Equality of Learning Opportunity via Individual Fairness in Personalized Recommendations

Abstract Online education platforms play an increasingly important role in mediating the success of individuals’ careers. Therefore, while building overlying content recommendation services, it becomes essential to guarantee that learners are provided with equal recommended learning opportunities, according platform principles, context, and pedagogy. Though importance ensuring equality opportunities has been well investigated traditional institutions, how this can be operationalized online ecosystems through recommender systems is still under-explored. In paper, we shape a blueprint decisions processes considered context based on principles need empirically-validated (no evaluation live performed). To end, first provide formalization educational model recommendations’ properties, novel fairness metric combines them monitor among learners. Then, envision scenario wherein should arranged such way generated recommendations meet each principle certain degree for all learners, constrained their individual preferences. Under view, explore by course platform, uncovering systematic inequalities. reduce effect, propose post-processing approach balances personalization opportunities. Experiments show our leads higher equality, negligible loss personalization. This paper provides theoretical foundation future studies learners’ preferences limits concerning