=Paper=
{{Paper
|id=Vol-1442/paper_24
|storemode=property
|title=The Multiple Applications of a Mature Domain Ontology
|pdfUrl=https://ceur-ws.org/Vol-1442/paper_24.pdf
|volume=Vol-1442
|dblpUrl=https://dblp.org/rec/conf/ontobras/AbelCFGR15
}}
==The Multiple Applications of a Mature Domain Ontology==
https://ceur-ws.org/Vol-1442/paper_24.pdf
The Multiple Applications of a Mature Domain Ontology
Mara Abel1, Joel Carbonera1, Sandro Fiorini1, Luan Garcia1, Luiz Fernando De
Ros2
1
Informatics Institute, 2Geoscience Institute
Universidade Federal do Rio Grande do Sul (UFRGS)
PO 15.064 – 91.501-970 – Porto Alegre – RS – Brazil
{marabel,jlcarbonera,srfiorini,lfgarcia,lfderos}@inf.ufrgs.br
Abstract. Ontologies have been growing in importance regarding their
reusability for distinct applications, since this allows amortizing the significant
cost of development of a knowledge base. Large portions of knowledge models
now are modelled as ontologies and these portions are shared through
several applications. Considering the immature stage of the methodologies
of Ontology Engineering and the considerable short space of time for
evolving fully operational domain ontology, few reports of real cases of
ontology reuse are found in the literature. This article describes a mature
domain ontology for Petrographic description and the several knowledge-
based applications that it supports. The ontology development started in the
90’s and it is still in evolution, both by extending vocabulary as by improving
the rigor of the conceptual modelling approaches. We analyze here the impact
that each new application has caused over the ontology, requiring
improvements and modifications in the original model.
1. Introduction
Building a fully operational domain ontology is a long time and resource-consuming
effort that can keep a team of professionals dedicated for years in refining and improving
the knowledge modelled. The team usually demands professionals of the domain along
with knowledge engineers and software analysts, whose combined profiles can cover the
requirements of expert knowledge, formal correctness, semantic richness and efficiency,
required for such knowledge-based applications.
This effort can be rewarded by the several uses that a heavy domain
ontology can support if its development has followed methodological approaches that
guarantee a high level of generality and modularity of the modelled ontology. Each
possibility of reuse brought by the development of a new knowledge-based system
in the same domain can amortize the cost of development and maintenance of the
domain ontology.
Much has been said about the advantages of building a well-founded domain
ontology regarding the potential software applications that can be supported by
ontologies. However, ontology engineering is still a recent area of research, and its
technological products are just starting to be delivered and evaluated.
Kop in [Kop 2011] discusses the limitations in adopting an existent domain
ontology as the basis for a new knowledge-based application. Different views over the
domain and ontological choices driven by diverse goals require significant adaptations
on the ontology, which are hard to be accomplished by knowledge engineers. The
author claims that the reuse can be assured by the involvement of the domain expert
in the ontology adaptation. Confirming the Kop claim, the adaptation of ontology to
support new applications was successful applied in the several Geology projects
described here, in this article. Still, the motivation for ontology reuse can go beyond the
reuse of an available formal piece of knowledge. Shah [Shah et al. 2014] has described
a framework to help the reuse of a biomedical ontology with the intention of helping the
integration of distinct specialties in Medicine thought a common knowledge-based
framework of software. Nevertheless, the cost or utility motivation for ontology reuse
and the possibilities of reducing the cost of knowledge-based applications by recycling
existent ontologies still face the problems of correctness of the ontology modelling
[Guarino & Welty 2002], quality of documentation [Simperl et al.2011] and the further
modifications of a shared ontology that can impact the maintenance of applications
[Tsalapati et al. 2009].
Our experience shows that, despite of the cost of developing fully operational
domain ontologies, the possibilities of reuse of the artifact outspreads the costs and effort
of the development.
In order to contribute to the understanding of the potential uses of domain
ontologies in knowledge-based applications, this article analyses the actual uses of a
mature domain ontology whose development started on 90’s and is being continuously
enhanced. We described the commercial and non-commercial software applications and
how each new application has affected the original definition of concepts and the
improvements that were done in order to keep compatibility and modularity among the
several supported software families.
