Formal ontological analysis is a methodology that builds on some philosophical notions in order to guide the process of building ontologies whose structure is correct and little or no tangled. This paper presents an ontology model that facilitates formal ontological analysis, by providing a set of metaproperties which characterise the behaviour of concept properties in a concept definition, while providing a richer semantics of the concept. We describe concepts in terms of their attributes (characterising features) and we also describe the role played by these features in the concept definition, whether they are prototypical or exceptional, whether they are permitted to change over time, and if so, how often this happens, how likely is a concept to show these features, etc. We show that these metaproperties can support a methodology, OntoClean [44] that uses formal ontological analysis to build cleaner taxonomies (which are thus more sharable). The set of metaproperties for attributes we propose can be used to guide in determining which metaproperties for concepts hold for an ontology and therefore can support the use OntoClean.
Many current applications such as e-commerce or the semantic web
rely on the ability of different resources or agents to interoperate
with each others and with users. In some cases, interoperation
becomes more complex, because agents may have been
independently developed, therefore the assumption that agents use the same
communication language and the same terminology in a consistent
way cannot be made. When dealing with independently developed
agents, their interoperability with humans and others depends on
the agents’ ability to understand them, which leads us directly
to ontologies. Ontologies are an explicit, formal specification of
a shared conceptualisation, where a ‘conceptualisation’ refers to
an abstract model of some phenomenon in the world by having
identified the relevant concepts of that phenomenon, ‘explicit’ means
that the type of concepts used, and the constraints on their use are
explicitly defined, ‘formal’ refers to the fact that the ontology should
be machine-readable, and lastly ‘shared’ reflects the notion that an
ontology captures consensual knowledge, that is it is not private to
some individual, but accepted by a group [
Ontology merging starts with the attempt to find the places in which
the source ontologies overlap [
Any consideration on ontology heterogeneity it is usually done
assuming that the ontologies involved in the merging process are
either built according to some kind of engineering methodology,
such as Methontology [
Methontology and OntoClean are complementary
methodologies in that Methontology provides the guidelines for building
or re engineering ontologies, whereas OntoClean can be used
either in the validation step (when ontologies are engineered or
restructured) or simultaneously with the ontology construction
(when ontologies are built from scratch). These two
methodologies are currently undergoing an integration process [
Methodologies to obtain well-built ontologies, however, are not
enough to support the semi-automatic coalescence process. In fact
in order to recognise whether two concepts (that can be affected
by heterogeneity) are similar, we cannot only rely on the the
terms denoting them, on the relationships with other terms, and on
their descriptions, but we need to have a full understanding of the
concepts. As noted by McGuinness [
This paper is organised as follows: Section 2 presents the OntoClean
methodology and the notions of formal ontological analysis, while
Section 3 introduces our proposal for an ontology model
encompassing a set of metaproperties for attributes which are discussed in
the following subsections. This ontology model was also presented
in [
OntoClean [
OntoClean is strongly based on the philosophical notions of identity, unity, essence (rigidity), and dependence. The attribute metaproperties we present in this paper are related to these notions, and we discuss them below.
Identity: Identity is the logical relation of numerical sameness,
in which a thing stands only to itself. Based on the idea that
everything is what it is and not anything else, philosophy has tried for a
long time to identify the criteria which allow a thing to be identified
for what it is even when it is cognised in two different forms, by
two different descriptions and/or at two different times [
This comprises both aspects of finding constitutive criteria (which features a thing must have in order to be what it is), and of finding re-identification criteria (which feature a thing has to have in order to be recognised as such by a cognitive agent). These are distinct, although equally important aspects of identity.In fact, while identity is not affected by the context and is based on the the intrinsic features of an object, whereas re-identification is affected by context and it is based on features that are external to the object. For example, an identity criterion for people is to have matching fingerprints, so two people are the same if they have the same fingerprints. Fingerprints are intrinsic to the individual, they are not assigned by an external agent. A re-identification criterion might depend on the role played by the object: one can be a student and an employee at the same time, and is re-identified as student by the student id, whereas is re-identified as employee by an employee number.
