CodePlexProject Hosting for Open Source Software

NosoiFighters

The project special attention on research of Complex Networks. Complex Networks have Scale Free and Small Word features, what make them accurate model of many networks such as social networks. These features, which appear to be very efficient for communication networks, favor at the same time the spreading of many diseases.

Based on defined centrality measures, we show how to discover the critical elements of any network. The Degree Centrality measure gives the highest score of influence to the vertex with the largest number of first-neighbours. This agrees with the intuitive way to estimate someone’s influence from the size of his immediate environment. If we need to find influential nodes in an area modeled by network it is quite natural to use the Radius Centrality measures. This measure chooses the vertex with the smallest value of shortest longest path starting in each vertex. Closeness Centrality focuses on the idea of communications between different vertices. The vertex which is ‘closer’ to all vertices gets the highest score. In effect this measures indicates which one of two vertices needs fewer steps in order to communicate with some other vertex. Betweenness Centrality (or Load Centrality) refines the concept of communications, introduced in Closeness Centrality. Informally, Load Centrality of a vertex can be defined as the percent of shortest paths connecting any two vertices that pass through that vertex. This definition of centrality explores the ability of a vertex to be ‘irreplaceable’ in the communication of two random vertices. It is of particular interest in the study of network attack, because at any given time the removal of the maximum betwenness vertex seems to cause maximum damage in terms of connectivity and mean distance in the network. Where degree centrality gives a simple count of the number of connection a vertex has, eigenvector centrality acknowledges that not all connections are equal. In general, connections to people who are themselves influential will lend a person more influence than connections to less influence people.

The identification and then vaccination of the critical elements of a given network should be the first concern in order to reduce the consequence of epidemics. We define dynamic model for the spreading of infections on networks and build application to simulate and analyse many epidemic scenarios for various diseases. Based on available data of some social networks, we show how and why epidemics are spreading in real networks and how could be halted.

Presenting idea is a new attempt at integrating theories and practices from many area, in particular: social networks, graph and network theory, decision theory, data mining and security. It utilizes that theoretical basis for very practical purpose of growing importance and demand: widely understood countering high contagious diseases like HIV/AIDS, SARS and others.

Additional information on starting a new project is available here: Project Startup Guide.

Last edited Mar 3, 2009 at 1:01 PM by baalazamon, version 3