Content-Type: text/html R-LAIR: Riverside Lab for Artificial Intelligence Research

UCR

Computer Science and Engineering



R-LAIR: Riverside Lab for Artificial Intelligence Research

A new criterion function for exploratory blockmodeling for structural and regular equivalence (2013)

by Juan I. Casse, Christian R. Shelton, and Robert A. Hanneman

Abstract: We present a new criterion function for blockmodeling two-way two-mode relation matrices when the number of blocks as well as the equivalence relation are unknown. For this, we specify a measure of fit based on data compression theory, which allows for the comparison of blockmodels of different sizes and block types from different equivalence relations. We arm an alternating optimization algorithm with this criterion and demonstrate that the method reproduces consensual blockings of three datasets without any pre-specification. We perform a simulation study where we compare our compression-based criterion to the commonly used criterion that measures the number of inconsistencies with an ideal blockmodel.

Download Information

Juan I. Casse, Christian R. Shelton, and Robert A. Hanneman (2013). "A new criterion function for exploratory blockmodeling for structural and regular equivalence." Social Networks, 35(1), 31-50.       ext

Bibtex citation

@article{CasSheHan13,
   author = "Juan I. Casse and Christian R. Shelton and Robert A. Hanneman",
   title = "A new criterion function for exploratory blockmodeling for structural and regular equivalence",
   journal = "Social Networks",
   year = 2013,
   volume = 35,
   number = 1,
   pages = "31--50",
}

full list

More Information

Address

University of California, Riverside
Chung Hall, room 368
Riverside, CA 92521

 


External Links