[MA] Knowledge Extraction Solution with Natural Language Processing and Knowledge Graphs

Start: 15.12.2020

End: 13.07.2021

Type: Master thesis

Student: Kiara Marnitt Ascencion Arevalo

Supervisor: Prof. Dr. Andreas Harth, Andreas Belger

Abstract: Organizations have a wide variety of available data coming from different sources. To generate value, it is essential for companies to exploit all this information and convert it into knowledge. However, the different data sets are rarely interoperable as the data varies greatly between sources in terms of type, scope, and structure. For example, an organization’s context information is often provided as unstructured texts like analyst reports, public tenders, or press mentions. This thesis aims to create a solution to extract information from unstructured texts with the support of state-of-the-art NLP methods and represent it in a structured form through the application of Knowledge Graphs and Ontologies.