ENGINEERING
TIME IN AN ONTOLOGY FOR POWER SYSTEMS
THROUGH THE ASSEMBLING OF MODULAR ONTOLOGIES
Jorge Santos, Lu
´
ıs Braga
Departamento de Engenharia Inform
´
atica, Instituto Superior de Engenharia, Porto, Portugal
Anthony G. Cohn
School of Computing, Leeds University, Leeds, U.K.
Keywords:
Power Systems, Ontologies, Knowledge Engineering, Temporal/Spatial Reasoning and Representation.
Abstract:
In this paper we investigate how timeless ontologies such as DFault, an ontology for fault diagnosis in power
transmission networks can be re-engineered to include temporal entities. We propose a methodology, FONTE
(Factorising ONTology Engineering complexity), that allows this complex process to be factored by dividing
the problem into parts: modelling the domain concepts ontology (atemporal and aspatial), modelling or ac-
quiring the temporal and /or spatial ontology, and finally producing the target ontology by assembling these
modular ontologies via a semi-automated process.
1 INTRODUCTION
In recent years many technologies have been de-
veloped allowing the capture and transmission
(e.g., smart sensors networks and wireless communi-
cations) of information about the state of a wide range
of devices connected to networks scattered over large
geographical areas (e.g., power transmission grids).
The emergence of technology standards lowered the
prices for these devices/technologies and eased its in-
tegration with existing systems. The result is that
the data owners are overloaded with large quantities
of heterogeneous information. One major challenge
therefore consists of developing applications able to
provide end-users with a system overview in a hu-
man readable, relevant and consistent manner, so as
to interpret and manage the system being monitored.
There are at least two issues obstructing this goal: the
information heterogeneity and the requirement for ap-
propriate models for spatial and temporal knowledge
(STK). The modelling of STK in intelligent systems
is a complex process. This work proposes a semi-
automatic method, FONTE (Factorising ONTology
Engineering complexity), based on the assembling of
multiple ontologies in order to obtain the target on-
tology. This method tackles both the above problem-
atic issues: information heterogeneity and modelling
STK. The first is addressed through the use of on-
tologies as building blocks, given that ontologies are
shared and common agreed models about a specific
domain. The second issue is addressed through the
use of a set of rules that drives the process of assem-
bling orthogonal categories, like space and time, in
a semi-automatic way, releasing the knowledge en-
gineer and the experts from the rendering of intri-
cate concepts related to complex theories of time and
space.
2 FONTE METHODOLOGY
In order to illustrate the assembly process two
ontologies will be used as building blocks
for the target ontology, a temporal ontol-
ogy and the timeless domain ontology DFault
(www.dei.isep.ipp.pt/jsantos/F2C/DFault) for fault
diagnosis in power transmission networks.
Modular Ontologies. The assembly process can
be used either for the development of ontologies
with time from scratch, or for re-engineering exist-
ing atemporal ontologies in order to include time. For
our case study we have used the time-less DFault on-
tology that captures the main concepts related to the
characterisation of the Portugese National Electricity
Transmission Grid (RNT) (www.ren.pt), and concepts
related to fault diagnosis in power transmission net-
works (previously used to develop the SPARSE sys-
255
Santos J., Braga L. and G. Cohn A. (2010).
ENGINEERING TIME IN AN ONTOLOGY FOR POWER SYSTEMS THROUGH THE ASSEMBLING OF MODULAR ONTOLOGIES.
In Proceedings of the 7th International Conference on Informatics in Control, Automation and Robotics, pages 255-258
DOI: 10.5220/0002952502550258
Copyright
c
SciTePress
tem (Vale et al., 2002)).
The DFault ontology (OWL version)includes 92
classes, 30 properties and 30 restrictions. The hi-
erarchy is divided in four sub hierarchies, with root
concepts: Event, Entity, Local and Object. The Event
class includes the different type of events that may oc-
cur in the electric network, such as breaker tripping or
power lines going out of service. The Entity class in-
cludes the organisations (e.g., suppliers, clients) and
persons involved in the management/exploration of
the network. The Object class is divided into phys-
ical objects, which include the electric network de-
vices and facilities, and abstract objects, which in-
clude the faults diagnosis and messages acquired from
the SCADA (Supervisory Control And Data Acquisi-
tion) system. The Local class represents a hierarchy of
physical locations, which are used to spatially charac-
terise the network devices and facilities.
