Martin-Luther-Universität Halle-Wittenberg

Weiteres

Login für Redakteure

TopicExplorer

TopicExplorer is a web-based system that employs topic models to  analyze data. Topic models have been successfully applied in a wide  range of domains, e.g. text mining, search, bioinformatics, image  and video analysis as well as recommender systems. However, no  guarantees are given on the quality of the learned topics. Therefore,  TopicExplorer allows users to interactively explore the learned topics   in the context of the data given by the particular application.

TopicExplorer is designed as middleware that connects machine learning and topic inference with databases and visual web-based user interfaces. It can be easily adapted to very different application domains through a novel workflow based plugin-mechanism.  The system stores the topic models training data, the inference output  and additional data depending on the application. It is designed to  scale to very large data. Different data stores can be mixed to give  optimal performance, e.g. different types of SQL and No-SQL databases.

For further details and demos, see the Topicexplorer-Blog.

Currently implemented application domains

  • Analysis of Japanese blogs on nuclear power and the Fukushima disaster
TopicExplorer für die Analyse of japanischer Blogs

TopicExplorer für die Analyse of japanischer Blogs

TopicExplorer für die Analyse of japanischer Blogs

Road Map

  • Wikipedia
  • Analysis of natural compounds, metabolites and two-dimensional nuclear magnetics resonance (NMR) spectra

Collaborations

  • Development is supported by a generous grant by Klaus-Tschira-Foundation in the joint project "Computer science methods for social sciences, topic model for a pilot study on public understanding of science and technology" with Christian Oberländer (Institute of Japanese studies).
  • Teleport Sachsen-Anhalt GmbH (TSA), Halle/Saale
  • Unister GmbH, Leipzig

Publications

Contact

Zum Seitenanfang