Transportation is a vital function for cities, to organize exchanges of persons or goods. It includes road, pedestrian, public transportation, or at larger scales airplanes. It is a crucial question to understand how transportation networks organize themselves, and how they interact with the city.
The scientific challenge specific to transportation is to understand how the various transportation systems work. The great increase of possibilities to collect data allows to have a deeper insight into the dynamics of these systems, on time scales ranging from the duration of displacements to the much larger time scale of network generation. Our aim is to develop a cooperative approach, between on the one hand modeling, including simple models that allow to evidence the undergoing mechanisms, or more complex and realistic models, and on the other hand data collection and analysis. Both microscopic and macroscopic approaches have to be considered, if one wants to address important societal questions, as security issues, capacity of networks, or impact on pollution.
- City growth: geographical and social issues
The global challenge is to integrate the knowledge about different features, such as transportation means, networks, cultural, economical and other socio-geographical features, in order to analyze city growth and propose models for the development of cities. It is important for city planning to understand which features have the strongest impact on the structures of cities. A better knowledge could help to define more efficient policies against segregation, or to minimize the ecological cost of cities in the future.
Criminality is a topic that is of clear importance to decision makers, and indeed the public, in modern society, especially in urban environment. In Europe, and especially in France, the use of quantitative approaches is much less advances than in the USA, where, e.g., the effects of carefully monitored police patrols are evaluated through both data collecting and analysis, and through modelling. There is a clear demand from police forces for realistic prediction of crime hot spots, both temporal and spatial. The analysis of organized crime, especially of illicit drug traffic, is another topic on which modelling and numerical simulations can help to provide insights for policy making. The understanding of the interplay between socio-economic factors, laws (e.g. how drug consumption is penalysed) and illegal activities requires a major multidisciplinary program.
The challenge is to develop a systematic quantitative approach to criminality using the computational and mathematical methodologies that are most appropriate, combined with the expertise of criminologists, economists an other social scientists as well as jurists, and through the analysis of empirical data. The methodologies may involve some or all of game theory, risk and statistics, population dynamics, network theory and multiscale analysis, optimal control theory, agent based simulations. One major aim is to provide decision markers with tools that can be used to anticipate the consequences of new policies before they are implemented in practice.
The spread of infectious diseases depends crucially on the pattern of contacts among individuals, and knowledge of these patterns is essential to inform models and computational effort. However, until recently, few empirical studies were available that provide estimates of the number and duration of contacts among social groups; these studies relied mostly on surveys and disregarded the dynamical aspect of the contacts. The use of ICT devices, as explained in the “Scientific challenges” page, allows to make experiments in order to investigate the dynamical networks of contacts between individuals.
The scientific challenges lies in the temporal dimension of the contact network data gathered, while network science has mostly dealt with static networks until now. Many questions remain open, such as how to characterize the network dynamics, how to define new, dynamical, centrality measures and to find important (vulnerable) nodes, which modeling frameworks to adopt, and how to study dynamical processes such as epidemic spreading which unfold on top of dynamical networks. These questions involve topics which range from fundamental graph theory to applied epidemiology. For instance, the study of the dynamics of contact patterns in particular settings such as hospitals or schools would allow to simulate spreading of diseases in such sensitive settings, and to better understand the effect of various containment scenarios.
In this application, we adress the major challenges connected with the risks of water scarcity and food security. Indeed, it is forcasted that the world population will grow to 9 billions individuals with a tendency to adopt progressively the consumption standards of western countries. In addition, the climate change is expected to modify the patterns of fresh water availability and land fertility. It is therefore of high importance to develop tools allowing us to anticipate these evolution and the potential effect of various policy options.
In the project, we shall particularly focus on modelling the socio-economic dynamics of consumers and producers, adopting or not different patterns of behaviour. Thus the problem will be to model these evolutions as a function of some variables characterising the agents and of their interactions. We shall use various sources of data, such about different types of consumption, farming practises general characteristics of the population (age, education…). We shall also use focus groups and participatory approaches. These evolutions will have to be crossed with:
- demographic evolution (including migrations)
- technology evolution
- scenarios of policy regulations
We shall develop a model applied to the whole French population and some Southern countries particularly connected to France.
It has long been argued that markets, if left to themselves, self organise in a stable manner. Furthermore the globalisation of these markets was supposed to provide increased liquidity and, as a result of increased diversification of risk, markets should become much less volatile than in the past. Such reasoning has been shown by recent events to be manifestly deficient. The governors of the European Central Bank, the Federal Reserve Board and of several of the major national central banks have all expressed their dissatisfaction with standard macroeconomic models. Indeed Governor Trichet has argued for the development of large scale agent based models in which the individuals use simple rules of thumb. The Bank of England has insisted on the importance of understanding the robustness of the international financial network and, in particular of the banking network.
The scientific challenge is to be able to examine the symptoms of network breakdown and analyse the sort of measures that can be taken to mitigate such events. To this aim, specific models will be explored within a complex system approach. In particular, macroeconomic models of the type envisaged by the participants in this project will incorporate naturally the possibility of crises and “phase transitions” which are not features of standard economic models, but which unfortunately are very real phenomena.
A knowledge community is a social system consisting of agents who interact or collaborate in order to produce knowledge, exchange information, discuss on topics — and where ICT plays an important role. This encompasses many knowledge networks stemming from so-called “Web 2.0” websites. Blogs, micro-blogs, collaborative platforms, tagging systems, rating groups, among others: all these online social systems broadly provide a large amount of open-access, very precise, dynamic, structural and semantic data about sizeable groups of individuals. Another source of dynamic, fine-grained information on knowledge dynamics stems from scientific collaboration networks: bibliographic databases, for instance, make it possible to study the epistemic evolution of large communities of scientists, along with the deformation of their social (collaboration-based) and affiliation (citation-based or organization-based) networks.
The scientific challenge is to the study the rich socio-semantic interactions occurring in these ICT-related knowledge networks. This, in turn, will make it possible to adopt a formal and quantitative point of view on several issues raised by social sciences, in terms of knowledge diffusion, structure of social networks of knowledge, online deliberation and e-democracy, and opinion formation (quality evaluations, content filtering and curation). They moreover provide the occasion to investigate further current questions in complex systems theory: morphogenesis phenomena, information broadcast process and interactions at several scales, levels and in various networks.