Maintaining SOA Systems of the Future
How Can Ontological Modeling Help?
Bilal Gonen
1
, Xingang Fang
1
, Eman El-Sheikh
1
, Sikha Bagui
1
, Norman Wilde
1
,
Alfred Zimmermann
2
and Ilia Petrov
2
1
Department of Computer Science, University of West Florida, 11000 University Parkway, Pensacola, FL 32514, U.S.A.
2
Faculty of Informatics, Reutlingen University, Alteburgstrasse 150, D-72762 Reutlingen, Germany
Keywords: Services Oriented Architecture (SOA), Ontology, Knowledge Modelling, Semantic Browsing, Software
Maintenance, Software Evolution.
Abstract: Many future Services Oriented Architecture (SOA) systems may be pervasive SmartLife applications that
provide real-time support for users in everyday tasks and situations. Development of such applications will
be challenging, but in this position paper we argue that their ongoing maintenance may be even more so.
Ontological modelling of the application may help to ease this burden, but maintainers need to understand a
system at many levels, from a broad architectural perspective down to the internals of deployed
components. Thus we will need consistent models that span the range of views, from business processes
through system architecture to maintainable code. We provide an initial example of such a modelling
approach and illustrate its application in a semantic browser to aid in software maintenance tasks.
1 INTRODUCTION
As computing resources have become pervasive,
with powerful networked computers in everything
from smart phones to smart cars, a new class of
information systems is emerging. Perhaps we can
best begin with an example.
Consider the driver of a future intelligent car.
A warning light appears on his dashboard as
he is travelling on an expressway. He selects
"investigate" and the car's maintenance
application looks up the error code and
identifies which engine component triggered
the warning. It then consults several online
databases that pull together information from
both the car's manufacturer and service
histories of other vehicles that use the same
component.
The application sends a summary of the results
to the driver's smart phone and recommends
an immediate replacement of the component. It
identifies several vendors and checks their
inventory. The driver can choose to schedule
service at a repair shop or to order the
component online and swap components
himself (Zimmermann, 2014).
Such systems, intended to support many
everyday situations and tasks, have sometimes been
called SmartLife applications. Their development
will involve advanced cloud infrastructures and
services, new tools and methods for software
development, and Big Data and open data
innovation. These are major current research
objectives of the National Science Foundation
(National Science Foundation, 2014) and the
European Union (European Commission, 2013).
In many cases the most reasonable structure for
such applications will use Services Oriented
Architecture (SOA) and thus the application will be
implemented by orchestrating loosely-coupled
services running on many nodes and communicating
via message passing. SOA applications often follow
the Web Services standards so that orchestration is
done with Business Process Execution Language
(BPEL), service interfaces are specified using Web
Services Description Language (WSDL) and
messages use XML described by an XML Schema
Definition (XSD).
Development of such applications will be
challenging, but as we look to the future their
ongoing maintenance may be even more so. For
some time researchers have pointed out the
difficulties of maintaining SOA (e.g. Gold, 2004;
Lewis, 2008). But at least in most cases earlier SOA
376
Gonen B., Fang X., El-Sheikh E., Bagui S., Wilde N., Zimmermann A. and Petrov I..
Maintaining SOA Systems of the Future - How Can Ontological Modeling Help?.
DOI: 10.5220/0005132903760381
In Proceedings of the International Conference on Knowledge Engineering and Ontology Development (KEOD-2014), pages 376-381
ISBN: 978-989-758-049-9
Copyright
c
2014 SCITEPRESS (Science and Technology Publications, Lda.)
systems have been implemented within the context
of some organizational framework, whether a
corporation or a government, that can provide the
necessary governance to manage change. However
these new SOA applications look to require
integration of services and data from multiple
partners with little organizational nexus.
The essential problem of software maintenance
has always been program comprehension: changes
made to imperfectly understood software can have
disastrous results. Maintainers are often not the same
as the original software developers so they require
time and effort to grasp the details that are relevant
for each change. In the case of these new
applications, maintainers may face events such as
unexpected partner-forced changes or the sudden
appearance of a security risk of unknown scope.
Maintainers will need to respond quickly and
accurately to such events, which will require rapid
analysis of both their own code and its place within
the wider application.
In this position paper we argue that maintenance
of future SOA systems will be a serious challenge.
