Date of Completion
8-6-2015
Embargo Period
8-5-2015
Major Advisor
Dipak K. Dey
Associate Advisor
Nitis Mukhopadhyay
Associate Advisor
Haim Bar
Field of Study
Statistics
Degree
Doctor of Philosophy
Open Access
Open Access
Abstract
This dissertation has its main focus on the development of social network community detection algorithms. Real world social networks are usually found to divide naturally into small communities. In the big data age, effective and scalable algorithms detecting network community structures are in demand in a wide range of business applications, such as marketing segmentation, friend recommendation in online social networks, and product recommendation for online retailers such as Amazon. We aim to leverage the power of statistical inference over uncertainty to scalable community detection algorithms. We developed three novel community detection algorithms: the first is a statistical model-based clustering approach, the second is performing optimization on a global objective function, and the third is based on the optimization of a localized objective function. These three algorithms may service for different purposes of applications.
Recommended Citation
Ouyang, Guang, "Social Network Community Detection" (2015). Doctoral Dissertations. 870.
https://digitalcommons.lib.uconn.edu/dissertations/870