Advertisement

Joining Data and Maps in the Government Enterprise Architecture by a Semantic Approach: Methodology, Ontology and Case Study

  • Daniela Giordano
  • Alfredo Torre
  • Carmelo Samperi
  • Salvatore Alessi
  • Alberto FaroEmail author
Conference paper
  • 832 Downloads
Part of the Lecture Notes in Business Information Processing book series (LNBIP, volume 190)

Abstract

The problem of managing data and maps within an ontological approach is little studied in the Government Enterprise Architecture. Aim of this paper is to present a methodology to solve this problem in case we would join municipal and cadastral data bases. In particular, we aim at linking the information contained in the local taxation registry to the urban territory to allow the Public Administration managers to check if the taxes have been paid, and the citizens to compute the correct amount to pay. The paper presents in detail the adopted ontology and the technological architecture, whereas a case study clarifies how the methodology works in practice. We plan to extend this methodology to manage e-gov services needing to interconnect data stores of city interest to vector layers derived from the Cadastre or other CAD systems.

Keywords

E-Government Government Enterprise Architecture Geographic information systems Ontology engineering 

1 Introduction

The Semantic Web is a mesh of information linked up in such a way to be processed easily by machines on a global scale [1]. The Semantic Web is built generally on syntaxes which use International Resource Identifiers (IRIs) to represent resources, i.e., subjects and objects, linked by properties. Subject-predicate-object relations are represented by triples, also called semantic web statements. Let us recall that IRI is an extension of the Uniform Resource Identifier (URI) that provides an encoding for Unicode character sets.

The semantic web statements are usually formalized by the Resource Description Framework (RDF), i.e., a directed multi-graph consisting of subjects, predicates and objects [2]. The RDF graph can be queried by means of the SPARQL query language to retrieve and manipulate the stored data [3]. The RDF Scheme (RDFS) is a collection of RDF resources that behaves as a vocabulary of terms and properties related to application-specific domain. Such vocabularies may range from controlled lists of terms to taxonomies and thesauri depending on the type of terms and relationships that can be expressed (e.g. parent-child relationships in a taxonomy).

Ontology refers to a formal specification of a shared vocabulary and allows us to define formally a set of terms, interconnections, constraints and rules of inference on a particular domain [4, 5]. A logical formalism is needed to represent an ontology such as Description Logic (DL) [6].

Ontology, with rule definition language and description logic, can also provide a new kind of data retrieving and mash up with the “backward chaining” concept to make possible the inference of data structures not present in the knowledge base at the moment of the query.

The use of an ontology with the intention of describing a particular aspect of reality, provides information reusable for all parties in the given domain. Regarding the e-government activities, the Linked Data group at the W3C and the Government Linked Data (GLD) are publishing data sets and knowledge bases (often in the form of light-weight ontologies and vocabularies) to support e-government services involving different organizations. These ontologies are under test and will be refined in the next future by incorporating novel global and local vocabularies.

The problem of managing data and maps within the mentioned ontological approach is little studied because it is necessary to study more complex problems that involve location based information, and because unifying terms of proprietary vocabularies such as road and street in view of a shared vocabulary implies only an equivalence between symbols, whereas equivalent drawings even if they are labeled by the same name, e.g. building, needs to be processed by complex conversion procedures when passing from the adopted Geographic Information System (GIS) to the one used by another organization to be sure that they deal with the same physical entity.

Therefore, in problems starting from personal and cadastral data based on maps, as well as for identifying the escape routes in case earthquake, we have to adopt not only a standard vocabulary that behaves as a bridge between equivalent terms used in the proprietary systems, but also conversion procedures to ensure that a physical vector in a GIS is the same in another one.

Of course, such problem would disappear if one adopts the same vocabulary and the same GIS in all the computing systems, however this is not only unrealistic but also not useful since proprietary codification of data and drawings may be more effective than the standard ones to carry out some basic operations such as storing and updating.

