Jennifer D'Souza

I work as a junior research group lead for the SCINEXT project and also serve as a postdoctoral researcher for the Open Research Knowledge Graph project within the Data Science and Digital Libraries group at TIB.

My work experience is in developing unsupervised or supervised machine learning techniques for natural language processing (NLP). My current research focus is on semantifying and synthesizing knowledge from scientific text by leveraging generative AI methods, particularly LLMs. I hold a PhD from the University of Texas at Dallas where my research focus was on relation mining from natural language text with a specific focus on time and space relations. I have since worked as a postdoctoral researcher at the University of California, Davis, where I studied the application of NLP techniques to software engineering tasks under the umbrella of the Naturalness of Software initiative. Following this and before joining TIB, I spent a brief period in a startup environment in industry developing concept discovery software solutions.

My research interests (and not exclusively limited to) broadly are:

• Computational Linguistics/NLP and Language Models.

• Machine learning models for precise extraction of scientific information, scaling knowledge graphs, and powering AI-driven scientific search engines. 

• Machine learning models for organizing knowledge to automate or semi-automate ontology generation and recommendation systems.

Information about publications and projects can be found in the TIB research information system, on the websites of the University of Dallas and at Google Scholar.

Last but not least :), if you're interested in learning more about the Open Research Knowledge Graph (ORKG) project—where I'm currently part of the research and development team—I recommend starting with the videos below. The ORKG's mission is to create a free and open knowledge graph that structures FAIR-compliant scholarly contributions, allowing them to be automatically compared, visualized, and reviewed. Recently, our team launched a new scientific search engine powered by large language models (LLMs). Check it out here: https://ask.orkg.org/!