Increasing Trust in (Big) Data Analytics

back to overview


Schneider, J., Handali, J. P., & vom Brocke, J. (2018). Increasing Trust in (Big) Data Analytics. Paper presented at the 30th International Conference on Advanced Information Systems Engineering (CAiSE) 2018: Advanced Information Systems Engineering Workshop, Tallin, Estonia.

Publication type

Paper in Conference Proceedings


Trust is a key concern in big data analytics (BDA). Explaining “black-box” models, demonstrating transferability of models and robustness to data changes with respect to quality or content can help in improving confidence in BDA. To this end, we propose metrics for measuring robustness with respect to input noise. We also provide empirical evidence by showcasing how to compute and interpret these metrics using multiple datasets and classifiers. Additionally, we discuss the state-of-the-art of various areas in machine learning such as explaining “black box” models and transfer learning with respect to model validation. We show how methods from these areas can be adjusted to support classical validity measures in science such as content validity.


Organizational Units

  • Institute of Information Systems
  • Hilti Chair of Business Process Management

Original Source URL