Andreas Rauber

The growing hierarchical self-organizing map: exploratory analysis of high-dimensional data

The self-organizing map (SOM) is a very popular unsupervised neural-network model for the analysis of high-dimensional input data as in data mining applications. However, at least two limitations have to be noted, which are related to the static architecture of this model as well as to the limited capabilities for the representation of hierarchical relations of the data. With our novel growing hierarchical SOM (GHSOM) we address both limitations. The GHSOM is an artificial neural-network model with hierarchical architecture composed of independent growing SOMs. The motivation was to provide a model that adapts its architecture during its unsupervised training process according to the particular requirements of the input data. Furthermore, by providing a global orientation of the independently growing maps in the individual layers of the hierarchy, navigation across branches is facilitated. The benefits of this novel neural network are a problem-dependent architecture and the intuitive representation of hierarchical relations in the data. This is especially appealing in explorative data mining applications, allowing the inherent structure of the data to unfold in a highly intuitive fashion.

Content-based organization and visualization of music archives

With Islands of Music we present a system which facilitates exploration of music libraries without requiring manual genre classification. Given pieces of music in raw audio format we estimate their perceived sound similarities based on psychoacoustic models. Subsequently, the pieces are organized on a 2-dimensional map so that similar pieces are located close to each other. A visualization using a metaphor of geographic maps provides an intuitive interface where islands resemble genres or styles of music. We demonstrate the approach using a collection of 359 pieces of music.

The growing hierarchical self-organizing map

We present the growing hierarchical self-organizing map. This dynamically growing neural network model evolves into a hierarchical structure according to the requirements of the input data during an unsupervised training process. We demonstrate the benefits of this novel neural network model by organizing a real-world document collection according to their similarities.

Using Smoothed Data Histograms for Cluster Visualization in Self-Organizing Maps

Several methods to visualize clusters in high-dimensional data sets using the Self-Organizing Map (SOM) have been proposed. However, most of these methods only focus on the information extracted from the model vectors of the SOM. This paper introduces a novel method to visualize the clusters of a SOM based on smoothed data histograms. The method is illustrated using a simple 2-dimensional data set and similarities to other SOM based visualizations and to the posterior probability distribution of the Generative Topographic Mapping are discussed. Furthermore, the method is evaluated on a real world data set consisting of pieces of music.

Uncovering hierarchical structure in data using the growing hierarchical self-organizing map

Abstract Discovering the inherent structure in data has become one of the major challenges in data mining applications. It requires stable and adaptive models that are capable of handling the typically very high-dimensional feature spaces. In particular, the representation of hierarchical relations and intuitively visible cluster boundaries are essential for a wide range of data mining applications. Current approaches based on neural networks hardly fulfill these requirements within a single model. In this paper we present the growing hierarchical self-organizing map ( GHSOM ), a neural network model based on the self-organizing map. The main feature of this novel architecture is its capability of growing both in terms of map size as well as in a three-dimensional tree-structure in order to represent the hierarchical structure present in a data collection during an unsupervised training process. This capability, combined with the stability of the self-organizing map for high-dimensional feature space representation, makes it an ideal tool for data analysis and exploration. We demonstrate the potential of the GHSOM with an application from the information retrieval domain, which is prototypical both of the high-dimensional feature spaces frequently encountered in todays applications as well as of the hierarchical nature of data.

Systematic planning for digital preservation: evaluating potential strategies and building preservation plans