Aim of this paper is to present an ontology based methodology to solve this problem by illustrating how it works in practice by a case study dealing with the computation of local taxes from municipal and cadastral data. In particular, we discuss how linking the information contained in the local taxation registry to the urban territory by geographic points (Points Of Interest - POI) to allow the Public Administration (PA) to check if the taxes have been paid, and the citizens to determine the correct amount to pay.

At more general level, by this work, we also aim at supporting the transition of the PA information systems from their current structure, often consisting of separated silos of data, towards a Government Enterprise Architecture (GEA) following the principles of the Connected Government Model that enables “the governments to connect seamlessly across functions, agencies, and jurisdictions to deliver effective and efficient services to citizens and businesses” [7].

In particular we take into account the dimensions of the Connected Government Model that may be improved by adopting a better data organization, i.e., the first three dimensions of ones featuring an effective Government Enterprise Architecture:
  1. 1.

    Common infrastructure and interoperability,

     
  2. 2.

    Collaborative services and business operations,

     
  3. 3.

    Citizen centricity,

     
  4. 4.

    Social inclusion,

     
  5. 5.

    Networked organizational model,

     
  6. 6.

    Public sector governance.

     
Consequently, aspects such as public sector governance, networked organizational model and social inclusion mainly depending on the PA organization model are outside the scope of the paper and will discussed in further studies.
More specifically, with reference to the GEA model, the paper is focused on presenting a methodology to integrate data and processes residing at the lower layers of the GEA layered model consisting of the following four main layers:
  1. 1.

    the technological infrastructure where all the application processes and data are implemented.

     
  2. 2.

    the data layer where we have the data required to study and define the strategies of the general business processes,

     
  3. 3.

    the specific management processes dealing with a well defined application domain, and

     
  4. 4.

    the business processes and related outcomes to carry out the strategies of the organization enabling the services required by citizens and enterprises.

     
This will be obtained by using the semantic web technologies that facilitate better than the ones available on the market the data and processes integration over the web [8]. To avoid of illustrating the proposed methodology at a too theoretical level, in the paper we discuss a specific case, i.e., the computation of local taxes that requires integration of municipal and cadastral data stored at different PA offices following the requirements of a project, named K-Metropolis, supported by the Regional Government of Sicily, whose principal aim is collecting data originating from databases of different organizations to offer suitable e-services to citizens and enterprises [9]. However, the proposed methodology may be followed to manage other problems involving data and maps integration.

Let us note that another relevant part of the K-Metropolis project, called Wi-City, aiming at supporting mobile people activities following a semantic approach will not be discussed in the paper. The interested reader may find detailed information on the methodology adopted in Wi-City and its implementation structure in [10, 11, 12].

Section 2 illustrates the ontology and technologies that allow the taxation registry and the land registry to be interconnected in a single RDF framework.

Section 3 points out, by a case study, the main steps to convert the proprietary SQL codification of the original data bases into standard RDF statements, as well as the map conversion to allow the physical entities associated to the terms of the ontology to be represented by the same physical entity in almost all the available open source GISs.

Section 4 presents an example of the SPARQL queries that allow the citizens to extract the geo-referenced reports on how much they should pay to PA for their estates and support the PA employers to check if the citizens are in arrears.

2 Joining Municipal Data and Cadastral MAPS: Ontologies and Technologies

As pointed out in the introduction, the final purpose of this work is to develop a new kind of distributed system architecture capable of aggregating heterogeneous data from multiple data sources that have their own storage and representation format. In particular, the paper aims at integrating municipal and cadastral data bases by using a specific ontology and suitable technologies, as illustrated in the following sections.

2.1 K-Metropolis Ontology: KMET

To identify the relevant ontology of an e-government problem, the first step is the one of classifying the entities involved in the specific domain of interest. With reference to the mentioned local taxation problem discussed in detail in the case study, the main conceptual elements are: the taxpayer and her/his personal data; the property tax data referred to a specific period of time; and the waste fee tax data.

