Database Reference
In-Depth Information
1%, this still means millions of geo-tagged messages per day. People can now be
considered as sensors, producing signals on events they are directly involved in
or they have witnessed. Finding, visualizing, and making sense of vast amounts
of geo-referenced information will lead to a multi-resolution, multi-dimensional
representation of the planet known as
Digital Earth
.
Such multi-modality and heterogeneity of online geo-referenced multimedia
has encompassed challenges not seen in traditional geographic data analysis and
mining and has attracted the attention of researchers from various communities
of knowledge discovery in databases, multimedia, digital libraries and com-
puter vision. However, there are clearly several challenges associated with such
information: the frequent changes in the data structure, the unstructured nature
of contents, the limited quality control of information, varying uncertainty of
geographic information, and the semantic aspect on the content published, to
mention a few issues. In the era of Web 2.0, the various geo-referenced media are
mostly socially generated, collaboratively authored and community contributed.
The temporal and geographical references, together with textual metadata, reflect
where and when the media were collected or authored, or the locations and time
described by the media content. The enriched online multimedia resources open
up a new world of opportunities to discover knowledge and information related
to location and our human society.
Social networks that also use and create geo-social information have grown
in importance and popularity, adopting names such as location-based mobile
social networks, or geographic social networks, or simply social networks with
geographic features. In general, there exist several types of media with tem-
poral and geographical references on the Internet: (1) geo-tagged photos on
photo-sharing websites like Flickr, (2) geo-referenced videos on websites like
Youtube, (3) geo-referenced web documents, such as articles in Wikipedia and
blogs in MySpace, (4) geo-referenced microblogging websites such as Twitter,
and (5) “check-in” services (users can post their location at a venue and connect
with friends) such as Foursquare. Most of these services publish unsupervised
(geo-spatial) content. Their importance has grown in such a way that several
terms are currently circulating:
crowd sourcing
, which considers users as sensors
for gathering data;
distributed intelligence
, where users are basic interpreters or
preprocessors in transmitting information;
participatory science
, when citizens
participate in problem definition, data collection, and data interpretation;
volun-
teered geographic information
(VGI), when the contributive aspect is crucial;
contributed geographic information
(CGI), when the geographic features are
activated by the user; or just
user generated geographic content
(UGGC), when
there is a geographic reference, such as a place name, but the user-active con-
tribution is unpredictable. Some ambiguity in the use of different terms exists,
such as crowd-sourced data being synonymous with volunteered geographic
