The goal of the geocrowd project is to establish a fertile research environment by means of a training network that will promote the GeoWeb 2.0 vision and advance the state of the art in collecting, storing, analyzing, processing, reconciling, and making large amounts of semantically rich user-generated geospatial information available on the Web. Specifically, activities will be centered on (i) exploiting user-contributed geospatial data, (ii) Web-geodata management and (iii) efficient means for data collection and dissemination, e.g., mobile computing. geocrowd will offer young researchers the opportunity to develop projects under one of three inter-related research themes briefly described in the following.
Research Theme 1: Integrating Geospatial Content Streams
With the proliferation of the Internet as the primary medium for data publishing and information exchange, we have seen an explosion in the amount of online content available on the Web. Thus, in addition to professionally-produced material being offered free on the Internet, everybody is able to make its content available online to everyone. The volumes of such User-Generated Content (UGC) are already staggering and constantly growing. Specifically in this project, our goal is to tame this data explosion, which applied to the spatial domain translates to massively collecting and sharing geospatial knowledge to ultimately digitize the world.
A fundamental problem in ingesting and effectively processing diverse incoming information is to specify the relationships and correspondences between data and/or metadata of different sources and proceed to reconcile them. The objective of this research theme is to develop a framework for matching and mapping different geospatial data streams. To generate better mappings, tools will be developed that allow for user involvement in the process. In the overall approach, special attention will be given to uncertainty, which plays a central role in un-structured data (streams).
Research Theme 2: GeoWeb Data Management
Tapping into large amounts of geospatial data streams will result in sizeable amounts of data. Hence efficient data management techniques are of outmost importance. The focus in our research will be on distributed data management schemes such as cloud computing, which are paramount in Web2.0 applications and services. In addition, we will investigate novel concepts such as dataspaces that represent evolving data management schemas for the specific case of geospatial information. As mobile devices are poised to become Web infrastructure nodes, we will investigate such devices in the context of distributed geospatial data management.
In terms of computing infrastructure have our current technological advances have been largely outpaced by the rate in which data is generated, success in data matching, both in terms of efficiency and in terms of effectiveness can only be achieved through collaborative effort, both in terms of engines and in terms of humans. As such, the network will investigate the benefits that cloud computing has to offer.
Ad-hoc networks for mobile devices have been a research topic for about a decade. Their purposes and goals have been shifting with technology and needs, without really ever reaching a critical mass or a well-defined application domain. Yet, the technology has matured and can be put to use in a wide range of applications. The key idea in this theme with respect to mobile computing is to explore and develop technologies that would allow users to create ad hoc (self organized) networks in a variety of settings, and to explore and develop applications executed over such networks that exploit locality and spatial data. The applications we have in mind include social networks, spontaneous gaming, and collaborative work.
Research Theme 3: Accessing Geospatial Content
Geospatial data collections need to be maintained and accessed. Dubbed “Accessing Content”, this research area addresses aspects of geospatial data collection and dissemination through respective services using Web-based as well as mobile interfaces. In addition, we will investigate cognitive and situational aspects of interaction with geospatial content.
Web-based services and tools can provide means for users through “attentional” (e.g., geo-wikis, geocoding photos) or “unattentional” efforts (e.g., routes from their daily commutes) to create vast amounts of data concerning the real world that contain significant amounts of information (“crowdsourcing”). In geocrowd we aim at developing Web-based methods that allow for the attentional collection of geospatial data that goes beyond that what is currently possible with Web sites and services.
Cognitive aspects can improve our means for interacting with geospatial data. Spatial data and spatial assistance is widely available, for example, in the form of navigation assistance (internet route planners, car navigation systems, etc.). However, such systems generally ignore most of the principles of how humans perceive and process spatial information as they have been identified in empirical research.
Location-based services can be improved by considering additional metadata in such services. The notions of context and situation have been studied in different fields of computer science. In geocrowd we aim at providing the foundations for “situationalized” ubiquitous information and services supply.