The solution of various tasks benefits from the fusion of heterogeneous data and information or presupposes it, as this is the only way to create a sufficiently high-quality information basis for subsequent decisions. Machine-implementable models and processes that enable IT systems and applications to exploit, to formalize, to combine and ultimately to make the best possible use of different information contributions in a similar "cognitive" way to humans for the solution of a concrete possess high potential, especially in the era of Big Data.
We develop tools for the IT-supported fusion of syntactically incompatible data and information, also under consideration of the semantic interoperability. Since the world of the future is increasingly digitalized and networked, special attention is paid to their applicability in system networks. Thereby, the use of Natural Language Processing (NLP) methods also enables the exploitation of unstructured textual information contributions.
In the EU projects MAGNETO and PREVISION, we develop text and data mining methods for the extraction of relevant information from large unstructured text and data collections in order to support police investigators in solving extensive and complex crime cases.