Although the problem of identifying what features an entity should have in order to be what it is and recognised as such has been central to philosophy, it did not have the same impact in conceptual modelling and more generally AI. The ability to identify individuals is central to the modelling process, more precisely, it is not the mere problem of identifying an entity in the world that is central to the ontological representation of the world, but the ability to re-identify an entity in all its possible forms, or more formally reidentification in all possible worlds. 2 That is, the problem is related to distinguishing a specific instance of a concept from its siblings on the basis of certain characteristic properties which are unique and intrinsic to that instance in its whole. Intrinsic properties correspond to the modelling primitive attributes. Extrinsic properties represent relations between classes, thus corresponding to the modelling primitive relationship.
This notion is, of course inherently time dependent, since time gives rise to a particular system of possible worlds where it is highly likely that the same instance of a concept exhibits different features 3.
This problem is known as identity through change: an instance of a
concept may remain the same while exhibiting different properties
at different instants of time. Therefore it becomes important to
understand which features or properties can change and which
cannot [
If we reformulate the identity problem as re-identification we realise that re-identification is also affected by time; how can we re-identify the same instance at different instant of times? We face the re-identification problem in everyday life; we are able to recognise the features that permits us to distinguish an instance from the others, and when intrinsic features are not available, we ‘attach’ artificial features, that permit us to establish identity. One example is the Student ID, which is assigned to university students, in order to identify students univocally.
Unity: the notion of unity is often included in a more generalised notion of identity, although these two notions are different.
While identity aims to characterise what is unique for an entity
of the world when considered as a whole, the goal of unity is
that of distinguishing the parts of an instance from the rest of the
world by means of a unifying relation that binds them together (not
involving anything else) [
Some philosophers, e.g. Lewis [21, page 39 ff], hold that there is no such thing as trans-world identity, although objects in one world can have counterparts in other worlds.
Here the counterpart theory does not hold, and so identity through time is
always accepted.
#
" holds in all possible worlds means that is possible, i.e. that
! operators according to the following meanings: means that
property that is necessary to all instances in any instant of time, that
is a property such that: . For
this formula, and in the remainder of this paper, we use the modal
notions of necessity and possibility quantified over possible
worlds (in Kripke’s semantics [
Rigidity strictly depends on the notions of time and modality [
Time is undoubtedly one partition of these worlds, but temporal
permanence means that the property is true in that world (time), with
no information concerning the other possible worlds, and this might
happen by pure chance.
# Inheritance and Distinction: inherited metaproperties regard those
# Prototypes and Exceptions: the metaproperties Prototypical and
# Attribute’s behaviour over time: The metaproperties Mutability,
%$%&’)(* %$%,+’ erty such that , that is there is possibly
Mutability Frequency, Event Mutability and Reversible Mutability
provide a better description of attributes by characterising their
behaviour over time, that is, whether they are allowed to change their
value during the concept lifetime (Mutability) and how often the
change occurs Mutability Frequency), whether the change is
reversible (Reversible Mutability), and what triggers change (Event
Mutability);
Modality: this meta-property is a qualitative description of the
degree of inheritability of a concept property by its subconcepts;
Exceptional aim to describe properties that are prototypical for
a concept, that is the properties that obtain for the prototypical
(from a cognitive viewpoint, according to Rosch [
Attributes are described in terms of their structural characteristics,
such as the concepts that they are defining, their allowed values, the
type of the values (string, integer, etc.), and the maximum and
minimum values (if attributes are numeric). Attributes are also described
in term of the following metaproperties:
Dependence: In OntoClean [
In order to establish whether these metaproperties hold, OntoClean is supported by a description logic based system that can help knowledge engineers to assign the metaproperties to concepts and to verify the taxonomic structure on the grounds of the modelling methodology. In this paper we focus our attention on the process of assigning the metaproperties. OntoClean guides knowledge engineers in this process by asking them to answer some questions such as “Does the property carry identity”. Knowledge engineers can answer yes, no or unsure, in this latter case more specific questions can be asked, such as “Are instances of the property countable?”.