While there will be no simple solution to this
challenge, ontological models may be able to ease
the maintainer's task. Such models must cover both
high level business and architectural views of the
whole application as well as a lower, code-focused
view, since ultimately most maintenance tasks
require changing code.
As an example, we describe how an existing
SOA ontology from the Open Group (2010) can be
extended to a SOA Evolution Ontology that better
meets the needs of a software maintainer. The Open
Group's ontology describes business processes,
services and their interfaces in a fairly abstract
manner. The maintainer needs that description, but
he also needs to deal with concrete implementation
details as may be found in design rationale, detailed
interface specifications and in code.
We show how the resulting ontology can support
semantic browsing to help a maintainer quickly
acquire the information he needs for a particular
maintenance task.
2 ARCHITECTURAL MODELING
AND THE ESARC CUBE
One methodology to help address the range of issues
facing SmartLife evolution is the ESARC Cube,
developed to support the assessment of the maturity
of SOA toolsets (Zimmermann, 2011). ESARC, the
Enterprise Services Architecture Reference Cube, is
an architecture reference model, which identifies an
integral view for the main interweaved architecture
domains such as: Architecture Governance,
Architecture Management, Business and Information
Architecture, Information Systems Architecture,
Technology Architecture, Operation Architecture,
and Cloud Services Architecture. ESARC provides a
coherent aid for the examination, comparison,
classification, quality evaluation and optimization of
architectures. ESARC abstracts from any specific
concrete business scenarios or technologies. The
Open Group Architecture Framework (Open Group,
2009) provides the basic blueprint and structure for
the extended service-oriented enterprise software
architecture domains, views and viewpoints.
This approach for architectural modelling
focuses on metamodels as abstractions for
architectural elements and relates them to
architecture ontologies (Zimmermann, 2013).
Metamodels define models of models and are used
within ESARC to define generic architecture model
elements and their relationships. Architecture
ontologies represent a common vocabulary that is
based on explicitly defined concepts for enterprise
architects who need to share their knowledge.
Ontologies include the ability to automatically infer
transitive knowledge. The metamodel of the
Business and Information Reference Architecture
consists of ESARC-specific concepts, which are
derived as specializations from generic concepts
such as Element and Composition from the
previously mentioned Open Group’s SOA Ontology
(Open Group, 2010). There are exemplary
metamodels and related ontologies for the following
main architecture domains of ESARC: Business and
Information Reference Architecture, Information
Systems Reference Architecture, and the
Technology Reference Architecture.
From the point of view of a software maintainer
the most relevant of these will be the Information
Systems Reference Architecture and the Technology
Reference Architecture. These describe the code and
deployment issues most likely to be relevant in
making a specific change. However the maintainer
must also be aware of the business processes,
business rules and the organizational concerns at the
higher levels. He thus must gather and use a wide
range of information, from his own organization,
from partner service provider organizations, and
from infrastructure vendors (database management
systems, enterprise service middleware, etc.). There
will be many opportunities for confusion and
misunderstanding as the maintainer tries to integrate
MaintainingSOASystemsoftheFuture-HowCanOntologicalModelingHelp?
377
this range of sources. Our hypothesis in this paper is
that ontologies could provide a significant aid to
software comprehension, provided they consistently
integrate information from different layers of the
ESARC Cube.
3 RELATED WORK
As well as the ESARC Cube background, the
research described in this paper draws upon earlier
work in three areas: SmartLife research, software
maintenance research on program comprehension,
and semantic web research related to the study of
software artefacts.
Applications and research focused on
SmartLife were introduced by Hitachi (2013),
SmartLife tracking and rescuing disaster
management (Nagashree, 2012), work-life
innovation (Mitchell, 2012), smart public
information system for public transport (Patinge,
2012), smart energy systems (B.A.U.M. Consult,
2012), which includes IT for smart grids,
supply/demand energy coordination, IT for smarter
buildings, security, green IT, and IT for novel
energy forms. Smart grids are advanced electricity
systems of networks, which enable a two-way
exchange of electricity power and information
between suppliers and consumers based on
intelligent communication, monitoring information
and management systems.
In the software maintenance literature most of
the research on understanding SOA applications has
focused on dynamic analysis approaches that start
from message logs or traces of execution (e.g. De
Pauw, 2005). However, some research has looked at
static analysis of the artefacts that describe services,
such as WSDL interface descriptions and XSD data
schemas (Coffey, 2012; Goehring, 2013).