2. The PetroGrapher project
The Petrography domain ontology was the main product of the PetroGrapher project
developed by the Intelligent Database Group of Federal University of Rio Grande do
Sul, Brazil, from 1995 to 2007 [Silva 1997; Abel 2001; Mastella 2004; Victoreti 2007].
The domain ontology was aimed to organize and represent the Geology vocabulary
required to support the quality evaluation of clastic and carbonate petroleum reservoirs
through petrographic analysis. An intelligent database application – Petroledge® system1
- was developed to support the petrographer through the task of reservoir description
and interpretation. The original ontology published in [Abel 2001] was a partonomy
of 21 geological terms (Figure 1), whose attributes and values added another 1500 terms
to the initial model. The terms were structured mainly through the part of relationship.
The more significant hierarchies refer to the mineral constituents: Detrital and
Diagenetic composition classes and subclasses (not detailed in Figure 1). The concept
Diagenetic composition, its attributes and domain of possible values are detailed in
Figure 2. The Figure illustrates the frame-based formalism adopted in the knowledge
representation and exemplifies the level of detail in which the ontology was formalized.
The knowledge representation formalism was chosen intended to facilitate the mapping of
the concept representation to a relational database model, since the database acts as the
repository of the domain ontology. The Figure 2, in particular, shows the attributes
Location and Paragenetic Relation, which express the spatial relationships that a
diagenetic mineral has with its neighborhood that can be visually recognized by the
geologist. The Diagenetic composition concept and attributes are essential for the
several interpretation tasks described in the Section 3.
1
Petroledge, Petroquery, Hardledge, RockViewer and Petrographypedia are trademarks of ENDEEPER
Company. The suite of ontology‐based applications described in this paper can be known in
www.endeeper.com/products.
Basically, petrographic evaluation refers to the formal description of visual
aspects of a rock sample, as they appear in naked-eye analysis and under an optical
microscopic. Starting from the petrographic features that are discerned, the
petrographer infers the possible geological interpretation(s) of the rock, which will
strongly influence the method of evaluation of the potential of the geological unit as an
oil reservoir. The geologist analyses the physicochemical conditions, called diagenetic
environment, in which the rock was possibly produced, according to the features that
would have been imprinted in the rock by the conditions of this environment.
Figure 1. Main concepts of the ontology of Petrography for petroleum reservoir.
The boxes describe the concepts and the arcs represent the part-of
relationship.
The greater challenge in building knowledge application in Geology is that the
explicit part of the knowledge that can be expressed through words is just a part of the
body of knowledge applied in interpretation. Most of the data relevant for geological
interpretation of oil reservoirs consist of visual information that have no formal
denomination and are learnt through an implicit process during training and field
experience. These features without names constitute the implicit body of knowledge,
also called tacit knowledge by Nonaka and Takeuchi [Nonaka et al. 1995] when
referring to the unarticulated knowledge that someone applies in daily tasks but is not
able to describe in words. The articulated or explicit knowledge that we call ontology
refers to the consciously recognized entities and how these entities are organized. Tacit
and explicit knowledge should be seen as two separate aspects of knowledge that
demands their own representational formalism and not different sorts of it.
