Do You Know Social Networks, Real Objects of Scientific Research | Antivirus

  Social Networks you know Social Networks, Real Objects of                                          Scientific Research

The 3 points to remember:

Social data opens up new avenues of research and helps speed up scientific time;

prevention of health risks: social media make it possible to better hear the voice of patients;

ethics and consent of Internet users: data science raises the question of data protection and transparency in the purpose of their use.  

Databases in the (almost) open sky, social networks have become real subjects of scientific research, offering unequaled material for the work of many researchers. Sociologists, epidemiologists, scientists use social interactions (their content and form) to feed their research… and to share it.

 To discuss it, we invited Stéphane Schück, CEO of Kap Code, Guy Fagherazzi, researcher specializing in e-epidemiology at CESP (INSERM) and Célya Gruson-Daniel, researcher in science and technology studies and CEO of DRISS.

 This Social Media Club session was moderated by David Reguer, Managing Director of RCA Factory and Guilhem Boyer, Social Media Manager at Safran.  

If the use of social data in the field of human, social and political sciences has long been recognized, its application in the field of medical research is increasingly scrutinized. 

“We have seen the emergence of more and more applications of social data in the health sector in recent years,” confirms David Réguer (RDC Factory) in the introduction.

 Social media represent an unprecedented wealth for health actors, both academically, institutionally and economically.

 As part of his scientific research in epidemiology, Guy Fagherazzi specializes in the identification of risk factors and complications of diabetes.

For the past two years, he has been conducting a worldwide study on diabetes-related distress in diabetic patients, and more specifically on psychological factors. “Historically, epidemiologists have done a lot of work on lifestyle determinants.

 Until now, the missing link has been the emotions and the psychological, which are very little studied in the cohorts.

 Thanks to social networks, we can now tackle these subjects and understand how the disease is experienced, ”develops the Inserm researcher.

Digitosome  

Two types of data from social media are of interest to health researchers: textual data and metadata. “Geolocation is very interesting for epidemiology, antivirus.

For example. This is a major contribution to this science ”, underlines Stéphane Schück, doctor and CEO of Kap Code. This start-up specializes in the extraction of social data and its applications in health: 

"social networks bring us a new perspective, which interests us enormously: Data in 'real life'", in opposition to the data collected as part of clinical studies.

 With regard to textual data, "what interests us is the way in which patients express themselves on a daily basis, in a natural way, on the news of their disease and what they feel", explains Guy Fagherazzi. , which works mainly on Twitter and on the online communities of patients known as Diabetes Online Community (DOC).

 It is important to note that when it comes to health, the big platforms (Facebook, Twitter, Instagram, Youtube) are not the only research field: specialized forums are also a privileged place for data collection. .  

The arrival of social data has completely changed the traditional methods of scientific research by validating a hypothesis. The speakers speak of digitosome or “numerisome” to designate this new aspect of research. Stéphane Schück (Kap Code) thus opposes “the old world of medical thought:

 I formulate a hypothesis, I experiment and I verify” to the “data driven” logic of digital technology, where the formulation of hypotheses happens a posteriori.

 “Digital technology has brought transversality to the way we approach data,” explains Célya Gruson-Daniel, CEO of DRISS. 

Iteration practices change by the very nature of the social data flow, which is less structured than others. »For the researcher, the analysis of social data with the appearance of what we call data sciences today is distinguished above all by an acceleration of iterations.

 An observation shared by Guy Fagherazzi: “Ten years can pass between the formulation of a hypothesis by a scientist and publication. 

Today, in many areas, this long time is no longer acceptable. We are therefore turning more and more to areas where data is already present. "