Research has also been reported on applying
semantic web techniques for maintaining traditional
(non-SOA) software systems. This research focused
on providing ontological support for software
artefacts such as source code and documentation. In
work reported by Witte (2007), customized
ontologies were populated automatically from
source code and documentation, and then queried to
provide support for source code security analysis,
for traceability links between source code and
documentation and for architecture analysis. In work
by Hyland-Wood (2008), an ontology was
developed to describe the relationship between
object-oriented software components. However,
very little research has been reported on the
application of semantic techniques for maintenance
and evolution of SOA systems in particular.
4 EXTENDING THE OPEN
GROUP ONTOLOGY
The starting point for our SOA Evolution Ontology
is the Open Group SOA Ontology (Open Group,
2010). This was developed in order to facilitate
understanding of SOA applications and improve
communications between business and information
technology experts. The Open Group SOA Ontology
seemed an appropriate point of departure given its
earlier use with ESARC and the maintainer's need to
comprehend the system at multiple levels. The Open
Group SOA Ontology is defined in the web ontology
language (OWL) and is ready for extension and
population for specific applications. In the ontology,
15 classes and 30 object properties are defined. The
class hierarchy is shown in Figure 1.
Figure 1: The Open Group SOA Ontology class hierarchy.
To develop the SOA Evolution Ontology, the
Open Group SOA ontology was extended to
improve support for software maintainers while
preserving consistency with any existing higher
level models. For example, the Service class was
subclassed into InternalService and ExternalService.
The maintainer must approach very differently those
internal services whose code may be modified, since
it is owned by his organization, and those external
services which he can only invoke through an
interface. Since tracing data usage is a very common
task in software maintenance, a DataItem class was
added to record the fields in each message. As a
final example, a ProcessingModule class was added
to allow Service instances to be linked to code
artefacts.
KEOD2014-InternationalConferenceonKnowledgeEngineeringandOntologyDevelopment
378
As an example of the use of the extended
ontology we populated it to provide a description of
WebAutoParts, a hypothetical online automobile
parts dealer that has been used in previous research
studies (Wilde, 2012). WebAutoParts models an
Internet start-up company using a SOA strategy for
rapid development. A small amount of BPEL code is
used to orchestrate real commercial web services
from well-known vendors such as Amazon. These
are represented by their WSDL and XSD interface
descriptions. WebAutoParts might be typical not of
a whole SmartLife application, but of one of its
major components.
5 USING SEMANTIC SEARCH
TO SUPPORT MAINTENANCE
TASKS
Software maintainers inevitably spend a great deal
of their time searching for the detailed information
they need to do their jobs. Finding information and
understanding relationships among software
components and documentation is critical for
software engineers to make timely decisions during
the software maintenance process. As we have
argued, maintainers of future SOA applications will
probably confront an even greater diversity of
information sources as compared with those found
for conventional software.
Most search systems support information access
through keyword-based search that often returns
ambiguous results putting the burden on the user to
select and filter a large volume returned results. In
contrast, semantic search, which has been a focus of
the Semantic Web initiative (Semantic Web, 2014),
improves information retrieval on the web by giving
machines the ability to reason about web content to
better serve user needs.
As a first application of the SOA Evolution
Ontology we are applying semantic web ideas to
explore its use in a Semantic Browser (Gonen 2011).
This specialized browser would support navigating
the large repositories of textual, semi-structured
artefacts describing a SOA system. These artefacts
are annotated through semantic labels which support
discovery of the semantic relations between different
artefacts. Textual artefacts include natural language
design rationale, design and code documentation,
semi-formal service interface specifications (e.g.
WSDLs), BPEL orchestration code, etc.
To use our current Semantic Browser, still under
development, the classes of the SOA Evolution
Ontology are first populated for a specific
application such as WebAutoParts, thus creating a
set of related "individuals" (instances) for each class.
The original artefacts are then annotated so that,
each time the name of an individual appears, a
semantic link is added showing the relations to other
individuals. These semantic links allow users to
navigate the artefacts by following these named
relations.
As an example, consider a scenario of a software
maintainer trying to diagnose a problem with the US
Postal Service shipping costs computed by
WebAutoParts. He needs to discover where in the
system shipping costs are computed and particularly
what data is involved. In the software maintenance
literature this kind of problem has traditionally been
known as "concept location" or "concept
assignment" (Biggerstaff, 1993). Suppose the
maintainer has located one identifier as a starting
point "GetUSPSRate".
