The Data Science for Social Good (DSSG) Initiative is a powerful example of applied research that has yielded profound direct and indirect impacts across multiple dimensions.
The DSSG was developed to maximise the benefits of data towards the empowerment of social impact organizations. It encapsulates multiple research projects conducted by data scientist and enthusiasts that aim to address real world questions, through the analysis of patterns or algorithm design.
Recently, the DSSG Europe was established in Portugal transporting its noble purpose to national institutions, in academia through the Data Science Knowledge Center of Nova School of Business and Economics, and then in Portugal to broader society through an independent non-profit association DSSG PT. The knowledge centrecapitalized on its faculty talent to further applied data science research and implement projects with numerous day-to-day applications.
The accomplishments of DSSG Europe founders were possible because of a dedication and passion for impact by a handful of talented researchers and a powerful computational infrastructure capable to store data and run intensive tasks. The DSSG-related work yielded several academic publications (20 conference papers, 5 journal papers, and 1 book chapter) and a free educational summer program at NOVA SBE, receiving its well-deserved recognition with the Golden Award at the 1st SAS Curiosity Data Science Iberian Award as a PI project in the category Data4Good, among other awards.
One of the most impactful examples of the work was the development and implementation of a long-term unemployment prediction algorithm, created in partnership with the Portuguese Public Employment Service (IEFP). As part of this project, the team implemented advanced machine learning algorithms and incorporated dynamic data, while also redesigning key aspects of the IT infrastructure. This initiative marked a pioneering use of AI in public administration in Portugal, and the algorithm has since been adopted by all IEFP counsellors across continental Portugal.
The IEFP project was a result of an extensive collaboration with the various members of the management team, several counsellors at their job site and unemployed citizens. After the initial implementation, DSSG team offered additional support to help IEFP staff re-train the existing model. It is beyond evident the beneficial outcomes that this initiative granted, besides its enormous achievement for the Portuguese context. This resulted in the selection by the Ministry for Administrative Modernization as one of the top 4 Portuguese projects for AI for public administration to be showcased at the European Parliament, in 2020.
Another great project was the big data analysis of dynamics of Portuguese small-scale fisheries seafood catch during the COVID-19 pandemic, conducted with Docapesca and ANP|WWF. Researchers analysed a unique dataset with data of 5 million first sales of seafood transactions by small-scale fishers in Portugal from 2017 to 2020. Data showed an overall decline of reported catch and fishers during lockdown, but a significant increase in reported catch volume in Lisbon and Algarve regions when compared with other areas. Reported patterns contributed to the argument that supports small-scale fishing communities and the maintenance of health levels of seafood populations, enriching policy evidence and discussion. This project was awarded a Green Prizes honoroble mention from Visao and Aguas de Portugal in 2021.
Lastly, it’s important to highlight the Tourist mobility studies performed for Italy and Portugal. Data scientists reutilized tourist mobility data for the Toscana Promozzione Turistica, gathering who are the tourists in Tuscany – but also when and why they are visiting – and proposed innovative methods to study tourism paths to collaborative municipalities. Their proven expertise in the topic inspired Turismo de Portugal to explore new partnership with telecom operators in Portugal, as a resource for new studies that unveil new methods to evaluate events or support destination-management-related decisions.
[Collaboration with DSKC] has contributed to improving our infrastructure and has helped us better understand how to integrate AI with our technicians. We can even say that the Data Science Knowledge Center and the team led by Leid Zejnilovic are, in practice, our AI Lab.
Carlos Sanatan, IEFP