The OntoClean methodology depends on the knowledge engineers understanding of the ontologies to analyse and can thus be problematic if used to evaluate independently designed ontologies.
Moreover, OntoClean does not take into account the structure of concept definitions, as it does not consider the characteristic features (or attributes) that might have been used to define concepts.
This work proposes an enriched ontology model whose aim is to complement the OntoClean methodology, by providing an additional way to determine metaproperties to concepts. In our proposal we describe concepts in terms of their characterising properties, which are in turn described not only in terms of their structural features (such as range, domain, cardinality etc.), but also in terms of their metaproperties, which describe the contribution given by these properties to the concept definition. We describe the enriched ontology model and the metaproperties for attributes in the next sections. 3
The ontology model we propose comprises concepts, attributes, relations, and instances. We do not consider here axioms. Concepts represent the entities of the domain and the tasks we want to model in the ontology. Concepts are described in terms of defining properties, which are represented by associating an attribute with either a single value or a set of values. Concepts are organised into an Is-a hierarchy, so that a concept attributes and their values are inherited by subconcepts. Multiple inheritance is permitted, so attributes and
These metaproperties provide means to distinguish between necessary and sufficient conditions for class membership. Indeed, the modality meta-property and those describing the behaviour over time permit the identification of essential (or rigid) properties and necessary properties are those that are essential to all instances of a concept. Prototypical properties are good candidates to identify sufficient conditions, as discussed in Section 3.3.
Relations between concepts are supported by the model as are instances. Finally, the ontology model supports roles. Concepts are also used to represent roles, which can be thought of describing the part played by a concept in a context, (a more complete discussion on roles is postponed to Section 4.3). Roles are described in terms of their context, and the formal role relationship holds, that is, roles are related to concepts by a ‘Role-of’ relations.
This ontology model enriches the traditional model proposed initially
by Gruber [
These metaproperties describe the behaviour of fluents over time, where the term fluent is borrowed from situation calculus to denote a property of the world that can change over time. Modelling the behaviour of fluents corresponds to modelling the changes in properties that are permitted in a concept description without changing the essence of the concept. Describing the behaviour over time also involves distinguishing properties whose change is reversible from those whose change is irreversible.
Property changes over time are caused either by the natural
passage of time or are triggered by specific event occurrences. We need,
therefore, to use a suitable temporal framework that permits us to
reason with time and events. In [
Changes of properties can be modelled as processes [
Any regular occurrence over time can be, however, expressed in form of an event, since most of the forms of reasoning for continuous properties require discrete approximations. Therefore in the ontology model we present here, continuous properties are thought of as discrete properties where the event triggering the change in property is the passing of time from the instant to the instant . Events are always thought of as point events, and we consider durational events (events which have a duration) as being a collection of point events in which the property whose mutability is modelled by the set of metaproperties hold as long as the event lasts. in a way which depends on concept definitions in terms of attributes.
This might result particularly useful when knowledge engineers need to assign formal properties to ontologies they have not designed.
On the other hand, this conceptual model for ontologies facilitates a better understanding of the concept semantics. Currently ontology merge is performed by hand based on the expertise of the knowledge engineers and on the ontology documentation. Even in this case the ontology model we propose can prove useful by providing a characterisation of the properties, which can help to identify semantically related terms. The following subsections describe all the metaproperties for attributes but Inheritance and Distinction (which are trivial) more in detail: # Event Mutability, which models the reasons why a property may
Mutability, which models the liability of a concept property to
change, a property is mutable if it can change during the concept
lifetime;
Mutability Frequency, which models the frequency with which a
property can change in a concept description;
change; Reversible Mutability, which models reversible changes
of the property.
