http://orcid.org/0000-0001-5828-9001
The “Frontiers of Geospatial Data Science” was the theme of a joint meeting of the Cartography and Geographic Information Society (CaGIS) and the University Consortium for Geographic Information Science (UCGIS), held 22–24 May 2018, in Madison, Wisconsin. At the event, the audience explored emerging opportunities and challenges including mapping social media data, mobile mapping, 3D mapping tools, interoperability, spatial flow algorithms, volunteered geographic information, geographic ontologies and geographic features, mapping and natural hazards, geovisualization and education, and research questions dealing with large databases.
In preparation for the event, the Call for Participation was circulated widely throughout the geospatial community as well as to all members of the two organizations. There were multiple options for participation, including poster presentations, oral presentations with an accompanying paper in the Proceedings, and oral presentations followed by possible publication in this journal. Of the 70 abstracts that were submitted, 24 were for research papers seeking journal publication. The Program Committee, with some additional assistance by reviewers, evaluated all the submissions and recommended a set of nine articles that underwent the journal’s regular peer-review process (managed by the Journal Editor). Once this was complete, we accepted five papers for publication in this special content section. Several of the others were invited to resubmit for consideration in later issues, pending more substantial revision. All decisions on manuscripts were made jointly by the Co-editors of this special issue (Dr. Diana Sinton, Professor Scott Freundschuh, and Dr. Nicholas Chrisman).
This collection of papers reflects the diversity of the geospatial data science agenda. Spatio-temporal issues continue to pose some of the greatest challenges (as they have for decades). Acker and Yuan analyzed traffic accident data with a temporal as well as spatial component. Chow tracked crowds as a temporal process across a complex city. Koylu, Dietrich, and Larson developed tools to examine Twitter discourse in time and space. Armstrong, Wang, and Zhang considered processing techniques for fast data streams created by sensors and other items in the Internet of Things. Armstrong also contributed what began as an opinion paper on the topic of “active symbolism” or the use of deep learning techniques to invert the classic arrangement of cartographic software. Through a contentious review process, this paper expanded (and moderated its presentation) into the form published here.
While not a complete coverage of all issues raised by data science, these cover a wide range of the current work of researchers in CaGIS and UCGIS. As the accepted papers from the conference did not fill a complete issue, the Editor has included another paper previously accepted.