English Language Proficiency Level Classification Of Students In A Massive Open Online Course

Creative Commons License

Duru İ. , Doğan G. , Demir A. S. , Ata H., Diri B. , White S.

International Conference On Advanced Technologies, Antalya, Turkey, 28 April 2018, pp.451-460

  • Publication Type: Conference Paper / Full Text
  • City: Antalya
  • Country: Turkey
  • Page Numbers: pp.451-460
  • Yıldız Technical University Affiliated: Yes


Student performance estimation, measurement and improvement are concepts that have gained importance in recent years. Measurement of learning performance are one of the most popular topics in educational data mining and learning analytics research areas. 75% of the courses on massive open online course platforms(MOOCs) are offered in English. The general problem with these platforms is that the same content is presented to all users in the same way. For this reason, the success rate of English as a second language students studying on the platform varies according to their English level. Within the scope of the study, it is planned to analyze online course data taken from the FutureLearn MOOC platform used in Southampton University, England. We have previous work on this topic. In this work, we will further analyze user comments on the forums. From the comments, we will deduce information about their language skills. We have divided users into groups such as i) English as official and primary language ii) English as official but not primary language iii) English as a second language learners. After analyzing students’ comments on the forum we will update the groups according to the results we gained. We aim to find number of words and characters in the user's comments, the average number of sentences, the number of conjunctions used, the negative-positivity status of the words, and the number of stop words (ineffective words). As a result, it is planned to provide special insight to classification of students according to their language proficiency level.