Ebook: Modular Ontologies

It is widely accepted that modularity will help to solve many problems in the construction and management of ontological systems, and researchers are actively investigating techniques for modularization and parameters for module assessment.

This note aims to diverge attention from technical issues to what is the basic goal of modularity. Modularization is a technique that can be devoted to solve different types of problems, depending on the type of ontology one works with. We start from a classification of ontology research in three rough classes, and discuss what modularization should achieve from each perspective. This observation gives already some indications on what modules one should look for.

We conclude with the optimistic view that the modular approach can radically change the way we build foundational ontologies. This, however, requires a further study of what should count as a module. Something we are still far from understanding.

Modular Ontologies Modularity has been and continues to be one of the central research topics in ontology engineering. The number of ontologies available, as well as their size, is steadily increasing. There is a large variation in subject matter, level of specification and detail, intended purpose and application. Ontologies covering different domains are often developed in a distributed manner; contributions from different sources cover different parts of a single domain. Not only is it difficult to determine and define interrelations between such distributed ontologies, it is also challenging to reconcile ontologies which might be consistent on their own but jointly inconsistent. Further challenges include extracting the relevant parts of an ontology, re-combining independently developed ontologies in order to form new ones, determining the modular structure of an ontology for comprehension, and the use of ontology modules to facilitate incremental reasoning and version control. Still catching up with 40 years of related research in software engineering, ontological modularity is envisaged to allow mechanisms for easy and flexible reuse, generalisation, structuring, maintenance, collaboration, design patterns, and comprehension. Applied to ontology engineering, modularity is central not only to reducing the complexity of understanding ontologies, but also to maintaining, querying and reasoning over modules. Distinctions between modules can be drawn on the basis of structural, semantic, or functional aspects, which can also be applied to compositions of ontologies or to indicate links between ontologies. In particular, reuse and sharing of information and resources across ontologies depend on purpose-specific, logically versatile criteria. Such purposes include ‘tight’ logical integration of different ontologies (wholly or in part), ‘loose’ association and information exchange, the detection of overlapping parts, traversing through different ontologies, alignment of vocabularies, module extraction possibly respecting privacy concerns and hiding of information, etc. Another important aspect of modularity in ontologies is the problem of evaluating the quality of single modules or of the achieved overall modularisation of an ontology. Again, such evaluations can be based on various (semantic or syntactic) criteria and employ a variety of statistical/heuristic or logical methods. Recent research on ontology modularity has produced substantial results and approaches towards foundations of modularity, techniques of modularisation and modular developments, distributed and incremental reasoning, as well as the use of modules in different application scenarios, providing a foundation for further research and development. Since the beginning of the WoMO workshop series, there has been growing interest in the modularisation of ontologies, modular development of ontologies, and information exchange across different modular ontologies. In real life, however, integration problems are still mostly tackled in an ad-hoc manner, with no clear notion of what to expect from the resulting ontological structure. Those methods are not always efficient, and they often lead to unintended consequences, even if the individual ontologies to be integrated are widely tested and understood. Topics covered by WoMO include, but are not limited to:

What is Modularity? - Kinds of modules and their properties - Modules vs. contexts - Design patterns - Granularity of representation

Logical/Foundational Studies - Conservativity and syntactic approximations for modules - Modular ontology languages - Reconciling inconsistencies across modules - Formal structuring of modules - Heterogeneity

Algorithmic Approaches - Distributed reasoning - Modularisation and module extraction - (Selective) sharing and reusing, linking and importing - Hiding and privacy - Evaluation of modularisation approaches - Complexity of reasoning - Reasoners or implemented systems

Application Areas - Modularity in the Semantic Web - Life Sciences - Bio-Ontologies - Natural Language Processing - Ontologies of space and time - Ambient intelligence - Collaborative ontology development

The WoMO 2011 workshop follows a series of successful events that have been an excellent venue for practitioners and researchers to discuss latest work and current problems. It is intended to consolidate cutting-edge approaches that tackle the problem of ontological modularity and bring together researchers from different disciplines who study the problem of modularity in ontologies at a fundamental level, develop design tools for distributed ontology engineering, and apply modularity in different use cases and application scenarios. Previous editions of WoMO are listed below. The links refer to their homepages and proceedings.

