Ebook: Formal Ontology in Information Systems

Ontology began life in ancient times as a fundamental part of philosophical enquiry concerned with the analysis and categorisation of what exists. In recent years, the subject has taken a practical turn with the advent of complex computerised information systems which are reliant on robust and coherent representations of their subject matter. The systematisation and elaboration of such representations and their associated reasoning techniques constitute the modern discipline of formal ontology, which is now being applied to such diverse domains as artificial intelligence, computational linguistics, bioinformatics, GIS, knowledge engineering, information retrieval, and the Semantic Web. Researchers in all these areas are becoming increasingly aware of the need for serious engagement with ontology, understood as a general theory of the types of entities and relations making up their respective domains of enquiry, to provide a solid foundation for their work.

The conference series Formal Ontology in Information Systems (FOIS) is intended to provide a meeting point for researchers from these and other disciplines with an interest in formal ontology, where both theoretical issues and concrete applications can be explored in a spirit of genuine interdisciplinarity. FOIS began with the first meeting in Trento, Italy, in June 1998, which was followed by meetings in 2001, 2004, 2006, and 2008. The current, sixth FOIS conference is being held in Toronto, Canada, during 11–14 May 2010. In our call for papers, we solicited contributions from a wide range of areas important to the development of formal ontologies:

Foundational Issues

• Kinds of entity: particulars vs universals, continuants vs occurrents, abstracta vs concreta, dependent vs independent, natural vs artificial

• Formal relations: parthood, identity, connection, dependence, constitution, subsumption, instantiation

• Vagueness and granularity

• Identity and change

• Formal comparison among ontologies

• Ontology of physical reality (matter, space, time, motion, …)

• Ontology of biological reality (genes, proteins, cells, organisms, …)

• Ontology of artefacts, functions and roles

• Ontology of mental reality and agency (beliefs, intentions and other mental attitudes; emotions, …)

• Ontology of social reality (institutions, organizations, norms, social relationships, artistic expressions, …)

• Ontology of the information society (information, communication, meaning negotiation, …)

• Ontology and Natural Language Semantics, Ontology and Cognition

Methodologies and Applications

• Top-level vs application ontologies

• Ontology integration and alignment; role of reference ontologies

• Ontology-driven information systems design

• Ontology-based application systems

• Requirements engineering

• Knowledge engineering

• Knowledge management and organization

• Knowledge representation; Qualitative modeling

• Computational lexicons; Terminology

• Information retrieval; Question-answering

• Semantic web; Web services; Grid computing

• Domain-specific ontologies, especially for: Linguistics, Geography, Law, Library science, Biomedical science, E-business, Enterprise integration, …

Out of 71 papers submitted to FOIS-2010, 28 were selected for inclusion in the final programme, covering both the foundational issues and more application-oriented topics. Every submission was reviewed by at least two (and in almost every case by three) experts from the Programme Committee, together with a number of extra reviewers as listed below.

Many people have contributed to the success of FOIS-2010. First and foremost, of course, are the authors who responded to the call for papers, without whom there would have been no FOIS. We would like to thank the members of the Programme Committee for undertaking the reviewing work to a tight deadline, as well as the additional reviewers whose help was solicited where necessary. We are particularly grateful to our three invited speakers, Francis Jeffry Pelletier, John Bateman, and Alan Rector, whose contributions in themselves range across the whole spectrum of ontology from philosophical foundations to applications. We would like to thank the Conference Chair, Nicola Guarino, and the local chairs, Michael Grüninger and Chris Welty, for their indispensible handling of organizational aspects of the conference. Finally we must acknowledge the support of the newly-formed International Association for Ontology and its Applications, which from 2010 has taken over the general oversight and management of FOIS.

February 2010

Antony Galton, Riichiro Mizoguchi

Mass terms are those such as ‘water’, ‘computer software’, ‘advice’, and ‘knowledge’. They are contrasted with count terms such as ‘person’, ‘computer program’, ‘suggestion’, and ‘belief’. Intuitively, mass terms refer to “stuff” while count terms refer to “objects”. Since mass terms refer to stuff, they (but not count terms) allow for measurement: ‘a liter of water’, ‘three CDs worth of computer software’, ‘too much advice’, ‘many books worth of deep knowledge’. Since count terms refer to objects, they (but not mass terms) allow for counting, quantifying and individuating: ‘a person’, ‘three computer programs’, ‘each suggestion’, ‘that belief of his’.

