=Paper=
{{Paper
|id=Vol-1690/paper116
|storemode=property
|title=Ontop-spatial for Geospatial Data Integration using GeoSPARQL-to-SQL Translation
|pdfUrl=https://ceur-ws.org/Vol-1690/paper116.pdf
|volume=Vol-1690
|authors=Konstantina Bereta,Guohui Xiao,Manolis Koubarakis,Martina Hodrius,Conrad Bielski,Gunter Zeug
}}
==Ontop-spatial for Geospatial Data Integration using GeoSPARQL-to-SQL Translation==
https://ceur-ws.org/Vol-1690/paper116.pdf
Ontop-spatial: Geospatial Data Integration using
GeoSPARQL-to-SQL Translation
Konstantina Bereta1 , Guohui Xiao2 , Manolis Koubarakis1 , Martina Hodrius3 ,
Conrad Bielski4 , and Gunter Zeug5
1
National and Kapodistrian University of Athens, Greece
2
Free University of Bozen-Bolzano, Italy
3
VISTA GmbH, Germany
4
EOXPLORE UG, Germany
5
Terranea UG, Germany
Abstract. We present Ontop-spatial, a geospatial extension of the well-
known OBDA system Ontop, that leverages the technologies of geospatial
databases and enables GeoSPARQL-to-SQL translation. We showcase
the functionalities of the system in real-world use cases which require
data integration of different geospatial sources.
1 Introduction
During the recent years, the amount of geospatial data in the Web of Data
has increased. This is because geospatial data practitioners coming from various
domains (e.g., earth scientists, geologists, civil engineers) that are involved in
the processing of geospatial data, also publish them as RDF to increase its value
by combining it with other data.
As a result, the Semantic Web community became active proposing data
models, query languages and applications for the representation, modeling and
visualization of linked geospatial data [5]. These efforts have been strength-
ened by the establishment of the Open Geospatial Consortium (OGC) standard
GeoSPARQL [1], a geospatial extension of RDF and SPARQL. At the same time,
similar extensions of RDF and SPARQL were also proposed, such as the frame-
work of stRDF and stSPARQL which extends RDF and SPARQL with space
and time [6,3]. RDF stores with geospatial support were also implemented, such
as Parliament, uSeekM and Virtuoso, that implement a subset of GeoSPARQL,
and Strabon [6], that implements both GeoSPARQL and stSPARQL. Despite
the long tradition of research in geospatial relational databases, and the ex-
istence of Ontology-Based Data Access (OBDA) systems that offer on-the-fly
SPARQL-to-SQL translation based on ontologies and mappings (e.g., Ontop [9],
Ultrawrap [10], Morph [8]), there was no OBDA system with GeoSPARQL sup-
port. In [2], we describe how we extended Ontop [9] with geospatial support and
implemented Ontop-spatial, a geospatial extension of the system Ontop.
The development of Ontop-spatial was initially motivated by the Statoil use
case in the context of the EU FP7 project OPTIQUE6 , in order to address the
6
http://optique-project.eu/
issue of creating virtual RDF graphs on top of large relational databases that
contain geometries and get frequently updated. Ontop-spatial is being used in
the Urban accountant, Land management, and Crisis Mapping services of the
EU FP7 project MELODIES7 . More recently, Ontop-spatial has been used in
the maritime security domain by the German BMBF project EMSEC [4].
In this demo paper we focus on how Ontop-spatial can be used to integrate
geospatial data from different sources and express rich queries that combine
them. In Section 2 we present a technical overview of Ontop-spatial, explaining
its compliance with other visualization tools that are useful for domain experts.
In Section 3 we give an overview of our demonstration, presenting how Ontop-
spatial is used in land management and crisis mapping real-world scenarios.
2 Ontop-spatial
Ontop-spatial8 extends Ontop to enable the on-the-fly GeoSPARQL-to-SQL
translation on top of geospatial databases and thus becomes the first OBDA
system with geospatial support. It is able to connect to a geospatial database
(currently PostGIS or Spatialite) and create virtual geospatial RDF graphs on
top of it, using ontologies and mappings. It supports the following components of
GeoSPARQL: Core, Topology Vocabulary, Geometry topology extension, RDFS
entailment and a subset of Geometry Extension. To the best of our knowledge,
it is also the first GeoSPARQL implementation that supports the query rewrite
extension of GeoSPARQL. In [2] we explain how GeoSPARQL queries are pro-
cessed by Ontop-spatial and are transformed into the respective spatial SQL
queries that are evaluated by geospatial databases.
For example, the GeoSPARQL query in Listing 1.1 retrieves buildings that
are affected by floods (i.e., they have intersecting geometries with the flood
geometries). The query described in Listing 1.2 can also be evaluated in Ontop-
spatial returning the same results, as it gets transformed internally to the query
in 1.1, according to the query rewrite extension of the GeoSPARQL specification.
Listing 1.1: Query 1 (Quantitative) Listing 1.2: Query 2 (Qualitative)
SELECT DISTINCT ? name ? build ? type SELECT DISTINCT ? name ? build ? type
WHERE {? s1 f : type ? type . WHERE {
? s1 geo : asWKT ? g1 . ? s1 f : type ? type .
? s2 geo : asWKT ? g2 . ? s2 rdf : type osm : Building .
? s2 rdf : type osm : Building . ? s2 osm : hasName ? name .
? s2 osm : hasName ? name . ? s2 osm : b u i l d i n g C a t e g or y ? build .