The element enabling the right connections among the various entities in the database is the taxpayer identification number subdivided in people and organization identification number. Therefore, the main taxation concepts are as follows:
  • County
    • City

  • Taxpayer
    • Citizen

    • Organization

  • Tax Return
    • real estate property tax

    • waste fee tax

Although the above classification has been used to manage the entities of a specific city, the entity City has been taken into account in our ontology as a concept, i.e., the class taxpayer consists of citizens and organizations of a city that is viewed as an autonomous entity rather one of citizens and organizations attributes.

For what concerns the Tax Return class, it is connected to the cadastral geographic entities to compute the local taxes, thus depending on the specific cadastral data organization adopted at the national level. Therefore, our ontology depends on how the Cadastre is organized in Italy and may differ for another country.

In Italy, the Cadastre consists of two main sections: the Cadastre of Land Properties and the one of Real Estates. The data deal with owners and holders of the estates or the lands whose relevant attributes are geographical location, size, intended use, earning capacity and consistency.

Since in the paper we are interested in the Real Estates Cadastre (REC), we have deepened only its structure and found that it is divided into sections and “pages”, each of which includes parcels associated to the basic entities, i.e., the Urban Real Estate Unit (UREU), defined as a portion of a building, an entire building or set of buildings that is capable of producing an independent income and has functional autonomy (access independence, self-sufficiency and autonomy in terms of use classification).

It should be emphasized that the UREU is no constrained to belong to a single owner. Consequently, a Real Estate Unit belonging to more owners will be reported with a single contextual registration and multiple identification. The identification of an UREU is made by the following identifiers:
  • Cadastral municipality, i.e., the municipality where the property is located;

  • Administrative Section, i.e., a portion of the municipality;

  • Page, i.e., a section of the municipality that is represented in the cartographic maps of the Cadastre Registry;

  • Parcel, i.e., a piece of land or building and any area of relevance within the Page;

  • Subordinate:, i.e., the actual element identifier of the UREU

Generally, each UREU is identified by its own subordinate, but, if the building is made up of a single UREU, then the subordinate may be missing. Considering the case of our interest and the above cadastral data structure, we attached the cadastral main entities (section, page, parcel, subordinate) to the fundamental geometrical ontology, as shown in Fig. 1.
Fig. 1.

Geographic and cadastral entities.

Also, two more identifiers can be found in addition to those above listed: Development and Attachment, denoted respectively in the national cadastre as “sviluppo” and “allegato”; the values of these identifiers indicate printings that represent on a larger scale some particular portion of the territory of a Page drawn on a smaller scale.

All the above concepts have been implemented using the Ontology Web Language (OWL) defined by the World Wide Web Consortium (W3C) in http://www.w3.org/2004/OWL/. Figure 2 shows this ontology named KMET since it has been adopted by the mentioned K-Metropolis project.
Fig. 2.

K-METropolis (KMET) ontology.

Such ontology is available at http://purl.org/net/kmet. On the left we have the entities related to taxpayer, whereas on right we have the ones related to tax returns and cadastral units.

2.2 Technologies

Figure 3 presents how the users may query the RDF triple store and the technologies involved in the deployment of the proposed distributed architecture. Each technology is a free and open source software to allow the Public Administration to reduce running expenses and maintenance costs.
Fig. 3.

Overview of the multi-tier system architecture and of the technologies adopted.

Also, this choice allows us to follow the cornerstone philosophy of the Open Data movement, whose description is given at http://opendefinition.org.

As shown on the right of Fig. 3, our model suggests that the data sources from the cadastral domain are converted from CXF format to shapefiles, i.e., ESRI SHP files; then they are imported into Quantum GIS to be exposed as vector layers provided with geo-referenced data.

These latter data are then imported to a PostGIS relational database to be available, together with the public administration data, to the end users through SQL queries (with automatic GeoJSON marshalling) via a RESTful Webservice.

Since our aim is to expose these data as RDF data structures through a SPARQL endpoint, in our model the relational databases at the center of Fig. 3, derived from the original municipal and cadastral databases drawn at the top of Fig. 3, are mapped in RDF triples using the D2RQ Platform (http://d2rq.org), that is an Open Source system that offers RDF based access to the contents of the relational databases.

