As part of our MigHANA research cooperation and in collaboration with our project partner OOWV, we aim to develop an innovative hybrid data model. The goal of this master's thesis is to create a data model where the data itself is stored in a uniform and very simple schema (one table per attribute of an entity). Additionally, technical and semantic metadata will be stored for each attribute. This approach aims to serve both classical Business Intelligence (BI) environments and Machine Learning (ML) applications from the same data set (without redundancies and using virtual, non-persistent information models).

• Analysis of Existing Modeling Paradigms: Study and evaluate current modeling approaches in the context of BI and ML.
• Development of the Hybrid Data Model: Create a uniform schema for data storage and define the associated technical and semantic metadata.
• Implementation of Virtual Information Models: Develop and implement virtual, non-persistent information models to meet both BI and ML requirements.
• Evaluation of the Model: Apply and test the developed model using sample data to assess its effectiveness and efficiency.

• Interest in data modeling and architecture
• Knowledge of database management systems and data analysis
• Basic programming skills (e.g., in Python or SQL)
• Experience with Business Intelligence and Machine Learning tools is advantageous
• Independent, structured, and analytical working style