Geolinking Service SoRa - A FAIR Social-Spatial Science Research Data Infrastructure
Zapilko, Benjamin; Goebel, Jan; Jünger, Stefan; Jung, Alexander; Lieth, Jonas; Meinel, Gotthard; Rieche, Theodor; Sikder, Sujit Kumar
- Publisher
- Zenodo
- Language
- en
Abstract
The integration of social and spatial data offers significant potential for interdisciplinary research, but it is constrained by infrastructural, legal, and methodological challenges. The Geolinking Service SoRa addresses these by providing a sustainable, privacy-compliant research data infrastructure that enables the linkage of social science survey data with georeferenced spatial information. Building on established platforms such as the IOER-Monitor, ALLBUS, and SOEP, the Geolinking Service SoRa supports complex analyses at the interface of the social and spatial sciences. A key innovation of SoRa is supporting the entire analysis process that involves geocoded survey data, including data preparation, linkage, and analysis within a secure environment. The infrastructure includes a user module, a public/private mode API, and a Geolinking API, along with R and Stata packages to support efficient workflows. Several geolinking methods are implemented, including point-to-point, buffer, isochrone-based, and routing to the nearest point of interest (e.g., schools or hospitals). Additionally, SoRa facilitates the preparation of analyses using structural datasets that mimic spatial distributions without exposing sensitive data and compromising data privacy. These synthetic datasets are tailored to specific survey datasets, it can enable researchers to test workflows and conduct preliminary, explorative analyses before accessing secure room environments. This poster presents the project’s progress, from defining infrastructure components to creating prototypes and implementing privacy-preserving features. A live demonstration will showcase how SoRa can be used to enhance social-spatial research applications. By promoting FAIR and inclusive data integration, SoRa lays a strong foundation for interdisciplinary research, empowering scholars to tackle innovative anddata-driven questions with confidence.
Full text
Survey data + spatial data = rich analysis potential Example: Is individual health status affected by urban green space? Objective: Analyse relationships between individual and environmental factors Method: Combine survey variables (e.g. health status) and spatial indicators (e.g. percentage of green spaces within 15 min walking distance) SoRa at a glance •Secure linking of social and spatial data •Privacy-by-design: geocoordinates remain hidden, yet spatial joins are precise •Designed for integration into Research Data Centres (RDCs) •R interface for researchers •Synthetic dataset for preparing the analysis (Structural Dataset) •Provides citable provenance records and transparent linking workflows to ensure reproducibility Contact: Benjamin Zapilko ([email protected]) A FAIR Social-Spatial Science Research Data Infrastructure Benjamin Zapilko, Jan Goebel, Stefan Jünger, Alexander Jung, Jonas Lieth, Gotthard Meinel, Theodor Rieche, Sujit Sikder Visit the SoRa website! Privacy, provenance, and reproducibility •Users cannot view respondents' original geocoordinates due to privacy restrictions •Users must trust correctness of linkage as geocoordinates remain invisible •Provenance record for each linkage to ensure reproducibility, which includes: •Record allows other researchers to repeat the exact same query R Package Data picker •Easy selection of datasets for intended linking •Automatic generation of reusable linking scripts •Executable in a web browser or within the R package Infrastructure •Secure and privacy compliant architecture •Supports public and private modes operation, inside and outside of secure environments •Separated processing of geocoordinates •Scalable and extensible for integration into RDCs Identifiers of all input datasets Hash values of all input data and output data Selected linking methods and parameters Identifier for the provenance record itself for citation Structural Dataset •Similar distribution/information as original data •Reflects regional structure of original data as accurately as possible •Households remain anonymized •Usage in scripts identical to original data / geocoordinates •Enables preparation of scripts for analyses •Provides initial overview of expected number of households by region or regional structure Challenges High risk of re-identification Data is distributed across RDCs GIS skills required Secure workflows for data access required
