On Connecting Network Analysis and Cluster Analysis

摘要: Actor equivalence is a fundamental concept for the analysis of network representations social structure. In this paper it shown that partitioning in terms some kind (e.g., structural or regular equivalence) essentially clustering problem. When searching best partition with standard algorithms, criterion function has to be defined, compatible chosen equivalence. Such can constructed indirectly as (dis)similarity measure between pairs actors directly measuring concordance given an ideal