Assisting education through real-time learner analytics

Abstract

Learning analytics has been dominating the education arena for these last few years as the efficacy and applicability of big data in general has been recognized, appreciated and fruitfully employed. Such analytics can assist and support higher education learners to optimize their interaction with available resources, while providing crucial insights on their learning behavior and study processes. The main advantage of learning analytics is their personalization capabilities that enables individual learners to gain specifically tailored knowledge, as well as adaptive academic support that supplements their own unique learning journey. What if these learning analytics were provided in real-time to both the learner and the educator within a context that enables the optimization of the entire learning process? What if the environment itself was conducive to complement the acquired knowledge and inferred recommendations extracted from the same analytics? Furthermore, what if the other learners together with the online crowdsourcing community were to contribute and collaborate to share their knowledge and experiences in an academic attempt to enrich the learning process? In this paper, we present a case study of how we attempt to answer the above research questions, as we engaged our undergraduate students within an Ambient Intelligent classroom. Apart from a number of strategically positioned sensors within the classroom, the learners had access to an associated online portal that interacted with the sensors as well as with the underlying virtual learning environment. The environment simulates the physical environment for remote learners who were not able to attend class in person, but also keeps track of the individual learners’ academic profile once they had logged into the system. Those learners who made it to class were still required to login, but the system was aware of their attendance as they were physically identified within the real classroom from their electronic university cards. The system coordinates the interaction between all the students and teacher who is present in class, but also between learners themselves, as such interactions are considered paramount to the receptive and education acquisition processes. We take full advantage of the ambient intelligent environment to employ non-intrusive sensors, displays and cameras, in combination with a real-time machine learning application to detect learners’ levels of attention and participation. This, combined with other learning data collected from the portal, assisted in the learning analysis process which looped back to the learners’ learning environment, as well as to the educator, in an attempt to improve the delivery and the entire educational process. We report on several issues encountered together with accuracy and reliability concerns as we draw a number of conclusions and offer recommendations.