Smart Cities combine diverse technologies to reduce their environmental impact and o er citizens a higher quality of life. In this paper we present an ongoing e ort, within the context of Ciudad2020 project, for overcoming the challenge of homogenizing the citizen's access to services o ered by heterogeneous, and independent entities within a Smart City scenario. We describe how we are applying the Linked Data Lifecycle, from speci cation to exploitation, within the vertical domains de ned in such Spanish project.
The number of Open Data portals is increasing, because of the demand of transparency and easy access to the data. In this context, there are also plenty of works related to
Smart Cities, though not all of them are related to Linked Data. In this section we present some of the approaches related to Smart Cities that follow the Linked Data paradigm.
Zaragoza Public Data Catalogue4 is an Open Data Portal that shows a Smart City as a city that allows mobility, knowledge and open access to the data. For this purpose, it includes twenty di erent datasets and mobile applications. It also provides a catalogue and a SPARQL End Point to the user.
Opendata Caceres5 is an Open Data Portal that o ers Linked Open Data datasets, allowing to the citizens and the companies access to the municipal data, facilitating reuse for developing applications.
United Kingdom Catalogue6 promotes Innovation as the key of and Smart City. This portal works with UK Public Sector information and data, encouraging the use and reuse of government datasets. It includes a directory of data avaliable, applications and a SPARQL EndPoint.
Liviu-Gabriel Cretu [4] de nes a Smart City as an event-oriented architecture, where digital devices allow interoperability between Internet of Services, Internet of Things and Internet of People. They explore the usability of the latest advances in SOA or Services-oriented Architecture and Semantic Technologies.
Lopez et al. [8] describe a Smart City as a complex system with heterogeneous data and present a Linked Data Platform for cataloguing, indexing and querying all the information.
Tallevi-Diotallevi et al. [9] aim to capture the pulse of the city of Dublin monitoring and decision-making with three aspects: Extending the SPARQL Language, processing heterogeneous data (streams and static) in real time and using a hybrid RDFS reasoner.
Balduini et al. [1] present a Streaming Linked Data framework to collect data streams, analyse and visualize the results using London Olympic Games and Milano Design Week as use-cases. This proposal is related to event analysis in the city, uses RDF for modelling and integrating data, SPARQL and sentiment analysis techniques for processing and analyse social data.
Table 1 summarizes the classi cation by domain, data and target audience of the Linked Data initiatives included in this survey.
The lack of Linked Data initiatives following a Multi-Domain, Multi-user and Multinature Data approach along with the needs of the Spanish Citizens and Public Administrations is what encouraged us to apply Linked Data Lifecycle within Ciudad2020 project along their vertical domains. 3
Generation and Exploitation of Linked Data within Ciudad2020 Project Ciudad2020 project7 is focussed on the three fundamental axes of a Smart City, which are Energy, Transport, Environment, and City. Within this project, the Linked Data Portal8 was created for integrating the data coming from Smart Cities in the four axes, as shown in Fig. 1. This portal has several datasets whose contents include bike sharing systems, restaurants, museums, energy performance certi cates and city tweets. For developers and for Public Administrations, it also provides a SPARQL endpoint and the possibility
7 http://www.innprontaciudad2020.es/
Reference Domain Used Data Target Users Zaragoza Public Mobility, Knowledge and Open Data Citizens and DevelData Catalogue4 Open Access opers Opendata Caceres5 Culture, Transport, Linked Open Data Citizens, Companies Environment, Society, and Developers
Healthy, Energy United Kingdom Environment, Govern- Open Data Citizens, Developers, Catalogue6 ment, Mapping, Society, and Administration Health, Education,
Bussiness and Justice
Cretu, L.[4] Event-driven Architec- Semantic Citizens
ture Smart Cities Web/Linked Data
Lopez et al. [8] Urban Monitoring Static and Streaming Citizens and
AdminData istration
Tallevi-Diotallevi et Transport, Environment Streams and Static Citizens
al. [9] and Energy
Balduini et al. [1] City-scale Events
of querying streaming data, showing results in real time. In the following sections we
are going to describe how we apply the Linked Data Life Cycle [7]; which consists of the
following activities (
Junta de castilla y Leon Catalogue10 is the Open Portal of the city of Leon. Zaragoza Public Data Catalogue11 is the Open Portal of the city of Saragossa.
El viajero [6] is a travel guide resulting from PRISA Group Data12. { Dynamic Data Sources. These data compress available bikes and slots in the di erent bike stations.
Citybike API 13 is the API for Bike Sharing Systems.
Data Conversion Services. We make di erent types of transformation depending on the nature of the data. In the case of Static Data Sources, we transform them by ETL processes (Extract, Transform and Load), generating data in RDF format.
On the other hand, regarding the Streaming Data Sources, we do not use an ETL transformation, since it would result in a hard and slow process. This Data Sources are shown as virtual RDF sources via streaming, by using the morph-streams technology, connecting the API REST of streaming services, web services and database producers based on complex events (CEP), establishing R2RML mappings [5].
