A Study of Repeat Collaboration in Social Affiliation Networks
Open Access
- Author:
- Carrino, Christopher N.
- Graduate Program:
- Industrial Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- August 10, 2006
- Committee Members:
- Soundar Rajan Tirupatikumara, Committee Chair/Co-Chair
Reka Z Albert, Committee Member
Leah C Newman, Committee Member
Tao Yao, Committee Member - Keywords:
- social network analysis
repeat collaboration
link prediction
link analysis
affiliation network
centrality
similarity
network data mining
pareto clique - Abstract:
- In recent years complex systems have often been analyzed from an abstract perspective using nodes that represent system entities and edges that represent the relationships among entities. The system itself is represented by the collection of nodes and entities, referred to as a network. This perspective forms the basis of a new field of study known as network science, and has been used to analyze complex systems across many diverse disciplines including sociology, biology, linguistics and engineering. One of the most critical questions that network scientists seek to answer is “Which processes determine the network’s topology of nodes and edges?” The purpose of this thesis is to extend the field of network science by providing an analysis of one such process that has yet to be analyzed in detail - the process of repeat collaboration. Repeat collaboration is a special case of collaboration where, in the case of a social network, people (nodes) choose to work with each other (edges) on more than one occasion. The amount of repeat collaboration that occurs among the participants in a social network varies across different networks and different social contexts. To fully understand this process, patterns of repeat collaboration are analyzed in this thesis across five different social contexts: corporate boards of directors, coauthors, creative artists, entertainment actors and casual acquaintances. Repeat collaboration is shown to significantly influence the formation of the topology of social networks of multiple types and to influence the performance of algorithms on these networks as well. The goals of this analysis of repeat collaboration are to: 1) develop metrics of repeat collaboration from various perspectives within a social affiliation network; 2) identify and explain patterns of repeat collaboration within various social contexts; 3) utilize these metrics and patterns to extrapolate knowledge from any given affiliation network topology about the participants and events that inhabit the network.