EDISON utilizes big data and data mining to perform social dynamics on networks. Dynamics on networked populations are useful in understanding social processes on networked populations.
This tool contains network analysis libraries to compute structural characteristics of networks. The libraries are GaLib, NetworkX, and SNAP.
CINET is actively being used in classrooms in several universities. The platform for research, teaching and collaboration offers an environment for resource sharing in network science.
This middleware will give Network Scientists access to an unparalleled computational and analytic environment for research, education and training. By harnessing new cloud-based resources in an easily accessible manner, this project will enable Network Science researchers to tackle larger, more complex problems. The project vision is to provide researchers, analysts and educators interested in Network Science with an easy-to-use cyber-environment that is accessible from their desktop and integrates into their daily work. A key goal is to greatly expand the size of networks that are routinely studied from hundreds or thousands of nodes to hundreds of millions of nodes. It will leverage the technology, data and experience of a multi-institutional team.
Goals
Features
For some of these algorithms, GaLib implements sampling based approximation algorithms known in the literature, but with error guarantees that can be controlled by the users, in contrast to exact implementations provided in other libraries, leading to significant speedups. The data structure and algorithms included in GaLib are carefully tuned to be capable for running on large graphs containing up to millions of nodes. GaLib is written in C++.
In addition to the sequential algorithms, GaLib contains a bunch of distributed-memory parallel algorithms for generating massive networks (e.g., Preferential-Attachment, Chung-Lu models, etc.), and for computing different graph algorithms (e.g., counting motifs or subgraphs, performing edge-switch in a simple graph, computing clustering coefficient, number of triangles, multinomial distribution, etc.). The parallel algorithms are able to deal with massive graphs (i.e., graphs with billions of nodes and hundreds of billions of edges) that the sequential algorithms are not even able to load into memory.
We have adapted a few graph algorithms (implementations) of GaLib for CINET. GaLib, along with NetworkX and SNAP, is used in CINET as computation engine for network analysis.
The GaLib manual contains a list of the graph algorithms currently adapted from GaLib to include them in CINET.
NetworkX is a powerful Python package for the creation, manipulation, and study of the structure, dynamics, and functions of complex networks. It was developed at Los Alamos National Laboratory by Aric Hagberg and his group, and was first released in 2005 for public use as an open source software package. NetworkX scales up to networks with hundreds of thousands of nodes, and provides algorithms for generating various kinds of random networks and for computing several properties of networks. For more information on NetworkX, please visit: http://networkx.github.io/
We have adapted a few graph algorithms (implementations) of NetworkX for CINET. NetworkX, along with GaLib and SNAP, is used in CINET as computation engine for network analysis.
The NetworkX manual contains a list of the graph algorithms currently adapted from NetworkX to include them in CINET.
Stanford Network Analysis Platform (SNAP) is a general purpose network analysis and graph mining library developed at Stanford University by Jure Leskovec and his group. It is written in C++ and easily scales to massive networks with hundreds of millions of nodes, and billions of edges. It efficiently manipulates large graphs, calculates structural properties, generates regular and random graphs, and supports attributes on nodes and edges. For more information on SNAP, please visit: http://snap.stanford.edu.
We have adapted a few graph algorithms (implementations) of SNAP for CINET. SNAP, along with GaLib and NetworkX, is used in CINET as computation engine for network analysis.
The SNAP manual contains a list of the graph algorithms currently adapted from SNAP to include them in CINET.
CINET provides implementation of 70+ network analysis algorithms with variety of types related to shortest path, sub graph and motif counting, centrality, graph traversal and so on.
In upcoming versions, CINET will also include the capability to simulate diffusion processes on networks. We will integrate multiple different simulation codes that have been developed at NDSSL, to provide different diffusion models and simulation capabilities.
EpiFast: A fast, scalable, distributed memory simulation tool capable of representing and reasoning about complex interventions and public policies.
Indemics: An interactive epidemic simulator that allows online interaction between a user and the simulation engine. It integrates a database with the simulation engine using abstractions and data models that allow efficient queries.
InterSim: A general-purpose flexible framework for simulating Graph Dynamical Systems and their generalizations. These include general kinds of (vector valued) update functions, interaction networks, update orders, and finite state machines (FSM) that describe state transitions. InterSim also integrates with Indemics.
CINET provides more than 110+ networks from various areas such as social networks, web/internet networks, biological networks, and infrastructure and transportation networks, artificial networks and so on. Networks can be visualized using different layout algorithms and feature based organizations, e.g., determining node size using degree, betweenness centrality, applying community detection algorithm. Users can add their own networks and make them either public or private. CINET supports the following two different representations of the networks:
Participating Institutions and Investigators:
Virginia Tech: Madhav V. Marathe, Keith R. Bisset, Edward A. Fox, Maleq Khan, Chris J. Kuhlman, Anil Vullikanti, Henning Mortveit, Samarth Swarup
Indiana University: Geoffrey C. Fox, Judy Qiu
University of Houston-Downtown: Ongard Sirisaengtaksin
University of Chicago and Argonne National Laboratory: Kamil Iskra
Northwestern University: Peter Beckman
Jackson State University: Richard A. Alo
North Carolina Agricultural and Technical State University: Albert Esterline
University at Albany, State University of New York: S. S. Ravi
External collaborators:
University of Illinois at Urbana-Champaign: Zsuzsanna Fagyal
Clemson University: Matthew Macauley
Virginia Tech: T. M. Murali, Rahul Kulkarni
2015 Workshop Organizing Committee:
Keith Bisset, Maleq Khan, Chris Kuhlman, Madhav Marathe, S.S. Ravi, Sherif Abdelhamid, S.M. Arifuzzaman, Md Hasanuzzaman Bhuiyan, S.M. Shamimul Hasan