- [
Here we use the term modality to denote the degree of surprise in encompassed in this meta-property is important for reasoning with statements that have different degrees of credibility. Indeed there is a difference in asserting facts such as ‘Cats are pets’ and ‘All felines are pets’, the former is generally more believable than the latter, for which many more counterexamples can be found. The ability to distinguish facts whose truth holds with different degrees of strength is important in order to find which facts are true in every possible world and therefore constitute necessary truth.
The ability to evaluate the degree of confidence in a property
describing a concept is also related to the problem of reasoning with
ontologies obtained by merge. In such a case, mismatches can arise if a
concept inherits conflicting properties. In order to be able to reason
with these conflicts some assumptions have to be made, concerning
on how likely it is that a certain property holds. In case of conflict the
property’s degree of credibility can be used to apply some forms of
non monotonic reasoning or belief revision. For example, we could
rank the possible alternatives on the grounds of the degree of
credibility following an approach similar to the one presented in [
The metaproperties which model the behaviour of the attributes over
time are:
In order to get a full understanding of a concept it is not sufficient
to list the set of properties generally recognised as describing a
typical instance of the concept but we need to consider the known
exceptions. In this way, we partially take the cognitive view of
prototypes and graded structures, which is also reflected by the
information modelled in the meta-property modality. In this view all
cognitive categories show gradients of membership which describe how
well a particular subclass fits people’s idea or image of the category
to which the subclass belong [
Ontologies typically presuppose context and this feature is a major source of difficulty when merging them, since information about context is not always made explicit.
Prototypes are also quite important in that they provide a frame of reference for linguistic quantifiers such as tall, short, old, etc. These quantifiers are usually defined or at least related to the prototypical instance of the concept which is being described, and indeed their definition changes if we change the point of reference.
Therefore including the notions of prototypes and exceptions permits us to provide a frame of reference for defining these qualifiers with respect to a specific concept. For the purpose of building ontologies, distinguishing the prototypical properties from those describing exceptions increases the expressive power of the description. Such distinctions do not aim at establishing default values but rather to guarantee the ability to reason with incomplete or conflicting concept descriptions.
The ability to distinguish between prototypes and exceptions helps to determine which properties are necessary and sufficient conditions for concept membership. In fact a property which is prototypical and that is also inherited by all the subconcepts becomes a natural candidate for a necessary condition. Prototypes, therefore, permit the identification of the subconcepts that best fit the cognitive category represented by the concept in the specific context given by the ontology. On the other hand, by describing which properties are exceptional, we provide a better description of the membership criteria in that it permits us to determine what are the properties that, although rarely holding for that concept, are still possible properties describing the cognitive category.
Prototypes and exceptions can prove useful in dealing with conflicts arising from ontology merging. When no specific information is made available about a concept and it inherits conflicting properties, then we can assume that the prototypical properties hold for it. 4
The ontology model presented in previous section could be
implemented in any kind of ontology representation formalisms. In [
By adding the metaproperties to the ontology model, we provide an
explicit representation of the attributes’ behaviour over time, their
prototypicality and exceptionality, and their degree of applicability
to subconcepts. This explicit representation may be used to support
and complement the OntoClean methodology [
Furthermore, the enriched ontology model we propose forces
knowledge engineers to make ontological commitments explicit, that is the
agreement on the meaning of the terms used to describe a domain
[
A possible drawback of enriching the ontology model is that knowledge engineers are required a deeper analysis of a domain. We realise that it makes the process of building an ontology even more time consuming but we believe that a more precise ontological characterisation of the domain at least balances the increased complexity of the task. Indeed, in order to include the attribute’s metaproperties to the ontology model, knowledge engineers need to have a full understanding not only of the concept they are describing, but also of the context in which the concept is used. Arguably, they need such knowledge if they are to perform the modelling task thoroughly. The evaluation of the cost to pay for this enriched expressiveness and of the kind of reasoning inferences permitted by this model are strictly dependent on the domain and the task at hand. We can imagine that the automatic coalescence of terms might require more sophisticated inferences whose cost we cannot evaluate a priori. In some other cases, the simple matching between properties’ charactersiations might help in establishing or ruling out the possiblity of semantic relatedness. For example, two concepts are described by the same properties but with different characterisations, this might indicate that the concepts have been conceptualised differently. 4.1
The idea of modelling the permitted changes for a property is strictly
related to the philosophical notion of identity. The metaproperties
modelling the behaviour over time are, thus, relevant for establishing
the identity of concept descriptions [
Prototypical and exceptional properties and the modality
metaproperties describing how the property is inherited in the hierarchy can all
contribute to determine what are the necessary and sufficient
conditions for class membership. Necessary and sufficient conditions are
ultimately the conditions that permit us to define the properties
constitutive of identity and to distinguish them from those that permit
re-identification.
