Catálogo de publicaciones - libros

International media have recognized the visual appeal of geo-browsers such as NASA World Wind and Google Earth, for example, when Web and television coverage on Hurricane Katrina used interactive geospatial projections to illustrate its path and the scale of destruction in August 2005. Yet these early applications only hint at the true potential of geospatial technology to build and maintain virtual communities and to revolutionize the production, distribution and consumption of media products. This chapter investigates this potential by reviewing the literature and discussing the integration of geospatial and semantic reference systems, with an emphasis on extracting geospatial context from unstructured text. A content analysis of news coverage based on a suite of text mining tools (webLyzard) sheds light on the popularity and adoption of geospatial platforms.

Geospatial data and geoprocessing techniques are now directly linked to business processes in many areas. Commerce, transportation and logistics, planning, defense, emergency response, health care, asset management and many other domains leverage geospatial information and the ability to model these data to achieve increased efficiencies and to develop better, more comprehensive decisions. However, the ability to deliver geospatial data and the capacity to process geospatial information effectively in these domains are dependent on infrastructure technology that facilitates basic operations such as locating data, publishing data, keeping data current and notifying subscribers and others whose applications and decisions are dependent on this information when changes are made. This chapter introduces the notion of infrastructure technology for the Geospatial Web. Specifically, the Geography Markup Language (GML) and registry technology developed using the ebRIM specification delivered from the OASIS consortium are presented as atomic infrastructure components in a working Geospatial Web.

As geospatial imagery becomes more available and more commonly in demand as an indispensable part of the geospatial community's workflows, new solutions must be found for overcoming the barriers that have marginalized image data in the past — in particular, the compression of massive image sets without loss of quality and inclusion of the geographic metadata that would make imagery “spatially aware”. The relatively new JPEG 2000 standard is ideally suited as a delivery technology for geo-referenced imagery, but a few features are still required for use in the Geospatial Web, including a mechanism for representing geo-spatial metadata and bandwidth-aware standards for client/server interchange of image data. This article discusses JPEG 2000 and gives examples of some of the emerging technologies surrounding it — largely from the Open Geospatial Consortium (OGC) — which together make it the right imaging format for the Geospatial Web.

Geospatial information can act as a thread that can be used to integrate information from heterogeneous sources. It does so by exploiting common location information components that often exist across different domains. As such it has the potential to be a valuable resource in the implementation of the Semantic Web. This chapter examines the challenges of adding a geospatial component to the Web, with particular reference to doing so in a way that also supports the current initiatives to semantically enable the Web. It identifies those that are largely peculiar to geography and those that, whilst issues within geography, are also likely to occur in many other domains.

Traditional queries require users to invoke specific data sources, manually integrate data across multiple sources and interpret the results. These costly operations are increased for geospatial data, which may have elaborate formats and require complex geospatial operations. Conceptual search solves these problems by leveraging semantic technologies to give a new paradigm for querying data sources. New semantic geospatial tools are also added to facilitate geospatial reasoning. Conceptual search can be implemented via a service-oriented architecture for further benefits.

In recent years, the relation of Web information to a physical location has gained much attention. However, Web content today often carries only an implicit relation to a location. In this chapter, we present a novel location-based search engine that automatically derives spatial context from unstructured Web resources and allows for location-based search: our focused crawler applies heuristics to crawl and analyze Web pages that have a high probability of carrying a spatial relation to a certain region or place; the location extractor identifies the actual location information from the pages; our indexer assigns a geo-context to the pages and makes them available for a later spatial Web search. We illustrate the usage of our spatial Web search for location-based applications that provide information not only right-in-time but also right-on-the-spot.

Pervasive computing technologies, together with the increasing participation of the Web community in feeding geo-located information within tools such as Google Earth, will soon make available a huge amount of real-time information about the physical world and its processes. This opens up the possibility of exploiting all such information for the ubiquitous provisioning of context-aware services for “browsing the world” around us. However, for this to occur, proper general-purpose data models and software infrastructures must be developed. In this chapter, we propose a simple, yet effective model for the representation of heterogeneous contextual information and the design and implementation of a general user-centric infrastructure for ubiquitous browsing of the world. The presentations of some exemplar services we have implemented over it and have made available to users via Google Earth interfacing complete the chapter.

This chapter presents practical approaches to data processing in the space, time and theme dimensions using existing Semantic Web technologies. It describes how we obtain geographic and event data from Internet sources and also how we integrate them into an RDF store. We briefly introduce a set of functionalities in space, time and semantics. These functionalities are implemented based on our existing technology for main-memory-based RDF data processing developed at the LSDIS Lab. A number of these functionalities are exposed as REST Web services. We present two sample client-side applications that are developed using a combination of our services with Google Maps service.

Local cultural heritage document collections are characterized by their content, which is strongly attached to a territory and its land history (i.e., geographical references). Our contribution aims at making the content retrieval process more efficient whenever a query includes geographic criteria. We propose a core model for a formal representation of geographic information. It takes into account characteristics of different modes of expression, such as written language, captures of drawings, maps, photographs, etc. We have developed a prototype that fully implements geographic information extraction (IE) and geographic information retrieval (IR) processes. All PIV prototype processing resources are designed as Web Services. We propose a geographic IE process based on semantic treatment as a supplement to classical IE approaches. We implement geographic IR by using intersection computing algorithms that seek out any intersection between formal geocoded representations of geographic information in a user query and similar representations in document collection indexes.

Internet users are flooded with information and are thankful for help in categorizing and visualizing textual content. Geographical categorization is one of the most important criterion for filtering, grouping and prioritizing information as users are naturally more interested in local information. We describe a way to extract geographical information from textual content using natural language processing, and we display the information within a geographical context on maps and satellite images. Using the widely supported RSS format as the input format, this approach allows us to process content from nearly all online news sites and blogs.