Towards Automatic Content-based Organization of Multilingual Digital Libraries: an English, French, and German View of the Russian Information Agency Novosti News

In this paper we present the application of the SOMLib digital library system to a multilingual document corpus from the Russian Information Agency Novosti. News articles in Russian, English, and German are automatically organized into separate topic hierarchies using a novel unsupervised neural network, namely the Growing Hierarchical Self-Organizing Map. Furthermore, machine translation is used to create a coherent corpus in a single target language. In spite of the “noise” introduced by the automatic translation a consistent topical structuring of the integrated document collection can be created by the neural network. This facilitates straightforward browsing and exploration of multilingual document collections in a given target language.