A probabilistic approach for discovering difficult course topics using clickstream data

Assma Boughoula, Chase Geigle, Cheng Xiang Zhai

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

One of the main factors affecting the success and effectiveness of Massive Open Online Courses is the ability of the instructor to acquire and incorporate student feedback in a timely manner, and preferably before assigning grades to student assessments. This research uses raw clickstream data from video watching sessions of the Coursera MOOC: "Text Retrieval and Search Engines"1 to discover which topics are difficult for the students. We introduce a measure for topic difficulty based on these clickstream events, and rank the topics according to this measure. The validity of our ranking is evaluated by comparing it with the ranking of topics based on student votes and find that our method agrees with the ranking based on student votes with > 63% accuracy.

Original languageEnglish (US)
Title of host publicationL@S 2017 - Proceedings of the 4th (2017) ACM Conference on Learning at Scale
PublisherAssociation for Computing Machinery
Pages303-306
Number of pages4
ISBN (Electronic)9781450344500
DOIs
StatePublished - Apr 12 2017
Event4th Annual ACM Conference on Learning at Scale, L@S 2017 - Cambridge, United States
Duration: Apr 20 2017Apr 21 2017

Publication series

NameL@S 2017 - Proceedings of the 4th (2017) ACM Conference on Learning at Scale

Other

Other4th Annual ACM Conference on Learning at Scale, L@S 2017
Country/TerritoryUnited States
CityCambridge
Period4/20/174/21/17

Keywords

  • Probabilistic clustering
  • Student feedback
  • Topic difficulty

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Education
  • Software
  • Computer Science Applications

Fingerprint

Dive into the research topics of 'A probabilistic approach for discovering difficult course topics using clickstream data'. Together they form a unique fingerprint.

Cite this