Publish and Enrich Geospatial Data as Linked Open Data
Claire Ponciano
1 a
, Markus Schaffert
, Falk W
and Jean-Jacques Ponciano
1 b
i3mainz – Institute for Spatial Information and Surveying Technology, Mainz University of Applied Sciences,
Lucy-Hillebrand-Straße 2, 55128 Mainz, Germany
Federal Agency for Cartography and Geodesy, Richard-Strauss-Allee 11, 60598 Frankfurt am Main, Germany
Semantic Interpretation, Linked Open Data, SPARQL, Ontology, Spatial Data, Semantic Web, Neural
Machine Translation.
The rapid growth of geospatial data (at least 20% every year) makes spatial data increasingly heterogeneous.
With the emergence of Semantic Web technologies, more and more approaches are trying to group these data
in knowledge graphs, allowing to link data together and to facilitate their sharing, use and maintenance. These
approaches face the problem of homogenisation of these data which are not unified in the structure of the data
on the one hand and on the other hand have a vocabulary that varies greatly depending on the application
domain for which the data are dedicated and the language in which they are described. In order to solve
this problem of homogenisation, we present in this paper the foundations of a framework allowing to group
efficiently heterogeneous spatial data in a knowledge base. This knowledge base is based on an ontology linked
to and DCAT-AP, and provides a data structure compatible with GeoSPARQL. This framework
allows the integration of geospatial data independently of their original language by translating them using
Neural Machine Translation.
Nowadays, we live in a world lead by the informa-
tion. In the geospatial domain, it exists a lot of various
sources (e.g. governmental data, crowd sourcing data,
Linked Open Data, etc). Several approaches (Guan
et al., 2012; Gannon et al., 2020; Simon and Fr
2007) have attempted to implement frameworks in or-
der to be able to gather and share geospatial data more
easily. With the emergence of Semantic Web tech-
nologies, new approaches based on knowledge graphs
(such as (Karam and Melchiori, 2013; Abbas and Ojo,
2013; Larkou et al., 2013)) have proposed to define
open data structures that can be linked together in or-
der to facilitate data integration and sharing. The ma-
jor issue of data integration is to allow to enrich data
with other data from other sources (such as Open-
StreetMap, DBpedia, and Wikidata). Each source has
its own way of structuring the information. Moreover,
this information is composed of vocabulary that can
be from different languages. The diversity of struc-
tures and vocabularies that can be used in geospatial
information have pushed approaches to specialize in
few sources to reduce the complexity of integration.
It is, therefore, necessary to homogenize the differ-
ent geospatial data structure and to homogenize the
different vocabularies in different languages. Such
a homogenization requires to define a data structure
and a vocabulary sufficiently flexible to take into ac-
count the evolution of data over time, language vari-
ations, and allowing to retrieve geospatial informa-
tion as quickly as possible. To meet these needs, in
this paper we propose the foundations of a framework
based on Semantic Web technologies, allowing us to
unify the different structures and vocabularies most
used in the geospatial domain in a single knowledge
base. This unification allows a universal sharing of
geospatial data and facilitates collaboration between
different states and organizations worldwide. The
goal of this project is to enable the Federal Agency
for Cartography and Geodesy in Germany to (i) bring
together the geospatial information contained in the
data held by the different Bundeslaender (data in dif-
ferent formats and structures) into a single platform
in RDF form, (ii) enrich this data and link it to the
Linked Open Data, and (iii) query and use this plat-
form to create the most informative maps possible. To
reach this goal, the main contributions made are the
Ponciano, C., Schaffert, M., Würriehausen, F. and Ponciano, J.
Publish and Enrich Geospatial Data as Linked Open Data.
DOI: 10.5220/0011550600003318
In Proceedings of the 18th International Conference on Web Information Systems and Technologies (WEBIST 2022), pages 314-319
ISBN: 978-989-758-613-2; ISSN: 2184-3252
2022 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
ontology defining a geospatial data structure that
can be universally used
method of homogenization of the different
geospatial vocabularies
integration of heterogeneous geospatial data into
linked data.
The article is organized as follows: Section 2
presents the state of the challenges related to the cre-
ation of such a framework. Section 3 presents the pro-
posed solutions and section 4 concludes with a discus-
sion and future work to be done.
Storing geospatial data for widespread use and shar-
ing is a challenging research problem, primarily be-
cause of the rapid growth of data, at least by 20% ev-
ery year (Lee and Kang, 2015).
