Kurzfassung

---

Data science is an interdisciplinary field that uses scientific methods, processes, algorithms and systems to gain knowledge from and insights into data that is available in various formats. On the one hand, the field is characterized by its high significance. Based on the results obtained, numerous important decisions are made that affect the individual or society as a whole: diagnoses, therapies, credit decisions, spatial planning, etc. On the other hand, data science is characterized by...
Data science is an interdisciplinary field that uses scientific methods, processes, algorithms and systems to gain knowledge from and insights into data that is available in various formats. On the one hand, the field is characterized by its high significance. Based on the results obtained, numerous important decisions are made that affect the individual or society as a whole: diagnoses, therapies, credit decisions, spatial planning, etc. On the other hand, data science is characterized by the iterative and empirical-heuristic approach by which knowledge is extracted and decisions are derived. What is typically neglected is a systematic, engineering-oriented approach that allows statements to be made about the quality of data analysis.

Today, the correctness of software systems is supported or even proven by methodical procedures and formal tools. For data science and the resulting data-intensive software, however, an appropriate inventory of methods, processes, algorithms and systems that contribute to correctness is missing. In particular, "correct behavior" of a data-intensive system cannot easily be described, because the result is not predetermined and should only be obtained through the data analysis process.

The goal of this research program is to explore methodological approaches and formal tools that support the engineering development of correct - or at least trustworthy - data-intensive software. This goal will be pursued in different subprojects, implemented in a data science development environment like Jupyter, and evaluated in application areas.


Subproject 1: Data provenance
Prof. Dr. Rälf Lämmel, Prof. Dr. Steffen StaabWhere does the data come from?Subproject 2: Compliance
Prof. Dr. Patrick Delfmann, Prof. Dr. Jan Jürjens, Dr. Marco Konersmann, Prof. Dr. Andreas Mauthe
What limitations must be considered when analyzing data and using their results?Subproject 3: Explainability
Prof. Dr. Viorica Sofronie-Stokkermans, PD Dr. Matthias ThimmHow does the software get its results?Subproject 4: Resilience
Prof. Dr. Andreas Mauthe, Prof. Dr. Jan Jürjens, Dr. Marco Konersmann
How is manipulation prevented? In the context of EnTrust, the following research projects are currently underway

• HIBPM - Handling Inconsistencies in Business Process Modelling (DFG)
• LISeQ - Language Integrated Semantic Queries (DFG)
• XAPS - Erklärbare KI für automatisierte Produktionssysteme (BMBF)

 
Sprecher
 


Prof. Dr. Jan Jürjens Vice Dean for Research B 118 +49 261 287-2722 juerjens at uni-koblenz.de

Projekt-Koordinator
 


Dr. Marco Konersmann Mitarbeiter der Arbeitsgruppe Jürjens B 119 +49 261 287 -2789 konersmann at uni-koblenz.de

Mitarbeitende
 


Prof. Dr. Patrick Delfmann A 109 +49 261 287-2516 delfmann at uni-koblenz.de



Prof. Dr. Ralf Lämmel Dean B 231 Dekan, B 214 Professor +49 261 287-2500 Telefon - Dekan dekanat4 at uni-koblenz.de



Prof. Dr. Andreas Mauthe Prodekan für Lehre Fachbereich 4: Informatik A 108 +49 261 287-2547 mauthe at uni-koblenz.de



Prof. Dr.-Ing. Dietrich Paulus Head of the Active Vision Group B 308 +49 261 287-2788 paulus at uni-koblenz.de



Prof. Dr. Viorica Sofronie-Stokkermans Leiterin der Arbeitsgruppe Sofronie-Stokkermans B 225 +49 0261 287-2780 sofronie at uni-koblenz.de



Prof. Dr. Steffen Staab B 108 +49 261 287-2761 staab at uni-koblenz.de



Dr. habil. Matthias Thimm B 122 +49 261 287 2715 thimm at uni-koblenz.de



Mahmood Al-Doori Mitarbeiter der Arbeitsgruppe Jürjens mahmoodaldoori at uni-koblenz.de



Marco Ehl Mitarbeiter der Arbeitsgruppe Jürjens B 103 2966

» weiterlesen» einklappen

---

---

---

---

---

---

---

---

---