=Paper=
{{Paper
|id=None
|storemode=property
|title=GetThere: A Rural Passenger Information System Utilising Linked Data & Citizen Sensing
|pdfUrl=https://ceur-ws.org/Vol-1035/iswc2013_demo_22.pdf
|volume=Vol-1035
|dblpUrl=https://dblp.org/rec/conf/semweb/CorsarEBMPN13
}}
==GetThere: A Rural Passenger Information System Utilising Linked Data & Citizen Sensing==
https://ceur-ws.org/Vol-1035/iswc2013_demo_22.pdf
GetThere: A Rural Passenger Information
System Utilising Linked Data & Citizen Sensing
David Corsar, Peter Edwards, Chris Baillie, Milan Markovic,
Konstantinos Papangelis, and John Nelson
dot.rural Digital Economy Hub,
University of Aberdeen, Aberdeen, UK
{dcorsar,p.edwards,c.baillie,m.markovic,k.papangelis,j.d.nelson}@abdn.
ac.uk
http://www.dotrural.ac.uk
Abstract. This demo paper describes a real-time passenger information
system based on citizen sensing and linked data.
Keywords: provenance, data quality, citizen sensing, linked data, se-
mantic infrastructure, transport
1 Introduction
Real-time passenger information (RTPI) systems provide details about public
transport, allowing passengers to plan and make decisions regarding their jour-
neys. Typical requirements for RTPI systems include: 1) listing available public
transport services; 2) providing timetable (schedule) information for those ser-
vices; 3) providing (real-time) vehicle locations; and 4) providing details of dis-
ruptions. However, few RTPI systems exist in rural areas for a variety of reasons,
including a lack of infrastructure for obtaining and providing real-time informa-
tion [7]. As part of the Informed Rural Passenger project1 , we are developing
GetThere, an RTPI system for rural areas. The GetThere system consists of a
smartphone app, supported by a semantic infrastructure that integrates data
from multiple sources (including users). This system has been deployed in the
Scottish Borders, UK in partnership with First Group.
This demonstration2 will show a typical use of the GetThere app to view
timetabled and real-time vehicle locations for a selected route, contribute vehi-
cle locations while making a journey, report a disruption event, and assess the
quality of real-time locations with and without the presence of disruption. The
demo will utilises the datasets and services shown in Fig. 1.
2 Information Ecosystem
GetThere is supported by a semantic information ecosystem (Fig. 1) itself under-
pinned by a series of ontologies. Semantic web and linked data technologies are
1
http://www.dotrural.ac.uk/irp
2
A video of the demo is available at http://www.gettherebus.com/iswcdemo/
used for data representation and storage within the ecosystem as they provide
an effective approach for large scale data integration [4]. Further, accessing and
storing data via SPARQL endpoints allows storage to be handled by technolo-
gies appropriate for the characteristics of individual datasets; for example, using
RDF streams or a database with a R2RML wrapper for high throughput data.
Clients Android Mobile Application
Web Services
Timetable User Sensor Location Observation Disruption Quality Provenance
Internal External
Datasets Timetable Infrastructure User Observation Disruption Datasets NaPTAN
Ontologies
Transport Travel
Transit Infrastructure LinkedGeoData User FOAF SIOC SSN Qual-O Prov-O
Sensors Disruption
Fig. 1. Real-time passenger information ecosystem.
Details of public transport services and timetables are stored in the Timetable
dataset3 and represented by the Transit ontology4 . This dataset is used by the
Timetable Service to provide details of available transport services, timetable,
and vehicle location information to the GetThere app. The Infrastructure dataset
provides details of the road networks used by public transport vehicles. This data
is extracted from openstreetmap.org, and is represented using the Infrastructure
ontology5 (which defines bus route maps) and the LinkedGeoData6 ontology.
NaPTAN7 provides details of bus stops, including their IDs and locations.
The User ontology8 and dataset describe user profiles using SIOC9 and
FOAF10 , a description of each user’s msobile device(s), along with details of
public transport journeys made while using the GetThere app. The Observation
dataset uses the Transport Sensors ontology11 which extends the W3C Seman-
tic Sensor Network (SSN) ontology12 to describe observations (e.g. of vehicle
occupancy level, vehicle location) obtained from users of the GetThere app. The
Sensor service provides an API for storing and retrieving sensor and observation
3
Timetable information is received in the ATCO-CIF format (http://www.
travelinedata.org.uk/CIF/atco-cif-spec.pdf); the RDF conversion program is
available at https://github.com/dcorsar/ecosystem.timetable.
4
http://vocab.org/transit/terms/
5
http://www.dotrural.ac.uk/irp/uploads/ontologies/infrastructure.owl
6
http://linkedgeodata.org
7
http://data.gov.uk/dataset/naptan
8
http://www.dotrural.ac.uk/irp/uploads/ontologies/user.owl
9
http://rdfs.org/sioc/spec/
10
http://xmlns.com/foaf/spec/
11
http://www.dotrural.ac.uk/irp/uploads/ontologies/sensors.owl
12
http://www.w3.org/2005/Incubator/ssn/ssnx/ssn
descriptions expressed using the SSN ontology; the Location Observation ser-
vice handles real-time locations provided by app users. The Travel Disruption
ontology describes different types of disruption, based on an analysis of existing
travel disruption information sources [5]. Disruption reports from app users are
managed and stored by the Disruption service and dataset.
