KISS - Artificial Intelligence Service and Systems
Sub-project: Data management
As part of the KISS project, the IDACUS Institute will be working on the topic of data management. Data management is a fundamental competence in AI and deals with the way in which data is processed, prepared and analyzed. Scientifically sound methods, processes, algorithms and systems can be used to derive insights and patterns from structured and unstructured data. Topics that are dealt with within this framework are
- Data integration in order to link data from a wide variety of sources, harmonize a wide variety of data formats and make them available. Application methods include queuing systems, streaming/batch data infrastructures, data pipelines, ETL tools and data preparation processes.
- Data quality to evaluate, prepare and integrate data. Application methods include data cleansing algorithms, matching rules, feature engineering methods, data quality metrics and augmentation approaches.
- Data Mining covers data mining algorithms and related methods of knowledge extraction, knowledge representation and reasoning. Mathematical models and algorithms that can be applied to large amounts of data are covered. Methods include processes for frequency and association analysis (rule mining), knowledge discovery in databases (KDD), information retrieval, text, web and process mining, time series analysis and the corresponding reporting.
- Data governance and IT security aspects discuss issues relating to security, protection, reliability, integrity, confidentiality and availability, verification options for the IT systems used.
- In addition, IDACUS has the task of designing and providing a Jupyterhub for the entire university.