In (Jung et al., 2013), the authors present an
ontology-enabled framework to find geographic ser-
vices. They aim to solve geospatial problems. The
presented ontology for GIS data has the benefit of re-
specting “OGC Abstract Specification Topic 5: Fea-
ture” that stipulates a geospatial feature should con-
tain a particular type of geometry with an SRS and
attributes. However, their ontology does not pro-
vide a common vocabulary to describe geographic
features and attributes. In (Sun et al., 2019), the
authors present the GeoDataOnt ontology that pro-
vides a general semantic basis for integrating and
sharing geospatial data. This ontology has the ben-
efit of allowing semantic issues to be resolved at a
coarse level. However, the descriptions of some fea-
tures of some geospatial data are not integrated or
not precise enough, which limits the resolution of the
corresponding feature semantic problems using this
ontology. In (Budak Arpinar et al., 2006), the au-
thors provide an interesting ontology for geospatial
semantic analytics but do not deal with geospatial at-
tributes integration and uniformization. In (Dsouza
et al., 2021), the authors present the knowledge graph
WorldKG, which aims at representing geographic en-
tities in OpenStreetMap semantically. This work
facilitates using OpenStreetMap, a rich source of
openly available geographic information, on the Se-
mantic Web and among Linked Open Data, thanks
to its links with Wikidata and DBpedia. However,
they do not deal with the integration of heterogeneous
spatial data. In (Karalis et al., 2019), the authors
present YAGO2geo, an extension of YAGO2 (knowl-
edge graph based on Wikipedia, WordNet, and GeoN-
ames). It aims at adding geospatial information rep-
resented by geometries encoded by Open Geospatial
Consortium standards. They also integrate other at-
tributes; however, they add properties specific to the
dataset into YAGO2geo and do not use a common
vocabulary. Although these research projects do not
group heterogeneous spatial data in a knowledge base
at an advanced level and using a common vocabulary,
projects such as (Jung et al., 2013; Budak Arpinar
et al., 2006; Dsouza et al., 2021) demonstrate that in-
telligent data provisioning using Semantic Web meth-
ods appears to be an effective solution for simplified
(from a technical point of view) linking of official and
unofficial data for geodata users. Moreover, projects
using ontology-based frameworks such as (Dsouza
et al., 2021; Sun et al., 2019; Karalis et al., 2019)
highlight the growing interest in structuring geospa-
tial data in the form of knowledge graphs. This struc-
turing method brings more flexibility than traditional
database structures, which is essential in view of the
fast evolving nature of data. Such graphs are stored
in triplestores (such as Apache Marmotta
, Apache
, Eclipse RDF4J
, Strabon
, Oracle Spatial and
, GraphDB
, Stardog
, and Virtuoso Univer-
sal Server
) allowing fast read and write access by
In the domain of geospatial, the GeoSPARQL
(Battle and Kolas, 2011) approach has greatly im-
proved access to geospatial data from a triplestore,
allowing to retrieve and update geospatial data with
simplified queries. However, GeoSPARQL requires
that the stored data be structured according to a par-
ticular form in order to be usable, but most of the
triplestores from the LOD have not added the struc-
ture proposed by GeoSPARQL. This constraint limits
the use of GeoSPARQL for the LOD.
The other main issue is the heterogeneity of the
geospatial data. Usually these data are stored in spe-
Apache Marmotta:, ac-
cessed on 2022-07-07
Apache Jena:
tdb/, accessed on 2022-07-07
Eclipse RDF4J:, accessed on 2022-
Strabon:, accessed on 2022-
Oracle Spatial and Graph:
atabase/technologies/spatialandgraph.html, accessed on
b/, accessed on 2022-07-07
Stardog:, accessed on 2022-
Virtuoso Universal Server: https://virtuoso.openlinks, accessed on 2022-07-07
Publish and Enrich Geospatial Data as Linked Open Data
cific formats (such as GeoJSON (Butler et al., 2016),
, OpenStreetMap (Ramm et al., 2014))
having their own structure and being in different lan-
guages. This variation in vocabulary poses a major
challenge for the integration of data from different
sources. It requires either the creation of queries spe-
cific to the original structure of the data and the lan-
guage used, thus limiting the sharing and exploitation
of these data, or transform the data after defining a
common data schema (such as INSPIRE) and convert
all other schema to the chosen one and do analyses
afterwards, which adds more steps to the processing.