Tripping
Breaker
Technician Device
hasBreaker
isA
operates
operatedBy
Figure 1: Excerpt of the DFault Ontology.
Fig. 1 presents an excerpt of the DFault ontol-
ogy that was used in order to elucidate the FONTE
method and the Protg plug-in that supports it. A cir-
cuit breaker is an electrical device designed to protect
a circuit from damage caused by an electrical event
such as overload or short circuit. Its basic function is
to detect a fault condition and, by interrupting conti-
nuity (tripping), to immediately discontinue electrical
flow. A circuit breaker can be reset (either manually
or automatically) to resume normal operation.
The temporal ontology used in our case study
(see Fig. 2 for the UML-like depiction of an
excerpt) embodies many concepts such as Instant
or Period found in ‘standard’ ontologies such as
OWL-Time (www.w3.org/TR/owltime/) or SUMO
(www.ontologyportal.org/) and assumes a standard
interpretation, mapping time points and intervals to
real numbers and intervals on the real line respec-
tively. A temporal representation requires the char-
acterisation of time itself and temporal incidence,
which are represented in our temporal ontology by
TemporalEntity and Eventuality, respectively. A fur-
ther notion, TimedThing, which is used during the as-
sembly process, bridges between temporal concepts
and domain concepts .
Temporal Entities. In the temporal ontology we
used for the case study there are two subclasses of
TemporalEntity: Instant and Period. The relations
before, after and equality can hold between Instants,
TimeRoot
TemporalEntity Eventuality TimedThing
Process
startedBy:Event
finishedBy:Event
duringAt:Period
Event
startes:Process
finishesBy:Process
atTime:Instant
Instant
starts:Interval
finishes:Interval
Period
startedBy:Instant
finishedBy:Instant
TimedConcept
TimedRelation
IntervalThing
PeriodThing
Figure 2: Excerpt of the Temporal Ontology.
respectively represented by the symbols:, Â, =, al-
lowing an algebra based on points to be defined(Vilain
et al., 1989). It is assumed that the before and after are
strict linear, i.e. irreflexive, asymmetric, transitive and
linear. The thirteen binary relations proposed in the
Allen’s interval algebra (Allen, 1983) can be defined
easily from , Â, and =. The starts and finishes are
relations from TemporalEntity to Instant. Also, there
are no null duration periods and each period is unique.
Processes and Events. There are two subclasses of
Eventuality, Process and Event, in order to be possible
to express continuous and instantaneous eventualities,
respectively. Event has a relation atTime to Instant
while Process has a relation duringAt to Period. The
relations starts and finishes can be used to state what
can start or finish a process.
The Assembly Method. FONTE (Santos and Staab,
2003) allows the targeted complex ontology to be
built by factorising concepts into their temporal, spa-
tial and domain (atemporal and aspatial) aspects, and
then assembling the temporally/spatially situated en-
tity from these primitive concepts. This is more sim-
ilar to a Cartesian Product than a union of ontolo-
gies. Each of these component ontologies can be
built/acquired independently, allowing a factorisation
of complexity. The ontologies assembly is performed
through an iterative and interactive process that com-
bines two types of inputs: i) human assembly ac-
tions between the component ontologies; and ii) auto-
matic assembly proposals obtained from semantic and
structural analysis of the ontologies. This process is
propelled by a set of rules and a set of constraints. The
set of rules drives a semi-automatic process proposing
assembly actions; the set of constraints allows to as-
sess which of the generated proposals are valid.
The proposed methodology divides the task of
building a temporalised ontology for power network
control and monitoring into the task of constructing
simpler sub-ontologies which can be built indepen-
dently, factorising the problem complexity. The do-
main concepts ontology DFault (DO) and the tempo-
ral ontology (TO) (both described in the previous sec-
tion) will be used to illustrate the assembly process.