A Jena/Fuseki framework contained into the domain translation, drawn at the bottom of Fig. 3, allows the end user to receive RDF formatted responses to the queries issued through external SPARQL endpoints. Such responses are also stored in a knowledge base to be reused to speed up the future queries.

Figure 4a shows the result of a query to visualize the UREUs of an area using a web GIS application as a layer superimposed to the OpenStreetMap raster layer. Such response has been obtained by using the GIS capabilities of our system, i.e., by using the right part of the model drawn in Fig. 3 devoted to manage the queries issued by the users to visualize a specific portion of the urban cadastral map. In particular, the queries of this type are directed to Quantum GIS to extract the shape files related to the specific cadastral map of user interest that will be superimposed to the raster background chosen by the user.
Fig. 4.

Responses to two typical user queries: a) cadastral map of user interest superimposed to a raster background (i.e., OpenStreetMap), and b) detailed administrative information about the specific graphical element selected by the user that is located at the center of the figure.

Figure 4b, shows how the public administration data mapped through the D2RQ Platform into RDF schemas may be visualized by another type of user query issued through the semantic layer i.e., the query managed by the mentioned multi-tier querying by RESTful Web Services, RDF marshalling and HTML/XHTML visualization. In particular, after received the shapefiles of her/his interest (e.g., urban blocks or districts) superimposed to the chosen raster background following the first type of query, the users may issue the other type of query to extract data coming from the different data stores to know detailed administrative information about a specific graphical element (e.g., a building or an apartment) located in the selected area, as shown in Fig.4b.

3 Case Study

In this section we point out the main problems that arise when one tries to implement the previous architecture to provide in practice a specific e-government service, i.e., the local estate taxation depending on both personal data of the owners and cadastral data of the real estate units.

3.1 Municipal Databases

The datasets about citizen and taxation were provided by the local administration in the form of IBM DB2 Databases. Therefore, before using the D2RQ Platform these databases were converted into MySQL databases to work on an open source format.

Then, after selecting the relevant data bases, e.g., the relational data base (RDB) at the top of Fig. 3, we have exported them as normalized tables into an additional MySQL database. Finally, the last step of the process was the one of mapping the contents of the databases into RDF triples through the on-the-fly translation obtained by the above mentioned D2RQ Platform.

Although the above procedure seems easy, several problems were encountered such as the large number of tables in the original databases and the lack of semantics in the table definition.

In particular, we have cut off the tables to a large extent since such RDBs were made up of 58 tables and 19,170,374 records containing many repetitions. Also, only a subset of the data that would be useful for the problem at hands were selected. In fact, since the taxation process has to be applied to citizens who own a property, only the table rows corresponding to estate owners were selected, whereas only the columns of these tables relevant for the taxation process were imported in the mentioned MySQL database, with the intent of mapping them in RDF triples to be used for supporting the taxation payment and checking using SPARQL queries.

To this aim, we have carried out a connection of the resulting municipal triple stores to the information stored in the cadastral database, using a multi-tier distributed system that gives the ability to expose cartographic representation (stored on a PostGIS database) and makes it responsive to the end user inputs treating the GIS data bases not as MySQL or PostGIS data bases, but as a RDF graph obtained using a custom D2R mapping system, such as G2R [13]. Thanks to these structured semantic connections we obtained a unique SPARQL endpoint where it is possible to attach many kinds of end-user application logic.

3.2 Cadastral Databases and GIS

Let us note that the cadastral datasets were provided by the provincial Land Administration as an extraction of the national cadastral map database from WEGIS, that is a licensed closed source powered by SOGEI (http://www.sogei.it) and used by the Italian Land Administration. In this extraction each Page is represented by a pair of ASCII files: a CXF file (Cadastral eXchange Format) containing all the graphical elements that compose the cadastral map, and a homonymous SUP file containing statistical data and parcel surfaces. Thus, a suitable conversion of the CXF and SUP files has been done to store the cadastral information within relational GISs that can be interconnected to the municipal data, as suggested in the previous section.