We can deduce from this study that (i) the use of
knowledge base (in the form of ontology and triple-
store) is the most flexible and promising way to store
large amounts of geospatial data, (ii) that there is cur-
rently no known platform to efficiently group hetero-
geneous spatial data in a knowledge base with a com-
mon vocabulary for all integrated datasets, (iii) that
the major challenge for the integration of geospatial
data is the diversity of structures and vocabularies
used to organize and express this information.
In order to meet the challenge of storing vast amount
of geospatial data for universal sharing, we outline
a framework for integrating heterogeneous geospatial
data from different sources and languages in a struc-
tured and consistent manner. Figure 1 illustrates the
goal of this framework. As a position paper, we focus
on how to implement this framework using existing
research, rather than explaining all the mechanisms
that serialized it, which is part of our future work.
In section 3.1, we highlight solution for the inte-
gration of heterogeneous geospatial data into a Uni-
versal Spatial Knowledge Base (USKB). In Section
3.2 we highlight possibility to create common vocab-
ulary, on the basis of which universal queries can be
performed on data in RDF format. This common vo-
cabulary aims to bring together vocabularies used in
data from different sources such as Linked Open Data
and geospatial formats.
3.1 Automatic Integration of
Heterogeneous Data into a
Universal Spatial Knowledge Base
Among the various approaches developed to unify
geodata structure and vocabularies, the Ordnance Sur-
om/x-fmt/235, accessed on 2022-07-07
Figure 1: Goal of the proposed framework.
vey’s Linked Data Platform proposes an approach
using OWL (Antoniou and Harmelen, 2004) to link
GeoSPARQL concepts to equivalent concepts in its
ontology. GeoSPARQL has thus become a stan-
dard for representing and querying linked geospatial
data for the Open Geospatial Consortium (OGC) (van
Rees, 2013) Semantic Web.
Inspired by this approach, we highlight the rele-
vance of creating an ontology combining the concepts
of GeoSPARQL with the work (Quarati et al., 2021).
Moreover, the work (Stadler et al., 2012) allows for
describing metadata quality concepts, and link them
with the LinkedGeoData.
Such an ontology aims to bring together geo-
information from different sources by converting
common file formats in the geodomain and their
application schemas, such as GeoJSON, Shapefile,
OpenStreetMap, RDF (Pan, 2009). The work (Pa-
troumpas et al., 2014) proposed such conversion with-
out considering other ontologies. Thus, they are lim-
ited to the vocabulary used in the data file for creating
RDF concepts. It is therefore necessary to adapt the
RDF vocabulary and structure to those of the ontol-
ogy for which we intend to integrate them.
The RDF structure adaptation can be performed
with the help of an alignment of the ontology struc-
ture as proposed in (Li et al., 2008) using automatic
integration approaches developed in the works (Prud-
homme et al., 2020; Prudhomme et al., 2017).
However, such approaches required different
parametrisation according to the data source qual-
ity. This quality mainly depends of the source type.
Therefore an evaluation of data quality is needed to
guide the integration process.
As an exemple, such evaluation can be based on
the ”5-star Open Data”
principle as follows:
1. 1 star: Data available on the Web (in any format)
under an open license,
2. 2 stars: Data available as structured data (spread-
10, accessed on 2022-07-07
WEBIST 2022 - 18th International Conference on Web Information Systems and Technologies
sheet) in a proprietary format,
3. 3 stars: Data available in a non-proprietary open
format (such as CSV),
4. 4 stars: Data available linked to URIs that to de-
note things,
5. 5 stars: Data available and linked to other data to
provide context.
Automatic integration of data into an ontology
without human supervision can lead to an inconsis-
tent structure due to the accumulation of errors. Thus,
to ensure the consistency of the added data, each con-
cept is described using OWL2 (Consortium et al.,
2012) constraints. These constraints allow to verify
that the added data is consistent with the structure
imposed by the ontology. This consistency check is
performed using reasoner (such as HermiT (Glimm
et al., 2014), Owlgres (Stocker and Smith, 2008), Pel-
let (Sirin et al., 2007), DeLorea (Bobillo et al., 2012)).
