Ontology for Seamless Integration of
Agricultural Data and Models
Ioannis N. Athanasiadis
1
, Andrea-Emilio Rizzoli
1
, Sander Janssen
2
,
Erling Andersen
3
, and Ferdinando Villa
4
1
Dalle Molle Institute for Articial Intelligence, USI-SUPSI, Lugano, Switzerland
2
Alterra, Wageningen University and Research Centre,
Wageningen, The Netherlands
3
Danish Centre for Forest, Landscape and Planning, University of Copenhagen,
Copenhagen, Denmark
4
Ecoinformatics Collaboratory, University of Vermont, Burlington, VT, USA
Abstract. This paper presents a set of ontologies developed in order to
facilitate the integration of a variety of combinatorial, simulation and op-
timization models related to agriculture. The developed ontologies have
been exploited in the software lifecycle, by using them to specify data
communication across the models, and with a relational database. The
Seamless ontologies provide with de nitions for crops and crop products,
agricultural feasibility lters, agricultural management, and economic
valuation of crop products, and agricultural and environmental policy,
which are in principle the main types of data exchanged by the models.
Issues related to translating data structures between model program-
ming languages have been successfully tackled by employing annotations
in the ontology.
1 Introduction
The study of agricultural systems requires data spanning across several domains,
including ecology, crop science, agronomy, meteorology, economy, policy and de-
mographics. Any modelling framework that aims to integrate crop biophysical
models and agro-economic models, at different scales of time and space, obvi-
ously needs to offer processes and tools for the seamless and sound management
of data. Accessing data is just one side of the problem, as different sources need to
be homogenized, documented and properly annotated, before been made avail-
able. The other side is persistent storage of simulation results, which again re-
quires rich meta-data to ensure transparency and provide some degree of quality
control. We faced such issues in the development of the Seamless-IF framework,
where a community of more than one hundred scientists were in need to achieve
consensus in their data and model conceptualizations.
This paper presents a remedy to tackle the complexity of agricultural data
management issues, by developing and utilizing a set of ontologies for the de-
velopment of knowledge bases related to agriculture. In the following section we
discuss in short the Seamless-IP project and its supporting software infrastruc-
ture, SeamFrame, from the perspective of data integration and annotation. Next,
F. Sartori, M.A. Sicilia, and N. Manouselis (Eds.): MTSR 2009, CCIS 46, pp. 282 293, 2009.
c Springer-Verlag Berlin Heidelberg 2009
In 3rd Intl Conf onMetadata and Semantics Research (MTSR 09), (Sartori, F., Sicilia, M. A., andManouselis,
N., Eds.), Springer Verlag, 2009, pp.282 293.

Ontology for Seamless Integration of Agricultural Data and Models 283
we document the empirical process followed in the development of the Seamless
ontologies by a community of scientists. In section 4 the main constructs of the
developed ontologies are presented along with their use to facilitate the integra-
tion of a set of constituent models with a relational database. We also report
how the ontology development was integrated in the software lifecycle.
2 Ontologies in Integrated Assessment Studies
2.1 The Seamless Integrated Project
The Seamless Integrated Project (Seamless-IP)
1
develops an integrated frame-
work for assessing and comparing, ex-ante, alternative agricultural and environ-
mental policy options, allowing analysis across different scales, dimensions of
sustainability and for a broad range of issues and agents of change [1].
A large community of more than a hundred scientists from different disci-
plines was involved in Seamless-IP to study the phenomena involved, develop
new (or adopt existing) computer models to quantify them, discover and orga-
nize appropriate data required for model calibration and execution, and develop
a computer-based integrated framework that is capable to execute the model
chain and apply it to various regions of Europe. Certainly, the goals of the
project are highly complex, as it is required to bring together an array of hetero-
geneous models, which are developed following different paradigms (continuous-
time simulation models, combinatorial models, market and farm optimization
models) accessing data provided by diverse sources. Agricultural, economic, me-
teorological and landscape data, at different temporal and spatial scales, are fed
into the models.
The wide diversity of modeling paradigms and data sources underline the
need for cross-disciplinary conceptual integration, by facing the challenge of sci-
entific integration, while providing with practical solutions that can be applied
in the software development process. The approach adopted in Seamless-IP was
to employ Semantic Web techniques for specifying the domain of agriculture.
Specifically, this was achieved by developing a set of domain ontologies in order
to:
build a shared view on the systems modeled, through identifying and resolv-
ing ambiguities in terms and data structures;
facilitate model integration in a sound way, by overcoming scaling problems
that are typically remain hidden in low levels (i.e. at the coding phase);
contribute with added value to the model development, by targeting reusabil-
ity, interoperability and extensibility of model components.
Mutual understanding across disciplines is often hindered by jargon, language,
past experiences and presumptions of what constitutes persuasive argument,
and different outlooks across disciplines or experts of what makes knowledge or
information salient for policy makers or policy assessments [2].
1
The Seamless-IP project website is: http://www.seamless-ip.org