Data Science Laboratory
The Data Science Laboratory (DS_LAB) researches new approaches to the processing, analysis and visualization of complex data to identify potential new services. Data Science is the ability to extract information from unstructured data, from large repositories or from data related to different domains, using a multidisciplinary approach. The Data Science Laboratory therefore wants to become a research environment for the analysis of complex systems, integrating competences in mathematics, statistics and informatics and exploiting the data heritage that will be gradually integrated into the smart data platform Yucca.
This type of research strengthens the knowledge of CSI on data science, also thanks to the organization of mixed teams with professionals from the academic world. This allows us to make the most of our clients' wealth of information and to enhance the range of services we offer.
WiFi networks in Emilia Romagna and Piedmont
In collaboration with Lepida, an in-house company of the Emilia Romagna Region, the data coming from the Wi-Fi network were analysed. They referred to 1,597 public hot spots active in Piedmont and Emilia Romagna. The survey was carried out with the smart data platform Yucca, which made it possible to analyse the data and return the results on a dashboard in a completely anonymous way, in compliance with the GDPR. We were thus able to know not only the total number of users connected to public Wi-Fi from February to October 2018, but also the most popular time slots, the days of the week with the highest number of connections, the most used hot spots and people's movement flows.
University of Turin: lectures, classrooms, didactics, outliers
We have structured a mixed work team together with the Departments of Philosophy and Educational Sciences and Law and the Information Systems, Portal, E-learning Directorate of the University of Turin to analyse some situations: lessons organization, classrooms occupation and the interest on didactics, students and the outliers critical path.
The data were ingested in the Yucca smart data platform and enriched with data available to the University and from other sources (weather, WiFi locations, cameras...).
Dashboards (power BI) have been created for consultation and descriptive analysis available to the University Departments involved. Two machine learning models have been developed for the prediction of success/failure of the students of the Law Department, five-year degree. The data have undergone a data quality process.