? s2 osm : b u i l d i n g C a t e g o r y ? build . ? s1 geo : sfIntersects ? s2
FILTER ( geof : sfIntersects (? g1 , ? g2 )) } }
7
http://www.melodiesproject.eu/software-tools
8
https://github.com/ConstantB/ontop-spatial
3 Demonstration Overview
The demonstration of Ontop-spatial will be based on the real-world scenar-
ios of land management and crisis mapping studied in the EU FP7 project
MELODIES. These use cases are led by the German company VISTA and the
companies EOXPLORE and Terranea respectively.
In the land management scenario, we are interested in discovering agricultural
fields that intersect with protected areas. To achieve this, we need to combine
information from the following geospatial datasets:
– Agricultural fields. This dataset contains information about fields, i.e., their
geographic location, their name, code, etc. This is in-house data, but we can
use a sample of it for the demonstration.
– Protected areas. This is a set of different datasets that describe categories
of protected areas. Most of this data is in-house.
– Corine Land Cover (CLC). This dataset is released as open data by the
European Environment Agency (EEA) and it contains information about
the land cover of various countries of Europe. In the context of this use case,
we focus on categories that represent protected areas.
In the crisis mapping scenario, we are mainly interested in information about
floods, which is stored originally in a PostGIS database9 maintained by EOX-
PLORE and Terranea. This database consists of a table that contains only basic
information about floods such as the location, date and some information about
the region and the country. We enrich this dataset by integrating the following
open datasets:
– The Global Administrative Areas dataset10 that contains information about
all levels of administrative divisions worldwide.
– The Corine Land Cover dataset, focusing in areas that are characterized as
“Water bodies”.
– The Open Street Maps dataset (OSM). OSM also contains some categories
for water bodies, such as rivers and lakes, as well as points of interest.
All datasets described in both use cases are originally in Shapefile format,
except for the floods data that are relational (stored in a PostGIS database).
In the demonstration we will show how we can integrate geospatial data
coming from different sources and then pose rich queries combining them, in
both of these use cases.
Integration. All Shapefiles are imported to a database. In case a database pre-
exists (e.g., the floods database), so we import the additional shapefiles there.
Every shapefile will be imported to a correspoding table, and the columns of
this table will be the same as the attributes of the shapefile.
Ontology. We need to construct an ontology to model the information that we
want to map to RDF. In order to exploit the geospatial features, this ontology
should be an extension of the GeoSPARQL ontology11 .
9
http://bit.ly/29otBQk
10
http://www.gadm.org/
11
http://schemas.opengis.net/geosparql/1.0/geosparql_vocab_all.rdf
Mappings. Mappings describe how relational data can be translated to RDF.
A mapping file is constructed, using either the R2RML mapping languages or
the native mapping language of Ontop. The Protègè plugin of Ontop provides a
user-friendly graphical interface for editing and managing mappings.
Posing rich geospatial queries. A user can create a repository using either
the embedded Sesame-based web interface of Ontop or the visual query interface
the Optique platform [11].
Visualization of results. Ontop-spatial inherits the ability of Ontop to be
used as a SPARQL endpoint. So the results can be visualized on the map using
Sextant [7], a web-based tool for browsing and visualizing linked geospatial data,
that is able to connect to (Geo)SPARQL endpoints and project the results of
the geospatial queries on the map.
The queries described in 1.1 and 1.2 are examples of the queries used in
the crisis mapping scenario. A more detailed description of the demonstration is
given in the online video: https://youtu.be/F5_2Zxi5_e8.
Acknowledgement. This work is partially supported by the EU FP7 projects
Optique (318338) and MELODIES (603525).
References
1. Open Geospatial Consortium. OGC GeoSPARQL - A geographic query language
for RDF data. OGC Candidate Implementation Standard (02 2012)
2. Bereta, K., Koubarakis, M.: Ontop of Geospatial Databases. In: International Se-
mantic Web Conference (ISWC) 2016. (To appear).
3. Bereta, K., Smeros, P., Koubarakis, M.: Representation and querying of valid time
of triples in linked geospatial data. In: ESWC. LNCS, vol. 7882, pp. 259–274 (2013)
4. Bruggemann, S., Bereta, K., Xiao, G., Koubarakis, M.: Ontology-Based Data Ac-
cess for Maritime Security, pp. 741–757. Springer International Publishing (2016)
5. Koubarakis, M., Karpathiotakis, M., Kyzirakos, K., Nikolaou, C., Sioutis, M.: Data
Models and Query Languages for Linked Geospatial Data. LNCS, vol. 7487, pp.
290–328. Springer (2012)
6. Kyzirakos, K., Karpathiotakis, M., Koubarakis, M.: Strabon: A Semantic Geospa-
tial DBMS. In: ISWC. LNCS, vol. 7649, pp. 295–311. Springer (2012)
7. Nikolaou, C., Dogani, K., Bereta, K., Garbis, G., Karpathiotakis, M., Kyzirakos,
K., Koubarakis, M.: Sextant: Visualizing time-evolving linked geospatial data. J.
Web Sem. 35, 35–52 (2015)
8. Priyatna, F., Corcho, O., Sequeda, J.: Formalisation and experiences of R2RML-
based SPARQL to SQL query translation using Morph. In: Proc. of the 23rd In-
ternational Conference on World Wide Web. pp. 479–490. ACM, NY, USA (2014)
9. Rodriguez-Muro, M., Rezk, M.: Efficient SPARQL-to-SQL with R2RML mappings.
Journal of Web Semantics (2015)
10. Sequeda, J., Miranker, D.P.: Ultrawrap: SPARQL execution on relational data.
Web Semantics: Science, Services and Agents on the World Wide Web 22(0) (2013)
11. Soylu, A., Giese, M., Jiménez-Ruiz, E., Vega-Gorgojo, G., Horrocks, I.: Experi-
encing OptiqueVQS: a multi-paradigm and ontology-based visual query system for
end users. Universal Access in the Information Society 15(1), 129–152 (2016)