Concept Diagenetic-Composition
Is-a Object
Part-of Concept Sample-Description
Mineral Name one-of [Diagenetic-Constituent]
one-of [Silica, Feldspar, Infiltrated clays,
Constituent Set Pseudomatrix clays, Authigenic clays, Zeolites, Carbonates,
Sulphates, Sulfides, Iron oxides/hydroxides, Titanium minerals,
Other diagenetic constituents]
Habit one-of [Habit-Name]
Amount range [0.0 - 100.00]
Nominal Amount one-of [abundant, common, rare, trace]
one-of [intergranular continous pore-lining, intergranular discontinous pore-
lining, intergranular pore-filling, intergranular discrete, intergranular
displacive, intragranular replacive, intragranular pore-lining, intragranular
Location pore-filling, intragranular discrete crystals, intragranular displacive, moldic
pore-lining, moldic pore-filling, oversized pore-lining, oversized pore-filling,
grain fracture-filling, grain fracture-lining, rock fracture-filling, rock fracture-
lining, concretions/nodules, massive beds/lenses]
Modifier one-of [dissolved, zoned, fractured, recrystallized]
one-of [Covering , Covering , Covered by ,
Replacing grain of , Replacing matrix of
, Replacing , Replaced by , Alternated
with , Engulfing , Engulfing , Engulfed by
, Intergrown with , Overgrowing ,
Paragenetic Overgrowing , Overgrown by , Expanding ,
Compacted from , Within intergranular
primary porosity, Within intergranular porosity after , Within intergranular porosity after detrital
matrix, Within intragranular porosity in ,
Within intracrystalline porosity in ,
Within moldic porosity after , Within
moldic porosity after , Within shrinkage
porosity of , Within shrinkage porosity of
, Within grain fracture porosity in , Within rock fracture porosity in ]
one-of [Silica, Feldspar, Infiltrated clays,
Pseudomatrix clays, Authigenic clays, Zeolites, Carbonates,
Paragenetic Sulphates, Sulfides, Iron oxides/hydroxides, Titanium minerals,
Relation Other diagenetic constituents, Detrital quartz, Detrital feldspar, Plutonic rock
Constituent Set fragments, Volcanic rock fragments, Sedimentary rock fragments,
Metamorphic rock fragments, Micas/chlorite, Heavy minerals, Intrabasinal
grains, Detrital matrix, Other detrital constituents]
Figure 2. A detail of the attributes and domain values of the Diagenetic
Composition concept represented in the ontology. The lists [Diagenetic-
Constituent] and [Habit- Name] describe the specialized vocabulary that
describes mineral names and formats of minerals modelled in a separated way
for a question of modularity and reusability.
The Petroledge application was conceived in order to allow a user with a medium
level of expertise to describe petrographic features in his/her own level of technical
language. The system has the role of applying knowledge to recognize, within those
ontologically described features, the items that can serve as diagnostic cues for higher
levels of expertise in interpretation, in some imitation of a process of visual
interpretation (but with even images being described symbolically). In order to achieve
that, the knowledge model represents the connection between the features described
using ontological vocabulary and those no-named features utilized by the experts to
support interpretation. In other words, the model explicity represents the way in which
the expert would see the same features seen and described by the user with support of
ontology. The knowledge acquisition process and the way in which the knowledge was
modeled and implemented in Petroledge system are described in [Abel et al. 1998].
3. Ontology-based applications
The long-term effort of building a detailed domain ontology in Petrography had the aim
of developing a software application to support the highly specialize task of quality
evaluation of petroleum reservoir. Petroledge features include an optimized support for
the petrographic description of clastic and carbonate reservoirs and other sedimentary
rocks. The system guides sample description, according to a systematic order, allowing
the standardization and easy access to petrographic terminology for all aspects of
description. The user will produce a structured description of the rock under analysis
according to the knowledge model. The knowledge base is composed by the ontology
and a set of distinct representational formalisms that describe the scheme of a description
and the inferential knowledge applied by problem-solving methods [Gómez-Pérez &
Benjamins 1999]. Each description is stored as a set of tuples of concept-attribute-value
or any logical combination of concept-attribute-value. Records within a relational
database are further processed by several problem-solving methods, each one intended
to extract geological interpretation, such as, rock provenance, diagenetic environment of
rock formation, original rock composition before diageneses, and others. This simple
structure (frames + inferential relationships) is the base for supporting multiple
applications.
The more powerful inferential formalism applied by Petroledge is the knowledge
graph, which plays the role of a rule type (in the sense defined in [Schreiber et al. 2000])
in defining the inference paths of the problem-solving process. They were built as an
AND/OR tree, where the root represents the interpretation and the leaves are instances
of no-named visual features. By its side, each no-named feature is associated to a set of
terms of the ontology that better describe the visual aspect of that evidence (Figure 3).
This aggregate structure of knowledge and its cognitive significance was firstly defined
as a visual chunk by [Abel 2001]. The k-graph as a whole represents how much each
feature influences the choice of some particular interpretation as a solution for the
interpretation problem. It also provides a connection between the expert-level knowledge
and the shared ontology applied by the professionals on communication and daily tasks.