The presumed fact that mass nouns are not true of individuals has seemed to some to show that traditional logical theory is inadequate as a medium for their semantic representation. It has seemed implausible to represent ‘Snow is white’ as ∀x(Snow(x)→White(x)), since there are no plausible values for x. (What could be the value of x in ‘For each x, if x is snow then x is white’?) In turn, this has been the driver for the use of mereological-based approaches to the semantics of mass terms. (‘Mereology’ is here to be understood widely to include not only classical mereology, but a range of approaches that share some features with the classical mereology.)

Further philosophical problems traditionally associated with mass terms include distinguishing mass from count terms (is it a syntactic or semantic distinction?), deciding the extent of the classification (does it include more than noun phrases?), describing the semantic underpinnings of mass terms (since they are not true of individuals, how can a model theory be developed?), and explaining the ontology presupposed by mass terms vs count terms.

Alongside these concerns, there is the meta-philosophical question of the extent to which the linguistic practices of a language can be used as evidence for how those speakers view reality, or indeed, what reality is. It is this issue, as applied to the question of the ontology presupposed by mass terms vs count terms, that I wish especially to address.

I am going to restrict my attention to the category of nouns and noun phrases, even though some writers have noted parallels between count terms and event-verbs and between mass terms and process-verbs. Others have noted how aspectual distinctions sometimes mirror the mass/count distinction in nouns: one can eat dinner for an hour (eating dinner is mass-like; and the temporal modification is not punctual), but it seems nonsensical to eat a dinner for an hour (eating a dinner is count-like, but this does not happily combine with non-punctual temporal modification). Some writers have argued that some adjectives can be classified as count or mass: for instance, ‘triangular’ might be count, ‘watery’ might be mass. I think many of my observations within the nominal domain might carry over to these other categories, but I don't wish to argue for that here.

Throughout the recent history of philosophical and linguistic reflections on the mass/count distinction, there have been skeptics who disparage the attempt to make the distinction carry any semantic weight or philosophical importance. Such skeptics point to the fact that the same meaning can be count in one language but mass in another, even in very closely related languages. Furthermore, some languages, especially the Sino-Japanese-Korean ones, have been argued to assign mass to all lexical nouns. There are also some Amerindian languages that seem to have no mass/count distinction at all. The skeptics also point to pairs of terms within one language where one is count and the other mass, yet they seem to designate very similar items of reality—for instance, ‘brownie’ vs. ‘baklava’—or sometimes even the same item of reality—for instance, ‘this suggestion’ vs. ‘this advice’. (Or, to consider a more philosophically relevant example: ‘belief’ is a count term while ‘knowledge’ is a mass term. Yet we commonly think that what is one and the same thing can be a belief of Sandy's but knowledge of Kim's; and that before Kim had enough evidence to make it be knowledge, it was a belief of Kim's.) Still others, myself included, have argued that (almost?) all nouns can be employed in both a mass and a count manner, showing that some very fancy footwork is required in order to give a reasonable semantic and ontological accounting for this.

I too am a skeptic about the viability of using the mass/count distinction to draw any firm metaphysical/ontological conclusions. However, I am not a skeptic about the individual/stuff distinction—it is just that I don't think it parallels the mass/count distinction. And I want to describe some ways that one might take to de-couple the two.

One way I want to proceed is to cast doubt on the assignment of mass/count to lexical nouns, and I will argue instead for the view that they should be assigned only at the level of noun phrases. (Following the suggestions made by Keith Allen (1980).) But in addition to these syntactic moves, we need to deal with the semantic consequence of (what I presume to be) the fact that every noun is sometimes used in a count manner and sometimes in a mass manner. Solutions in the literature postulate a number of ‘conversion functions’ or ‘coercions’ that will convert the meaning of ‘chicken’ (+count) to the meaning of ‘chicken’ (+mass). I think these are all misguided, and wish to give a plausible replacement account.

The situated interpretation of natural language concerning space, spatial relationships and spatial activities is a complex problem spanning contributions from several disciplines. Space plays a central role in many theories of cognition and the spatial language observed in actual contexts of use is extremely flexible. In our work on spatial representations of all kinds, principles drawn from ontological engineering play a central role. Moreover, we have been led to augment those principles in particular ways: most specifically with respect to ontological modularity, ontological heterogeneity, and multiple ontological levels or strata. In the presentation accompanying this position statement, I give an overview of our work on space and the ontologically-relevant conclusions that we have drawn. I also suggest that these conclusions are relevant for ontological work in general, and not just for that concerned with spatial issues.

