Testing geographical information systems: a case study in a fire prevention support system
| Published date | 10 August 2012 |
| Date | 10 August 2012 |
| Pages | 184-199 |
| DOI | https://doi.org/10.1108/13287261211255310 |
| Author | Mark Taylor,Emma Higgins,Paulo Lisboa |
| Subject Matter | Information & knowledge management |
Testing geographical information
systems: a case study in a fire
prevention support system
Mark Taylor, Emma Higgins and Paulo Lisboa
School of Computing and Mathematical Sciences,
Liverpool John Moores University, Liverpool, UK
Abstract
Purpose – The purpose of this paper is to describe the development and evaluation of a geographical
information system (GIS) testing framework that was used to test a fire prevention support GIS.
Design/methodology/approach – A year-long case study was undertaken concerning the testing
of a fire prevention support GIS in a UK fire and rescue service.
Findings – The GIS testing framework developed involved testing the different components of a GIS,
testing their interactions, and then testing the system as a whole. Since GISs contain different
components such as spatial analyses and map-based output, this supports the adoption of a different
testing framework compared to existing types of information systems.
Research limitations/implications – GISs will typically be used by organisations for decision
making. Clearly if the information presented by a GIS is inaccurate, unrepresentative, or unreliable,
then the decision-making process can be undermined.
Practical implications – This is particularly important with regard to GISs used by emergency
services (such as the fire and rescue service studied) where lives could potentially be put at risk by
erroneous information provided by such systems.
Originality/value – Previous research had indicated that GISs may be inadequately tested.
The framework developed for GISs testing provided a systematic testing approach, reducing the
likelihood of errors in such systems.
Keywords United Kingdom,Fire services, Information systems,Decision support systems, Testing,
Geographicalinformation systems
Paper type Research paper
Introduction
Geographical information systems (GISs) provide map based display of information at
a variety of map scales and resolutions as well as providing both data and spatial
query facilities (Johansson et al., 2010; Longley et al., 2005). GISs have been used in a
variety of organisations from utility companies to councils, emergency services to the
armed services. GISs can provide a valuable insight into the spatial and geographic
aspects of organisational activities. For example, GISs can assist in identifying crime
“hotspots” in a policing context, or areas of higher fire incidence in a fire and rescue
service context.
GISs are different to other existing forms of information systems in a number of
ways. GISs may use data sets from a variety of external sources, and such data sets
may not necessarily be easily integrated with each other. For example, data sets may
potentially contain data at a variety of different levels or spatial detail. In ad dition data
may not always be available for certain geographical areas, or may not be available at
the required level of detail. In some instances data sets may be incomplete and have
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1328-7265.htm
JSIT
14,3
184
Journal of Systems and Information
Technology
Vol. 14 No. 3, 2012
pp. 184-199
qEmerald Group Publishing Limited
1328-7265
DOI 10.1108/13287261211255310
missing values, which may hinder meaningful analysis. GISs typically involve spatial
analysis and the presentation of processed information on map based output at a
variety of levels of geographic detail. In addition GISs may allow users to generate
their own geographic analyses based upon a wide variety of spatial, statistical and
mathematical models. For these reasons, the development of GISs is different to the
development of other types of information systems (Giradin et al., 2010; Church, 2002).
The testing of GISs is different to the testing of other forms of information systems
due to the type of data held by GISs (spatial data such as geo-coded data, and
shapefiles used to geometrically describe points, polylines and polygons), the manner
in which the data is processed and analysed, and the manner in which the resultant
information is displayed.
GISs hold spatial data in addition to the other types of data held by existing types of
information systems, such as descriptive data concerning people and properties.
The spatial data may be obtained from a variety of different (and potentially external)
sources, and may be provided in different formats. This implies that testing the quality
and compatibility of such spatial data is an important step in the testing of a GIS. GISs
process and analyse spatial data in conjunction with the processing and analysis of
other types of data. The spatial processing and analysis may typically be performed by
built-in functions within a GISs development tool.
It is important to ensure that the manner in which such spatial processing and
analysis functions is used is appropriate, and that such functions provide correct
results. Spatial processing and analysis functions are unique to GISs, an d therefore this
distinguishes the testing of GISs from the testing of other types of existing information
systems. Spatial analysis involves the use of mathematical techniques to examine the
geographic properties of spatial data.
GISs may include user developed mathematical and statistical models. Although
some existing types of information systems such as those used in marketing and
financial services may utilise mathematical and statistical models, most existing form s
of information systems include limited mathematical or statistical modelling. GISs are
distinguished from other forms of information systems that utilise mathematical or
statistical models, since GISs incorporate mathematical or statistical models concerned
with spatial analysis (Chen et al., 2010; De Smith et al., 2007). In this paper we outline a
testing framework for the testing of GISs, and demonstrate and evaluate its application
with regard to the testing of a GIS for fire prevention support based upon a year long
case study in a UK fire and rescue service.
The importance of evaluating GISs
The testing process for software applications typically involves a number of stages
including unit testing, component testing, system testing, performance testing an d user
testing (Sommerville, 2010; Stikkel, 2006). Bryce and Colbourn (2006) commented
that software testing is an expensive and time-consuming activity that is often
restricted by limited project budgets. Kelly and Oshana (2000) advocated the use of
statistical testing techniques to improve unit and unit integration testing. Ali et al. (2007)
stressed the importance of integration testing with regard to the integration of classes
in object oriented systems. The testing of GISs has common features with the testing
of other types of information systems in that a structured testing approach involving
unit, component, system, volume and user acceptance testing would be required.
Geographical
information
systems
185
To continue reading
Request your trial