ICINCO 2010 - 7th International Conference on Informatics in Control, Automation and Robotics
256
The methodology hence proposes to obtain the
target ontology (power network control and monitor-
ing with a temporal) through a semi-automatic pro-
cess of assembling. The concept DO.Device.Breaker
will be linked with TO.TemporalEntity, while the
property DO.Device.Breaker.triggering will be linked
with TO.Process. In this way the instances of
DO.Device.Breaker concept will have a time span
of use. Instances of DO.Device.Breaker.triggering
process have a time span in which they occur
and additionally the user will be asked to assem-
ble the events that define the beginning and end-
ing of this process (e.g., DO.Device.Breaker.opening
and DO.Device.Breaker.closing). After this step,
related concepts, properties and axioms are pro-
posed for assembling, through a cascade process:
e.g. sibling concepts of DO.Device.Breaker such
as DO.Device.Line or DO.Device.Transformer will be
proposed for the assembly process; properties and
axioms related to DO.Device.Breaker.triggering like
. . . .triggering.monophasic or . . . .triggering.triphasic
would be proposed for assembly depending on their
characteristic event sequence.
Protg Plug-in. In order to support the iterative and in-
teractive process used in FONTE, a Protg plug-in was
developed. An assembly task consists of the defini-
tion of the actions to be performed in the target on-
tology after the performance of an assembly action
(e.g., creation of a relation isA between a domain and
a temporal concept). Due to the characteristics of the
platform (Protg), two types of tasks were defined:
Internal Tasks, which allow basic operations to ma-
nipulate the ontologies to be performed (e.g., create,
delete and modify classes or properties), and provide
access to the API functionalities of the Protg platform
in a transparent mode;
External Tasks. (also called assembly rules), which
are procedures written in a pseudo-code language
that includes common program language instructions
(e.g., if, then, else) and special keywords (e.g., do,
propose, check) whose semantics has been previously
provided. In order to facilitate the edition/creation of
these tasks, a specific tool supported in a graphic in-
terface was developed (see below).
The FONTE plug-in architecture (see Fig. 3) has
different abstraction levels which present several ad-
vantages for the knowledge engineer: i) the knowl-
edge engineer does not need to know the specifics
of the Protg API to manipulate the ontologies. In
addition, the Internal Tasks provide an abstraction
level between External Tasks and Protg API assur-
ing independency between the External Tasks and the
Protg API version; ii) the External Tasks may be cre-
ated/edited during the execution time and do not re-
quire the alteration of the application and consequent
compilation; iii) Different rules set (stored in distinct
files) allow different temporal/spatial theories in the
assembly process to be used in a flexible way.
Internal Task
Task
isA
External Task
uses
isA
Protégé API
uses
FONTE plug-in
provides
Rule Editor
create/edit
Figure 3: FONTE plug-in architecture.
The plug-in provides a set of functionalities, such
as: linking concepts of the domain and tempo-
ral/spatial ontology; accepting, rejecting or even de-
laying the execution of a task; and visualising statis-
tics of the assembly process. As depicted in Fig. 4,
the plug-in has two panels for the manipulation of on-
tologies (on the left-hand side) and a list of propos-
als (on the right-hand side). The panel further to the
left contains the domain ontology (DFault, which is
timeless and spaceless); from this panel it is possible
to access the classes and properties hierarchies. The
other panel contains the temporal/spatial ontologies
to be used as construction blocks for the production
of the target ontology. The list of proposals contains
the records of the task instances generated by the sys-
tem. Details of this list are presented below.
To promote the assembly process the knowledge
engineer needs to select the ontologies that will par-
ticipate in the assembly process as well as the files
containing the assembly rules for each ontology;
these can be selected using the setup window (trig-
gered by the setup button shown in Fig. 4).
All the tasks that are successfully performed (ei-
ther triggered manually by user-driven action or auto-
matically by the structural analysis module) are added
to a list containing the instance tasks historic. Associ-
ated with each task instance proposal there is: a trig-
ger list (the elements that triggered the proposal); the
task weight (an indication of the importance of each
proposal, which influences the likelhood of proposal
acceptance during the assembly process); and a ques-
tion in natural language (a phrase that summarises
the proposal objective, instantiated with the elements
contained in the instance task).