2 in [
Identity through change is also relevant to determine rigidity. In Section 2 a rigid property is defined as a property that is essential to all its instances.
In [
Indeed, since an ontology defines a vocabulary, we can restrict
ourselves to the set of possible worlds which is defined as the set of
maximal descriptions obtainable using the vocabulary defined by the
ontology [
It is important to note that, although in [
Roles dependence on identity and rigidity .+@<44(A+?/B<44 .@<4’ < ’, where denotes that is a part of /B<4’ < while denotes that is a constituent of . In other words a concept is a role if its individuals stand in relation to other individuals, and they can enter or leave the extent of the concept without losing their identity. From this definition it emerges that the ability of recognising whether rigidity holds for some property is essential in order to distinguish whether is a role.
Roles may be ‘naturally’ determined when social context is taken into account, and the social context determines the way in which a role is acquired and relinquished. For example, the role of President of the country is relinquished differently depending on the context provided by the country. So, for example, in Italy the role may be acquired and relinquished only once in the lifetime of an individual, whereas if the country is the United Sates, the role can be acquired and relinquished twice, because a president can be re-elected. Social conventions may also determine that once a role is acquired it cannot be relinquished at all. For example, the role Priest in a catholic context is relinquished only with the death of the person playing the role. The ability to distinguish roles gives also a deeper understanding of the possible contexts in which a concept can be used. Recognising a role can be equivalent to defining a context, and the notion of context is the basis on which prototypes and exceptions are defined.
In [
For the aforementioned reasons, ways of representing roles must be supported by some kind of explicit representation of time and events.
Indeed the proposed model provides a way to model roles as fluents; moreover, by modelling the reason for which a property change, we provide knowledge engineers the ability to model the events that constrain the acquisition or the relinquishment of a role.
We have illustrated and discussed a ontology model which is enriched with metaproperties providing a better characterisation of attribute. This characterisation is meant to help in disambiguating heterogeneous concepts when merging ontologies, since we assume that two concepts can be matched if :
# Scalability: The addition of a new resource to the architecture # Accommodation of diverse formalisations: A cluster can be # Division of ontology authoring efforts: Ontologies composing a # Impact of change minimisation: If a concept description needs # Composability: Different clusters are composed by generalising
Modularity/separability: Each cluster is like a module in
software engineering and represents a specific aspect of the domain;
the concepts that are common to them. This is the first step to
permit heterogeneous resources to communicate;
requires only the production of the mapping rules between the
ontology associated to the new resource and the cluster to which this
resource belongs;
to be changed only the mapping rules between the updated
ontology and the cluster to which this ontology belongs need to be
rewritten;
cluster do not need to be authored by the same people as long as
their concepts can be mapped into the concepts of the cluster.
comprised of ontologies representing different formalisations of
the same domain, such as different temporal ontologies.
# candiCdate matching attributes show the same behaviour in
modelling the concept, that is, the same metaproperties hold for the
attributes.