Given the open nature of this data, issues such as data quality and trust
naturally arise [6]. Examples range from malicious users and inaccurate devices
to out-of-date information (e.g. timetables). As part of addressing these issues,
the ecosystem features a service that can evaluate data quality. The quality
ontology (Qual-O13 ), and its associated quality assessment service are discussed
in detail elsewhere [1]. Briefly, the service is configured with a set of quality
metrics encoded as SPARQL rules expressed against the relevant ontologies.
These guide a SPIN reasoner [3] to perform quality assessment, producing quality
scores which can be utilised by other services to filter low quality data.
Our current quality metrics are focused on real-time locations, and have been
developed following several deployments of the system. They include: Timeliness
- timely observations are under 1 minute old; Accuracy - accurate observations
have a GPS error margin of less than 25 metres; Relevance - relevant observations
are no further than 500 metres from the expected route of travel; Availability -
observations with a high availability score have no more than a 1 minute delay
between being created on the device and published by the ecosystem.
The provenance service uses the W3C Prov-O ontology14 to maintain a record
of the entities, agents, and activities involved in producing data within the
ecosystem. Uses of provenance include: associating users with location obser-
vations generated by their mobile device, which can support detection of po-
tentially malicious users; and recording dataset provenance to ensure the latest
timetable information is provided to users [2].
2.1 The GetThere Smartphone App
The ecosystem has been designed to support a range of applications through
the creation of relevant application services. At present we have used the ecosys-
tem to support the GetThere RTPI system, which is provided via an Android
smartphone app (see Fig. 2). The app invokes the web services, which execute
SPARQL queries against relevant datasets, process the results, and send a re-
sponse to the app. Users are presented with a list of available bus routes; after
selecting a route (and direction, either inbound or outbound), vehicle locations
are displayed. These locations include both estimates based on the timetable and
real-time locations obtained from other users on that route (Fig. 2, left screen-
shot). Bus stops along the route are also shown. The user can access timetable
information for the previous and next arrivals at a particular stop. When the
user boards the bus, they have the option of pressing a button to have their
13
http://sensornet.abdn.ac.uk/onts/Qual-O.ttl
14
http://www.w3.org/TR/prov-o/
Fig. 2. Screenshots of the GetThere app showing (left to right): vehicle locations; the
results of invoking the quality assessment service; and creating a disruption report.
location uploaded to the ecosystem every minute. The uploaded location is then
used as the vehicle’s real-time location provided to other users.
Users can view quality assessment results for a real-time vehicle location
by tapping its icon. We are working with users to determine an appropriate
visualisation of quality results. Currently each assessed dimension is shown with
a colour-coded bar representing its quality score (Fig. 2, centre screenshot).
Acknowledgements The research described here is supported by the award
made by the RCUK Digital Economy programme to the dot.rural Digital Econ-
omy Hub; award reference: EP/G066051/1
References
1. C. Baillie, E. Edwards, P. Pignotti, and D. Corsar. Short paper: Assessing the
quality of semantic sensor data. In Proc. of The 6th International Workshop on
Semantic Sensor Networks, page to appear, October 2013.
2. D. Corsar, P. Edwards, N. Velaga, J. Nelson, and J. Pan. Exploring provenance
in a linked data ecosystem. In P. Groth and J. Frew, editors, Provenance and
Annotation of Data and Processes, volume 7525 of LNCS, pages 226–228. Springer
Berlin Heidelberg, 2012.
3. C. Furber and M. Hepp. Swiqa - a semantic web information quality assessment
framework. In 19th European Conference on Information Systems, pages 922–933,
2011.
4. V. Lopez, S. Kotoulas, M. Sbodio, M. Stephenson, A. Gkoulalas-Divanis, and
P. Aonghusa. Queriocity: A linked data platform for urban information manage-
ment. In The Semantic Web – ISWC 2012, volume 7650 of LNCS, pages 148–163.
Springer Berlin Heidelberg, 2012.
5. M. Markovic, P. Edwards, D. Corsar, and J. Pan. Demo: Managing the provenance
of crowdsourced disruption reports. In Provenance and Annotation of Data and
Processes, volume 7525 of LNCS, pages 209–213. Springer Berlin Heidelberg, 2012.
6. S. D. Ramchurn, T. D. Huynh, and N. R. Jennings. Trust in multiagent systems.
The Knowledge Engineering Review, 19(1):1–25, 2004.
7. N. R. Velaga, M. Beecroft, J. D. Nelson, D. Corsar, and P. Edwards. Transport
poverty meets the digital divide: accessibility and connectivity in rural communities.
Journal of Transport Geography, 21(0):102 – 112, 2012.