Data analytics can support social security institutions to improve their administrative effectiveness and efficiency by enabling them to understand the past, explain the cause of events, anticipate what is likely to happen and suggest some actions to take. Institutions can apply data analytics in a wide diversity of areas such as healthcare, detecting and preventing error, evasion and fraud, proactive social policy and programme design, actuarial projections, improving service delivery, among others.

Data analytics is mainly based on institution’s data and potentially external ones, which, after preparation, are analyzed to derive insights using various analytic approaches, in particular:

• Descriptive analytics, which tries to answer “what has happened”. It provides an understanding of the past transactions that occurred in the organization;
• Diagnostic analytics, which tries to answer the question “why or how did it happen”. It involves an understanding of the relationship between relatable data sets and identification of specific transactions along with their behaviour and underlying reasons;
• Predictive analytics, which tries to predict “What, When, where will it happen” based on past data. Forecasting techniques can be used to predict, to a certain extent, the future outcome of an activity;
• Prescriptive analytics, which allows to “prescribe” a range of possible actions as inputs such that outputs in the future can be directed towards desired outcomes or solutions. In prescriptive analytics, multiple future scenarios can be identified based on different input interventions.

In turn, big data analytics leverages on very large volumes of data usually beyond institutions’ transactions. Big data is characterized by the “4 Vs”: Volume, Variety, Velocity and Veracity. For instance, a potential source of big data could be medical home devices monitoring patients’ vital signs. Big data analytics requires a revisit of data analysis techniques in fundamental ways at all stages from data acquisition and storage to data transformation and interpretation and, in particular, the task of collecting and analyzing data, which is at the heart of the big data analytics pipeline.

Concerning the support to decision making through Machine Learning, the main types of techniques are:

• Inductive learning in which models are built from the generalization of examples;
• Deductive learning in which deduction is applied to obtain generalizations from a solved example and its explanation;
• Genetic learning in which algorithms are inspired in the theory of evolution are applied to find general description to groups of examples;
• Connexionist learning in which generalization is performed by the adaptation mechanisms of artificial neural networks.

The main goals of the section are to support social security institutions on applying data analytics as well as on adopting emerging technologies.

These guidelines are primarily intended to provide guidance to the ICT unit on implementing and providing adequate enabling tools and services to the business areas. They also aim at providing guidance to the institution’s management on applying cutting-edge and emerging technologies. In addition, these guidelines may mean that the technical development and operational teams will have to adapt their skills, and they will help identify new skills requirements.