The Semantic Web - ISWC 2008

The process of authoring ontologies requires the active involvement of domain experts who should lead the process, as well as providing the relevant conceptual knowledge. However, most domain experts lack knowledge modelling skills and find it hard to follow logical notations in OWL. This paper presents

ROO

, a tool that facilitates domain experts’ definition of ontologies in OWL by allowing them to author the ontology in a controlled natural language called

Rabbit

.

ROO

guides

users through the ontology construction process by following a methodology geared towards domain experts’ involvement in ontology authoring, and exploiting intelligent user interfaces techniques. An evaluation study has been conducted comparing

ROO

against another popular ontology authoring tool. Participants were asked to create ontologies based on hydrology and environment modelling scenarios related to real tasks at the mapping agency of Great Britain. The study is discussed, focusing on the usability and usefulness of the tool, and the quality of the resultant ontologies.

Ontologies are becoming so large in their coverage that no single person or a small group of people can develop them effectively and ontology development becomes a community-based enterprise. In this paper, we discuss requirements for supporting collaborative ontology development and present Collaborative Protégé—a tool that supports many of these requirements, such as discussions integrated with ontology-editing process, chats, and annotations of changes and ontology components. We have evaluated Collaborative Protégé in the context of ontology development in an ongoing large-scale biomedical project that actively uses ontologies at the VA Palo Alto Healthcare System. Users have found the new tool effective as an environment for carrying out discussions and for recording references for the information sources and design rationale.

More and more ontologies have been published and used widely on the web. In order to make good use of an ontology, especially a new and complex ontology, we need methods to help understand it first. Identifying potentially important concepts and relations in an ontology is an intuitive but challenging method. In this paper, we first define four features for potentially important concepts and relation from the ontological structural point of view. Then a simple yet effective Concept-And-Relation-Ranking (

CARRank

) algorithm is proposed to simultaneously rank the importance of concepts and relations. Different from the traditional ranking methods, the importance of concepts and the weights of relations reinforce one another in

CARRank

in an iterative manner. Such an iterative process is proved to be convergent both in principle and by experiments. Our experimental results show that

CARRank

has a similar convergent speed as the PageRank-like algorithms, but a more reasonable ranking result.

Controlled Language (CL) for Ontology Editing tools offer an attractive alternative for naive users wishing to create ontologies, but they are still required to spend time learning the correct syntactic structures and vocabulary in order to use the Controlled Language properly. This paper extends previous work (CLOnE) which uses standard NLP tools to process the language and manipulate an ontology. Here we also generate text in the CL from an existing ontology using template-based (or shallow) Natural Language Generation (NLG). The text generator and the CLOnE authoring process combine to form a RoundTrip Ontology Authoring environment: one can start with an existing imported ontology or one originally produced using CLOnE, (re)produce the Controlled Language, modify or edit the text as required and then turn the text back into the ontology in the CLOnE environment. Building on previous methodology we undertook an evaluation, comparing the RoundTrip Ontology Authoring process with a well-known ontology editor; where previous work required a CL reference manual with several examples in order to use the controlled language, the use of NLG reduces this learning curve for users and improves on existing results for basic ontology editing tasks.

Navigational features have been largely recognized as fundamental for graph database query languages. This fact has motivated several authors to propose RDF query languages with navigational capabilities. In particular, we have argued in a previous paper that

nested regular expressions

are appropriate to navigate RDF data, and we have proposed the nSPARQL query language for RDF, that uses nested regular expressions as building blocks. In this paper, we study some of the fundamental properties of nSPARQL concerning expressiveness and complexity of evaluation. Regarding expressiveness, we show that nSPARQL is expressive enough to answer queries considering the semantics of the RDFS vocabulary by directly traversing the input graph. We also show that nesting is necessary to obtain this last result, and we study the expressiveness of the combination of nested regular expressions and SPARQL operators. Regarding complexity of evaluation, we prove that the evaluation of a nested regular expression

E

over an RDF graph

G

can be computed in time

O

(|

G

|·|

E

|).

