Learned Abstraction: Knowledge Based Concept Learning for Cyber Physical Systems.

  • Andreas Bunte
  • Peng Li
  • Oliver Niggemann
Conference paper
Part of the Technologien für die intelligente Automation book series (TIA, volume 11)


Machine learning techniques have a huge potential to support humans, some impressive results are still achieved, such as AlphaGo. Until now the results are on a sub-symbolic level which is hard to interpret for humans, because we think symbolically. Today, the mapping is typically static which does not satisfy the needs for fast changing CPSs which prohibit the usage of the full machine learning potential. To tackle this challenge, this paper introduces a knowledge based approach of an automatic mapping between the sub-symbolic results of algorithms and their symbolic representation. Clustering is used to detect groups of similar data points which are interpreted as concepts. The information of the clusters are extracted and further classied with the help of an ontology which infers the current operational state. Data from wind turbines is used to evaluate the approach. The achieved results are promising, the system can identify the operational state without an explicit mapping.


Clustering Ontology Knowledge Reasoning Classification 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.



The work was supported by the German Federal Ministry of Education and Research (BMBF) under the projects “Semantics4Automation” (funding code: 03FH020I3) and “Provenance Analytics” (funding code: 03PSIPT5B).


  1. 1. I. Ali, N. A. Madi, and A. Melton. Using text comprehension model for learning concepts, context, and topic of web content. In 2017 IEEE 11th International Conference on Semantic Computing (ICSC), pages 101–104, Jan 2017.Google Scholar
  2. 2. T. Araki, T. Nakamura, and T. Nagai. Long-term learning of concept and word by robots: Interactive learning framework and preliminary results. In International Conference on Intelligent Robots and Systems, pages 2280–2287, Nov 2013.Google Scholar
  3. 3. Lucas Drumond and Rosario Girardi. A survey of ontology learning procedures. In WONTO, volume 427 of CEUR Workshop Proceedings., 2008.Google Scholar
  4. 4. Brenden M Lake. Towards more human-like concept learning in machines: Compositionality, causality, and learning-to-learn. PhD thesis, Massachusetts Institute of Technology, 2014.Google Scholar
  5. 5. M. Mahmoodian, H. Moradi, M. N. Ahmadabadi, and B. N. Araabi. Hierarchical concept learning based on functional similarity of actions. In First International Conference on Robotics and Mechatronics (ICRoM), pages 1–6, Feb 2013.Google Scholar
  6. 6. I. Ocampo-Guzman, I. Lopez-Arevalo, and V. Sosa-Sosa. Data-driven approach for ontology learning. In 2009 6th International Conference on Electrical Engineering, Computing Science and Automatic Control (CCE), pages 1–6, Jan 2009.Google Scholar
  7. 7. T. Suma and Y. S. K. Swamy. Email classication using adaptive ontologies learning. In 2016 IEEE International Conference on Recent Trends in Electronics, Information Communication Technology (RTEICT), pages 2102–2106, May 2016.Google Scholar
  8. 8. M. Zhu, Z. Gao, J. Z. Pan, Y. Zhao, Y. Xu, and Z. Quan. Ontology learning from incomplete semantic web data by belnet. In 2013 IEEE 25th International Conference on Tools with Articial Intelligence, pages 761–768, Nov 2013.Google Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2020

Authors and Affiliations

  • Andreas Bunte
    • 1
  • Peng Li
    • 1
  • Oliver Niggemann
    • 1
    • 2
  1. 1.Ostwestfalen-Lippe University of Applied Sciences, Institut Industrial ITLemgoGermany
  2. 2.Fraunhofer IOSB-INALemgoGermany

Personalised recommendations