For this reason, the format of these two files was converted to ESRI shapefile, i.e., a geospatial vector data format for geographic information system software developed by ESRI using CXFToShape, i.e., a free CXF to ESRI shapefile converter. Then, the shapefile formatted files was imported into Quantum GIS, i.e., a cross-platform free and open source GIS application that provides capabilities of data visualization, editing, and analysis and may be viewed as a set of graphical elements linkable to the virtual RDF store obtained through the mentioned D2R Platform.

Let us note that in this way we may use the textual information contained in the Cadastre as RDF triples, but for using the related drawings in any GIS it is necessary to represent them into a reference system known by all the GISs available on the market. Thus, we have geo-referenced the shapefile imported into Quantum GIS by means of the Cassini-Soldner geo-coordinate system through the definition of a custom projection algorithm.

In this way the cadastral data may be processed as triples and the related vector drawings can be overlapped to any raster data layers (e.g., Google Street Maps satellite, Bing Map Aerial layers and cartographic regional data provided by the Province Bureau) or added to existing vector data layers (such as OpenStreetMap, Google Streets and Bing Road).

Figure 5 shows the almost perfect overlapping obtained in Quantum GIS of the Cassini-Soldner layer dealing with the coast area and of some buildings close to the sea derived from the Cadastre over the OpenStreetMap raster data layer.
Fig. 5.

Almost perfect overlapping of the Cassini-Soldner layer of a coast area (on the left), and of a resort area (on the right) to the OpenStreetMap raster layer.

This means that not only a SPARQL query allow us to join the municipal and cadastral data using the relevant shared fields, e.g., the fiscal codes of the owner or the UREU codes of the estates, but also that the vector layers of the cadastral entities involved in the problem at hands (e.g., buildings, roads, zones) may be represented in any GIS as entities colored depending on the map theme.

In this way, we may visualize on OpenStreetMaps in red the buildings registered in the cadastre data base that contain commercial activities that are in arrears if one is checking tax evasions, in yellow the roads that are not covered by regular waste collection and in orange the downsized schools if one is studying the quality of the services offered to the citizens. For example, from the fact that all the buildings in Fig. 6 are in green and some streets in yellow we can easily deduce that all the administrative taxes related to the buildings of the area under consideration were paid, but that some problems about the regular waste collection were signaled by the citizens.
Fig. 6.

Thematic map displaying tax payments and waste collection services for an urban area chosen by the user.

4 Querying Data and Maps

The semantic web gives to a developer two major choices when using distributed databases for querying in a semantic manner: the former is to copy the entire amount of data in a unique knowledge base, the latter is to perform a distributed query.

Each of them has two main disadvantages: latency and scale, but for the Public Administrations that have usually to manage large databases and big amount of data, copying an entire data store is hardly feasible.

Thus, in our approach we adopted the distributed model already shown in Fig. 3 to allow the users to query two or more RDF knowledge bases using a formula expressed in SPARQL, or to combine relational databases to be queried with a D2RQ mapping language that translates the SPARQL query to SQL on the fly.

As said before, the user interface allows the user to select a single Parcel showed on a multi layered map through a mouse click on a single polygon that represents a building as shown in Fig. 7. These polygons are recreated by processing the GeoJSON vector layers obtained from a SQL query against PostGIS database and produced by a PHP Web Service.
Fig. 7.

Building chosen by the end user using the interface of the web GIS application.

The mouse click is detected on the client-side via Leaflet JavaScript library (available at http://leafletjs.com/) and then the following SPARQL query is produced to know the payment situation of each apartment of the selected building:
  • prefix kmet:<http://purl.org/net/kmet#>

  • SELECT DISTINCT ?CadastralUnit ?Page ?Parcel

  • ?Subordinate ?LocalPropertyTaxReturn

  • WHERE {

  • ?CadastralUnit a kmet:CadastralUnit;

  • kmet:page ?Page;

  • kmet:parcel ?Parcel;

  • kmet:subordinate ?Subordinate.

  • FILTER(str(?Page)=‘0069’).

