Term-weighting approaches in automatic text retrieval
Information Processing and Management: an International Journal
Neural computation and self-organizing maps: an introduction
Neural computation and self-organizing maps: an introduction
Laying out and visualizing large trees using a hyperbolic space
UIST '94 Proceedings of the 7th annual ACM symposium on User interface software and technology
A focus+context technique based on hyperbolic geometry for visualizing large hierarchies
CHI '95 Proceedings of the SIGCHI Conference on Human Factors in Computing Systems
Self-organizing maps
An improved boosting algorithm and its application to text categorization
Proceedings of the ninth international conference on Information and knowledge management
An Evaluation of Statistical Approaches to Text Categorization
Information Retrieval
Text Categorization with Suport Vector Machines: Learning with Many Relevant Features
ECML '98 Proceedings of the 10th European Conference on Machine Learning
Indexing and mining audiovisual data
AM'03 Proceedings of the Second international conference on Active Mining
Hi-index | 0.00 |
This paper introduces a new type of Self-Organizing Map (SOM) for Text Categorization and Semantic Browsing. We propose a "hyperbolic SOM" (HSOM) based on a regular tesselation of the hyperbolic plane, which is a non-euclidean space characterized by constant negative gaussian curvature. This approach is motivated by the observation that hyperbolic spaces possess a geometry where the size of a neighborhood around a point increases exponentially and therefore provides more freedom to map a complex information space such as language into spatial relations. These theoretical findings are supported by our experiments, which show that hyperbolic SOMs can successfully be applied to text categorization and yield results comparable to other state-of-the-art methods. Furthermore we demonstrate that the HSOM is able to map large text collections in a semantically meaningful way and therefore allows a "semantic browsing" of text databases.