3.2 Universal Vocabulary
The data vocabulary provided varies depending on
the source of the Linked Open Data (e.g. Wikidata
(van Veen, 2019), DBpedia
) and the data format
(e.g. GML, KML, GeoJSON). The establishment of
a common vocabulary allows for gathering geospatial
knowledge in one place and thus, facilitating the re-
trieval of geographic data. Access to the data knowl-
edge can then be done via a web interface.
To be universal, this language can be based on the
English lexicon provided by WordNet (Miller, 1995).
Related works such as (Frontini et al., 2016; Bond
and Bond, 2019) have succeed in linking GeoNames
Ontology and WordNet. Other approaches such as (Guha et al., 2016) and KBpedia
fully exploiting the power of linked open data to pro-
vide a varied and structured vocabulary in the En-
glish language. Recently DCAT- AP (Kirstein et al.,
2019) and GeoDCAT-AP
are planning to link their
knowledge to
, making a
promising base for the geospatial domain. Moreover, currently support complex geometries
and WKT literals, which would make the adoption of
GeoSPARQL straightforward.
Since the vocabulary comes from various lan-
guages, we propose to use a Neural Machine Trans-
DBpedia:, accessed on 2022-
12, accessed on 2022-07-07
accessed on 2022-07-07
14 19115 -
DCAT - mapping, accessed on 2022-07-07
lation approach to automatically translate each term
into English while being able to automatically detect
the original language. The works (Koehn, 2020) and
(Stahlberg, 2020) provide a review of different Neural
Machine Translation approaches. Recently, the work
of (Zhao et al., 2021) proposes to improve these ap-
proaches by combining them with a knowledge graph,
thus improving the accuracy of the translation.
The translated vocabulary is then aligned with the
knowledge base vocabulary using an ontology align-
ment approach such as (Zhang et al., 2014). Follow-
ing this alignment, the aligned terms between the vo-
cabulary used in the RDF files and the vocabulary of
the knowledge base is submitted to the user for ap-
proval. The user can then validate and complete the
vocabulary match in order to allow the integration of
the data to enrich the knowledge base. The alignment
of the terms is then stored to allow for better automa-
tion in the future.
Figure 2 summarizes this process of supervised
learning and alignment.
Figure 2: Overview of supervised learning and alignment
Let us take as an example a spatial data con-
taining the geometries and attributes of schools in
the Bundesland Rheinland-Pfalz in Germany in or-
der to illustrate the proposed solution. In this
dataset, a school has a geometry represented by a
point and the following list of attributes in Ger-
Offnungszeiten, FaxNummer, and Name. The
names of these attributes will be translated into En-
glish as: opening hours, fax number and name
respectively. An individual of type dcat:Dataset
will be created to represent the data. Then for
each school contained in the data, an individual
will be created. This individual will be of type
geo:Feature (from GeoSPARQL) and schema:School
(from Then, an equivalence to the at-
tributes will be matched with in order to
add the information related to this individual. In our
example, the values of the attributes
FaxNummer, and Name will be added respec-
tively with the properties schema:openingHours,
schema:faxNumber, schema:name thanks to the En-
glish translation done before. Finally, the individ-
ual will be linked to its geometry with the prop-
erty geo:hasGeometry. Its geometry will be of type
geo:Point and will have a value linked by the prop-
erty geo:asWKT.
Publish and Enrich Geospatial Data as Linked Open Data
In this position paper, we propose conceptual guide-
lines for the development of a framework to unify
the integration of geospatial data within a Universal
Spatial Knowledge Base (USKB) and to easily link
geospatial data to Linked Open Data. This frame-
work is composed of an ontology, whose concepts
are mainly based on standards (such as GeoSPARQL)
and well-known ontologies (as GeoNames). The var-
ious data sets contained in files (such as GeoJSON,
Shapefile, OpenStreetMap, INSPIRE) are integrated
into this ontology using structured mapping. The con-
sistency of the data integration is ensured by a rea-
soning applied on the constraints defined for each
concept, which ensures a continuous consistency of
the data. The data vocabulary is translated into En-
glish from its original language by a Neural Machine
Translation approach. The translated vocabulary is
mapped to the ontology vocabulary. The resulting
mapping is submitted to the user for verification, who
can modify or approve the proposed mapping. The
decisions taken are then saved as a mapping table to
allow continuous learning of the geospatial vocabu-
lary and thus improve the mapping. The future work
consists in enriching the basic concepts of the ontol-
ogy and the mapping of the vocabulary by supervised
learning in order to make the integration process as
automated as possible. Furthermore, the quality of
the data integration must be thoroughly evaluated by
experts. The exploitation of expert knowledge mod-
elled in an ontology can be considered in order to au-
tomate this evaluation, e.g. by generating correspon-
dence concepts to semantically map the different data
systems used to the respective INSPIRE systems.
