spatial@linkedscience – Exploring the Research Field of GIScience with Linked Data

Keßler Carsten, Janowicz Krzysztof, Kauppinen Tomi

Abstract

Metadata for scientific publications contain various explicit and implicit spatio-temporal references. Data on conference locations as well as author and editor affiliations – both changing over time – enable insights into the geographic distribution of scientific fields and particular specializations. At the same time, these byproducts of scientific bibliographies offer a great opportunity to integrate data across different bibliographies to get a more complete picture of a domain. In this paper, we demonstrate how the Linked Data paradigm can assist in enriching and integrating such collections. Starting from the bibliographies of the GIScience, COSIT, ACM GIS, and AGILE conference series, we show how to convert the data to Linked Data and integrate the previously separate datasets. We focus on the spatio-temporal aspects and discuss how they help in matching and disambiguating entities such as authors or universities. We introduce a novel user interface to explore the integrated dataset, demonstrating the potential of Linked Data for innovative applications using spatio-temporal information, and discuss how more complex queries can be addressed. While we focus on bibliographies, the presented work is part of the broader vision of a Linked Science infrastructure for e-Science.