Efficient RDF data management is one of the cornerstones in realizing the Semantic Web vision. In the past, different RDF storage strategies have been proposed, ranging from simple triple stores to more advanced techniques like clustering or vertical partitioning on the predicates. We present an experimental comparison of existing storage strategies on top of the SP

2

Bench SPARQL performance benchmark suite and put the results into context by comparing them to a purely relational model of the benchmark scenario. We observe that (1) in terms of performance and scalability, a simple triple store built on top of a column-store DBMS is competitive to the vertically partitioned approach when choosing a physical (predicate, subject, object) sort order, (2) in our scenario with real-world queries, none of the approaches scales to documents containing tens of millions of RDF triples, and (3) none of the approaches can compete with a purely relational model. We conclude that future research is necessary to further bring forward RDF data management.

We present a technique for answering queries over RDF data through an evolutionary search algorithm, using fingerprinting and Bloom filters for rapid approximate evaluation of generated solutions. Our evolutionary approach has several advantages compared to traditional database-style query answering. First, the result quality increases monotonically and converges with each evolution, offering “anytime” behaviour with arbitrary trade-off between computation time and query results; in addition, the level of approximation can be tuned by varying the size of the Bloom filters. Secondly, through Bloom filter compression we can fit large graphs in main memory, reducing the need for disk I/O during query evaluation. Finally, since the individuals evolve independently, parallel execution is straightforward. We present our prototype that evaluates basic SPARQL queries over arbitrary RDF graphs and show initial results over large datasets.

This paper studies the expressive power of SPARQL. The main result is that SPARQL and non-recursive safe Datalog with negation have equivalent expressive power, and hence, by classical results, SPARQL is equivalent from an expressiveness point of view to Relational Algebra. We present explicit generic rules of the transformations in both directions. Among other findings of the paper are the proof that negation can be simulated in SPARQL, that non-safe filters are superfluous, and that current SPARQL W3C semantics can be simplified to a standard compositional one.

The Web Ontology Language (OWL) provides a modelling paradigm that is especially well suited for developing models of large, structurally complex domains such as those found in Health Care and the Life Sciences. OWL’s declarative nature combined with powerful reasoning tools has effectively supported the development of very large and complex anatomy, disease, and clinical ontologies. OWL, however, is not a programming language, so using these models in applications necessitates both a

technical

means of integrating OWL models with programs and considerable

methodological

sophistication in knowing how to integrate them. In this paper, we present an analytical framework for evaluating various OWL-Java combination approaches. We have developed a software framework for what we call

hybrid modelling

, that is, building models in which part of the model exists and is developed directly in Java and part of the model exists and is developed directly in OWL. We analyse the advantages and disadvantages of hybrid modelling both in comparison to other approaches and by means of a case study of a large medical records system.

Various semantic web service discovery techniques have been proposed, many of which perform the profile based service signature (I/O) matching. However, the service I/O concepts are not sufficient to discover web services accurately. This paper presents a new method to enhance the semantic description of semantic web service by using the semantic constraints of service I/O concepts in specific context. The semantic constraints described in a constraint graph are extracted automatically from the parsing results of the service description text by a set of heuristic rules. The corresponding semantic web service matchmaker performs not only the profile’s semantic matching but also the matching of their semantic constraints with the help of a constraint graph based matchmaking algorithm. The experiment results are encouraging when applying the semantic constraint to discover semantic web services on the service retrieval test collection OWLS-TC v2.

Currently proposed Semantic Web Services technologies allow the creation of ontology-based semantic annotations of Web services so that software agents are able to discover, invoke, compose and monitor these services with a high degree of automation. The OWL Services (OWL-S) ontology is an upper ontology in OWL language, providing essential vocabularies to semantically describe Web services. Currently OWL-S services can only be developed independently; if one service is unavailable then finding a suitable alternative would require an expensive and difficult global search/match. It is desirable to have a new OWL-S construct that can systematically support substitution tracing as well as incremental development and reuse of services. Introducing inheritance relationship (IR) into OWL-S is a natural solution. However, OWL-S, as well as most of the other currently discussed formalisms for Semantic Web Services such as WSMO or SAWSDL, has yet to define a concrete and self-contained mechanism of establishing inheritance relationships among services, which we believe is very important for the automated annotation and discovery of Web services as well as human organization of services into a taxonomy-like structure. In this paper, we extend OWL-S with the ability to define and maintain inheritance relationships between services. Through the definition of an additional “inheritance profile”, inheritance relationships can be stated and reasoned about. Two types of IRs are allowed to grant service developers the choice to respect the “contract” between services or not. The proposed inheritance framework has also been implemented and the prototype will be briefly evaluated as well.

Re-using and combining multiple ontologies on the Web is bound to lead to inconsistencies between the combined vocabularies. Even many of the ontologies that are in use today turn out to be inconsistent once some of their implicit knowledge is made explicit. However, robust and efficient methods to deal with inconsistencies are lacking from current Semantic Web reasoning systems, which are typically based on classical logic. In earlier papers, we have proposed the use of

syntactic relevance functions

as a method for reasoning with inconsistent ontologies. In this paper, we extend that work to the use of semantic distances. We show how Google distances can be used to develop

semantic relevance functions

to reason with inconsistent ontologies. In essence we are using the implicit knowledge hidden in the Web for explicit reasoning purposes. We have implemented this approach as part of the PION reasoning system. We report on experiments with several realistic ontologies. The test results show that a mixed syntactic/semantic approach can significantly improve reasoning performance over the purely syntactic approach. Furthermore, our methods allow to trade-off computational cost for inferential completeness. Our experiment shows that we only have to give up a little quality to obtain a high performance gain.

A novel family of parametric language-independent kernel functions defined for individuals within ontologies is presented. They are easily integrated with efficient statistical learning methods for inducing linear classifiers that offer an alternative way to perform classification w.r.t. deductive reasoning. A method for adapting the parameters of the kernel to the knowledge base through stochastic optimization is also proposed. This enables the exploitation of statistical learning in a variety of tasks where an inductive approach may bridge the gaps of the standard methods due the inherent incompleteness of the knowledge bases. In this work, a system integrating the kernels has been tested in experiments on approximate query answering with real ontologies collected from standard repositories.

This paper describes the first steps towards developing a methodology for testing and evaluating the performance of reasoners for the probabilistic description logic P-

${\ensuremath{\mathcal{SHIQ}}(D)}$

. Since it is a new formalism for handling uncertainty in DL ontologies, no such methodology has been proposed. There are no sufficiently large probabilistic ontologies to be used as test suites. In addition, since the reasoning services in P-

${\ensuremath{\mathcal{SHIQ}}(D)}$

are mostly query oriented, there is no single problem (like classification or realization in classical DL) that could be an obvious candidate for benchmarking. All these issues make it hard to evaluate the performance of reasoners, reveal the complexity bottlenecks and assess the value of optimization strategies. This paper addresses these important problems by making the following contributions: First, it describes a probabilistic ontology that has been developed for the real-life domain of breast cancer which poses significant challenges for the state-of-art P-

${\ensuremath{\mathcal{SHIQ}}(D)}$

reasoners. Second, it explains a systematic approach to generating a series of probabilistic reasoning problems that enable evaluation of the reasoning performance and shed light on what makes reasoning in P-

${\ensuremath{\mathcal{SHIQ}}(D)}$

hard in practice. Finally, the paper presents an optimized algorithm for the non-monotonic entailment. Its positive impact on performance is demonstrated using our evaluation methodology.

Human-defined concepts are fundamental building-blocks in constructing knowledge bases such as ontologies. Statistical learning techniques provide an alternative automated approach to concept definition, driven by data rather than prior knowledge. In this paper we propose a probabilistic modeling framework that combines both human-defined concepts and data-driven topics in a principled manner. The methodology we propose is based on applications of statistical topic models (also known as latent Dirichlet allocation models). We demonstrate the utility of this general framework in two ways. We first illustrate how the methodology can be used to automatically tag Web pages with concepts from a known set of concepts without any need for labeled documents. We then perform a series of experiments that quantify how combining human-defined semantic knowledge with data-driven techniques leads to better language models than can be obtained with either alone.