# Lack of tools that can support the building of the ontology clusters.
Matching similar concepts plays a pivotal role in those approaches
to knowledge sharing which rely on shared ontologies in order to
perform the translation between concepts in heterogeneous
ontologies. Usually, knowledge sharing is obtained by creating one shared
ontologies to which all the agents commit. However, such an
approach has been compared to imposing a standard and suffers from
the same drawbacks [
In contrast to an approach in which all resources share one body
of knowledge here we propose to locate shared knowledge in
multiple but smaller shared ontologies. This approach is referred to as
ontology-based resource clustering, or shortly, ontology clustering
[
Each cluster can be thought of as a micro-theory shared by all the
agents that conform to that cluster. Each micro-theory is in turn
generalised and they are all eventually generalised by the top-level
ontology which is a standard upper ontology like the Upper-Cyc [
This approach is analogous to modularisation in software
engineering and is thought of having the same advantages, which are:
This approach has not been tested yet, therefore we can only foresee
some disadvantages:
Ontology clustering is based on the similarities between the
concepts known to different resources, where each resource represents
a different aspect of the domain knowledge. We assume that the
ontologies modelling the resources are consistent, non-redundant,
and well structured. We also assume that the ontologies have been
built with a methodology including a formal evaluation step, such as
Methontology [
Since our resources need to communicate in a sensible fashion they
are all supposed to be familiar with some high level concepts. We
group these concepts in an ontology rooted at the top of the
hierarchy of ontologies. As it describes concepts that are specific to the
domain and tasks at hand we refer to this ontology as the application
ontology (following Van Heijst and colleagues, [
A cluster is referred to as a group of consistent ontologies (possibly one) in our structure and is described by an ontology which is shared by those composing the cluster. Both ontology clusters and ontologies within each cluster are organised in a hierarchical fashion where each sibling cluster specialises the concepts that are in its parent cluster. However, while multiple inheritance is permitted within the ontologies, it is not permitted between ontologies, therefore the structure of clusters is a tree. In this structure, the lower level clusters have more precise concept definitions than the higher levels, making the latter more abstract.
Clusters are linked by restriction or overriding relations, that is
concepts in one parent ontology are inherited by its children cluster,
but overriding is permitted [
Figure 1 illustrates an example of this structure, where Local Ont. are the local ontologies.
Since different siblings can extend their parent cluster concepts in different ways the cluster hierarchy permits the co-existence of heterogeneous (sibling) ontologies. Figure 1 illustrates this particular already contained in its parent ontology. Ultimately, ontologies are likely to have concepts that are not shared with any other ontology.
In our ontology structure, we then create a separate, domain-specific ontology as sub ontology of the cluster in which the ontology resides. We refer to these ontologies as local ontologies. The local ontologies are the leaf nodes of our ontology hierarchy. In each of the ontologies in the structure, concepts are described in terms of attributes and inheritance relations holding in the ontology’s structure. Concepts are hierarchically organised and the inheritance (with exceptions) allows the passing down of information through the hierarchy. Multiple inheritance is only permitted within the ontologies.
Concepts are expressed in terms of inherited and distinguishing attributes. To the set of inherited attributes other attributes are added to distinguish the specific concept from the more general one. These attributes describe the characteristic differences between a concept and its siblings. The distinguishing attributes are used to map concepts from a source ontology into a target ontology preserving the meaning of the concept. 5.2
The structure of ontology clusters introduced in Section 5.1 builds on the ability of identifying similar concepts in different ontologies. Identifying which concepts are similar and assessing the degree of semantic similarity between them are, thus, two essential steps in the process of building ontology clusters. However, assessing the similarity between concepts in diverse ontologies is not a trivial task because of the heterogeneity that can affect concepts and their descriptions.
