This research area deals with the analysis of data using ‘data science’ and related methods from the fields of artificial intelligence and machine learning. The insights gained from these methods can be incorporated into the improvement and redesign of processes, services, and products.

Such improvements and redesigns require in-depth technical and algorithmic understanding in such rapidly evolving areas as artificial intelligence and the ability to identify and understand challenges in various industries and organizations quickly.

Key questions include:

  • How can data be used to address problems?
  • How can advances in machine learning and data science be applied and implemented in practice?
  • How can we understand the results of ‘black box’ methods?
  • What are the implications of Big Data and artificial intelligence for governance, workforce capabilities, and human-machine interactions from a technical perspective?

Selected Projects

Large-Scale Pattern Mining: Finding and Analysing Data Patterns

The goal of this project, undertaken in cooperation with an industrial partner, is the further development of new intelligent analysis functions from the field of machine learning. Today’s companies operate in a complex environment in which a multitude of factors interact, as reflected in the patterns in data. Business decisions are based on a deep understanding of such patterns. An example is the analysis of supermarket purchases, where the search for patterns is usually computationally intensive and complex, not to mention cost-intensive. Research has provided a number of solutions to these issues that appear to be suitable for commercial use. The project’s objective is to evaluate several of these solutions for integration into the industrial partner’s new products.

Personalized Explanations

The goal of this project is to find personalized explanations for and to improve the efficiency of the interface between humans and artificial intelligence (AI) systems. Black-box models like state-of-the-art models for ‘deep learning’, are inherently difficult to understand. Personalizing explanations to a user's needs and abilities by, for example, adjusting complexity, are one way to facilitate understanding. This project also investigates how to reduce errors in an AI system’s interpretation of human input. We focus not only on improving the AI system but also on ‘co-adaptation’, where humans are taught how to interact with AI systems.


Selected Publications

Schneider, J., & Vlachos, M. (2017). Scalable density-based clustering with quality guarantees using random projections. Data Mining and Knowledge Discovery, 31(4), 972-1005.

Damevski, K., Shepherd, D. C., Schneider, J., & Pollock, L. (2016). Mining sequences of developer interactions in visual studio for usage smells. IEEE Transactions on Software Engineering, 43(4), 359-371.

Schneider, J. (2020, April). Human-to-AI Coach: Improving Human Inputs to AI Systems. In International Symposium on Intelligent Data Analysis (pp. 431-443). Springer, Cham.

Vlachos, M., Schneider, J., & Vassiliadis, V. G. (2015). On data publishing with clustering preservation. ACM Transactions on Knowledge Discovery from Data (TKDD), 9(3), 1-30.

Abraham, R., Schneider, J., & vom Brocke, J. (2019). Data governance: A conceptual framework, structured review, and research agenda. International Journal of Information Management, 49, 424-438.

Fusco, F., Vlachos, M., Vasileiadis, V., Wardatzky, K., & Schneider, J. (2019). RecoNet: An Interpretable Neural Architecture for Recommender Systems. In International Joint Conferences on Artificial Intelligence (IJCAI) (pp. 2343-2349).

Barenboim, L., Elkin, M., Pettie, S., & Schneider, J. (2016). The locality of distributed symmetry breaking. Journal of the ACM (JACM), 63(3), 1-45.

Schneider, J., & Handali, J. (2019). Personalized explanation in machine learning: A conceptualization. European Conference of Information Systems. 

Universität Liechtenstein
Dr. Johannes Schneider
9490 Vaduz

Telefon +423 265 13 23
Fax +423 265 11 12