Skip to main content

Evaluating the Effectiveness of Active Learning in Classifying Educational Forum Posts

Primary supervisor

Guanliang Chen

Students, particularly those are enrolled in online courses, often communicate with others by posting questions in the discussion forum provided by the Learning Management System (e.g., Moodle and Canvas). It is worth noting that, nowadays, the number of students in a course can be up to from hundreds to thousands. Considering the diverse background of the students, the topic covered in their forum posts can be various, e.g., expressing confusion, asking for social support, exchanging ideas, and so on. As indicated before, the student-to-instructor ratio in a course can be very high, it is very challenging for the instructors to efficiently and effectively track the whole forum, identify all of the urgent issues, provide necessary help, and better assist students with their learning. 

Student cohort

Double Semester


Therefore, this project will utilize the feedback data stored on the Moodle platform of Monash University for analysis and aims to testify the effectiveness of different Active Learning techniques when being used together with traditional Machine Learning and Deep Learning models to construct a subject-independent classifier to automatically categorize forum posts. In particular, this project will investigate the impact of various Active Learning techniques on predictive fairness, e.g., whether the posts generated by male students would be more likely to be accurately classified than those generated by female students?


Required knowledge

  • Strong programming skills (e.g., Python)
  • Basic knowledge in Data Science, Natural Language Processing, and Machine Learning
  • The following can be a plus: (i) prior experience in applying Deep Learning models; (ii) good at academic writing; and (iii) strong motivation in pursing a quality academic publication.