As the assembly process progresses, more propos-
als are generated. If different concepts happen to pro-
pose the same task instance, all the elements that have
triggered that proposal are included in the trigger list
and the proposal weight is increased.
All the proposed task instances are stored in the
list of proposals, which can be sorted by different
criteria (e.g., id, trigger or weight). The user can
then accept, reject, or even delay for later analysis,
ENGINEERING TIME IN AN ONTOLOGY FOR POWER SYSTEMS THROUGH THE ASSEMBLING OF MODULAR
ONTOLOGIES
257
Figure 4: FONTE plug-in for Protg.
each of the proposals. In order to avoid overload-
ing the knowledge engineer with useless proposals,
rejected proposals are never automatically proposed
again(though they may be manually retrieved).
In addition to the functionalities described above,
the plug-in also provides statistics about the assembly
process (results of the tool performance, including the
initial and current status of the domain ontology, the
number of tasks that has been initiated by the user and
how many proposals have been accepted or rejected).
A facility for saving a sequence of performed tasks
as a script file is also provided to allow a knowledge
engineer to easily totally or partially repeat a certain
set of tasks.
An application was developed to facilitate the cre-
ation of external files of rules. This supports the
knowledge engineer through a simple and interactive
graphic interface. The file management system pro-
vides a graphic visualisation of the rules included in
each file and offers various functionalities, such as:
sorting the list through different criteria; modifying
the order in which the rules are interpreted during
the assembly process; visualising the rules in XML
or pseudo-code; removing, editing or creating new
assembly rules. The knowledge engineer is alerted
about potential consistency errors (e.g., non declared
variables) or warnings (e.g., to declaring a variable
that is not used).
3 CONCLUSIONS AND FUTURE
WORK
In this paper we discussed the engineering of time in
DFault, an ontology for fault diagnosis in power trans-
mission networks. We proposed FONTE, a method
that supports the engineering of complex ontologies
including temporal and/or spatial knowledge that al-
lows process complexity to be factored by dividing
the problem in parts: modelling the domain concepts
ontology (atemporal and aspatial), modelling or ac-
quiring the temporal and/or spatial ontology, and fi-
nally producing the target ontology by assembling
these modular ontologies through a semi-automated
rule based approach.
A Protg plug-in developed to support method
FONTE, which allows FONTE to be used in an in-
tegrated form in the development of ontologies, was
also described. The FONTE methodology works in-
dependently of the temporal/spatial theory since it
allows different sets of assembly rules to be used
for each specific theory. A tool to support the cre-
ation/edition of these rule sets was also summarised.
Future work includes: i) evaluating the generic
characteristics of the proposed method with differ-
ent spatial/temporal ontologies (including 4D ontolo-
gies); ii) developing a functionality to predict the im-
pact of the acceptance of a particular proposal; iii) im-
proving the generation of automatic proposals during
the assembly process; (ii) and (iii) may be achieved
through the use of semantic analysis, previously suc-
cessfully used in diverse processes of ontology engi-
neering (e.g., merging, mapping and alignment).
REFERENCES
Allen, J. (1983). Maintaining knowledge about temporal
intervals. Communication ACM, 26(11):832–843.
Santos, J. and Staab, S. (2003). FONTE - Factorizing ON-
Tology Engineering complexity. In The 2nd Int. Conf.
on Knowledge Capture (K-Cap’03), pages 146–153.
Vale, Z., Ramos, C., Faria, L., Malheiro, N., Marques,
A., and Rosado, C. (2002). Real-time inference for
knowledge-based applications in power system con-
trol centers. Journal on Systems Analysis Modelling
Simulation (SAMS), Taylor&Francis, 42:961–973.
Vilain, M., Kautz, H., and Beek, P. (1989). Constraint
propagation algorithms: a revised report. Readings
in Qualitative Reasoning about Physical Systems.
ICINCO 2010 - 7th International Conference on Informatics in Control, Automation and Robotics
258