A weight assigned to each feature assets the relevance of that feature to a particular
geological interpretation. Twelve k-graphs represent the knowledge required by
Petroledge system to automatically interpret the six possible diagenetic environments for
clastic reservoirs. The reasoning mechanism of the Petroledge system exams the
description of the user in the database searching for described features that match to each
knowledge graph. When the weights of features are enough to support that interpretation,
the diagenetic environment and the founded features are shown to the user.
Figure 3. The knowledge graph describes the evidences that support geological
interpretation and also links the expert level features to the set of terms in the
ontology that describes the content of the evidence.
Figure 4 shows the visual chunk in the petrographic application that describes
Diagenetic Dissolution and its internal representation as it is manipulated by the system.
Figure 4. A representation of a visual chunk that describes the inference
for Diagenetic Dissolution interpretation and its internal representation as
manipulated by Petroledge.
Several other methods of reasoning were developed and applied over the
Petrography ontology-based model. Each method requires its own inferential
knowledge model and is called or not by the system in an independent way.
Compositional classification and provenance interpretation apply numerical methods
based on the proportion of minerals. Inferential rules can deconstruct the diagenesis and
retrieve the original composition of sediments. Textural classification is based on the
proportion of the size of grains. Geological rules can infer the proportion of intrabasinal
and extrabasinal sediments.
A further expansion of the ontology model has allowed the modelling of
diagenetic sequences, enabling new inference methods to extract the sequence of
physicochemical events that has generated a reservoir rock from the spatial relations
among mineral constituents [Mastella et al. 2007]. In order to support that, new concepts
describing events and temporal relations were included in the model and their instances
were defined. In addition, the paragenetic relations (showed in Figure 2) that describe
mineral constituent associations had their spatial attributes detailed. A set of inference
rules describes the relation between the mineral association and the event that has
happened with the rock. A reasoning method reads the features described by the user
and stored in the database and orders the events that have happened with the rock since
the deposition of sediment and later consolidation of the rock. Figure 5 shows the
graphical representation of the inference rule that allows ordering the generation of the
mineral dolomite as being happened before the generation of mineral anidrite.
Figure 5. The model of inference rules for extracting sequence of events from the
Petrography ontology model.
The flexibility of the ontology model allows each method being based on different
inferential knowledge models that are applied by independent modules of software,
according to the needs of a particular use of rock data.
Besides the several inference methods that were associated to the Petrography
knowledge model, several other applications had been developed getting advantage of
the strong and complete formalized vocabulary, even without being part of the
Petroledge suite of software.
The Petroquery® application implements a query system over the rock description
based on the ontology. Getting advantage of the vocabulary, the application offers to the
user his/her own vocabulary for consultation restricting the option of words that are
actually present in the database. The user builds SQL consultations by selecting the
controlled vocabulary and retrieving the rock descriptions that includes the query
arguments. With this support, the geologist can build domain specific consultations
like “Retrieve all rock samples that has dolomite replacing feldspar grains and anidrite
within intergranular porosity”.
The controlled vocabulary of the domain ontology was also applied for labeling
and indexing microscopic images of rocks in the RockViewer® system, developed in
2010. An editor allows the geologist to associate ontology-controlled text describing
images of the rock. After the images being labelled, usually for an experienced
petrographer, they are shared through a distributed database to be consulted. The
system is used in corporate environment for geologist consultation of the many
aspects of rocks that affect the quality of a petroleum reservoir. Figure 6 shows the
interaction with RockViewer®. The terms of ontology describing the content of image
and used for consultation are highlight in the image label.
Figure 6. Domain ontology allows to indexing and recovering image content.
The original domain ontology covers the domain of rock-reservoir description.
The knowledge schema models the structure of a reservoir description, while the mineral
names and characteristics and textural aspects, that constitute the bulk part of ontology,
were captured from the more general vocabulary of the Geology community, which
supports several other Geology interpretation tasks. Based on this assumption, the
ontology of Petroledge was extended to cover all types of rocks and a related
knowledge-based application – Hardledge® system - was developed to support mining
rock interpretation problems. This 2010’s developed ontology was already extended, in
2013, to support the interpretation of magma placement history in sedimentary basins
affected by tectonic events.
