A Supervised Learning framework for Learning Management Systems

Monllao Olive, David and Huynh, Du Q. and Reynolds, Mark and Dougiamas, Martin and Wiese, Damyon (2018) A Supervised Learning framework for Learning Management Systems. Proceedings of the First International Conference on Data Science, E-learning and Information Systems. ISSN 978-1-4503-6536-9

[img] Text

Download (1MB)
Official URL: https://dl.acm.org/citation.cfm?id=3279996


Educational Data Mining (EDM) and Learning Analytics (LA) focus on data analysis of learners in the context of educational settings like Moodle, a Learning Management System (LMS). Both EDM and LA aim to understand learners and optimise learning processes. Predictive modelling serve a key role in optimising learning processes. Learning Analytics in an LMS covers many different aspects: finding students at risk of abandoning a course, predicting students failing a quiz or students not reaching the end of a lesson in less than 15 minutes. Thus, there are multiple prediction models that can be explored. The prediction models can target at the course also. For instance, will this course engage learners? Will this forum be useful to the students of this course? To ease the evaluation and usage of Supervised Learning prediction models in LMS, we abstract the key elements of prediction models and we build an analytics framework for Moodle, one of the most popular Learning management Systems available in the market. Our software framework manages the complete cycle that predictive models follow until they are used in production, which includes calculations of features and labels from the LMS database raw data, normalization, feature engineering, model evaluation and a production-ready mode to generate insights for users from predictions. Apart from the software framework we also present a use case that serves as an example: A prediction model which is able to identify students at risk of abandoning a course with a 92\% in accuracy using past versions of the course as training data.

Item Type: Article
Subjects: Educational technology > Early warning systems
Educational technology > Experimental research involving methods and tools
Educational technology > Learning analytics
Educational technology > Open source
Divisions: Higher education, Universities, Vocational training, Colleges
Primary, Secondary, K-12
Workplace, Government, Health, Non-profit
Depositing User: David Monllao Olive
Date Deposited: 04 Dec 2018 16:20
Last Modified: 04 Dec 2018 16:20
URI: http://research.moodle.org/id/eprint/453

Actions (login required)

View Item View Item