“Big data” arises when volumes, variety and speed in the production and circulation of data are increasing. More and more data, relating to an ever increasing number of processes, is generated and stored, processed and used for ever faster decisions. According to Ben Williamson (Williamson, 2017), the expansion of the use of ‘big data’ in education is essentially due to two interrelated processes: dating and digitization.
By dating, Williamson explains, we mean the transformation of education into digital data. Test results, qualitative reports, streaming of data from online courses, etc .: everything is digitized and rendered in numerical forms, so that it can be further analyzed and processed in cognitive structures, which can be viewed through tables, graphs and images.
Digitization, on the other hand, concerns the coding of educational practices and policies into software codes and products, algorithms and applications. The two aspects are strongly correlated, so that one supports the other. In fact, digitization takes place so that educational software is capable of generating digital data, which in turn can be measured and processed. Then, the data that derive from tests or other forms of evaluation can in turn be easily collected in digital databases, which will be processed by other specific software and so on.
The sectors in which we are today facing the growing challenge posed by Big Data are many and even in the field of digital training the amount of data processed by learning management systems (LMS) has grown considerably. Each day, these systems accumulate an increasing amount of data on user interactions, personal data, system information and educational information.
The ability to find, understand, describe, use and produce structured information is part of the most typical cross-cutting competence of our years: Learning to learn. In a scenario where Big Data are so important, this competence becomes even more strategic and delicate. Intelligent data analysis, therefore, is a resource that will become increasingly important to support training processes. Machine learning, with its techniques for analyzing large amounts of data and producing intelligent syntheses, will be our precious ally. “