Other classes of software application can benefit by the reuse of available domain
ontology. The web-based application PetrographypediA [Castro 2012] applies the
ontology of minerals and their characteristics on microscope to build a visual all-type-
of-rock atlas on-line to be freely consulted by the Geology community. As for
RockViewer® application, the ontology of Petroledge and Hardledge® was used to label
and index rock pictures taken in optical microscope.
A remarkable application of the Petrography domain ontology in the last year is
related to the development of conceptual solutions to provide interoperability between
reservoir modelling applications along with petroleum chain. The ontology is being used
to make explicit the meaning of the geological concepts embedded in the software code
and models in order to allow these objects to be recognized and applied to anchor the
models of distinct suppliers [Abel et al. 2015b]. This initiative is being conducted by the
Energistics2 consortium in the definition of RESQML interchange standard [King et
al. 2012]. Also, the PPDM association is applying the well-founded ontology for
anchoring the concepts of data models and providing better support for data mapping
among different application models [Abel et al. 2015a].
2
ENERGISTICS is a global consortium that facilitates the development, management and adoption of
data exchange standards for petroleum industry. RESQML is the data exchange standard for reservoir data.
www.energistics.org.
4. The Petroledge Ontology Evolution
The knowledge model of Petrography was initially defined using a frame-based
formalism whose general aspect was showed in Figure 2. Two requirements oriented the
modelling definition: the understanding of the expert about the information required to
produce a qualified rock description and the data management requirements for storage
and retrieving a large number of descriptions in a corporate environment. The knowledge
acquisition was strongly based on the collection of cases of previous descriptions. As a
result, the original model was a flat representation of a rock description instead of
focusing in the rigid geological concepts and the hierarchy that structure the world in the
geologist mind.
The inadequacy of the original model was soon evidenced as much as the
reasoning method for diagenetic environment interpretation was developed. To cope
with the reasoning, the model was separated in three parts: the knowledge schema of the
domain (the partonomy that aggregates each aspect of a rock that needs to be described,
showed in Figure 1), the implicit visual knowledge applied by expert in supporting
interpretation (later on, it was modelled through visual chunks and knowledge graphs),
and the explicitly knowledge or the extensive list of mineral names, textural aspects,
lithology nomenclature and the structural relationships that had further grown as the
Petrography ontology. Although the knowledge model of rock description and the
further extracted visual chunks are still in use in Petroledge and Hardledge® systems,
most of maintenance done over the original knowledge model refers to the vocabulary
extension and quality improvement of the ontology.
The subsequent evolution was demanded by the interpretation of event
sequence that has generated the rock. It was necessary to identify through the domain
ontology the upper level classes of the modelled concepts, such as event, temporal
relation and spatial relation. This was done by aligning the ontology with other
upper ontologies described in literature [Sowa 1995; Scherp et al. 2009] and then using
the concepts of upper ontology to classify and organize the related concepts in the
domain ontology. As a result, the study of the paragenetic relationships described in the
Petrography model shows those that represent the spatial relationship between minerals
that express the occurrence of an event. Formal definitions of temporal relations based
on Allen relations [Allen 1991] were included in the ontology, as well as the definition
of events in terms of Geology phenomena. The Allen relations and the definition of
diagenetic events allow the extraction and ordering of the events that have transformed
the sediments in a consolidated rock from the information described by the user in the
rock description.
The RockViewer and PetrographypediA applications were the first Petroledge
independent systems that were based on the ontology. As a consequence, these projects
have required the ontology rebuilt as an independent artifact, stored in a separated
database for further consultation. This reconstruction has produced a new model for
the same domain knowledge expressed in the ontology. The rigid concepts (rock and
mineral constituent) and their attributes have built the main framework of restructured
ontology. New terms were added to expand the domain of application to new kinds of
rock and new rock features
The more significant advance for the ontology development came with the use of
ontology for improving the interoperability in the petroleum modelling chain by
embodying geological explicit concepts and rock properties in RESQML standard. The
previously described projects were developed under supervision of the original team of
knowledge engineers. For the application into petroleum standards, the ontology needs
to be used for several engineers from many distinct software suppliers around the world.
