BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Diffusion in networks: fake news and random walks
DTSTART:20190906T133000
DTEND:20190906T143000
DTSTAMP:20260920T172531Z
UID:c1568277a56b99e39ca0c41234637698e1dbef3459ec5e8a2011de9f
CATEGORIES:Conferences - Seminars
DESCRIPTION:Alexandre Bovet\nWe combined machine learning\, network scienc
 e\, statistical physics and causality analysis to measure the importance o
 f fake news compared to traditional news in Twitter during the 2016 US pre
 sidential election and also to understand their influence and the mechanis
 ms of their diffusion.\nUsing a dataset of more than 170 million tweets co
 vering the five months preceding election day and concerning the two main 
 candidates of the 2016 US presidential election\, we find that 25% of the 
 tweets linking to a news spread either fake or extremely biased news. We a
 nalyzed the networks of information flow and found the most important news
  spreaders by using the theory of optimal percolation and used a multivari
 ate causal network reconstruction to uncover how fake news influenced Twit
 ter activity during the presidential election.\n\nMany social\, biological
  or economic systems can be described as complex networks representing the
  interactions between multiple agents.\nOne important task to understand t
 hese systems is the problem of community detection\, i.e. finding a simpli
 fied view of a complex systems' components and how they interact.\nCommuni
 ty detection has been used productively in many applications\, including i
 dentifying allegiances or personal interests in social networks\, biologic
 al function in metabolic networks\, the modular organization of the brain 
 or fraud in financial transaction networks.\nHere\, we show that by modeli
 ng random walks on networks we can build a principled framework to describ
 e and detect communities in complex networks and how this approach can be 
 extended to the case of temporal networks to provide a principled method t
 o study communities in non-stationary temporal networks.\n\nAlexandre Bove
 t is a researcher at the Institute of Information and Communication Techno
 logies\, Electronics and Applied Mathematics (ICTEAM) of the Université C
 atholique de Louvain working on complex systems and in their modelling usi
 ng complex networks. He is interested in interdisciplinary approaches to a
 nswer biological\, economic and social questions using tools from physics 
 and data science.\nHe obtained his PhD in physics in 2015 from EPFL in Lau
 sanne\, Switzerland\, for his work on the transport of particles in turbul
 ent plasmas. He then moved to the Theoretical Biology lab at the ETHZ\, Zu
 rich\, to investigate the spread of diseases in a contact network of a pop
 ulation of wild mice as well as its social and behavioral organization. In
  2016\, he moved to the Levich Institute of the City College of New York\,
  USA\, with a fellowship of the SNSF to work on the diffusion of informati
 on and opinion dynamics in social networks.\n 
LOCATION:MXF 1 https://plan.epfl.ch/?room==MXF%201
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR