Trust-based community assessment

We present a trust model that computes group opinion w.r.t. a given community member.We apply it to the learning domain, and we compute group opinion w.r.t. the teacher.The model decreases the assessment load of teachers by relying on students' assessments.We conduct our experiments on real classroom datasets as well as simulated data.Experimental results show that the model is sound and scales for large numbers of students. In this paper we present Community Assessment (COMAS), a trust-based assessment service that helps compute group opinion from the perspective of a specific community member. We apply COMAS in the context of communities of learners, and we compute the group opinion from the perspective of the teacher. Specifically, our model relies on teacher assessments, aggregations of student assessments and trust measures derived from student assessments to suggest marks to assignments that have not been assessed by the teacher. The proposed model intends to support intelligent online learning applications by (1) encouraging students to assess one another, and (2) benefiting from students' assessments. We believe the task of assessing massive numbers of students is of special interest to online learning communities, such as massive open online courses (MOOCs). Experimental results were conducted on a real classroom datasets as well as simulated data that considers different social network topologies (where we say students assess some assignments of socially connected students). Results show that our method (1) is sound, i.e. the error of the suggested assessments decreases for increasing numbers of teacher assessments; and (2) scales for large numbers of students.