Tools and Techniques for Social Network Analysis
Pajek Program for Analysis and Visualization of Large Networks
Pajek: What is it Pajek is a program, for Windows and Linux (via Wine) Developers: Vladimir Batagelj Department of Mathematics, FMF Andrej Mrvar Faculty of Social Sciences University of Ljubljana, Slovenia In Slovenian language: Pajek means a spider
Pajek: What is it Pajek is freely available at: http://vlado.fmf.uni-lj.si/pub/networks/pajek/ Pajek is implemented in Pascal No installation required Only double-click and enjoy :
Pajek: Goals The main goals in the design of Pajek are: to provide the user with some powerful visualization tools; to implement a selection of efficient algorithms for analysis of large networks.
Pajek: The main Window
Pajek: Data Objects Networks main objects (vertices and lines). Default extension:.net. Network can be presented on input file in different ways: using arcs/edges (e.g. 1 line from 1 to ) using arcslists/edgeslists (e.g. 1 3 line from 1 to and from 1 to 3) matrix format Example: *Vertices 4 1 "Andrej" 0.101 0.849 0.5000 ellipse "Vlado" 0.8188 0.458 0.5000 box 3 "Pajek" 0.3688 0.779 0.5000 diamond 4 "Book" 0.8359 0.8333 0.5000 triangle *Edges 1 1 3 1 4 3 4 3 4
Pajek: Data Objects Permutations reordering of vertices. Default extension:.per. Example: *vertices 0 1 3 4 5 10 11 6 1 9 7 (vertices ranked by their degree/cc) 8 18 19 0 13 14 15 16 17
Pajek: Data Objects Clusters subset of vertices (e.g. one class from partition). Default extension:.clu *vertices 0 1 1 1 1 1 Female auxes o twe Nl Male
Pajek: Data Objects Vectors they tell for each vertex some numerical property (real number). Default extension:.vec. *vertices 0 15 10 10 8 7 15 5 11 8 9 16 10 14 5 7 ge D eer 11 7 5 15 4
Main Window Tools File Input/Output manipulation with the data objects. Network N.net Read network from Ascii file. Edit network. Choose vertex, show its neighbors and then: add new lines to/from selected vertex (by left mouse double clicking on Newline); delete lines (by left mouse double clicking); change value of line (by single right mouse clicking); Save selected network to Ascii file. Export Matrix to EPS write matrix in EPS format Dispose selected network from memory.
Main Window Tools Add additional vertices, lines or vertices/lines labels to network Source and Sink If network is acyclic, add unique first and last vertex (new network has two artificial vertices) Edges Arcs Convert all edges to arcs (in both directions) (make directed network). Arcs Edges Convert all arcs to edges (make undirected network).
Components Main Window Tools Strong Strong Components of selected network Weak Weak Components of selected network. Hierarchical Decomposition Clustering* Hierarchical clustering procedure. Input is dissimilarity network (matrix), which can be obtained using Operations/Dissimilarity or read from input file. Run Hierarchical clustering procedure. Result is hierarchy with nested clusters and dendrogram in EPS.
Main Window Tools Vector Get vector from network Betweenness centrality (Freeman). Clustering Coefficients Compute clustering coefficients in undirected network
Main Window Tools Nets Union of lines Intersection of lines Info Network Information about network General General information about network number of vertices, number of arcs, edges and loops density of lines sort line according to their values (ascending or descending) tofind the most/least important lines. Line Values Frequency distribution of line values.
Visualization Draw GraphOnly Options Move Info
Community Detection Louvain Algorithm https://sites.google.com/site/findcommunities/ Infomap algorithm http://www.tp.umu.se/~rosvall/code.html
Few more. SNAP (Stanford Network Analysis Platform) igraph Library c / c++ / python/r modules https://snap.stanford.edu/ http://igraph.sourceforge.net/ NetworkX -- Python http://networkx.github.io/ EgoNet http://egonet.softpedia.com/
Important sites Mark Newman: http://www-personal.umich.edu/~mejn/ Santo Fortunato https://sites.google.com/site/santofortunato/ Albert-László Barabási http://www.barabasi.com/ Lada Adamic http://www.ladamic.com/ Jure Leskovec https://snap.stanford.edu/