You are here :

Learning tracking: Big Data and Artificial Intelligence in the service of student success

Published on June 11, 2019 - Updated on February 25, 2020

With Kosmos Next, its research and innovation programme, Kosmos has chosen to assemble a team of experts to deliberate about various education-related subjects. The goal is to improve student success and build the school of tomorrow based on new technologies and the immense field of possibilities that they offer.

Learning tracking: data in the service of student success

It is possible today to mine a wealth of data, a process called learning tracking. These traces comes from the use of digital tools or online services allowing information about developments and the learning behaviour of students to be consolidated, which is a major lever for the future of education.

Learning Analytics is a discipline dedicated to the measurement, the collection, the analysis and the presentation of reports based on student data in the context of learning with the goal of understanding and optimising learning and the context.

The field of Learning Analytics comes from Data Analytics applied to the education sector where the concepts of Data Mining and Machine Learning have been adapted to it.

The analysis of learning data can generally be seen from two points of view: that of teachers trying to improve their knowledge of their student profiles (in particular students in difficulty) or from the point of view of the student his or herself, who could improve their ability to learn and their personal progress.

The digital environment, like the EMS (Education Management Software) used in educational establishments, are totally adapted to learning traces because of the high volume of data collected. EMSs are capitalisable platforms whose audience attendance amount to 50 million visits per month.

Kosmos en-route for Adaptive Learning

Through our partnership with Loria (Lorraine laboratory for research in computer science and applications) and our joint work on gathering and analysing learning traces coming from Skolengo services, the Kosmos EMS (Education Management Software) is part of the deliberations on these innovative research subjects.

Indeed, the data coming from the user and Skolengo’s educational services will be able to be stored in LRSs (Learning Record Stores), currently using the xAPI standard (standard for processing learning traces). In this framework, Kosmos guarantees the collection and use of data strictly in compliance with the GDPR.

With our experts and partners, it will study how to best use this information. The difficulty of analysing this data is part of the discussions of artificial intelligence in a context of Big Data: from a large quantity of data, apparently uncorrelated, how can we retrieve relevant information?

The project entitled METAL, brings together players in teaching, schools and universities, research laboratories, and municipalities and proposes to “design, develop and evaluate all the individualised tracking tools intended for students and teachers (learning analytics) and innovative technologies for the personalised learning of languages.”

Even if Adaptive Learning is just one of the tentative steps, this subject suggests that there are great possibilities for simulating individual, 100% adapted learning paths.

On the same basis as recommendation technologies, Adaptive Learning can allow personalised curricula to be offered to students using the analysis of data already gathered on earlier pieces of work. This approach borrows many of the characteristics from suggestion tools of similar content. Just like Google, Amazon or YouTube, who collect data in order to learn the preferences of their users, Adaptive Learning could allow data to be stored in order to offer content adapted to each person in order to improve their performances and thus to take into account the unique character of each student.

One can then imagine the future with the automation of these practices in order to easily manage tailored curricula for students. But it is also a precious tool for teachers and administration personnel for identifying profiles requiring assistance and special attention in order to best accompany all students in their learning paths.