 |
 |
***** To join INSNA, visit http://www.sfu.ca/~insna/ ***** Hi All - I wanted to feedback the information I gathered regarding the software comparisons / recommendations question posted to the UCINET and the SOCNET groups. (Sorry for the overlap if you are in both groups.) My specific question was about the software available for visualization and analysis of longitudinal network data. Below is a compilation of the different responses I received. Thank you all for your help! Joan Allatta Doctoral Candidate The Wharton School [log in to unmask] <mailto:[log in to unmask]> **************************************************************************** ******* Mark Huisman and Marijtje van Duijn wrote a paper on software for social network analysis; this paper provides a comparison of a number of different software pkgs. It will appear next Fall as a chapter in the new SNA book of Carrington, Scott and Wasserman: Huisman, M. & Van Duijn, M.A.J. (2004). Software for social network analysis. In P.J. Carrington, J. Scott, & S. Wasserman (Eds.) Models and methods in social network analysis. New York: Cambridge University Press. Write directly to Mark to obtain an advanced copy. ([log in to unmask] <mailto:[log in to unmask]> ) ******************************************************** Zoomgraph at HP You can try project: http://www.hpl.hp.com/research/idl/projects/graphs/ <http://www.hpl.hp.com/research/idl/projects/graphs/> Take a look at the movie links. It demonstrates interactions in a company (on a weekly basis) between individuals. The system accepts data with any kind of attributes, the ability to declare "states," and a way to move smoothly between states. Contact Eytan Adar at HP [[log in to unmask]] if you need any more info, ******************************************************** The combination of SoNIA and Siena could be a powerful means of analyzing your longitudinal network data. SoNIA was created so users could construct animations of dynamic networks (whether you have continuous or discrete notions of time), browse attribute-rich network data, and have a platform for testing and comparing different layout techniques. Hence, SoNIA allows you to use a variety of graphing algorithms, see the underlying process by which layouts are made (i.e., cooling functions, fitness of distances to graph-coordinate space, etc - so at least some comparison and discussion of "better" or "worse" graphs can be had), and it minimizes artificial movement that arises from animating across successive graphs. The end result is a dynamic network visualization with meaningful movement. The documentation for SoNIA can be found here http://www.stanford.edu/group/sonia/ <http://www.stanford.edu/group/sonia/> , here http://www.stanford.edu/group/sonia/papers/PiranVizTalk.pdf <http://www.stanford.edu/group/sonia/papers/PiranVizTalk.pdf> , and in a forthcoming American Journal of Sociology article that explains the theory behind it http://www.sociology.ohio-state.edu/jwm/NetMovies/ <http://www.sociology.ohio-state.edu/jwm/NetMovies/> . Multiple empirical studies using SoNIA are well under way and already under review. Siena is a nice compliment because it enables you to analyze the interrelation of networks and attitudes over time (or more broadly, relational and nodal mechanisms). Documentation for Siena can be found with the StocNet software here http://stat.gamma.rug.nl/ <http://stat.gamma.rug.nl/> , the Sunbelt satellite symposium, and in recent Sociological Methodology (2003) pieces by Tom Snidjers and others (see further down on the aforementioned webpage). Yours, -Dan Contact Daniel A McFarland [[log in to unmask]] for more info. **************************************************************************** ***************** you might have a look at the 'sna' package for the R statistical computing environment - available from http://erzuli.ss.uci.edu/R.stuff/ <http://erzuli.ss.uci.edu/R.stuff/> R has a non-trivial learning curve, but the folks on the R listserv are very friendly and helpful. --elijah Contact elijah wright [[log in to unmask]] for more info. **************************************************************************** *************** If you want to get a dynamic analysis/visualization over time, you can also try our software http://www.ickn.org <http://www.ickn.org/> (similar to Sonja, but more interactive). ************************************************************ from Ning Yu at Indiana University Here is a visualization I just generated for a Weblog research using Pajek. There are original 5516 nodes(unique Weblog URLs that manually collected, starting from 4 random chosen Weblogs) and I use cut-off at 10 in-degrees to make the visualization readable. Actually I had the 5 in-degrees cut-off and the main pattern is almost the same to cut-off at 10. The reason we choose in-degree is that this research group concerns more about the in-bound link to answer their research questions. -- Here is a snapshot with full legend information: http://ella.slis.indiana.edu/~nyu/Susan/blog/img/blog_in_10_0608.gif <http://ella.slis.indiana.edu/%7Enyu/Susan/blog/img/blog_in_10_0608.gif> -- Now comes a more dynamic one. If you have SVG viewer installed in you computer (if not, you can download from Adobe website), open the following page and you can see some check boxed in your right hand. Checking or un-checking them will show or hide particular classes. http://ella.slis.indiana.edu/~nyu/Susan/blog/img/blog_in_10_classes.htm <http://ella.slis.indiana.edu/%7Enyu/Susan/blog/img/blog_in_10_classes.htm> -- If you cannot read the file, here is a series of snapshots. http://ella.slis.indiana.edu/~nyu/Susan/blog/img/classes_1.gif <http://ella.slis.indiana.edu/%7Enyu/Susan/blog/img/classes_1.gif> http://ella.slis.indiana.edu/~nyu/Susan/blog/img/classes_2.gif <http://ella.slis.indiana.edu/%7Enyu/Susan/blog/img/classes_2.gif> http://ella.slis.indiana.edu/~nyu/Susan/blog/img/classes_3.gif <http://ella.slis.indiana.edu/%7Enyu/Susan/blog/img/classes_3.gif> http://ella.slis.indiana.edu/~nyu/Susan/blog/img/classes_4.gif <http://ella.slis.indiana.edu/%7Enyu/Susan/blog/img/classes_4.gif> http://ella.slis.indiana.edu/~nyu/Susan/blog/img/classes_5.gif <http://ella.slis.indiana.edu/%7Enyu/Susan/blog/img/classes_5.gif> The classification is based on in-degree again. For example, in classes_4.gif as , only node have over 59.71(well, 60 in this case) wiill be shown You can easily make animation using those snapshots.:) For comparing different visualization software, I never did a literature review on it.:P But I did tried a lot and find that actually those famous software have similar function and all can be very powerful if you really KNOW it. So it becomes a problem of whether it has good documentation and whether you feel comfortable with the language it used if it is command line based (e.g.GraphVis). Basically, the more users it owns, the more powerful it is. Take Graphvis for example again, There are Perl class and Java library provided by the users. You can find some social network analysis and visualization tools under "research and resources" and "tools" on this page: http://ella.slis.indiana.edu/~nyu/link.html <http://ella.slis.indiana.edu/%7Enyu/link.html> Ning Yu at Indiana University _____________________________________________________________________ SOCNET is a service of INSNA, the professional association for social network researchers (http://www.sfu.ca/~insna/). To unsubscribe, send an email message to [log in to unmask] containing the line UNSUBSCRIBE SOCNET in the body of the message.