COMET project proKNOW (2015-2018)
Discovery of Knowledge and Structure in Industrial Processes
Aims / Research Topics
This project focuses on the development of approaches for automated knowledge extraction and structure learning for novel industrial use cases such as optimization in discrete manufacturing, maintenance work and social fraud detection.
Together with one of our company partners we work on the development of a set of fitting indicators for social fraud detection. Based on these indicators clustering methods are used to generate fraud detection models for various economic sectors. The models are integrated into a monitoring tool that serve the experts for their decision making process. Major challenges are the incomplete and obviously faulty labelling of frauds, the proper combination of data based feature extraction approaches with already existing expert knowledge and the heterogeneity of data due to regional differences.
Concerning the use cases for the production industry we are researching and developing systems for monitoring processes and analysing process data in order to detect anomalies in processes and to point them out at an early stage.
Methods / Software / Proof of Concepts
- Development of a semi-automatic feature extractor for social fraud detection
- Development of data and knowledge based fraud detection models
- Evaluation and conceptualization of process monitoring systems
This project is subsidized in the frame of COMET – Competence Centers for Excellent Technologies by BMVIT, BMDW, State of Upper Austria and its scientific partners. The COMET program is handeled by FFG.