Combined Clustering of Graph and Attribute Data

Beschreibung

This thesis presents combined clustering approaches for clustering graph data and attribute data simultaneously in order to detect clusters that are densely connected in the graph and at the same time show similarity in the attribute space. As for high-dimensional vector data, clusters usually exist only in subspaces of the attribute space, we follow the principle of subspace clustering to enable the detection of clusters which show similarity only in a subset of the attributes.
This thesis presents combined clustering approaches for clustering graph data and attribute data simultaneously in order to detect clusters that are densely connected in the graph and at the same time show similarity in the attribute space. As for high-dimensional vector data, clusters usually exist only in subspaces of the attribute space, we follow the principle of subspace clustering to enable the detection of clusters which show similarity only in a subset of the attributes.

Autor*in

Brigitte Boden

Themen in »Combined Clustering of Graph and Attribute Data«

Informatik Clustering Data Mining

Stimmen zu »Combined Clustering of Graph and Attribute Data«

Details

Link teilen