Provenance Analytics:
Technologies for Interpretation of Provenance, Cause and Source in Complex, Data Driven and Connected Application

Project Duration: 1.10.2016 to 30.9.2019 (Finished)
Area of Research:


  

Technologies for Interpretation of Provenance, Cause and Source in Complex, Data Driven and Connected Application

Motivation


Data Analytics in the age of Big Data is combined with a wide range of intelligent technologies and therefore has been becoming more and more complex.  Although the success of data analytics is impressive, the trust of users in the results of the data analysis should be fostered that is nowadays generally questionable. Provenance plays a key role in building such trust with user through presenting analysis results to user in a comprehensible manner.

 

Challenges

 

There exist already systems, frameworks and initial proposals of standards for modelling, representing and generating provenance information. However, these are both for users and developers often not practical and comprehensive, so that the further development of an “actionable provenance” is necessary. Since the concrete provenance technologies are mostly domain-specific, different applications will be covered in this project, of which only few or no aspects of provenance were considered. Especially, the provenance technology for:

 

  • Data analysis in industry 4.0 environment with focus on application of diagnosis,
  • 3D digitization in the area of monument conservator and archaeology,
  • Analysis of message flow, detection of reuse and forensics as well as
  • Social semantic web of things with focus on exploration of relationships

Bauhaus-Universität Weimar,

Universität Passau,

ArcTron 3D Vermessungstechnik & Softwareentwicklungs GmbH

Image: Technologies for Interpretation of Provenance, Cause and Source in Complex, Data Driven and Connected ApplicationImage: Technologies for Interpretation of Provenance, Cause and Source in Complex, Data Driven and Connected ApplicationImage: Technologies for Interpretation of Provenance, Cause and Source in Complex, Data Driven and Connected ApplicationImage: Technologies for Interpretation of Provenance, Cause and Source in Complex, Data Driven and Connected Application


Publications:

Bunte, Andreas; Li, Peng; Niggemann, Oliver: Mapping Data Sets to Concepts Using Machine Learning and a Knowledge Based Approach. In: International Conference on Agents and Artificial Intelligence (ICAART) SCITEPRESS, Madeira, Portugal, Jan 2018 (More)

Chen, Baotong; Wan, Jiafu; Shu, Lei; Li, Peng; Mukherjee, Mithun; Yin, Boxing: Smart Factory of Industry 4.0: Key Technologies, Application Case, and Challenges. In: IEEE Access Mar 2018 (More)

Li, Peng; Niggemann, Oliver: A Data Provenance based Architecture to Enhance the Reliability of Data Analysis for Industry 4.0. In: 23th IEEE International Conference on Emerging Technologies and Factory Automation (ETFA) Sep 2018 (More)

Li, Peng; Niggemann, Oliver; Hammer, Barbara: A Geometric Approach to Clustering Based Anomaly Detection for Industrial Applications. In: 44th Annual Conference of the IEEE Industrial Electronics Society (IECON) Oct 2018 (More)

Li, Peng; Niggemann, Oliver: Non-convex hull based anomaly detection in CPPS. In: Engineering Applications of Artificial Intelligence(87) Oct 2019 (More)

Li, Peng; Niggemann, Oliver: A Non-Convex Archetypal Analysis for One-class Classification based Anomaly Detection in Cyber-Physical Systems. In: Transactions on Industrial Informatics(IEEE) Jul 2020 (More)


Funded by: BMBF
Project Management: Deutsches Zentrum für Luft- und Raumfahrt e.V. (DLR)
Grant ID: 03PSIPT5B
Contact Person: Dipl.-Math. Natalia Moriz , Dr.-Ing. Peng Li
Research Assistant: Dr.-Ing. Peng Li