Abstract. We think of Match as an operator which takes two graph-like structures and produces a mapping between semantically related nodes. We concentrate on classifications with tree structures. In semantic matching, correspondences are discovered by translating the natural language labels of nodes into propositional formulas, and by codifying matching into a propositional unsatisfiability problem. We distinguish between problems with conjunctive formulas and problems with disjunctive formulas, and present various optimizations. For instance, we propose a linear time algorithm which solves the first class of problems. According to the tests we have done so far, the optimizations substantially improve the time performance of the system. 1.

Abstract. While current approaches to ontology mapping produce good results by mainly relying on label and structure based similarity measures, there are several cases in which they fail to discover important mappings. In this paper we describe a novel approach to ontology mapping, which is able to avoid this limitation by using background knowledge. Existing approaches relying on background knowledge typically have one or both of two key limitations: 1) they rely on a manually selected reference ontology; 2) they suffer from the noise introduced by the use of semi-structured sources, such as text corpora. Our technique circumvents these limitations by exploiting the increasing amount of semantic resources available online. As a result, there is no need either for a manually selected reference ontology (the relevant ontologies are dynamically selected from an online ontology repository), or for transforming background knowledge in an ontological form. The promising results from experiments on two real life thesauri indicate both that our approach has a high precision and also that it can find mappings, which are typically missed by existing approaches.

We view match as an operator that takes two graph-like structures (e.g., classifications, XML schemas) and produces a mapping between the nodes of these graphs that correspond semantically to each other. Semantic matching is based on two ideas: (i) we discover mappings by computing semantic relations (e.g., equivalence, more general); (ii) we determine semantic relations by analyzing the meaning (concepts, not labels) which is codified in the elements and the structures of schemas. In this paper we present basic and optimized algorithms for semantic matching, and we discuss their implementation within the S-Match system. We evaluate S-Match against three state of the art matching systems, thereby justifying empirically the strength of our approach.

Abstract. A new paradigm in Semantic Web research focuses on the development of a new generation of knowledge-based problem solvers, which can exploit the massive amounts of formally specified information available on the Web, to produce novel intelligent functionalities. An important example of this paradigm can be found in the area of Ontology Matching, where new algorithms, which derive mappings from an exploration of multiple and heterogeneous online ontologies, have been proposed. While these algorithms exhibit very good performance, they rely on merely syntactical techniques to anchor the terms to be matched to those found on the Semantic Web. As a result, their precision can be affected by ambiguous words. In this paper, we aim to solve these problems by introducing techniques from Word Sense Disambiguation, which validate the mappings by exploring the semantics of the ontological terms involved in the matching process. Specifically we discuss how two techniques, which exploit the ontological context of the matched and anchor terms, and the information provided by WordNet, can be used to filter out mappings resulting from the incorrect anchoring of ambiguous terms. Our experiments show that each of the proposed disambiguation techniques, and even more their combination, can lead to an important increase in precision, without having too negative an impact on recall.

Abstract. Classifications have been used for centuries with the goal of cataloguing and searching large sets of objects. In the early days it was mainly books; lately it has also become Web pages, pictures and any kind of digital resources. Classifications describe their contents using natural language labels, an approach which has proved very effective in manual classification. However natural language labels show their limitations when one tries to automate the process, as they make it very hard to reason about classifications and their contents. In this paper we introduce the novel notion of Formal Classification, as a graph structure where labels are written in a propositional concept language. Formal Classifications turn out to be some form of lightweight ontologies. This, in turn, allows us to reason about them, to associate to each node a normal form formula which univocally describes its contents, and to reduce document classification and query answering to reasoning about subsumption. 1

“Ontology matching” is the process of finding correspondences between entities belonging to different ontologies. This paper describes a set of algorithms that exploit upper ontologies as semantic bridges in the ontology matching process and presents a systematic analysis of the relationships among features of matched ontologies (number of simple and composite concepts, stems, concepts at the top level, common English suffixes and prefixes, ontology depth), matching algorithms, used upper ontologies, and experiment results. This analysis allowed us to state under which circumstances the exploitation of upper ontologies gives significant advantages with respect to traditional approaches that do no use them. We run experiments with SUMO-OWL (a restricted version of SUMO), OpenCyc and DOLCE. The experiments demonstrate that when our “structural matching method via upper ontology” uses an upper ontology large enough (OpenCyc, SUMO-OWL), the recall is significantly improved while preserving the precision obtained without upper ontologies. Instead, our “non structural matching method” via OpenCyc and SUMO-OWL improves the precision and maintains the recall. The “mixed method” that combines the results of structural alignment without using upper ontologies and structural alignment via upper ontologies improves the recall and maintains the F-measure independently of the used upper ontology.

Abstract. The paper represents a first attempt to formalize the getspecific document classification algorithm and to fully automate it through reasoning in a propositional concept language without requiring user involvement or a training dataset. We follow a knowledge-centric approach and convert a natural language hierarchical classification into a formal classification, where the labels are defined in the concept language. This allows us to encode the get-specific algorithm as a problem in the concept language. The reported experimental results provide evidence of practical applicability of the proposed approach. 1

Ontology matching aims to find semantic correspondences between a pair of input ontologies. A number of matching techniques have been proposed recently, however, we may benefit more from a combination of such techniques as opposed to just a single method. This is more appropriate, but very often the user has no prior knowledge about which technique is more suitable for the task at hand. However, it remains a labour intensive and expensive task to perform. Further, the complexity of the matching process as well as the quality of the result is affected by the choice of the applied matching techniques. We study this problem and propose a framework for finding suitable matches. A main feature of this is that it improves the structure matching techniques and the end result accordingly. We have developed a running prototype of the proposed framework and conducted experiments to compare our results with existing techniques. While being comparable in efficiency, the experimental results indicate our proposed technique produces better quality matches.

Abstract. Matching has been recognized as a plausible solution for the semantic heterogeneity problem in many traditional applications, such as schema integration, ontology integration, data warehouses, data integration, and so on. Recently, there have emerged a line of new applications characterized by their dynamics, such as peer-to-peer systems, agents, web-services. In this deliverable we extend the notion of ontology matching, as it has been understood in traditional applications, to dynamic ontology matching. In particular, we examine real-world scenarios and collect the requirements they pose towards a plausible solution. We consider five general matching directions which we believe can appropriately address those requirements. These are: (i) approximate and partial ontology matching, (ii) interactive ontology matching, (iii) continuous ”design-time ” ontology matching, (iv) community-driven ontology matching and

Providing efficient ontology matching algorithms is one of the means for pursuing semantic interoperability. In this paper we discuss an algorithm that exploits natural language processing techniques for matching ontologies and that post-processes the obtained alignment in order to find semantic inconsistencies. The algorithm has been entirely implemented in Prolog, whose usefulness was mainly evident in the post-processing phase. A careful analysis of the recent stateof-the art witnesses the originality of our matching algorithm which is based on the “Adapted Lesk Algorithm” for word sense disambiguation. The experiments we carried out, although in their early stages, are encouraging.