The problem of assessing semantic similarity has received much
attention in the artificial intelligence field [
The ontology merging environment Chimaera [
In [
The method proposed by Rodr´iguez and Egenhofer is based on
Tversky [
It could be argued that enriching the concept structure by distinguishing between parts, functions and attributes can give rise to the articulation of new types of mismatches associated with the classifications of features. However, the authors claim that the advantages of enriching the concept structure, namely a matching process that compares corresponding characteristics of concepts, and the ability to distinguish different aspects of the context, modelled by the features, overweights the possible disadvantages deriving from a higher number of mismatches.
We believe that Rodr´iguez and Egenhofer approach to assess
semantic similarity rises an important issue, which is that, in order to be
able to have a better assessment of semantic similarity (that gives
also gradients of similarity and not only a binary function) it is
necessary to provide a richer description of the structure of the concepts
in the source ontologies. However, we believe that the distinguishing
features proposed in [
Sharing ontologies independently developed is a burning issue that
needs to be solved. This paper presents a set of metaproperties
describing concept characteristic features (attributes) that can be used
to support both the process of building correct ontologies (by
complementing and supporting the formal ontological analysis performed
by the OntoClean methodology [
The metaproperties we propose, namely Mutability, Mutability Frequency, Reversible Mutability, Event Mutability, Modality, Prototypicality, Exceptionality, Inheritance and Distinction encompass semantic information aiming to characterise the behaviour of properties in the concept description. We have argued that such a precise characterisation might help to disambiguate among concepts that only seem similar, and in turn can support mappings across the structure of multiple shared ontologies that we have devised as alternative to the current approaches to knowledge sharing. We claim that this characterisation of the concept properties is also very important in order to provide a precise specification of the semantics of the concepts. Such characterisation is essential if we want to perform a formal ontological analysis, in which knowledge engineers can precisely determine which formal tools they can use in order to build an ontology which has a taxonomy that is clean and not very tangled. The novelty of this characterisation is that it explicitly represents the behaviour of attributes over time by describing the permitted changes in a property that describe a concept. It also explicitly represents the class membership mechanism by associating with each attribute (represented in a slot) a qualitative quantifier representing how properties are inherited by subconcepts. Finally, the model does not only describe the prototypical properties holding for a concept but also the exceptional ones. By providing this explicit characterisation, we are asking knowledge engineers to make more hidden assumptions explicit, thus providing a better understanding not only of the domain in general, but also of the role a concept plays in describing a specific domain. This paper has also presented a structure of multiple shared ontologies for knowledge sharing. Although this is still on going research, we believe that such a structure has advantages over the others especially if considered in the context of an open environment such as the Internet. We believe that this kind of modularisation is the key to applications where intelligent agents (whose knowledge is represented by ontologies) interoperate dynamically, by agreeing on the vocabulary (and shared knowledge) which is closer to the conceptualisations of only those agents which are involved in the interoparation and not of all agents that can be potentially involved. We realise that we have not investigated in sufficient detail the issues related to building such structure in an efficient and cost effective manner, and the relationships existing within and between the ontologies composing the structure (both topics are future research directions that we will consider, see next section); but we think that we have laid the basis for future research. 7
Future research on ontology clusters concerns the relationships
between and within ontologies, which need to be clarified with respect
to previous work presented in the literature. Two candidate sets of
relations have been identified, these are Borst’s ontology projections:
include and extend, include and specialise, include and map [
From the viewpoint of the ontology conceptual metamodel, future work include understanding the kind of inferences and the reasoning mechanisms that are supported by the additional semantics included in the ontology metamodel. In order to support complex reasoning inferences, we will consider the implementation of the metamodel in some description logic’s based language, which should provide the capabilities to perform the inferences. This model is also quite demanding to use, future work should concentrate also on identifying the kinds of applications that can benefit from the expressive power provided by this model.
In order to test the effectiveness of the conceptual metamodel, we are planning to include the metaproperties in tools to build ontolo
The authors wish to thank Asunci o´n G o´mez-Pe´. The PhD presented in this paper was funded by BT plc.