Catálogo de publicaciones - libros

Currently, the introduction of an appropriate rule representation layer for the semantic web stack is discussed. However, with the inclusion of rule-based knowledge new verification issues for rule-augmented ontologies arise.

In this paper we investigate the detection of anomalies as an important subtask of verification. We extend and revise existing approaches for the syntactic verification of ontologies with respect to the existence of rules, and we introduce new anomalies considering the understandability and maintainability of such ontologies.

Robust mechanisms for ontology selection are crucial for the evolving Semantic Web characterized by rapidly increasing numbers of online ontologies and by applications that automatically use the associated metadata. However, existing selection techniques have primarily been designed in the context of human mediated tasks and fall short of supporting automatic knowledge reuse. We address this gap by proposing a selection algorithm that takes into account 1) the needs of two applications that explore large scale, distributed markup and 2) some properties of online ontology repositories. We conclude that the ambitious context of automatic knowledge reuse imposes several challenging requirements on selection.

The call for a “focus on content” in ontology research by Nicola Guarino and Mark Musen in their launching statement of the journal Applied Ontology has quite some implications and ramifications. We reflectively discuss ontology engineering as a scientific discipline, and we put this into the wider perspective of debates in other fields. We claim and argue that ontology is a new scientific method for theory formation. This positioning allows for stronger concepts and techniques for theoretical, empirical and practical validation that in our view are now needed in the field. A prerequisite for this is an emphasis on ontology as a (domain) content oriented concept, rather than as primarily a computer representation notion. We propose that taking domain theories and the associated substantive or content reference of ontologies really seriously as first-class citizens, will actually increase the contribution of ontology engineering to the development of scientific method in general. Next, ontologies should develop from the current static representations of relatively stable domain content into actionable theories-in-use, and a possible way forward is to build in capabilities for dynamic self-organization of ontologies as service-oriented knowledge utilities that can be delivered over the Web.

The contribution of this paper is to provide a methodology for automatic ontology enrichment and for document annotation with the concepts and properties of a domain core ontology. Natural language definitions of available glossaries in a given domain are parsed and converted into formal (OWL) definitions, compliant with the core ontology property specifications.

To evaluate the methodology, we annotated and formalized a relevant fragment of the AAT glossary of art and architecture, using a subset of 10 properties defined in the CRM CIDOC cultural heritage core ontology, a recent W3C standard.

The acquisition of explicit semantics is still a research challenge. Approaches for the extraction of semantics focus mostly on learning hierarchical hypernym-hyponym relations. The extraction of co-hyponym and co-meronym sibling semantics is performed to a much lesser extent, though they are not less important in ontology engineering.

In this paper we will describe and evaluate the XTREEM-SG (Xhtml TREE Mining – for Sibling Groups) approach on finding sibling semantics from semi-structured Web documents. XTREEM takes advantage of the added value of mark-up, available in web content, for grouping text siblings. We will show that this grouping is semantically meaningful. The XTREEM-SG approach has the advantage that it is domain and language independent; it does not rely on background knowledge, NLP software or training.

In this paper we apply the XTREEM-SG approach and evaluate against the reference semantics from two golden standard ontologies. We investigate how variations on input, parameters and reference influence the obtained results on structuring a closed vocabulary on sibling relations. Earlier methods that evaluate sibling relations against a golden standard report a 14.18% F-measure value. Our method improves this number into 21.47%.

Pattern-based approaches for knowledge identification in texts assume that linguistic regularities always characterise the same kind of knowledge, such as semantic relations. We report the experimental evaluation of a large set of patterns using an ontology enrichment tool: . Results underline the strong corpus influence on the patterns efficiency and on their meaning. This influence confirms two of the hypotheses that motivated to define as a support used in a supervised process: (1) patterns and relations must be adapted to each project; (2) human interpretation is required to decide how to report in the ontology the pieces of knowledge identified with patterns.

In the context of the Semantic Web, many ontology-related operations, e.g. ontology ranking, segmentation, alignment, articulation, reuse, evaluation, can reduced to one fundamental operation: computing the similarity and/or dissimilarity among ontological entities, and in some cases among ontologies themselves. In this paper, we review standard metrics for computing distance measures and we propose a series of semantic metrics. We give a formal account of semantic metrics drawn from a variety of research disciplines, and enrich them with semantics based on standard Description Logic constructs. We argue that concept-based metrics can be aggregated to produce numeric distances at ontology-level and we speculate on the usability of our ideas in potential areas.

Existing ontology matching algorithms use a combination of lexical and structural correspondence between source and target ontologies. We present a realistic case-study where both types of overlap are low: matching two unstructured lists of vocabulary used to describe patients at Intensive Care Units in two different hospitals. We show that indeed existing matchers fail on our data. We then discuss the use of background knowledge in ontology matching problems. In particular, we discuss the case where the source and the target ontology are of poor semantics, such as flat lists, and where the background knowledge is of rich semantics, providing extensive descriptions of the properties of the concepts involved. We evaluate our results against a Gold Standard set of matches that we obtained from human experts.

In today’s world there is a need for knowledge infrastructures that can support several autonomous knowledge bases all using different ontologies and constantly adapting these to their changing local needs. Moreover, these different knowledge bases are expressing their unique points of view and constitute different local contexts. At the same time interoperability is needed in order to connect these semantically dispersed knowledge bases, and we formalized this as a type of consistency. Both these aspects are included in our definition of semantic autonomy. We present a layered framework that shows how to design a scalable system having this property. In our approach both ontology and mapping evolution take place, at the same time as the whole system is kept coherent using lightweight methods for maintaining global consistency. However, in order to achieve this several restrictions are necessary and the logical language used by the individual ontologies is kept simple. Finally, we present some experimental results that demonstrate the scalability of our approach.

Ontology evolution becomes extremely important with the tremendous application of ontology. Ontology’s size and complexity change a lot during its evolution. Thus it’s important for ontology developers to analyze and try to control ontology’s complexity to ensure the ontology is useable. In this paper, an evaluation method for analyzing ontology complexity is suggested. First, we sort all the concepts of an ontology according to their (a definition we will give below), then by using a well-defined metrics suite which mainly examines the concepts and their hierarchy and the quantity, ratio of concepts and relationships, we analyze the evolution and distribution of ontology complexity. In the study, we analyzed different versions of GO ontology by using our evaluation method and found it works well. The results indicate that the majority of GO’s complexity is distributed on the minority of GO’s concepts, which we call “important concepts” and the time when GO’s complexity changed greatly is also the time when its “important concepts” changed greatly.