Ontologies have been highly successful in applications involving annotation and data fusion. However, ontologies as the core of “Knowledge Driven Architectures” have not achieved the same influence as “Model Driven Architectures”, despite the fact that many biomedical applications require features that seem achievable only via ontological technologies – composition of descriptions, automatic classification and inference, and management of combinatorial explosions in many contexts. Our group adopted Knowledge Driven Architectures based on ontologies to address these problems in the early 1990s. In this paper we discuss first the use cases and requirements and then some of the requirements for more effective use of Knowledge Driven Architectures today: clearer separation of language and formal ontology, integration with contingent knowledge, richer and better distinguished annotations, higher order representations, integration with data models, and improved auxiliary structures to allow easy access and browsing by users.

The paper examines the semantics of the terms ‘environment’ and ‘habitat’ and presents a semi-formalised ontological framework in which these concepts are related to the spatial and material structure of the world. Since habitats are essentially associated with lives and behaviour of animals and plants the framework incorporates an abstract model of an organism. This enables the spatial extensions of habitats to be characterised in terms of the possible ‘life trajectories’ of organisms constrained by the physical properties of a geographic region and by the biological requirements for their survival.

Generally, part-whole relations are modeled using fragments of first-order logic(FOL) and difficulties arise when meta-reasoning is done over their properties, leading to reason outside the logic. Alternatively, classical languages for ontological reasoning such as Description Logics + Logic Programming lack of expressive formal foundations resulting in ambiguous interpretations of relations. Moreover, they show some difficulties to prove that a given meta property is logically correct. In order to address these problems, we suggest a formal framework using a dependent (higher-order) type theory such as those used in program checking and theorem provers (e.g., Coq). All properties of part-whole relations are formalized through abstract constructs called parameterized specifications (p-specifications). We detail their content and explain how they are suitable to build an ontology of formal properties that can be further used for reasoning about higher-order properties.

In the context of developing formal theories of commonsense psychology, or how peole think they think, we have developed a formal theory of goals. In it we explicate and axiomatize, among others, the goal-related notions of trying, success, failure, functionality, intactness, and importance.

Considerations regarding predication in ordinary language as well as the ontology of relations suggest a refinement of the Ontological Square, a conceptual scheme used in many foundational ontologies and which consists of particular substrates as well as their types on the one hand and particular attributes as well as their types on the other hand. First, the distinction between particulars and universals turns out to be one of degree, since particulars are merely the least elements in the subsumption hierarchy. Second, relations may be analysed in terms of roles as ways of participating in events. In consequence, the Logic of the Ontological Square proposed in [1,2] has to be revised accordingly.

Taking for granted an ontological standpoint independent of any empirical or epistemological perspective, philosophical theories of properties are actually quite rarely adopted in the knowledge representation community. The theory of qualities introduced in the DOLCE-CORE ontology [4] allows for representing different viewpoints on the world in a unified framework but it does not refer to any process used in empirical or epistemic investigations. In this paper, I will found the theory of qualities on the measurement theory proposed in [8]. In this new perspective, the stability of properties is not assumed a priori but it is founded on the stability of specific (physical) objects: the measurement systems and standards.

In a series of publications, we have employed ontological theories and principles used to evaluate and improve the quality of conceptual modeling grammars and models. In this article, we continue this work by providing an ontological interpretation and sound modeling guidelines for a traditionally neglected notion in the conceptual modeling literature, namely, the representation of types whose instances are quantities (amounts of matter, masses). Here, we analyze different alternatives for the adequate representation of quantities as well as their parts in conceptual models. Moreover, we advance a number of metamodeling constraints that can be incorporated in a UML 2.0 metamodel extension, thus, allowing for the suitable representation of these notions.

Affordances elude ontology. They have been recognized to play a role in categorization, especially of artifacts, but also of natural features. Yet, attempts to ontologize them face problems ranging from their presumed subjective nature to the fact that they involve potential actions, not objects or properties. We take a fresh look at the ontology of affordances, based on a simple insight: affordances are perceived by agents and may lead to actions, just like qualities are perceived and may lead to observations. We understand perception as a process invoking a quale in an agent. This quale can then be expressed as an action, if it stems from an affordance, or as an observation, if it stems from an other quality. Thus, we see affordances as qualities of the environment, perceived and potentially expressed by agents. We extend our recently proposed ontology of observations to include affordances and show how the parallel between observations (producing values) and affordances (producing actions) provides a simple and powerful ontological account of both.

