TaToo: tagging environmental resources on the
web by semantic annotations
Andrea E. Rizzoli
a
, Gerald Schimak
b
, Marcello Donatelli
c
, Jiri Hrebicek
d
, Giuseppe
Avellino
e
, Jose Lorenzo Mon
f
, Ioannis Athanasiadis
1
a
IDSIA, Lugano, Switzerland, (andrea, ioannis @ idsia.ch)
b
AIT Austrian Institute of Technology, Austria (gerald.schimak@ait.ac.at)
b
JRC, Ispra, Italy (marcello.donatelli@jrc.it)
d
Masaryk University, Brno, Czech Republic
e
Elsag Datamat, Italy
f
Atos Origin, Spain
Abstract: The web is rapidly evolving and its traditional role of repository of static
information is changing into a hub for collaboration among people. Web resources tend to
become more and more complex, and to offer services that include access to remote
databases, and computational power. All of this becomes very interesting not only for the
common user, but especially for scientists and researchers which actually see their
computers "disappear" into the web “cloud", getting back an unprecedented access to
services and computational resources.
Yet, to exploit these new facilities new tools are needed. The TaToo project aims at
exploiting a common practice among web user: search, discovery and tagging of interesting
resources. The practice of tagging allows user groups to label and classify resources
enabling aggregators to display the most relevant ones according to the context. TaToo aims
to take the core idea of tagging and adding the ability to add valuable information in the
form of semantic annotations, thus facilitating future usage and discovery, and kicking off a
beneficial cycle of information enrichment. Thus, the production of semantic meta
information will improve the discovery process, but also interpretation in a larger sense
(verification that its the information I was looking for, assessment of usefulness for a given
situation, understanding of how to use the information correctly etc.).
Keywords: semantic annotation; semantic tagging; model search and discovery; web
services; environmental information enrichment.
1. Introduction
Technological progress has been constantly providing new tools and techniques for
environmental scientists. If we look back at the not so distant past, geologists, geographers,
hydrologists, ecologists and other environmental scientists they all had to manually and
painstakingly collect data by surveys, which were expensive, labour intensive, and required
lots of time. Data were then stored in paper based repositories and archives, and processing
data was also limited by the availability of data. Perhaps the most brilliant use of spatial
data and analytic reasoning was the discovery of the source of a cholera epidemic in
International Environmental Modelling and Software Society (iEMSs)
2010 International Congress on Environmental Modelling and Software
Modelling for Environment’s Sake, Fifth Biennial Meeting, Ottawa, Canada
David A. Swayne, Wanhong Yang, A. A. Voinov, A. Rizzoli, T. Filatova (Eds.)
http://www.iemss.org/iemss2010/index.php?n=Main.Proceedings

London by Dr John Snow. In 1854 he identified the water pump of Broad Street in London
by plotting deaths on a map of the borough, as represented in Figure 1.
Figure 1. A section of the original map published by Dr Snow in 1854. The black bars
represent deaths at a given house number.
This type of representation would require only a few minutes work using a modern GIS
tool, but it took Dr Snow several days of work to put together the above map, and a
considerable effort in collecting the data.
Nowadays data can be automatically collected by remote sensing or by sensor networks;
they are stored in information systems relying on advanced database techniques and data
storage facilities. Data representation is facilitated by desktop Geographical Information
Systems (e.g. ArcGis, MapWindows) or even web-based GIS-like applications (Google
Maps, Bing Maps). Data processing is provided by sophisticated and complex models,
supported by advanced computer architectures, exploiting distributed and parallel
processing.
Given the state of things, the outlook for the environmental scientist should be particularly
bright and rosy: vast amounts of data to process, a great variety of models available to
process and elaborate data, and powerful visualisation tools. While we must regard with
awe what we have achieved in the past 30/40 years, we should also be wary of the threat
posed by having access to too much information, which we cannot neither discern nor make
sense of.
Scientists and researchers have been aware of such a threat for a long time, and various
approaches and mitigation measures have been devised. We can enumerate a few: data
catalogues, metadata, model bases and model repositories (such as EIONET, UN FAO,
etc.). At the same time, the pressure towards the integration and exchange of data pushed
towards the creation of standards for data representation, such as HDF (hierarchical data
format), the OGC standards for GIS data, and standards for model integration, such as
OpenMI as described in Gregersen [2007].
A.E. Rizzoli et al. / TaToo: tagging environmental resources on the web by semantic annotations