The ontology needs to embody all restrictions requested to express the semantic of each
geological term in order to avoid a flexible use with another meaning, which is one of
the main sources of errors.
In order to support RESQML integration, each geological concept in the ontology
was studied based on the metaproperties proposed by Guarino and colleagues in
[Guarino & Welty 2001, 2002; Gangemi et al. 2003]. Physical objects, such as
lithological unit, and amounts of matter, like rock, were identified and modeled in a
separated way in the geological model. Usually these objects are collapsed or partially
merged in the geological models resulting in the main source of problems in reservoir
information integration, since many properties related to the substance, such as
permeability, are associated to bodies of rock and incorrectly extrapolated by the
simulation systems. In addition, the relevant attributes of the concepts that allow
defining the identity of each entity were specified as well as their domain of values.
The approach of conceptual spaces became the theoretical framework for modeling
domain of attributes aiming reusability in other areas of applications into the Geology
domain [Fiorini et al. 2015]. The ontological analysis of the main concepts of the
ontology that are being integrated into RESQML standard can be found in [Abel et al.
2015b].
In addition, the problem of scale of analysis that was never an issue for the
Petrography domain became central to support applications where the data is generated
and consumed in distinct scale of analysis. Basin (105 meters), reservoir (103 meters)
and well (10 meters) scales of studies have required that the range of numerical attributes
and the symbolic values were extended to cover the new possibilities of the domain.
5. Conclusion
The Petrography ontology has been continuously evolving since it was proposed. From
the initially two applications based on a set of twelve concepts, the model embodied a
vocabulary as large as 7000 terms split in two idioms which is shared by more than a
dozen applications.
This successful grown have been requiring continuous expansion in the number of
modelled concepts. Keeping the consistency and integrity of the knowledge base after
the inclusion of new concepts have requested periodic restructuring of the ontology
organization, sometimes followed by deep changes in the philosophical view that orients
the ontological decisions. These changes were especially significant on the first stages
of ontology-based application developments and now, when the ontology is going
to be integrated to the reservoir interchange standards. The rigor in making explicit
the semantic of each vocabulary for a large group of users of diverse specialties driven
by many distinct objectives is showing to be a challenge in terms of Ontology
Engineering. Some studies about the modularity of ontologies and the possibility of
offering specialized partial “views” to users according to their professional profile
[Aparicio et al. 2014] have indicate some new directions for the ontology evolution.
Acknowledgments: PetrograGrapher project were supported by CNPQ and CAPES.
The creation of commercial version of Petroledge and the Endeeper Co. was possible
thanks to the grants of FINEP and FAPERGS. We thank Endeeper for providing the
software detailed information described in this article.
6. References
Abel M. (2001). "The study of expertise in Sedimentary Petrography and its significance
for knowledge engineering (in Portuguese)". In: Informatics Institute, UFRGS,
Porto Alegre, p. 239.
Abel M., Castilho J.M.V. & Campbell J.A. (1998). "Analysis of expertise for
implementing geological expert systems". In: World Conference in Expert
Systems. Cognizant Communication Offices Mexico City, pp. 170-177.
Abel M., Lorenzatti A., Fiorini S.R. & Carbonera J. (2015a). "Ontological analysis of
the lithology data in PPDM well core model". In: 19th International Conference
on Petroleum Data Integration and Data Management. Pennwell Houston.
Abel M., Perrin M. & Carbonera J. (2015b). Ontological analysis for information
integration in geomodeling. Earth Science Informatics, 8, 21-36.
Allen J.F. (1991). Time and time again: The many ways to represent time. International
Journal of Intelligent Systems, 6, 341-355.
Aparicio J.M.L., Carbonera J.L., Abel M. & Pimenta M.S. (2014). "Ontology View
extraction: an approach based on ontological meta-properties". In: IEEE
international conference on tools with artificial intelligence - ICTAI 2014. IEEE
Limassol.
Castro E.S.E.D. (2012). PetrographypediA. The portal of Petrography. URL
http://www.Petrographypedia.com/
Fiorini S., Abel M. & Carbonera J. (2015). Representation of part-whole similarity in
geology. Earth Science Informatics, 8, 77-94.
