Dr. Jannik Stroetgen

Towards the Bosch Materials Science Knowledge Base

Today, a lot of knowledge is hidden in large amounts of unstructured data available in various forms (e.g., scientific publications, patents, corporate documents ...). Thus, text mining and natural language processing approaches are required to extract valuable information from textual resources, to manage extracted information in a meaningful way, and to build new knowledge by reasoning over extracted information. For manufacturing companies, a sample area of interest is the materials science domain: There are thousands of materials available for production purposes within the automotive industry, for consumer goods, energy solutions, or building technology. Developing new materials or substituting particular materials in existing products, to just name two use cases, critically depends on the ability to find high quality answers about existing materials and their relations in a timely manner – a task which cannot be solved with keyword-based search. To address this issue, we are developing the Bosch materials science knowledge base to offer ontology-based data access with complex query answering facilities that support aggregation of information and multi-hop reasoning.

In this talk, I will present recent developments of the Natural Language Processing and Semantic Reasoning research group of the Bosch Center for Artificial Intelligence on explainable and robust AI methods with a particular focus on tools and techniques used towards the development of the Bosch Materials Science Knowledge Base.

Jannik Stroetgen is group head of the Natural Language Processing and Semantic Reasoning research group at the Bosch Center for Artificial Intelligence (BCAI). His research interests lie in the fields of natural language processing, text mining, knowledge graphs, and information retrieval. Before joining BCAI, he worked as a senior researcher at the Max Planck Institute for Informatics, where he headed the Text Analysis group. Jannik holds a PhD in Computer Science from Heidelberg University, where he worked on spatio-temporal information extraction and event-centric information retrieval, with a special focus on multilingual and domain-sensitive temporal tagging.