Unformatted text preview: r o
we show later is t
arises as to whether
obscuring underlyin FIG. 1. which participate sample network (a) before preference can
• High degree nodes Community structures of awith out communityand
(b) after violator removal, showing the effect of two violators out of
obscure the behavior of ordered communities four communities at
40 nodes. Vertices are initially assigned to one of •
random (identiﬁed via node shape and shading). The two vertices with
In metabolic networks: Water, ATP, hydrogen (“currency” versus “commodity” metabolites),
the largest degree are violators (hexagons) and do not preferentially
M. Huss and P. Holme, any single Biol. 1, 280The remaining edges fall within
connect to IET Sys. community. (2007).
communities with 85% probability and between them with 15%.
In email networks the boss the communities detected usingbroadly (for discussion in Open
Boxes illustrate (and others) communicate the algorithm in
Source Software, . C. Bird, et al. ICSE (ACM Press, 2008)
HTTP server, Pytho
Ref. see, • In software call graphs, low-levelviolators. The methodsmalloc, etc). herein
and identity of the functions (e.g. printf, we develop
allow detection of vertices acting as violators without needing constructed by com
respective OSS proj
years [30–32]. In th So networks can break into communities but what does it
• Different algorithms can give different results:
Trust the communities that result consistently across algorithms
• In social systems, communities seem to be topically related.
• In biological systems, some evidence that communities relate to function,
some evidence they do not relate.
• In mechanical systems proper function seems to span communities. (See
Dan Whitney slides)
• Communities help us do good visualization layout (Porter et al AMS Notices
2009; and visualization lecture upcoming by Muelder)....
View Full Document
- Winter '11
- Sort, Complex network, community structure, Laplacian matrix, Physical Review E