The framework ontology and source code is available
at from the au-
This research is funded by the Federal Agency for
Cartography and Geodesy in Germany.
Abbas, S. and Ojo, A. (2013). Towards a linked geospa-
tial data infrastructure. In International Conference
on Electronic Government and the Information Sys-
tems Perspective, pages 196–210. Springer.
Antoniou, G. and Harmelen, F. v. (2004). Web ontology
language: Owl. In Handbook on ontologies, pages
67–92. Springer.
Battle, R. and Kolas, D. (2011). Geosparql: enabling
a geospatial semantic web. Semantic Web Journal,
Bobillo, F., Delgado, M., and G
omez-Romero, J. (2012).
Delorean: A reasoner for fuzzy owl 2. Expert Systems
with Applications, 39(1):258–272.
Bond, F. and Bond, A. (2019). Geonames wordnet (geown):
extracting wordnets from geonames. In Proceedings
of the 10th Global Wordnet Conference, pages 387–
Budak Arpinar, I., Sheth, A., Ramakrishnan, C.,
Lynn Usery, E., Azami, M., and Kwan, M.-P. (2006).
Geospatial ontology development and semantic ana-
lytics. Transactions in GIS, 10(4):551–575.
Butler, H., Daly, M., Doyle, A., Gillies, S., Hagen, S.,
Schaub, T., et al. (2016). The geojson format. Internet
Engineering Task Force (IETF).
Consortium, W. W. W. et al. (2012). Owl 2 web ontology
language document overview.
Dsouza, A., Tempelmeier, N., Yu, R., Gottschalk, S., and
Demidova, E. (2021). Worldkg: A world-scale ge-
ographic knowledge graph. In Proceedings of the
30th ACM International Conference on Information &
Knowledge Management, pages 4475–4484.
Frontini, F., Del Gratta, R., and Monachini, M. (2016).
Geodomainwordnet: Linking the geonames ontology
to wordnet. In Vetulani, Z., Uszkoreit, H., and Ku-
bis, M., editors, Human Language Technology. Chal-
lenges for Computer Science and Linguistics, pages
229–242, Cham. Springer International Publishing.
Gannon, B. M., Thompson, M. P., Deming, K. Z., Bayham,
J., Wei, Y., and O’Connor, C. D. (2020). A geospa-
tial framework to assess fireline effectiveness for large
wildfires in the western usa. Fire, 3(3):43.
Glimm, B., Horrocks, I., Motik, B., Stoilos, G., and Wang,
Z. (2014). Hermit: an owl 2 reasoner. Journal of
Automated Reasoning, 53(3):245–269.
Guan, W. W., Bol, P. K., Lewis, B. G., Bertrand,
M., Berman, M. L., and Blossom, J. C. (2012).
Worldmap–a geospatial framework for collaborative
research. Annals of GIS, 18(2):121–134.
Guha, R. V., Brickley, D., and Macbeth, S. (2016). Evolution of structured data on the web.
Commun. ACM, 59(2):44–51.
Jung, C.-T., Sun, C.-H., and Yuan, M. (2013). An ontology-
enabled framework for a geospatial problem-solving
environment. Computers, Environment and Urban
Systems, 38:45–57.
Karalis, N., Mandilaras, G., and Koubarakis, M. (2019).
Extending the yago2 knowledge graph with precise
geospatial knowledge. In International Semantic Web
Conference, pages 181–197. Springer.
Karam, R. and Melchiori, M. (2013). A crowdsourcing-
based framework for improving geo-spatial open data.
WEBIST 2022 - 18th International Conference on Web Information Systems and Technologies
In 2013 IEEE International Conference on Systems,
Man, and Cybernetics, pages 468–473. IEEE.
Kirstein, F., Dittwald, B., Dutkowski, S., Glikman, Y.,
Schimmler, S., and Hauswirth, M. (2019). Linked data
in the european data portal: A comprehensive plat-
form for applying dcat-ap. In Lindgren, I., Janssen,
M., Lee, H., Polini, A., Rodr
ıguez Bol
ıvar, M. P.,
Scholl, H. J., and Tambouris, E., editors, Electronic
Government, pages 192–204, Cham. Springer Inter-
national Publishing.
