Serendipity in Text and Audio Information Spaces: Organizing and Exploring High-Dimensional Data with the Growing Hierarchical Self-Organizing Map

Abstract

While tools exist that allow us to search through vast amounts of text within seconds, most systems fail to assist the user in getting an overview of the information available or maintaining orientation within an information space. We present a neural network architecture, i.e. the Growing Hierarchical Self-Organizing Map, providing content-based organization of information repositories, facilitating intuitive browsing and serendipitous exploration of the information space. To show the universal potential of this architecture, we present the automatic, content-based organization of two different types of repositories with diverse characteristics, the first being a collection of newspaper articles and the second being a music collection.