UNSUPERVISED ARTIFICIAL NEURAL NETWORKS FOR CLUSTERING OF DOCUMENT COLLECTIONS

摘要

The Self-Organizing Map (SOM) has shown to be a stable neural network model for high- dimensional data analysis. However, its applicability is limited by the fact that some knowledge about the data is required to define the size of the network. In this paper the Growing Hierarchical SOM (GHSOM) is proposed. This dynamically growing architecture evolves into a hierarchical structure of self-organizing maps according to the characteristics of input data. Furthermore, each map is expanded until it represents the corresponding subset of the data at specific level. We demonstrate the benefits of this novel model using a real world example from the document-clustering domain. Comparison between both models (SOM & GHSOM) was held to explain the difference and investigate the benefits of using GHSOM.