Koehn, P. (2020). Neural machine translation. Cambridge
University Press.
Larkou, G., Metochi, J., Chatzimilioudis, G., and
Zeinalipour-Yazti, D. (2013). Cloda: A crowdsourced
linked open data architecture. In 2013 IEEE 14th In-
ternational Conference on Mobile Data Management,
volume 2, pages 104–109. IEEE.
Lee, J.-G. and Kang, M. (2015). Geospatial big data: Chal-
lenges and opportunities. Big Data Research, 2(2):74–
81. Visions on Big Data.
Li, J., Tang, J., Li, Y., and Luo, Q. (2008). Rimom: A dy-
namic multistrategy ontology alignment framework.
IEEE Transactions on Knowledge and data Engineer-
ing, 21(8):1218–1232.
Miller, G. A. (1995). Wordnet: A lexical database for en-
glish. In Communications of the ACM, volume 38,
pages 39–41, Cham.
Pan, J. Z. (2009). Resource description framework. In
Handbook on ontologies, pages 71–90. Springer.
Patroumpas, K., Alexakis, M., Giannopoulos, G., and
Athanasiou, S. (2014). Triplegeo: an etl tool for trans-
forming geospatial data into rdf triples. In Edbt/Icdt
Workshops, pages 275–278. Citeseer.
Prudhomme, C., Homburg, T., Ponciano, J.-J., Boochs, F.,
Cruz, C., and Roxin, A.-M. (2020). Interpretation and
automatic integration of geospatial data into the se-
mantic web. Computing, 102(2):365–391.
Prudhomme, C., Homburg, T., Ponciano, J.-J., Boochs, F.,
Roxin, A., and Cruz, C. (2017). Automatic integra-
tion of spatial data into the semantic web. In WebIST,
pages 107–115.
Quarati, A., De Martino, M., and Rosim, S. (2021). Geospa-
tial open data usage and metadata quality. ISPRS In-
ternational Journal of Geo-Information, 10(1):30.
Ramm, F., Names, I., Files, S., Catalogue, F., Features, P.,
Features, N., and Cars, C. (2014). Openstreetmap data
in layered gis format. Version 0.6, 7.
Simon, R. and Fr
ohlich, P. (2007). A mobile application
framework for the geospatial web. In Proceedings of
the 16th international conference on World Wide Web,
pages 381–390.
Sirin, E., Parsia, B., Grau, B. C., Kalyanpur, A., and Katz,
Y. (2007). Pellet: A practical owl-dl reasoner. Journal
of Web Semantics, 5(2):51–53.
Stadler, C., Lehmann, J., H
offner, K., and Auer, S. (2012).
Linkedgeodata: A core for a web of spatial open data.
Semantic Web, 3(4):333–354.
Stahlberg, F. (2020). Neural machine translation: A review.
Journal of Artificial Intelligence Research, 69:343–
Stocker, M. and Smith, M. (2008). Owlgres: A scalable owl
reasoner. In OWLED, volume 432.
Sun, K., Zhu, Y., Pan, P., Hou, Z., Wang, D., Li, W., and
Song, J. (2019). Geospatial data ontology: the se-
mantic foundation of geospatial data integration and
sharing. Big Earth Data, 3(3):269–296.
van Rees, E. (2013). Open geospatial consortium (ogc).
Geoinformatics, 16(8):28.
van Veen, T. (2019). Wikidata. Information technology and
libraries, 38(2):72–81.
Zhang, Y., Wang, X., Lai, S., He, S., Liu, K., Zhao, J.,
and Lv, X. (2014). Ontology matching with word
embeddings. In Sun, M., Liu, Y., and Zhao, J., edi-
tors, Chinese Computational Linguistics and Natural
Language Processing Based on Naturally Annotated
Big Data, pages 34–45, Cham. Springer International
Zhao, Y., Zhang, J., Zhou, Y., and Zong, C. (2021). Knowl-
edge graphs enhanced neural machine translation. In
Proceedings of the Twenty-Ninth International Con-
ference on International Joint Conferences on Artifi-
cial Intelligence, pages 4039–4045.
Publish and Enrich Geospatial Data as Linked Open Data