- Information diffusion and language evolution on dynamical social networks
Position type: PhD training (3 years) or Postdoc position (1+1 year)
Location: Lyon (Ecole Normale Supérieure de Lyon, Inria)
Research theme: Complex networks, dynamical networks, data science, computational modelling
Duration: 3 years
HR Contact: firstname.lastname@example.org
Application deadline: 15/08/2016
Founded in 1880, the Ecole Normale Supérieure de Lyon (www.ens-lyon.fr) is one the most prestigious “Grand Écoles” in France with students selected from the top 5% among all students in the country and among the best students from abroad. The mission of the university is training national and international future professors, researchers, senior civil servants as well as business and political leaders, and to advance research in several disciplines ranging from humanity to natural sciences.
Established in 1967, Inria (www.inria.fr) is the only public research body fully dedicated to computational sciences. Combining computer sciences with mathematics, Inria’s 3,500 researchers strive to invent the digital technologies of the future. Educated at leading international universities, they creatively integrate basic research with applied research and dedicate themselves to solving real problems, collaborating with the main players in public and private research in France and abroad and transferring the fruits of their work to innovative companies.
The DANTE (team.inria.fr/dante/) team is and Inria team located in Lyon hosted by the ENS Lyon and the IXXI Complex System Institute. The goal of DANTE is to study and model the dynamics of large-scale complex networks, e.g. social networks, technological networks, social-communication networks, etc.. Building on the advancements of the digital data revolution, our main challenge is to propose generic methodologies and concepts to develop relevant formal tools to model, and analyse the dynamics of dynamical networks and ongoing dynamical phenomena. Our main focus areas are:
– Access and collect data with adapted and efficient tools. This includes a reflexive step on the biases of the data collected and their relations to real activities/application domain.
– Characterise and model the dynamics of complex temporal network structures (a) through the development of novel representation and metrics; (b) through the analysis of sizeable real world datasets; and (c) through large-scale numerical simulation based on novel computational models. We are interested in the structural and temporal properties of dynamical networks and their co-evolution with ongoing dynamical processes.
– Invent original approaches combining graph theory with signal processing. A key point is to capture temporal features in the data, which may reveal meaningful insights on the evolution of the networks.
– Find solutions in the applications domain related to communication networks. In particular, we study the dynamic of the network (available bandwidth, traffic, etc.) and propose solutions to adapt its parameterisation accordingly.
For decades studies of social behaviour have been limited to the analysis of small-scale datasets, typically collected by surveying a sample of individuals, or by direct observations after placing them in a carefully controlled experimental setting. Although these techniques have proven to be useful, they were limited by their methodology and the size of the data collected. However, through the advent of the digital data revolution automatically and unobtrusively collected information on the language, habits, interests, and interactions of millions of people allows now to make “in vivo” observations on the behaviour of single egos. At the same time industrial tools for large-scale distributed online experiments, such as Amazon’s Mechanical Turk, have been developed allowing us to test earlier hypothesis and directly ask novel questions. The combination of these new methodologies together with large datasets, state of the art techniques for data analysis, and data-driven modelling provides us with an unprecedented opportunity to understand the dynamics and formation of society and its coevolution with the ongoing language evolution in realistic settings. This approach calls for cooperation and methodologies borrowed from various fields of science as Linguistics, Statistical Physics, Computer Science, Complex Network Theory, and Social Science.
Network science is concerned with graphs that map entities and their interactions to nodes and links. Most networks/graphs describing real-world systems show the presence of complex properties and heterogeneities, which cannot be neglected in their topological and dynamical description. We are currently seeing the birth of a new field, Computational Social Science (Lazer et al., 2009), that combines the best of large-scale data analysis and methods of computational modelling to obtain a novel view of human behaviour. Computational Social Science relies on the data generated by our interactions with and through Information and Communication Technologies to characterise not only the way in which individuals interact with these systems (with the goal of improving them) but, primarily, the way in which individuals interact and influence each other. This field joins the empirical findings of system-wide, longitudinal data with the methodology of complex networks, and the theoretical understanding provided by computational models, to let us describe and predict complex patterns of collective phenomena in modern online societies.
During the program the successful candidate will analyse and model human interactions and dynamical processes intermediated by them. Social contagion phenomena, information diffusion, and most importantly language itself are the best examples within our focus. To achieve these goals the candidate is going to carry out research in the following directions:
– Observing, analysing, and modelling temporal patterns of interactions from the level of individuals to the level of globally connected social networks. This direction goes along with theoretical contribution to the fundamental description of temporal networks. Our aim is to provided metrics and computational models to describe and predict complex patterns of collective phenomena in modern online societies.
– Study the relation between the spatially and temporally detailed network structure and and ongoing dynamical processes. Special focus will be put on language evolution with the aim to design computational sociolinguistic approaches to study synchronic variations, diachronic change in correlation with social network structure.
This data-driven program will strongly rely on a large corpus of Twitter dataset enriched with precise information about the underlying social network. The candidate will investigate how the information dissemination and linguistic variation is coupled with the topology and the socio-demographic features of the social structure. The candidate will develop computational methods at the crossing of computational social science, sociolinguistics, machine learning and network science. During the training the candidate will closely collaborate with sociolinguists from the ICAR laboratory (ENS Lyon), Lalic Team (University Paris-Sorbonne, INRIA), and LIDILEM laboratory (Université Grenoble Alpes). The candidate will benefit form a rich and dynamic environment, working at the intersection of a computer science, network science, computational linguistic and computational social science.
The candidate will be hosted by the Inria DANTE team at LIP ENS Lyon. The program will be directed by Dr. Eric Fleury (Professor, Computer Science Department, ENS Lyon) and co-directed by Dr. Márton Karsai (Assistant Professor, Computer Science Department, ENS Lyon).
Skills and profile:
Applicants at PhD level should have a Master degree in computer science, physics, or related discipline with strong interest in social phenomena, while at Postdoc level a PhD degree is required in these fields.
Background in network science, data science, computational modelling is expected, proven by excellent grades or publications.
Efficiency in programming, data collection and analysis are required.
Good academic writing and presentation skills in English are required.
There are no teaching obligations but opportunities.
Salary is according to the ENS Lyon payroll system
Possibilities of on-site catering
Partial coverage of the transport costs in common
Candidates should send their detailed
– CV (with publication records)
– Motivation letter
– Transcripts of grades from the last two years (in case of PhD applications only)
– 2 recommendation letters
For more information on the post: email@example.com
For information of administrative order of application contact firstname.lastname@example.org
Location: ENS Lyon, IXXI Complex System Institute
Duration of the contract: Phd level – 3 years, Postdoc level – 1+1 year
Predictable date of hiring: in October 1st, 2016