Geographic Information Retrieval: Progress and Challenges in Spatial Search of Text

Abstract

Significant amounts of information available today contain references to places on earth. Traditionally such information has been held as structured data and was the concern of Geographic Information Systems (GIS). However, increasing amounts of data in the form of unstructured text are available for indexing and retrieval that also contain spatial references. This monograph describes the field of Geographic Information Retrieval (GIR) that seeks to develop spatially-aware search systems and support user's geographical information needs. Important concepts with respect to storing, querying and analysing geographical information in computers are introduced, before user needs and interaction in the context of GIR are explored. The task of associating documents with coordinates, prior to their indexing and ranking forms the core of any GIR system, and different approaches and their implications are discussed. Evaluating the resulting systems and their components, and different paradigms for doing so continue to be an important area of research in GIR and are illustrated through a number of examples. The article concludes by setting out a range of future challenges for research in this field.

GeoTxt: A scalable geoparsing system for unstructured text geolocation

“…Overall, some recurrent issues persist in geoparsing research: methods are not evaluated or evaluated on non‐public datasets or using proprietary systems (inaccessible, or behind paywalls); demonstration and evaluation rely on small tasks, or small gazetteers, or small geographic scope; and/or focused harvested corpora are used in evaluations that greatly simplify the toponym resolution task (Buscaldi, ; DeLozier et al, ; Gritta et al, ; Karimzadeh, ; Purves et al, ; Wallgrün et al, ).…”

“…About 60% of all data (textual and otherwise) are geospatially referenced (Hahmann & Burghardt, ). Although a decade of research in geographic information retrieval (GIR) has made substantial progress toward leveraging geographic information contained in text (Purves, Clough, Jones, Hall, & Murdock, ), geographic information systems (GIS) and software packages that are tasked with handling spatial data are still not capable of ingesting, detecting, structuring, storing, and (geo)visualizing place‐related information embedded in free‐form text.…”

Understanding the spatial dimension of natural language by measuring the spatial semantic similarity of words through a scalable geospatial context window

“…Moreover, due to the increase in geo-related information searches on the Internet, precise similarity measurements of geo-related information [ 18 – 22 ] are needed. In traditional geographic information retrieval (GIR) tasks, this problem has been defined and generalized as a measurement of the similarity between triplets, <theme><relationship><location>, formalized from documents [ 23 , 24 ]. On the basis of thematic similarity measurements in the textual context using standard traditional IR techniques (such as TF-IDF, Word-Net, and other embedding methods), GIR-related studies typically apply a consensus approach, that is, they add spatial location similarity as a constraint rule to address spatial losses in the ranking problem [ 24 – 27 ].…”

Enhancing spatial and textual analysis with EUPEG: An extensible and unified platform for evaluating geoparsers

“…Geoparsing is a research topic often studied in GIR (Jones & Purves, ; Purves et al., ). The goal of geoparsing is to recognize place names mentioned in texts and resolve them to the corresponding place instances and location coordinates (Barbaresi, ; Freire et al., ; Gritta et al., ).…”

“…Geoparsing is a critical process for extracting spatial information from textual data. It is recognized as an important research topic in the broader field of geographic information retrieval (GIR; Jones & Purves, ; Purves, Clough, Jones, Hall, & Murdock, ). A geoparsing system which takes unstructured textual data as the input and outputs a set of recognized place names and their spatial footprints is called a geoparser (Freire, Borbinha, Calado, & Martins, ; Leidner, ).…”