Modelling Learners in Crowdsourcing Educational Systems

Traditionally, learner models estimate a student’s knowledge state solely based on their performance on attempting assessment items. This can be attributed to the fact that in many traditional educational systems, students are primarily involved in just answering assessment items. In recent years, the use of crowdsourcing to support learning at scale has received significant attention. In crowdsourcing educational systems, in addition to attempting assessment items, students are engaged with other various tasks such as creating resources, creating solutions, rating the quality of resources, and giving feedback. Past studies have demonstrated that engaging students in meaningful crowdsourcing tasks, also referred to as learningsourcing, has pedagogical benefits that can enhance student learning. In this paper, we present a learner model that leverages data from students’ learnersourcing contributions alongside attempting assessment items towards modelling of students’ knowledge state. Results from an empirical study suggest that indeed crowdsourced contributions from students can effectively be used in modelling learners.