Building Knowledge Graphs from Survey Data: A Use Case in the Social Sciences (Extended Version)

• DOI: 10.1007/978-3-030-32327-1_48
• Publication Date: 2019-10-30T15:24:59Z
• Authors: Lars Heling; Felix Bensmann; Benjamin Zapilko; Maribel Acosta; York Sure-Vetter
• Keywords: Knowledge Graph Embedding; Signal Processing on Graphs; Knowledge graph; Relational Data Modeling; Temporal Data Mining; Knowledge Discovery

Many research endeavors in the social sciences rely on high-quality empirical data. Survey data is often used as a foundation to investigate social behavior. The GESIS Panel is a probability-based mixed-mode panel survey in Germany providing high-quality survey and statistical data about e.g. political opinions, well-being, and other contemporary societal topics. In general, the integration and analysis of relevant data is a time-consuming process for researchers. This is due to the fact that search, discovery, and retrieval of the survey data requires accessing various data sources providing different information in different file formats. In this paper, we present our architecture for building a Knowledge Graph of the GESIS Panel data. We present the relevant heterogeneous data sources and demonstrate how we semantically lift and interlink the data in a shared RDF model. At the core of our architecture is a Knowledge Graph representing all aspects of the surveys. It is generated in a modular fashion and, therefore, our solution can be transferred to the existing infrastructure of other survey data publishers.