Scalable Fine-Grained Document Clustering via Ranking
Taufik Edy Sutanto · Queensland University of Technology · 2017
Dealing with big data, this thesis presents a novel and effective approach of scalable document clustering via ranking. The proposed clustering methods address the high-dimensionality problem in clustering analysis by introducing effective and computationally efficient cluster representations. The clustering via ranking approach is applicable to semi-supervised and unsupervised clustering problems. The proposed methods are applicable to static data as well as streaming data. The methods have been successfully tested with big social media data providing interesting insight.