본문 바로가기
  • Home

Exploring Educational Needs to Improve the Dropout of Cyber Universities’ Learners Based on Topic Modeling

  • Global Creative Leader: Education & Learning
  • Abbr : GCL
  • 2021, 11(3), pp.79-110
  • DOI : 10.34226/gcl.2021.11.3.79
  • Publisher : Research Institute for Gifted & Talented Education, Soongsil University
  • Research Area : Social Science > Education
  • Received : August 15, 2021
  • Accepted : September 1, 2021
  • Published : September 30, 2021

Seo Hee-Jeong 1 Suna Kyun 2

1한국방송통신대학교 원격교육연구소
2숙명여자대학교

Accredited

ABSTRACT

The purpose of the present study is to explore the educational needs of cyber university learners, in order to obtain the basic data for curriculum improvement or development, which contributes to the improvement of learners’ dropout rates in the cyber university. To do this, 4,880 study plans of H cyber university's 1st year and transfer students were analyzed. 4,880 study plans were analyzed over two stages by faculties (Engineering, Humanities & Social Sciences, and Arts) and gender (male, female): ① frequency analysis of keywords, ② LDA-based topic modeling. The results are as follows. First, as a result of frequency analysis, the most frequently mentioned keywords were those related to career development such as ‘knowledge’, ‘job’, ‘practice’, ‘professionalism’, ‘graduation’, ‘certificate’, and ‘acquisition of degree’. Second, as a result of topic modeling, in the case of male and female students from the Engineering department and male students from the Humanities & Social Sciences department, ‘improving job competency in the major field’ was derived as the most important topic. Also, while female students from Humanities & Social Sciences department, showed ‘basic learning in their major’ as the most important topic, male and female students from the Arts department showed ‘intensive learning in their major’. Finally, based on these results of the learner's’ educational needs analysis confirmed through this study, the implications were discussed.

Citation status

* References for papers published after 2023 are currently being built.