Semantic Computing Group
The Semantic Computing Group researches and develops methods that enable machines to acquire relevant knowledge as well as linguistic capabilities. Using methods from natural language understanding and machine learning, we are aiming at machines that are capable of knowledge acquisition by reading unstructured textual data. In particular, the group focuses on methods for information extraction, semantic parsing, ontology learning, sentiment analysis, entity linking, as well as question answering.
Further, the group investigates how machines can acquire linguistic knowledge following a developmental approach. Recently, we have looked into the question as to how intelligent systems can acquire action concepts, while at the same time learning how they are linguistically realized.
Recently, the group has started to investigate how machines can be endowed with sufficient world knowledge so that they can understand argumentative structures expressed in natural language, but also generate coherent argumentative structures by their own. The group coordinates the priority program "RATIO: Robust Argumentation Machines" that has been granted by the German Science Foundation (SPP 1999).
We are also interested in the question how knowledge can be modeled using ontologies so that machines can exploit and reason upon this knowledge, in addition to exploring how the management of research data can be improved by applying semantic technologies. Further, the group investigates how the management of research data and exploitation of open data can be improved by applying semantic technologies.
By following these research lines, the group contributes to the Cognitive Interaction Technology Excellence Center (CITEC).