Understanding knowledge areas in curriculum through text mining from course materials

Curriculum analysis is attracting widespread interest in educational field. There are two main approaches: (i) human-based and (ii) text-based assessments. Although an evaluation by teachers and learners are widely used, it is inconvenient and time-consuming. Also, the results absolutely rely on individual attitude. The text-based approach aims to directly evaluate the course syllabus; however, there is only a course description in the syllabus, so this cannot really express the actual course contents. In this paper, we present an automatic text-based curriculum analysis that straightforwardly assesses entire course materials. Our approach employs a well-known text-mining technique that extracts keywords using TF-IDF. The analysis is based on keywords from the course materials matching to the keywords from online documents, which is similar to the domain expert. Moreover, a new measurement is proposed to quantify associations between course materials and online documents using amounts of matching keywords. The experiment was conducted on materials of three subjects collected from five top universities mapping to the latest Computer Engineering Curricular Guideline (CE2016). The results illustrate significant relations among courses from different universities and CE2016. To further analyze the courses, each of them are visualized using radar charts.

[1]  Masatoshi Yoshikawa,et al.  Course Content Analysis: An Initiative Step toward Learning Object Recommendation Systems for MOOC Learners , 2016, EDM.

[2]  Hideki Mima,et al.  Machine Learning-based Syllabus Classification toward Automatic Organization of Issue-oriented Interdisciplinary Curricula , 2011 .

[3]  Herman Lam,et al.  CE2016: Updated computer engineering curriculum guidelines , 2015, 2015 IEEE Frontiers in Education Conference (FIE).

[4]  Vincent Ng,et al.  Automatic Keyphrase Extraction: A Survey of the State of the Art , 2014, ACL.

[5]  Judy Kay,et al.  PROGOSS: Mastering the curriculum , 2012 .

[6]  Linda Marshall,et al.  A comparison of the core aspects of the ACM/IEEE Computer Science Curriculum 2013 Strawman report with the specified core of CC2001 and CS2008 review , 2012, CSERC.

[7]  W. Pirie.  Spearman Rank Correlation Coefficient , 2006 .

[8]  Gerard Salton,et al.  Term-Weighting Approaches in Automatic Text Retrieval , 1988, Inf. Process. Manag..

[9]  Brian Lott,et al.  Survey of Keyword Extraction Techniques , 2012 .

[10]  Herbert J. Walberg,et al.  A National Experiment in Curriculum Evaluation , 1972 .

[11]  Kazunori Yamaguchi,et al.  Mapping analysis of CS2013 by supervised LDA and isomap , 2014, 2014 IEEE International Conference on Teaching, Assessment and Learning for Engineering (TALE).

[12]  John Impagliazzo,et al.  Toward a modern curriculum for computer engineering , 2014, 2014 IEEE International Conference on Teaching, Assessment and Learning for Engineering (TALE).

[13]  Kazunori Yamaguchi,et al.  Curriculum analysis of CS departments based on CS2013 by simplified, supervised LDA , 2015, LAK.

[14]  Jiawei Han,et al.  Data Mining: Concepts and Techniques , 2000 .