A number of approaches have been proposed for the problem of digital preservation, and the number of tools offering solutions is steadily increasing. However, the decision making procedures are still largely ad-hoc actions. Especially, the process of selecting the most suitable preservation action tool as one of the key issues in preservation planning has not been sufficiently standardised in practice. The Open Archival Information Systems (OAIS) model and corresponding criteria catalogues for trustworthy repositories specify requirements that such a process should fulfill, but do not provide concrete guidance. This article describes a systematic approach for evaluating potential alternatives for preservation actions and building thoroughly defined, accountable preservation plans for keeping digital content alive over time. In this approach, preservation planners empirically evaluate potential action components in a controlled environment and select the most suitable one with respect to the particular requirements of a given setting. The method follows a variation of utility analysis to support multi-criteria decision making procedures in digital preservation planning. The selection procedure leads to well-documented, well-argued and transparent decisions that can be reproduced and revisited at a later point of time. We describe the context and foundation of the approach, discuss the definition of a preservation plan and describe the components that we consider necessary to constitute a solid and complete preservation plan. We then describe a repeatable workflow for accountable decision making in preservation planning. We analyse and discuss experiences in applying this workflow in case studies. We further set the approach in relation to the OAIS model and show how it supports criteria for trustworthy repositories. Finally, we present a planning tool supporting the workflow and point out directions for future research.

The SOM-enhanced JukeBox: Organization and Visualization of Music Collections Based on Perceptual Models

The availability of large music repositories calls for new ways of automatically organizing and accessing them. While artist-based listings or title indexes may help in locating a specific piece of music, a more intuitive, genre-based organization is required to allow users to browse an archive and explore its contents. So far, however, these organizations following musical styles have to be designed manually. With the SOM-enhanced JukeBox (SOMeJB) we propose an approach to automatically create an organization of music archives following their perceived sound similarity. More specifically, characteristics of frequency spectra are extracted and transformed according to psychoacoustic models. The resulting psychoacoustic Rhythm Patterns are further organized using the Growing Hierarchical Self-Organizing Map, an unsupervised neural network. On top of this advanced visualizations including Islands of Music (IoM) and Weather Charts offer an interface for interactive exploration of large music repositories.

Integrating automatic genre analysis into digital libraries

With the number and types of documents in digital library systems incr easing, tools for automatically organizing and presenting the content have to be found. While many approaches focus on topic-based organization and structuring, hardly any system incorporates automatic structural analysis and representation. Yet, genre information (unconsciously) forms one of the most distinguishing features in conventional libraries and in information searches. In this paper we present an approach to automatically analyze the structure of documents and to integrate this information into an automatically created content-based organization. In the resulting visualization, documents on similar topics, yet representing different genres, are depicted as books in differing colors. This representation supports users intuitively in locating relevant information presented in a relevant form.

How to choose a digital preservation strategy: evaluating a preservation planning procedure

An increasing number of institutions throughout the world face legal obligations or business needs to collect and preserve digital objects over several decades. A range of tools exists today to support the variety of preservation strategies such as migration or emulation. Yet, different preservation requirements across institutions and settings make the decision on which solution to implement very diffcult. This paper presents the PLANETS Preservation Planning approach. It provides an approved way to make informed and accountable decisions on which solution to implement in order to optimally preserve digital objects for a given purpose. It is based on Utility Analysis to evaluate the performance of various solutions against well-defined requirements and goals. The viability of this approach is shown in a range of case studies for different settings. We present its application to two scenarios of web archives, two collections of electronic publications, and a collection of multimedia art. This work focuses on the different requirements and goals in the various preservation settings.

Automatically Analyzing and Organizing Music Archives

We are experiencing a tremendous increase in the amount of music being made available in digital form. With the creation of large multimedia collections, however, we need to devise ways to make those collections accessible to the users. While music repositories exist today, they mostly limit access to their content to query-based retrieval of their items based on textual meta-information, with some advanced systems supporting acoustic queries. What we would like to have additionally, is a way to facilitate exploration of musical libraries. We thus need to automatically organize music according to its sound characteristics in such a way that we find similar pieces of music grouped together, allowing us to find a classical section, or a hard-rock section etc. in a music repository. In this paper we present an approach to obtain such an organization of music data based on an extension to our SOMLib digital library system for text documents. Particularly, we employ the Self-Organizing Map to create a map of a musical archive, where pieces of music with similar sound characteristics are organized next to each other on the two-dimensional map display. Locating a piece of music on the map then leaves you with related music next to it, allowing intuitive exploration of a music archive.