The software maintainer performs an initial
query on "GetUSPSRate," and is offered several
files containing that term as shown in Figure 2. One
such file is "OrderProcessing.bpel" and the
maintainer selects this file.
Figure 2: Search using Semantic Browser.
The Semantic Browser displays the file contents
(Figure 3). The named individuals, which exist in
the ontology, appear highlighted and underlined.
The software engineer clicks on the "GetUSPSRate"
in the text, and its relationships, such as "is a" and
"has interface" are displayed. Immediately the
maintainer would discover that "GetUSPSRate" is an
ExternalService and that fact will condition possible
maintenance fixes.
The software maintainer then selects the "has
interface" relationship from the list, and is offered a
list of interfaces, based on the ontology. He then
selects "USPS.GetUSPSRate.Interface" from the list,
and is offered all of the files that contain that term as
shown in Figure 3. He may click on any of those file
names to navigate to documentation that will help
explain how this service is invoked. So he is well on
MaintainingSOASystemsoftheFuture-HowCanOntologicalModelingHelp?
379
Figure 3: Sample search results using Semantic Browser.
his way to understanding what he may need to fix.
6 CONCLUSIONS AND FUTURE
WORK
In this paper we addressed the problem of the
evolution of future SOA applications. Software
maintainers of such SOA systems will confront great
challenges to keep them in continuous service in the
face of a rapidly changing environment, continually
emerging security risks, and a dynamic mix of
partner organizations. We argued that an ontology-
based approach could ease the difficulties of
maintenance and introduced a SOA Evolution
Ontology that can be populated with information
about any specific SOA-based system. This ontology
extends the Open Group SOA Ontology so it should
be compatible with other tools based on this
standard. As a first application of the SOA Evolution
Ontology we proposed a Semantic Browser to aid a
maintainer in navigating the many artefacts that
describe a SOA system. We illustrated the approach
by populating the SOA Evolution Ontology to model
WebAutoParts, a small SOA system which could be
typical of SmartLife components.
In the short run, future work will include
additional evaluation of the SOA Evolution
Ontology and the Semantic Browser both within our
academic environments and in cooperation with
industrial and other scientific partners.
The longer run vision for SmartLife maintenance
is that an "ecosystem" of ontologies will emerge to
describe these heterogeneous and complex software
systems (Zimmermann, 2014). Hopefully, consistent
modelling approaches can be found to bridge
architectural levels and address the different
concerns of business experts, developers and
maintainers. The task of supporting the evolution of
such systems will always be challenging, but such
models could greatly ease the burden on software
maintainers.
ACKNOWLEDGEMENTS
Work described in this paper was partially supported
by the University of West Florida Foundation under
the Nystul Eminent Scholar Endowment, and by the
SOA Innovation Lab, Germany.
REFERENCES
B.A.U.M. Consult, 2012. Smart Energy made in Germany:
Interim results of the E-Energy pilot projects towards
the Internet of Energy, http://www.e-energy.de/
documents/E-Energy_Interim_results_Feb_2012.pdf.
Biggerstaff, T. J., Mitbander, B. G., and Webster, D.,
1993. The concept assignment problem in program
understanding. In Proceedings of the 15th
International Conference on Software Engineering
(ICSE '93). IEEE Computer Society Press, Los
Alamitos, CA, USA, 482-498.
Coffey, J. W., Reichherzer, T., Owsnick-Klewe, B. and
Wilde, N., 2012. Automated Concept Map Generation
from Service-Oriented Architecture Artifacts, CMC
2012, Fifth International Conference on Concept
Mapping, Malta, Sept. 17-20, 2012.
De Pauw, W., Lei, M., Pring, E., Villard, L., Arnold, M.
Morar, J. F., 2005. Web Services Navigator:
Visualizing the execution of Web Services, IBM
Systems Journal, Volume 44, Number 4, Page 821,
DOI:10.1147/sj.444.0821.
KEOD2014-InternationalConferenceonKnowledgeEngineeringandOntologyDevelopment
380
European Commission, 2013. Horizon 2020 Work
Programme 2014-2015, 5i. Information and
Communication Technologies, http://ec.europa.eu/
research/participants/portal/doc/call/h2020/common/1
587758-05i._ict_wp_2014-2015_en.pdf.