  • FILTER(str(?Parcel)=‘20317’).

  • ?LocalPropertyTaxReturn kmet:hasCadastralUnit ?CadastralUnit.

    }

The following table shows the results of the previous SPARQL query to be appended to the map shown previously.

Cadastral unit

Page

Parcel

Subordinate

LocalPropertyTaxReturn

<cadastralunit/16076>

“0069”

“20317”

“0029”

< WasteFeeTaxReturn /9213>

<cadastralunit/36713>

“0069”

“20317”

“0028”

<WasteFeeTaxReturn/29215>

<cadastralunit/36710>

“0069”

“20317”

“0019”

<WasteFeeTaxReturn/29310>

<cadastralunit/230222>

“0069”

“20317”

“0073”

<RealEstateTaxReturn/96445>

<cadastralunit/230227>

“0069”

“20317”

“0055”

<RealEstateTaxReturn/98014>

<cadastralunit/230228>

“0069”

“20317”

“0047”

<RealEstateTaxReturn/98019>

<cadastralunit/36720>

“0069”

“20317”

“0110”

<RealEstateTaxReturn/119214>

Let us note that the same result may be obtained by executing the above query in either the centralized or distributed scenario. A performance analysis to compare these two scenarios is outside the scope of the paper and will be discussed in future works, even if better performance may be achieved certainly if the queries may reuse previous results stored on a RDF knowledge base such as the one shown at the bottom of Fig. 3.

5 Conclusions

Related works to the subject of the paper deal mainly with cadastral system interconnection and introducing spatial dimensions to RDF schemas.

An ontology architecture for the land administration domain, targeted to achieve semantic interoperability between cadastral systems is proposed in [14]. The architecture complies with both Geospatial and Land Administration standards. An example of extension dealing with the specificities of their national cadastre is also provided. In [15], a Semantic Web approach is proposed to customize the applications of the Land Administration Domain. Also, an OWL layered architecture adaptable across jurisdictions is outlined. Both these works differ from our proposal in the conversion from relational data models to ontology and in the presence of an explicit ontology alignment step that has been illustrated in detail in Sect. 2.

The work of introducing spatial dimensions to the semantic web is described in [16] where it is demonstrated how crowd sourced geographical data transformed into RDF can be interlinked (mapped) with other (spatial data) sets to enable spatial data web applications.

A similar work, i.e., [13] focuses explicitly on some computational issues influencing the use of GIS from the semantic web standpoint, and proposes to treat GIS as virtual RDF graphs instead of re-implementing GIS functionalities in semantic web frameworks. This is achieved through an extension of the D2RQ mapping language to include spatial data types.

Therefore, the application scenario addressed in this work and clarified by many examples and a suitable case study, i.e., integration of cadastral systems with citizen data, has not been tackled before.

Let us note that the proposed ontology based methodology to manage data and maps using RDF schemas is being experimented in the project promoted by the Sicily Region, named K-metropolis, to support the municipal Governments not only in defining the local taxation policy but also in the transition from the current not interoperable GIS platforms to the implementation of a spatial data infrastructure to integrate data and maps of different municipalities.

For example, we are studying how the proposed methodology may support the municipal Governments to define city master plans and civil protection policies by interconnecting both public and private data stores to the drawings derived not only from the Cadastre but also from any relevant institutional CAD system.

This will allow the municipal Government not only to draw the intended land use and the emergency plans over any raster background, such as aerial photogrammetry or satellite images, but also to represent these plans by a standard graphical notation immediately understandable by any other involved organization.

