This paper proposes a novel approach to negotiation, in which the negotiation protocol to adopt is not coded within the agents but it is expressed in terms of a common shared ontology that is shared by the agents in order to participate to a negotiation session. The negotiation ontology is defined in a way general enough to support a wide variety of market mechanisms, thus being particularly suitable for exible applications such as electronic commerce. The paper describes the negotiation ontology and provides a walkthrough example describing how the proposed approach could be used to model protocols for auctions.

This paper presents a novel approach to automated negotiation that is particularly suitable to open environments, such as the Internet. In this approach agents can negotiate in any type of marketplace regardless of the negotiation mechanism that these adopt. In order to support a wide variety of negotiation mechanisms, protocols are no longer hard-coded in the agents participating to negotiations, but are now expressed in terms of a shared ontology, thus making this approach particularly suitable for flexible applications such as electronic commerce.

Abstract: Metadata is used to describe documents and applications, improving information seeking and retrieval and its understanding and use. Metadata can be expressed in a wide variety of vocabularies and languages, and can be created and maintained with a variety of tools. Ontology based annotation refers to the process of creating metadata using ontologies as their vocabularies. We present similarities and differences with respect to other approaches for metadata creation, and describe languages and tools that can be used to implement these annotations.

: This paper introduces "Wilbur", an RDF and DAML toolkit implemented in Common Lisp. Wilbur exposes the RDF data model as a frame-based representation system; an object-oriented view of frames is adopted, and RDF data is integrated with the host language by addressing issues of input/output, data structure compatibility, and error signaling. Through seamless integration we have achieved a programming system well suited for building "Semantic Web" applications. 1.

This article will discuss issues of interoperability of Web Services. Semantic Web technologies, when applied to current Web Service architectures, will enable true automation and interoperation, and will “futureproof” systems unlike a priori standardization approaches. We will extend the Semantic Web approach to Web Services to Ubiquitous Computing: By abstracting device functionality via software agents, and using Semantic Web technologies to describe agent services and capabilities, we can achieve a rich and deep form of service discovery. By discovering partially matching services, and piecing these together into “virtual value chains”, we can automatically form device coalitions which operate within a dynamically changing environment. 1

Abstract not found

Abstract: Ontology and schema matching are well established techniques, which have been applied in various integration scenarios, e.g., web service composition and database integration. Consequently, matching tools enabling automatic matching of various kinds of schemas are available. In the field of model-driven engineering, in contrast to schema and ontology integration, the integration of modeling languages relies on manual tasks such as writing model transformation code, which is tedious and error-prone. Therefore, we propose the application of ontology and schema matching techniques for automatically exploring semantic correspondences between metamodels, which are currently the modeling language definitions of choice. The main focus of this paper is on reporting preliminary results and lessons learned by evaluating currently available ontology matching tools for their metamodel matching potential. 1