Goehring, G., Reichherzer, T., El-Sheikh, E., Snider, D.,
Wilde, N., Bagui, S., Coffey, J., White, L. J., 2013. A
Knowledge-Based System Approach for Extracting
Abstractions from Service Oriented Architecture
Artifacts. (IJARAI) International Journal of Advanced
Research in Artificial Intelligence, Vol. 2, No.3, 2013,
pp. 44-52.
Gold, N., Bennett, K., 2004. Program Comprehension for
Web Services, 12th IEEE International Workshop on
Program Comprehension (IWPC'04) p. 151.
Gonen, B. 2011. Traversing Documents by Using
Semantic Relationships. The 2011 International
Conference on Semantic Web and Web Services, Las
Vegas, Nevada, USA, July 2011.
Hitachi, 2013. Hitachi’s Vision for Smart Cities,
http://www.hitachi.com/products/smartcity/download/
pdf/whitepaper.pdf.
Hyland-Wood, D., Carrington, D., Kaplan, S., 2008.
Towards a software maintenance methodology using
Semantic Web techniques and paradigmatic
documentation modelling, IET Software, 2008, 2(4),
pp. 337-347.
Lewis, G. A., Smith, D. B., 2008. Service-Oriented
Architecture and its implications for software
maintenance and evolution. Proceedings Frontiers of
Software Maintenance. IEEE Computer Society:
Washington, DC, pp 1-10.
Mitchel, S., Spencer, P., 2012. Work-Life innovation. The
Role of Networked Technologies, Cisco Internet
Business Solution Group, http://www.cisco.com/web
/about/ac79/docs/ps/WLI-and-Technology_020312
FINAL.pdf.
Nagashree, C., Kavya Rao, B., Jyothi Lobo, M., Harshitha,
B. S., Antony, P. J., 2012. Smart Life Tracking and
Rescuing Disaster Management System, International
Journal of Computer Applications (0975-8887),
Volume 45, No. 23, May 2012, pp. 10-17.
National Science Foundation, 2014. Big Data Research
Initiative, http://www.nsf.gov/cise/news/bigdata.jsp.
Open Group, 2009. TOGAF - The Open Group
Architecture Framework, Version-9, The Open Group.
Open Group, 2010. Service-Oriented Architecture
Ontology, Open Group, ISBN 1931624887,
https://www2.opengroup.org/ogsys/catalog/C104.
Patinge, P. D., Kolhare, N. R., 2012. Smart Onboard
Public Information System using GPS and GSM
Integration for Public Transport, International Journal
of Advanced Research in Computer and
Communication Engineering Vol. 1, Issue V, July
2012, pp. 308- 312.
Semantic Web, 2014. http://semanticweb.org/wiki/
Main_Page.
Wilde, N., Coffey, J., Reichherzer, T., White, L., 2012.
Open SOALab: Case Study Artifacts for SOA
Research and Education, Principles of Engineering
Service-Oriented Systems, PESOS 2012, Zurich,
Switzerland, pp. 59-60, June 4, 2012, doi:
10.1109/PESOS.2012.6225941.
Witte, R., Zhang, Y., Rilling, J., 2007. Empowering
Software Maintainers with Semantic Web
Technologies, Lecture Notes in Computer Science
2007, 4519, pp. 37-52.
Zimmermann, A., Pretz, M., Zimmermann, G., Firesmith,
D. G., Petrov, I., El-Sheikh, E., 2013. Towards
Service-oriented Enterprise Architectures for Big Data
Applications in the Cloud, 17th IEEE International
EDOC Conference (EDOCW 2013): The Enterprise
Computing Conference with SoEA4EE, 9-13
September 2013, Vancouver, BC, Canada, pp. 130-135.
Zimmermann, A., Zimmermann, G., 2011. ESARC -
Enterprise Services Architecture Reference Cube for
Capability Assessments of Service-oriented Systems,
SERVICE COMPUTATION 2011, pp. 63-68,
September 25 - 30, 2011, Rome, Italy.
Zimmermann, A., Gonen, B., Schmidt, R., El-Sheikh, E.,
Bagui, S., Wilde, N., 2014. Adaptable Enterprise
Architectures for Software Evolution of SmartLife
Ecosystems, to appear, SoEA4EE 2014, Ulm,
September 2014.
MaintainingSOASystemsoftheFuture-HowCanOntologicalModelingHelp?
381