WoMO 2006 The 1st workshop on modular ontologies, co-located with ISWC 2006, Athens, Georgia, USA. Invited speakers were Alex Borgida (Rutgers) and Frank Wolter (Liverpool). http://www.cild.iastate.edu/events/womo.html http://sunsite.informatik.rwth-aachen.de/Publications/CEUR-WS/Vol-232

WoMO 2007 The 2nd workshop, co-located with K-CAP 2007, Whistler BC, Canada. The invited speaker was Ken Barker (Texas at Austin). http://webrum.uni-mannheim.de/math/lski/WoMO07 http://sunsite.informatik.rwth-aachen.de/Publications/CEUR-WS/Vol-315

WoRM 2008 The 3rd workshop in the series, co-located with ESWC 2008, Tenerife, Spain, entitled ‘Ontologies: Reasoning and Modularity’ had a special emphasis on reasoning methods. http://dkm.fbk.eu/worm08 http://sunsite.informatik.rwth-aachen.de/Publications/CEUR-WS/Vol-348

WoMO 2010 The 4th workshop in the series, co-located with FOIS 2010, Toronto, Canada. Invited speakers were Simon Colton (London) and Marco Schorlemmer (Barcelona). http://www.informatik.uni-bremen.de/ okutz/womo4 http://www.booksonline.iospress.nl/Content/View.aspx?piid=16268

This keynote addresses modularity issues focusing on biomedical ontologies, with a particular consideration of SNOMED CT. Emphasis is put on approaches that carve out high-coverage subsets addressing specialty-specific documentation needs.

The OWL 2 QL profile of OWL, based on the DL family of description logics, is emerging as a major language for developing new ontologies and approximating existing ones. Its main application is ontology-based data access, where ontologies are used to provide background knowledge for answering queries over data. We give a survey of recent results on the computational complexity of checking query inseparability for OWL 2 QL ontologies and analyse the impact of various OWL 2 QL constructs in the context of query inseparability. We also discuss practical query inseparability checking and minimal module extraction algorithms, as well as experimental results.

Reusing existing Semantic Web ontologies is necessary to avoid heterogeneity as well as redundant modeling efforts, because ontology engineering is a time-consuming and cost-intensive task. In order to decide whether a candidate ontology comprises the right concepts, an analysis process is necessary to understand the conceptual model of the ontology. Driven by the idea that concept grouping simplifies understanding the content of an ontology we investigate the applicability of community algorithms from the field of Social Network Analysis on the graph structure of RDF/XML based OWL documents to identify concept groups. In this paper, we present our experiments with different community algorithms on popular ontologies and compare our results with manually created concept groups.

Extracting a subset of a given ontology that captures all the ontology's knowledge about a specified set of terms is a well-understood task. This task can be based, for instance, on locality-based modules. However, a single module does not allow us to understand neither topicality, connectedness, structure, or superfluous parts of an ontology, nor agreement between actual and intended modeling.

The strong logical properties of locality-based modules suggest that the family of all such modules of an ontology can support comprehension of the ontology as a whole. However, extracting that family is not feasible, since the number of locality-based modules of an ontology can be exponential w.r.t. its size.

In this paper we report on a new approach that enables us to efficiently extract a polynomial representation of the family of all locality-based modules of an ontology. We also describe the fundamental algorithm to pursue this task, and report on experiments carried out and results obtained.

