Automated analysis of cognitive presence in online discussions written in Portuguese

Valter Neto, Vitor Rolim, Rafael Ferreira, Vitomir Kovanović, Dragan Gašević, Rafael Dueire Lins, Rodrigo Lins

Research output: Chapter in Book/Report/Conference proceedingConference PaperResearchpeer-review

10 Citations (Scopus)


This paper presents a method for automated content analysis of students’ messages in asynchronous discussions written in Portuguese. In particular, the paper looks at the problem of coding discussion transcripts for the levels of cognitive presence, a key construct in a widely used Community of Inquiry model of online learning. Although there are techniques to coding for cognitive presence in the English language, the literature is still poor in methods for others languages, such as Portuguese. The proposed method uses a set of 87 different features to create a random forest classifier to automatically extract the cognitive phases. The model developed reached Cohen’s κ of.72, which represents a “substantial” agreement, and it is above the Cohen’s κ threshold of.70, commonly used in the literature for determining a reliable quantitative content analysis. This paper also provides some theoretical insights into the nature of cognitive presence by looking at the classification features that were most relevant for distinguishing between the different phases of cognitive presence.

Original languageEnglish
Title of host publicationLifelong Technology-Enhanced Learning
Subtitle of host publication13th European Conference on Technology Enhanced Learning, EC-TEL 2018, Leeds, UK, September 3–5, 2018 Proceedings
EditorsViktoria Pammer-Schindler, Mar Perez-Sanagustin, Hendrik Drachsler, Raymond Elferink, Maren Scheffel
Place of PublicationCham Switzerland
Number of pages17
ISBN (Electronic)9783319985725
ISBN (Print)9783319985718
Publication statusPublished - 2018
EventEuropean Conference on Technology Enhanced Learning (EC-TEL) 2018 - Leeds, United Kingdom
Duration: 3 Sep 20185 Sep 2018
Conference number: 13th (Proceedings)

Publication series

NameLecture Notes in Computer Science
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


ConferenceEuropean Conference on Technology Enhanced Learning (EC-TEL) 2018
Abbreviated titleEC-TEL 2018
CountryUnited Kingdom
Internet address


  • Community of Inquiry (CoI) model
  • Content analytics
  • Online discussions
  • Text classification

Cite this