Welcome to the MINE research group 




We believe in a cascade of data->information->insights(knowledge)->data and work on the different aspects to discover valuable insights in data (masses). The term insights hereby refers to information that has a certain worth and importance, for example a reliable model to predict anomalies or outliers, a topic engine to identify financial terms, or a conversational pattern within texts, which explains a characteristic feature. Insights are different to information, since information also implies structures that may be redundant, worthless, and suboptimal. Discovered insights, on the other side, affect a behavioral change with respect to the data, quasi an inherent adaptation (and optimization). We call systems, which support such an adaptive cascade, intelligent. An inherent data (and information) complexity can be reduced to its essence and temporal changes encountered.



  • Data Science
  • Text Mining and Information Extraction
  • Data Mining and Knowledge Discovery
  • Intelligent Databases
  • Artificial Companions

Current research

  • Supporting ESCAPE (lead by Prof. Dr. Theoharry GrammatikosLSF; ESCAPE = Efficiency and Spreading of the Financial Crisis and Algorithmic Policy Evolution), where we analyze Thomson Reuters news messages, headlines, and alerts to identify financial topics and sentiments. ContactDimitrios KampasChristoph Schommer
  • Supporting the Pain and Suffering project (lead by Dr. Smadar BustanDept. of Integrative Research Unit on Social and Individual Development, UL). We are working on models to characterize, describe, and predict the occurrence of suffering as well as the subject's communicated pain intensity and unpleasantness. ContactChristoph Schommer
  • Detection of anomalies in satellite telemetry data. An example are batteries, which do not charge for 100% because of pollution in the space (the position of the wings, satellite changes after a contact). The point is: given telemetry data, can we predict such a situation reliably and (possibly) earlier than the engineers on ground? ContactFabien Bouleau (external PhD candidate, working for SES Engineering), Christoph Schommer
  • Modeling and Analysis of Human-to-Human Conversations towards an artificial companion. The personalization of an Artificial Companion should fundament on a real basis, therefore, several experiments on human-to-human conversations have been made and a linguistic analysis performed. Examples are switches from 'Sie' (french: Vous) to 'Du' (french: tu) and self-repairs. Contact: Sviatlana HoehnChristoph Schommer

Beside thinking about and working on research-related concerns and educating undergraduates and graduates, we understand teaching as a third essential cornerstone. Here, we do not only focus on the transfer of content, but also follow a logical play-together of the different aspects. With respect to this, all given courses are logically related to each other, meaning that the circuit Data -> Information -> Knowledge -> Data is mirrored and taught. All courses are 'standard courses' and rely on internationally recognized books. We are against a student's consumerism but expect a student's automotive and self-responsibility.

  • ICC: Inventing Communities of Communication (2006 - 2010).
  • TRIAS: Logic of Trust and Reliability of Information Agents in Science (2005-2008).
  • ADAM: Adaptive Associative Memories for Active Data Streams (2005-2008).
  • Evo-Business: Evolutionary Computing for E-business (2003-2006; lead by Prof. Dr. Pascal Bouvry).
  • INTRA: Internet Traffic Management and Computer Network Protection (2003-2006; lead by Prof. Dr. Ulrich SorgerProf. Dr. Massimo Malvetti).



  • Rahmi Emre Ocakdan: Implementation of k-Anonymity to protect sensitive data. 2014.
  • Cristina Ghet: Preserving innocence and traps perspective in linguistic steganography. 2014.
  • Sergio Sousa: Intention Detection for BCI-controlled Robots. 2011.
  • Francois Beckius: A Method for Closed-Loop Brain-Computer-Interface Training Using a 14-Channel EEG. 2011.
  • Ronny Heinz: Do you know who you are - using scientific paper abstracts for author profiling. 2010.
  • Cailing Dong: Text Mining and Characterization of Bibliographic Communities. 2008.
  • Naipeng Dong: A Fingerprint Engine for Author Profiling. Master Thesis, 2008.
  • Yafang Wang: Associative Mind-maps for Trust Modeling. Master Thesis. 2008.
  • Sviatlana Danilava: Attitude Mining zur Erkennung von Modalitaeten. 2008.
  • Claudine Brucks. Semantic Network Learning from Textual Streams. Master Thesis. 2008.
  • Cynthia Wagner: CoZo+: A framework for Content Zoning of Text Streams. Master Thesis. 2008.
  • Ejikeme Uzoghukwu: Adaptive Netting in Data Streams. Master Thesis. 2007.
  • Conny Uhde. Untersuchungen zur Autorenprofilierung in Texten mit Hilfe eines Clusteringansatzes. 2007.

Last updated on: 11 Nov 2015