What is a network backbone?
Weighted networks are networks in which each link has a specific value. These weights might capture the strength of a friendship, or the capacity of a road. Because weighted networks contain a lot of information, they can also be difficult to analyze and visualize. Therefore, it is often helpful to focus on the backbone, which is an unweighted binary network that contains only the most significant links. Backbones are especially useful for bipartite projections, where the weights represent the number of things two people share (e.g. how many of the same bills two legislators supported, how many of the same events two people attended). For an introduction to bipartite projections and backbones, check out the quick overview video (top, 13 minutes) or full-length workshop (bottom, 50 minutes; workshop materials are available here).
The R backbone package
The backbone package for R makes it easy to identify the most significant links in a weighted network and extract its backbone. The package is freely available from CRAN and can be installed by typing install.packages("backbone") into the R command line. Once the package is installed, load it by typing library(backbone) and view the documentation by typing ?backbone or vignette("backbone"). If you find a bug or want to request a few feature, please let us know via GitHub. For help using the package, don't hesitate to contact me.
The models in the backbone package have been used to study political coalitions and polarization, gastrointestinal dysbiosis in zebrafish, X-men collaborations, and Christmas movies. If you have used backbone in a publication or pre-print, let us know and we'll send you a cool backbone hex decal!
Papers about backbone
for more, please visit this page
Neal, Z. P., Domagalski, R., & Sagan, B. (2021). Analysis of spatial networks from bipartite projections using the R backbone package. Geographical Analysis. [pre-print]
Domagalski, R., Neal, Z. P., & Sagan, B. (2021). backbone: An R package for backbone extraction of weighted graphs. PLoS ONE. [pre-print]
Neal, Z. P. (2014). The backbone of bipartite networks: Inferring relationships from co-authorship, co-sponsorship, co-attendance and other co-behaviors. Social Networks, 39, 84 – 97. https://doi.org/10.1016/j.socnet.2014.06.001 [pre-print]
Collaborators on backbone
Research on backbone has been supported by the National Science Foundation (#1851625 & #2016320) and is a collaboration with Dr. Bruce Sagan and Rachel Domagalski. The project has also benefited enormously from conversations with David Schoch, Giovanni Strona, Bruce Desmarais, and many others.