Abstract There is a prominent trend to augment and improve the formality of biomedical ontologies. For example, this is shown by the current effort on adding description logic axioms, such as disjointness. One of the key ontology applications that can take advantage of this effort is the conceptual (functional) similarity measurement. The presence of description logic axioms in biomedical ontologies make the current structural or extensional approaches weaker and further away from providing sound semantics-based similarity measures. Although beneficial in small ontologies, the exploration of description logic axioms by semantics-based similarity measures is computational expensive. This limitation is critical for biomedical ontologies that normally contain thousands of concepts. Thus in the process of gaining their rightful place, biomedical functional similarity measures have to take the journey of finding how this rich and powerful knowledge can be fully explored while keeping feasible computational costs. This manuscript aims at promoting and guiding the development of compelling tools that deliver what the biomedical community will require in a near future: a next-generation of biomedical similarity measures that efficiently and fully explore the semantics present in biomedical ontologies. Keywords: Ontologies; Semantic Similarity; Functional Similarity; Description Logics Electronic version of an article published as Journal of Bioinformatics and Computational Biology, 11, 5, 1371001, 2013, dx.doi.org/10.1142 We have a natural tendency to compare biomedical entities based on their "looks", i.e. based on the digital representation of their primary structure. For example, sequence similarity measurement, like BLAST Divergence of function with sequence conservation is an exception rather than a rule, so in general sequence similarity remains as a reliable technique to determine the functional similarity of proteins. However, this gap between structural and functional similarity creates an opportunity to develop similarity measures that can juxtapose, combine and/or complement structural similarity measures with a degree of shared functional characteristics. For example, when searching for proteins with an oxidoreductase activity, we may also be interested in proteins with similar activities, such as monooxygenase activity, independently of their structural similarity. This analysis of similar activities has become computationally possible due to the prevailing usage of ontologies to functionally characterize biomedical entities. Ontologies and Similarity Measures Ontologies can be loosely defined as "a vocabulary of terms and some specification of their meaning" Recently, conceptual similarity was defined as a function that, given two ontology terms or two sets of terms annotating two entities, returns a numerical value reflecting the closeness in meaning between them Terminological approaches focus on the names of the classes. For example, the term Cats is a morphological derivation of Cat, and thus classes bearing such names are likely equivalent. These approaches can be complemented with thesauri and dictionaries, to explore lexical relations such as synonymy. Structural approaches explore the structure of the classes, i.e. their relations to other classes. The sub-class and super-class relations provide a taxonomic backbone that can be explored using graph matching techniques. Ontology-specific relations, along with their properties (or facets), such as domain, range and cardinality, can also be explored. Extensional approaches can only be applied when there is a large set of instances. The intuition behind these approaches is that the more instances two classes share, the more likely they have a high similarity. In case there are no shared instances between ontologies, distance metrics between individuals can be computed. Semantics-based approaches are sensitive to the semantics of the logical formalism in which the ontologies are formalized, and are thus enabled to resort to inference techniques. The goal of this type of measures is to fully explore the available description logic axioms. Note that ontology similarity, or global similarity Computational complexity: semantics-based measures usually require some sort of deduction, which usually requires exponential time, even in less expressive formalisms. 3 Incomplete knowledge: classes in an ontology do not need to be, and indeed rarely are, completely defined. The incomplete knowledge problem can only be effectively addressed if we improve the quality of the process of ontology development. For example, the omission of simple and obvious relationships in the January 2003 release of SNOMED-CT R , did not allow the inference of "uterus" as part-of "female genital tract", nor the inference that "uterus" as a role in "pregnancy" The computational complexity problem is due to the large size of biomedical ontologies. They usually contain thousands of concepts that are used to annotate an even larger set of entities. For example, studies show that consistency checking using description logics is EXPTIME-Complete Biomedical Ontologies Etymologiae was one of the first attempts to systematize medicine knowledge 4 As the number of people developing and using ontologies continues to grow, their size will rise too. However, to maintain the quality of ontologies we have now to pay special attention to important features of ontology languages that were neglected in their initial specification, such as the lack of description logical axioms in GO [54] and SNOMED-CT R Current Biomedical Functional Similarity A successful application of terminological approaches to biomedical ontologies has been to find the most similar ontological concepts to the terms recognized in biomedical literature by text mining methods. For example, in BioCreAtIvE 2004 [23] a measure based on the textual descriptors of GO concepts was used to resolve references to GO in biomedical literature In Molecular Biology, as in other biomedical areas, mainstream methods for functional characterization of genes and proteins are based on ontological annotation, for example using GO. This enabled the successful development and application of conceptual similarity measures based on extensional approaches. These measures compare two proteins according to the amount of ontological information their annotations share. The shared ontological information can be inferred from the most informative common ancestor of the annotated concepts, or from all the disjunctive common ancestors Inspired on Tversky's contrast model Nowadays, many conceptual similarity measures use structural or extensional approaches and are usually designated as semantic similarity measures

Abstract: Metamodels are abstractions that are used to specify characteristics of models. Such metamodels are gen-erally included in specifications or framework descriptions. A metamodel is for instance used to inform the generation of enterprise architecture content in the Open Group’s TOGAF 9 Content Metamodel description. However. the description of metamodels is usually done in an ad-hoc manner with customised languages and this often results in ambiguities and inconsistencies. We are concerned with the question of how the quality of metamodel descriptions, specifically within the enterprise architecture domain, could be enhanced. There-fore we investigated whether formal ontology technologies could be used to enhance metamodel construction, specification and design. For this research, we constructed a formal ontology for the TOGAF 9 Content Meta-model, and in the process, gained valuable insight into metamodel quality. In particular, the current TOGAF 9 Content Metamodel contains ambiguities and inconsistencies, which could be eliminated using ontology technologies. In this paper we argue for the integration of formal ontologies and ontology technologies as tools into meta-model construction and specification. Ontologies allow for the construction of complex conceptual models, but more significant, ontologies can assist an architect by depicting all the consequences of a model, allowing for more precise and complete artifacts within enterprise architectures, and because these models use standardized languages, they should promote integration and interoperability. 1

ii In this thesis, we will present methods for introducing ontologies in information retrieval. The main hypothesis is that the inclusion of conceptual knowledge such as ontologies in the information retrieval process can contribute to the solution of major problems currently found in information retrieval. This utilization of ontologies has a number of challenges. Our focus is on the use of similarity measures derived from the knowledge about relations between concepts in ontologies, the recognition of semantic information in texts and the mapping of this knowledge into the ontologies in use, as well as how to fuse together the ideas of ontological similarity and ontological indexing into a realistic information retrieval scenario. To achieve the recognition of semantic knowledge in a text, shallow natural language processing is used during indexing that reveals knowledge to the level of noun phrases. Furthermore, we briefly cover the identification of