Exploitation Use Case - Zaragoza Bizi. The Zaragoza Bizi Use Case14 combine static data and real time in the context of two cities: Saragossa and Leon. For instance, in Saragossa city there are 1300 bikes and more than 100 Km of cycle paths, serving citizens and tourists. This sums a total of 4,5 millions of uses and 13 millions of kilometres travelled in the last two years. At the same time, it avoided 2000 Tons of CO2 emissions.
Within this use case, citizens and tourists can check the number of available bikes as well as the number of free slots. They can also check in the map the location of the stations (the exact latitude and longitude are provided), nearby points of interest (including restaurants and museums) by choosing a determinate distance, routes between di erent stations, sharing the resource via Twitter, and send a suggestion for updating a resource. Finally, for developers, they can also access the traditional RDF information associated to a resource or consult de SPARQL endpoint15. 3.2
Speci cation We collect weather data streams from the weather API of OpenWeathermap16. It provides data from more than 40,000 weather stations. All weather data are obtained in JSON format.
Data Conversion Services In this case, the transformation process of the Streaming Data is the same we previously described in the Mobility and Transport section, we use the morph-streams technology [2] .
Exploitation Use Case - Weathermap Meteo The Weathermap Meteo Use Case17 shows the current weather and it is available for 200,000 cities.
This Use-Case provides a streaming querying service. The users can get the current weather data for any location on the Earth in real time by consulting the SPARQLStream endpoint. 10 http://www.datosabiertos.jcyl.es/ 11 http://www.zaragoza.es/ciudad/risp/ 12 http://www.prisa.com/es/ 13 http://api.citybik.es/ 14 http://transporte.linkeddata.es/browser.html 15 http://transporte.linkeddata.es/sparql.html 16 http://openweathermap.org/API 17 http://streams.linkeddata.es/register/weathermap 3.3
Speci cation. The Energy Portal18 combines Static Data Sources from di erent cities using Linked Data: { La Rioja city Energy E ciency Certi cates19: They provide an API to consult de database of certi cates of buildings and projects located in the city, providing the rating, the consum, the emission and the address of the buildings. { Navarra city Energy E ciency Certi cates20: They provide a web search service to obtain the Energy information associated to new projects, new buildings and existing buildings. Filters by ranking and type of building are available.
Data Conversion Services. In this particular case, we transform the Static Sources by ETL processes (Extract, Transform and Load), generating data in RDF format. Exploitation Use Case - Navarra Energy E ciency Certi cates In this use case21 we show the statistics of the di erent Spanish provinces, which publish these certi cates in Open Data portals.
We display a comparison of the provinces of La Rioja and Navarra, thereby the citizens can consult the energy performance certi cates and check the number of available certi cates with each rank, from A to G, where A means the most e cient buildings, and G means the least e cient buildings.
For instance, if the citizen choose the province of Navarra in the \Province" pulldown menu, the statistics of this city are shown. We can check that the most repeated rankings are D and E, with 1075 and 2368 buildings respectively. 3.4
Speci cation. The City Portal22 combine Static Data Sources and Streaming Data Sources, disambiguating and publishing those sources under the Linked Data paradigm.
Streaming Data Sources. Twitter API23. We collect the tweets geolocated in the city of Saragossa and published during the rst trimester of 2014.
Data Conversion Services We extract the named entities from Saragossa tweets. Once we get these named entities, we disambiguate them using the NERD tool 24. We associate a Dbpedia25 resource to each entity. We use the domain and semantic data ontology for disambiguating purposes. Finally, we generate RDF, we publish it and for the exploitation we use Flot Charts 26 to display the graphics associated to the statistics. Regarding the publication activity, note that we cannot publish the tweet text, since the API Terms of Twitter specify that we may only return tweet IDs and user IDs. 18 http://energia.linkeddata.es/ 19 http://www.larioja.org/npRioja/default/defaultpage.jsp?idtab=772883 20 https://administracionelectronica.navarra.es/webCertificacionesEnergeticas/
BuscarCertificado.aspx 21 http://energia.linkeddata.es/browser.html 22 http://ciudad.linkeddata.es/ 23 http://www.twitter.com/ 24 http://nerd.eurecom.fr/analysis 25 http://www.dbpedia.org/ 26 http://www.flotcharts.org/ Exploitation Use Case - Zaragoza Tweets This use case27 aims to show to the citizens a graphical visualization of statistics for the named entities in tweets geolocated in the city of Saragossa, during the rst quarter of 2014, disambiguated and published in Linked Data. In this period, we observe that some of the most repeated entities named in the city are Bands, Zaragoza city, Spain, Physical exercise, etc. Each of these entities has an associated DBPedia resource. DBpedia de nes an unique global identi er, including natural languages de nitions and relations to another resources. For instance, the Zaragoza entity has the associated resource http://dbpedia.org/resource/Zaragoza. 4
In this paper we have presented (
The implementation of the use cases using RDF and Linked Data Principles has
many bene ts: (
Regarding the future work we plan to (
Acknowledgments This work is partially supported by the Ciudad2020 INNPRONTA project (IPT-20111006). We would like to thank Ontology Engineering Group-UPM.