Gangemi A., Guarino N., Masolo C. & Oltramari A. (2003). Sweetening WordNet with
Dolce. AI Magazine, 24, 13-24.
Gómez-Pérez A. & Benjamins V.R. (1999). "Overview of knowledge sharing and reuse
components: Ontologies and problem-solving methods". In: International Joint
Conference on Artificial Intelligence(IJCAI-99), Workshop on Ontologies and
Problem-Solving Methods (KRR5) (eds. Benjamins VR, Chandrasekaran B,
Gomez-Perez A, Guarino N & Uschold M) Stockolm, Sweden.
Guarino N. & Welty C. (2001). "Identity and Subsumption". In: The Semantics of
Relationships. An Interdisciplinary Perspective (eds. Green R, Bean CA &
Myaeng SH). Springer Netherlands, pp. 111-126.
Guarino N. & Welty C. (2002). Evaluating ontological Decisions with Ontoclean.
Communications of the ACM 45, 61 – 65.
King M.J., Ballin P.R., Bennis C., Heath D.E., Hiebert A.D., McKenzie W., Rainaud J.-
F. & Schey J.C.S.P.E. (2012). Reservoir Modeling: From RESCUE To
RESQML. SPE Reservoir Evaluation & Engineering, Society of Petroleum
Engineers.
Kop C. (2011). "Domain expert centered ontology reuse for conceptual models". In.
Springer Verlag Hersonissos, Crete, Greece, pp. 747-762.
Mastella L. (2004). "An event-based knowledge model for temporal sequence
acquisition and representation in Sedimentary Petrography". In: Informatics
Institute, UFRGS, Porto Alegre.
Mastella L.S., Abel M., Ros L.F.D., Perrin M. & Rainaud J.-F. (2007). "Event Ordering
Reasoning Ontology applied to Petrology and Geological Modelling". In:
Theoretical /Advances and Applications of Fuzzy Logic and Soft Computing.
(eds. Castillo O, Melin P, Ross OM, Cruz RS, Pedrycz W & Kacprzyk J).
Springer-Verlag, pp. 465-475.
Nonaka I., Takeuchi H. & Takeuchi H. (1995). The knowledge-creating company: how
Japanese companies create the dynamics of innovation. Oxford University Press,
New York.
Scherp A., Franz T., Saathoff C. & Staab S. (2009). "F - A model of events based on the
foundational ontology DOLCE+DnS ultralite". In: K-CAP'09 - 5th International
Conference on Knowledge Capture. ACM SIGART Redondo Beach, CA, United
states, pp. 137-144.
Schreiber G., Akkermans H., Anjewierden A., Hoog R.d., Shadbolt N., Velde W.v.d. &
Wielinga B. (2000). Knowledge engineering and management:The
CommonKADS Methodology. The MIT Press, Cambridge.
Shah T., Rabhi F., Ray P. & Taylor K. (2014). "A guiding framework for ontology
reuse in the biomedical domain". In. IEEE Computer Society Waikoloa, HI,
United states, pp. 2878-2887.
Silva L.A.L. (1997). "Intelligent database for petroghaphic analysis (In Portuguese)".
In: Informatics Institute. UFRGS Porto Alegre, p. 114.
Simperl E., Sarasua C., Ungrangsi R. & Burger T. (2011). Ontology metadata for
ontology reuse. International Journal of Metadata, Semantics and Ontologies,
6,126-145. Inderscience Enterprises Ltd.
Sowa J.F. (1995). Top-level ontological categories. International Journal of Human
Computer Studies, 43, 669-669. Academic Press Ltd.
Tsalapati E., Stamou G. & Koletsos G. (2009). "A method for approximation to
ontology reuse problem". In. Inst. for Syst. and Technol. of Inf. Control and
Commun. Funchal, Madeira, Portugal, pp. 416-419.
Victoreti F.I. (2007). "Mapping and documentation of diagnostic visual features for
interpretation in a knowledge-based system in Petrography domain". In:
Informatics Institute, UFRGS, Porto Alegre.