There are many reasons for measuring a distance between ontologies. In particular, it is useful to know quickly if two ontologies are close or remote before deciding to match them. To that extent, a distance between ontologies must be quickly computable. We present constraints applying to such measures and several possible ontology distances. Then we evaluate experimentally some of them in order to assess their accuracy and speed.

The growing popularity of social tagging systems promises to alleviate the knowledge bottleneck that slows down the full materialization of the Semantic Web since these systems allow ordinary users to create and share knowledge in a simple, cheap, and scalable representation, usually known as folksonomy. However, for the sake of knowledge workflow, one needs to find a compromise between the uncontrolled nature of folksonomies and the controlled and more systematic vocabulary of domain experts. In this paper we propose to address this concern by devising a method that automatically enriches a folksonomy with domain expert knowledge and by introducing a novel algorithm based on frequent itemset mining techniques to efficiently learn an ontology over the enriched folksonomy. In order to quantitatively assess our method, we propose a new benchmark for task-based ontology evaluation where the quality of the ontologies is measured based on how helpful they are for the task of personalized information finding. We conduct experiments on real data and empirically show the effectiveness of our approach.

We introduce the notion of the mixed DL and entailment-based (DLE) OWL reasoning, defining a framework inspired from the hybrid and homogeneous paradigms for integration of rules and ontologies. The idea is to combine the TBox inferencing capabilities of the DL algorithms and the scalability of the rule paradigm over large ABoxes. Towards this end, we define a framework that uses a DL reasoner to reason over the TBox of the ontology (hybrid-like) and a rule engine to apply a domain-specific version of ABox-related entailments (homogeneous-like) that are generated by TBox queries to the DL reasoner. The DLE framework enhances the entailment-based OWL reasoning paradigm in two directions. Firstly, it disengages the manipulation of the TBox semantics from any incomplete entailment-based approach, using the efficient DL algorithms. Secondly, it achieves faster application of the ABox-related entailments and efficient memory usage, comparing it to the conventional entailment-based approaches, due to the low complexity and the domain-specific nature of the entailments.

An approach to improve an RCC-derived geospatial approximation is presented which makes use of concept inclusion axioms in OWL. The algorithm used to control the approximation combines hypothesis testing with consistency checking provided by a knowledge representation system based on description logics. Propositions about the consistency of the refined ABox w.r.t. the associated TBox when compared to baseline ABox and TBox are made. Formal proves of the divergent consistency results when checking either of both are provided. The application of the approach to a geospatial setting results in a roughly tenfold improved approximation when using the refined ABox and TBox. Ways to further improve the approximation and to automate the detection of falsely calculated relations are discussed.

We analyze the datatype system of OWL and OWL 2, and discuss certain nontrivial consequences of its definition, such as the extensibility of the set of supported datatypes and complexity of reasoning. We also argue that certain datatypes from the list of normative datatypes in the current OWL 2 Working Draft are inappropriate and should be replaced with different ones. Finally, we present an algorithm for datatype reasoning. Our algorithm is modular in the sense that it can handle any datatype that supports certain basic operations. We show how to implement these operations for number and string datatypes.

A justification for an entailment in an OWL ontology is a minimal subset of the ontology that is sufficient for that entailment to hold. Since justifications respect the syntactic form of axioms in an ontology, they are usually neither

syntactically

nor

semantically

minimal. This paper presents two new subclasses of justifications—

laconic justifications

and

precise justifications

. Laconic justifications only consist of axioms that do not contain any superfluous “parts”. Precise justifications can be derived from laconic justifications and are characterised by the fact that they consist of flat, small axioms, which facilitate the generation of semantically minimal repairs. Formal definitions for both types of justification are presented. In contrast to previous work in this area, these definitions make it clear as to what exactly “parts of axioms” are. In order to demonstrate the practicability of computing laconic, and hence precise justifications, an algorithm is provided and results from an empirical evaluation carried out on several published ontologies are presented. The evaluation showed that laconic/precise justifications can be computed in a reasonable time for entailments in a range of ontologies that vary in size and complexity. It was found that in half of the ontologies sampled there were entailments that had more laconic/precise justifications than regular justifications. More surprisingly it was observed that for some ontologies there were fewer laconic justifications than regular justifications.