References

  1. 1.
    Sheth, A.: Enterprise applications of semantic web. In: IFIP International Conference on Industrial Applications of Semantic Web (IASW2005), Jyväskylä, Finland (2005)Google Scholar
  2. 2.
    Hayes, P.: RDF Semantics. W3C Recommendation 10 (2004). http://www.w3.org/TR/rdf-mt
  3. 3.
    Prud’hommeaux, E., Seaborne, A.: SPARQL Query Language for RDF, W3C Rec. 15.1.2. (2008). http://www.w3.org/TR/rdf-sparql-query/
  4. 4.
    Zhai, J., Jiang, J., Yu, Y., Li, J.: Ontology-based integrated information platform for digital city. In: IEEE Proceedings of Wireless Communications, Networking and Mobile Computing, WiCOM ’08 (2008)Google Scholar
  5. 5.
    Faro A., Giordano D., Musarra A.: Ontology based mobility information systems. In: Proceedings of Systems, Men and Cybernetics Conference, SMC’03, vol. 3, pp. 4288–4293. IEEE (2003)Google Scholar
  6. 6.
    Baader, F., Calvanese, D., McGuinness, D.L., Nardi, D., Patel-Schneider, P.F.: The Description Logic Handbook: Theory, Implementation, Applications. Cambridge University Press, Cambridge (2003)Google Scholar
  7. 7.
    Saha, P.: Government Enterprise Architecture Research Project, NUS Systems Science Institute (2010). http://unpan1.un.org/intradoc/groups/public/documents/unpan/unpan039390.pdf
  8. 8.
    Langegger, A., Wöß, W., Blöchl, M.: A semantic web middleware for virtual data integration on the web. In: Bechhofer, S., Hauswirth, M., Hoffmann, J., Koubarakis, M. (eds.) ESWC 2008. LNCS, vol. 5021, pp. 493–507. Springer, Heidelberg (2008)Google Scholar
  9. 9.
    Costanzo, A., Faro, A., Giordano, D., Venticinque, M.: Wi-City: a federated architecture of metropolitan databases to support mobile users in real time. In: International Conference on Computer and Information Science, ICCIS, A Conference of World Engineering, Science and Technology Congress, ESTCON (2012)Google Scholar
  10. 10.
    Nebot, V., Berlanga, R.: Building data warehouses with semantic data. Decis. Support Syst. 52(4) (2012). http://sparql-wrapper.sourceforge.net/
  11. 11.
    Costanzo, A., Faro, A., Giordano, D.: WI-CITY: living, deciding and planning using mobiles in Intelligent Cities. In: 3rd International Conference on Pervasive and Embedded Computing and Communication Systems, PECCS, Barcelona, INSTICC (2013)Google Scholar
  12. 12.
    Faro, A., Giordano, D., Spampinato, C.: Integrating location tracking, traffic monitoring and semantics in a layered ITS architecture. Intell. Transport Syst. IET 5(3), 197–206 (2011)CrossRefGoogle Scholar
  13. 13.
    Della Valle E., Qasim, H.M., Celino, I.: Towards treating GIS as virtual RDF graphs. In: Proceedings of the 1st International Workshop on Pervasive Web Mapping, Geo-processing and Services, WebMGS (2010)Google Scholar
  14. 14.
    Boškovic, D., Ristic, A., Govedarica, M., Przulj, D.: Ontology development for land administration. In: IEEE 8th International Symposium on Intelligent Systems and Informatics, SISY, (2010)Google Scholar
  15. 15.
    Hay, G.C, Hall, G.B.: A semantic web approach to application configuration in the land administration domain. FIG Congress 2010 Facing the Challenges - Building the Capacity Sydney, (2010)Google Scholar
  16. 16.
    Auer, S., Lehmann, J., Hellmann, S.: LinkedGeoData: adding a spatial dimension to the web of data. In: Bernstein, A., Karger, D.R., Heath, T., Feigenbaum, L., Maynard, D., Motta, E., Thirunarayan, K. (eds.) ISWC 2009. LNCS, vol. 5823, pp. 731–746. Springer, Heidelberg (2009)CrossRefGoogle Scholar

Copyright information

© Springer International Publishing Switzerland 2014

Authors and Affiliations

  • Daniela Giordano
    • 1
  • Alfredo Torre
    • 1
  • Carmelo Samperi
    • 1
  • Salvatore Alessi
    • 1
  • Alberto Faro
    • 1
    Email author
  1. 1.Department of Electrical, Electronics and Computer EngineeringUniversity of CataniaCataniaItaly

Personalised recommendations