In the life sciences researchers are working with large amount of data from different domains that frequently overlap. Overlapping information can be utilized at the moment the domains are integrated. A typical case is the drug discovery process in which the information from different domains, e.g. diseases, proteins, pathways, drugs, etc. need to be integrated in order to connect a disease with genes, pathways and find a potential chemical compound that can be active as a drug. However, information from different domains is often available in different ontologies. In order to combine these data an ontology integration approach is required.

In this paper we demonstrate an approach in which a new integrated ontology is created from modules that are extracted from different ontologies. Module extraction is based on well defined notions of modularity, locality and conservative extensions. The signature of the modules is based on symbols of the user interest. Subsequently, the mappings between the similar concepts are generated. Finally, on the basis of these mappings we integrate modules in one ontology.

Within knowledge representation, ontologies are logical theories that support software integration and decision support systems. Ontology verification is concerned with the relationship between the intended structures for an ontology and the models of the axiomatization of the ontology. To verify a particular ontology, we ideally characterize all the models of the ontology up to elementary equivalence and prove that these models are equivalent to the intended structures for the ontology. In this paper, we investigate the use of automated theorem provers and model finders to assist in the interactive verification of first-order ontologies. We identify the reasoning tasks that are associated with different aspects of ontology verification and discuss challenges for the application of automated reasoning systems to support these tasks.

Bio-ontologies such as the Gene Ontology and ChEBI are characterized by large sizes and relatively low expressivity. However, ongoing efforts aim to increase the formalisation of these ontologies by adding full definitions (equivalent classes). This increase in complexity results in a decrease of performance for standard reasoning tasks. In this paper, we explore the contribution which modularization can play in the evolution of bio-ontologies. In particular, we focus on ChEBI, the ontology of chemical entities of biological interest. ChEBI consists of around 25,000 classes, organised into a structure-based chemical classification and enriched with a role-based classification of their biologically properties. Ontology modularization – partitioning large ontologies into smaller, more manageable chunks – provides the only feasible mechanism for sustainably maintaining the large-scale and ever-growing ontologies in the biomedical domain. We evaluate available ontology partitioning tools.

Ontologies can specify spatial information based on different aspects, for different purposes, with different formalization details, with different granularity levels, and thus from different perspectives. Spatial systems that formally describe spatial information often have to take into account this diversity in spatial representations. We propose that such a system's ontological representation should preserve the spatial diversity by making the different perspectives explicit. We use modularly designed spatial ontologies for reflecting different spatial perspectives and analyze which existing modularity techniques are relevant for combining the different modules.

The standard semantic web languages and reasoning tools do not explicitly take into account the contextual dimension of knowledge, i.e., the fact that a certain statement (RDF triple) is not universally true, but true only in certain circumstances. Such contextual information, which includes for instance, the time interval, the spatial region, or the sub-domain in which a certain statement holds, are of foremost importance to determine correct answers for a user query. Rather than proposing a new standard, in this work, we introduce a framework for contextual knowledge representation and reasoning based on current RDF(S) standards, and we provide a sound and complete set of forward inference rules that support reasoning in and across contexts. The approach proposed in this paper has been implemented in a prototype, which will be briefly described.

We present an overview of the landscape of ontology languages, mostly pertaining to the first-order paradigm. In particular, we present a uniform formalisation of these languages based on the institution theoretical framework, allowing a systematic treatment and analysis of the translational relationships between the various languages and a general analysis of properties of such translations. We also discuss the importance of language translation from the point of view of ontological modularity and logical pluralism, and for the borrowing of tools and reasoners between languages.

Ontology repositories stand to benefit through the connecting of stored ontologies via the meta-theoretic relationships they share. Creating this repository framework facilitates ontology reuse and design by allowing users to integrate different ontologies related in this manner. In this paper we construct such a repository by utilizing an automated theorem prover to identify and verify the relationships between three different ontologies of time intervals (two introduced by Hayes in his Catalog of Temporal Theories and one by van Benthem introduced in A Logic of Time). We identify the translation axioms and provide an account of the relationships found between ontologies.