The principle challenge for information semantics lies in the degrees of freedom to interpret symbols in terms of thoughts and experiences which leads to incompatible views on the world. Consequently, incompatible information ontologies and interpretations of the described data will remain. Even though there is usually a common experiential ground, it stays often unknown to users of semantically annotated data. This symbol grounding problem is a bottleneck of information semantics, which remains largely unsolved in ontological practice. In this paper, we suggest – in the spirit of Jeremy Bentham – to introduce formal primitives which are directly grounded in inter-subjective experience, and which serve to expose and construct complex qualities in information ontologies.

Current optimization techniques for answering queries over Semantic Web data use realization to precalculate the individuals associated with every concept in the given ontology. However, this technique does not take into account the type of queries, written for example in nRQL or SPARQL-DL, that will arrive at the system. In this paper we propose how this additional knowledge can be used to create query-specific indices. We include experimental results that show how our approach can be used to improve the performance of the Pellet query engine for the popular LUBM benchmark.

Our long-term goal is to build a query answering system that can answer questions on a wide variety of topics and explain the answers. In such a situation, a designer faces the challenge of how to specify the KR&R requirements that are needed to answer questions. In this paper, we introduce a categorization of KR&R methods, and apply it to specifying the requirements for answering questions in six different domains: Physics, Chemistry, Biology, Environmental Science, Microeconomics, and U.S. Government & Politics. Drawing from the corpus of about 500 questions that we analyzed, we consider an example question in each domain and show the analytical process that we used to derive the requirements in terms of the KR&R categorization. We analyze the effectiveness of the current KR&R categorization, and identify directions for future work suggesting how this categorization can be further evolved by community participation.

Finding appropriate caries treatments is of paramount importance in dental decision making. That is, finding restorative treatment alternatives predicated on the dental disease and findings are advantageous and gainful in dental restorative decision making. The most immediate problem in clinical decision support systems in dentistry is to capture a doctor's clinical knowledge of treatments. This study is to specify the inter-relations among disease, anatomy, and treatment for restorative treatment decision support, and to conceptualize restorative treatment. As an explanatory example, we expound the developmental process of our ontology, and the formal approach used, for caries treatment.

Neurosciences are progressively moving into a mass computational intensive era with the fusion of numerous large heterogeneous data sets from cellular to system level. To process and share this mass of information in a consistent and computational amenable form, computerized techniques – among them, ontologies – are currently designed to store, analyze and access this information. Recently, we proposed a multi-layered and multi-components ontology to deal with MR images and regions-of-interest that can be represented onto the images. In the present paper, we extend our initial ontology by adding a core ontology of subject data acquisition instruments modeling neuroclinical, neuropsychological and neurobehavioral tests used for neurological, behavioral and cognitive skills assessment. This ontology deals with instruments per se as specific artefacts, their variables and measured scores, and actions performed using instruments. In the paper, we underline the major aspects of our approach and emphasize its potential interest as a semantic reference for various neuroscience applications.

We introduce a process-centric ontological approach to relate observed properties to geo-processes that influence those observations. These relations are used to handle semantic heterogeneities that impede the integration of geo-sensor data. Our approach comprises a process-centric hydrology ontology and its alignment with the DOLCE foundational ontology. We describe how DOLCE assists in classifying two hydrological processes and their participating entities. Our preliminary results indicate that the process-centric ontological approach can be used to resolve process and property naming ambiguity within a domain. We consider this as a first step toward a process-centric semantic integration of geo-sensor data.

Software measurement is a relatively young discipline. As a consequence, it is not well defined yet, making the terminology used diverse. In order to establish a basic conceptualization regarding this domain, in this paper we present a Software Measur ement Ontology, developed grounded in the Unified Foundational Ontology.

By axiomatizing the semantics of Petri nets as a first-order (mostly) formal ontology called SCOPE, we propose in this paper a framework for the analysis of the structural and dynamical properties of Petri nets. More precisely, SCOPE is built as a Basic Action Theory in Reiter's version of situation calculus. In addition, we show the satisfiability of SCOPE, by virtue of the Relative Satisfiability Theorem. Fundamental structural and dynamical properties of Petri nets described in SCOPE are also presented, with two example uses of SCOPE given.