: JSPS KAKENHI.
Abstract (Originalsprache)
We developed a method for collecting computer science-related syllabi from university websites. Using many crawled web pages, we quantitatively and statistically analyzed university curricula. A syllabus defines the contents of a course and other information such as resources and assignments. Therefore, syllabus analysis is also used to understand trends in educational content. A large number of syllabi must be collected automatically and regularly with minimal effort to ensure that the results of the statistical analysis are more convincing. Our research questions are as follows: Can we obtain syllabus candidate pages semi-automatically? Can we analyze the data received as syllabus candidate pages using the CS2013 BOK-based ssLDA model? What knowledge can we gain from this analysis? To crawl the webpages, we used the Google Custom Search API and Scrapy, a Python-based general-purpose crawler. We extracted text from crawled web pages using Trafilatula. We also use Bidirectional Encoder Representations from Transformers (BERT) to detect a syllabus and analyze syllabi with a simplified version of supervised LDA (ssLDA) using Computer Science Curricula 2013 (CS2013). We examined approximately semi-automatically collected 7,700 web pages and found that approximately 66% were syllabus pages. When we divided the group of universities that obtained the syllabi into “Baccalaureate,” “Master” and “Doctor” according to the Carnegie Classification and analyzed the curricula using the Body of Knowledge (BOK) of CS2013, we discovered that the syllabi of universities classified as Doctor-level universities were statistically significantly more varied than those of universities classified as Master or Baccalaureate.
Förderung
- JSPS KAKENHI
Zitieren
APA
Sekiya, T., Matsuda, Y., & Yamaguchi, K. (2026). Analysis of CS Curriculum Using Automatic Syllabus Crawling Methodology. IEEE Access, 14, 78231-78241. https://doi.org/10.1109/access.2026.3694893
Vancouver
Sekiya T, Matsuda Y, Yamaguchi K. Analysis of CS Curriculum Using Automatic Syllabus Crawling Methodology. IEEE Access. 2026;14:78231-78241. doi:10.1109/access.2026.3694893
BibTeX
@article{Sekiya2026,
author = {Sekiya, Takayuki and Matsuda, Yoshitatsu and Yamaguchi, Kazunori},
title = {Analysis of CS Curriculum Using Automatic Syllabus Crawling Methodology},
journal = {IEEE Access},
year = {2026},
volume = {14},
pages = {78231--78241},
doi = {10.1109/access.2026.3694893}
}
EUVIMED