Towards the alignment of computer-based assessment outcome with learning goals: the LAERS architecture

Computer-based assessment (CBA) measures learner's progress in an automated manner. A challenge is to align the CBA outcomes with predefined learning goals. In this paper we propose an architecture that feeds the machine-readable results of learning analytics processes into an educational recommender system in order to produce personalized applicable feedback to guide learner during assessment. We describe the architecture, its inner structure, its separate modules, the data-flow between them and the basic mechanisms for data preparation and processing. Our goals are threefold: a) to meaningfully interpret learner's interaction-generated data into goal-oriented feedback, b) to implicitly link task-level cognitive feedback to learning goals and performance, and c) to automate the delivery of personalized assessment tasks to examinee. Finally, we conclude with our planning on future work.