Public health Data and gouvernment Social networks Journalism
Public health :
Data Science arises today as a major answer to important world-wide problems of health and health economics. It attracts the largest international companies that wish to invest in this field, and mobilizes all international governments in the management of their public health policies, to face extremely high challenges to improve their current health care systems and reduce substantial budgets (In France, 240 billion Euros are assigned to health expenditure, the most important item of the French national budget). The improvement of health care systems faces immense challenges such as detecting weak signals in pharmaco-epidemiology; optimizing the medical-economic protocols for particular pathologies; carrying out comparative evaluations of the effectiveness of therapeutic strategies, among others.
The Initiative team has many institutional partners within the public health system, in particular the French National Health Insurance Fund, CNAMTS (Caise Nationale de l’Assurance Maladies des Travailleurs Salariés); the Public Hospitals of Paris AP-HP (Assistance Publique - Hôpitaux de Paris); the Georges Pompidou Hospital, HEGP (Hôpital Européen Georges Pompidou); the Brain and Spinal Cord Institute and the ICM (Institut du Cerveau et de la Moëlle Epinière).
Open data and government data:
Public data, in particular data from public institutions, is definitely a source of amazing wealth that needs to be understood in order to improve the mechanics of our society and its institutions.
The application of Data Science to the study of government databases should provide responses to current problems of modern societies (employment, retirements, health, and social inequality, among others)
In this way, for example, within the framework of the Etalab government mission, in charge of creating an interdepartmental single portal for French public data, the Initiative collaborates regularly by researching these databases. It also works in collaboration with Tomas Piketty’s team at the WID data lab of the Paris School of Economics in order to implement a Big Data-type infrastructure (data bases, tools for analysis, etc) for the processing of the world database on wealth and income.
Social networks :
Twitter, Facebook, Instagram, SnapChat, … to quote only some examples. Social networks are part of people’s everyday life. Each month, the world records over 2 billion social network users, and 30% of the time spent online is devoted to the use of these networks. A simple hobby, essential component of current social links, an exceptional source of information, but also a tool for propaganda; social networks are omnipresent and they have attained a central place in today’s society.
Understanding and characterizing their community structure and their underlying time dynamics (each with its own characteristics), the possibility of tracking the diffusion of information, quantify its impact, and other issues, are exciting topics for the members of the Initiative.
Journalisme and fact-checking :
The profusion of data produced and exchanged by public or private actors is a goldmine to analyze the mechanisms of entities and organizations in spheres such as economics, politics, culture... The ability to quickly integrate heterogeneous data both by their structure and semantics, conditions the understanding of the world that surrounds us and provides the basis for democratic debate. Therefore, structured sources such as relational databases or tables can be related to less structured sources, such as semantic graphs or text. Data analysis and integration tools have proved their worth recently, for example in the analysis of the "Panama Papers" by an international consortium of investigative journalists. Tools for interconnecting and interpreting data also make it easier for journalists that specialize in fact-checking. We collaborate with the "Decoders" team of the newspaper Le Monde, by developing automatic tools for data analysis, to be used in journalism, and in particular, for fact-checking.