Finding mappings between compatible ontologies is an important but difficult open problem. Instance-based methods for solving this problem have the advantage of focusing on the most active parts of the ontologies and reflect concept semantics as they are actually being used. However such methods have not at present been widely investigated in ontology mapping, compared to linguistic and structural techniques. Furthermore, previous instance-based mapping techniques were only applicable to cases where a substantial set of instances was available that was doubly annotated with both vocabularies. In this paper we approach the mapping problem as a classification problem based on the similarity between instances of concepts. This has the advantage that no doubly annotated instances are required, so that the method can be applied to any two corpora annotated with their own vocabularies. We evaluate the resulting classifiers on two real-world use cases, one with homogeneous and one with heterogeneous instances. The results illustrate the efficiency and generality of this method.

The emergence of web based systems in which users can annotate items, raises the question of the semantic interoperability between vocabularies originating from collaborative annotation processes, often called folksonomies, and keywords assigned in a more traditional way. If collections are annotated according to two systems, e.g. with tags and keywords, the annotated data can be used for instance based mapping between the vocabularies. The basis for this kind of matching is an appropriate similarity measure between concepts, based on their distribution as annotations. In this paper we propose a new similarity measure that can take advantage of some special properties of user generated metadata. We have evaluated this measure with a set of articles from Wikipedia which are both classified according to the topic structure of Wikipedia and annotated by users of the bookmarking service del.icio.us. The results using the new measure are significantly better than those obtained using standard similarity measures proposed for this task in the literature, i.e., it correlates better with human judgments. We argue that the measure also has benefits for instance based mapping of more traditionally developed vocabularies.

Several ontology repositories provide access to the growing collection of ontologies on the Semantic Web. Some repositories collect ontologies automatically by crawling the Web; in other repositories, users submit ontologies themselves. In addition to providing search across multiple ontologies, the added value of ontology repositories lies in the metadata that they may contain. This metadata may include information provided by ontology authors, such as ontologies’ scope and intended use; feedback provided by users such as their experiences in using the ontologies or reviews of the content; and

mapping metadata

that relates concepts from different ontologies. In this paper, we focus on the ontology-mapping metadata and on community-based method to collect ontology mappings. More specifically, we develop a model for representing mappings collected from the user community and the metadata associated with the mapping. We use the model to bring together more than 30,000 mappings from 7 sources. We also validate the model by extending BioPortal–a repository of biomedical ontologies that we have developed—to enable users to create single concept-to-concept mappings in its graphical user interface, to upload and download mappings created with other tools, to comment on the mappings and to discuss them, and to visualize the mappings and the corresponding metadata.

Correspondences in ontology alignments relate two ontology entities with a relation. Typical relations are equivalence or subsumption. However, different systems may need different kinds of relations. We propose to use the concepts of algebra of relations in order to express the relations between ontology entities in a general way. We show the benefits in doing so in expressing disjunctive relations, merging alignments in different ways, amalgamating alignments with relations of different granularity, and composing alignments.

Grounded conjunctive query answering over OWL-DL ontologies is intractable in the worst case, but we present novel techniques which allow for efficient querying of large expressive knowledge bases in secondary storage. In particular, we show that we can effectively answer grounded conjunctive queries without building a full completion forest for a large Abox (unlike state of the art tableau reasoners). Instead we rely on the completion forest of a dramatically reduced summary of the Abox. We demonstrate the effectiveness of this approach in Aboxes with up to 45 million assertions.

Revision of a description logic-based ontology deals with the problem of incorporating newly received information consistently. In this paper, we propose a general operator for revising terminologies in description logic-based ontologies. Our revision operator relies on a reformulation of the kernel contraction operator in belief revision. We first define our revision operator for terminologies and show that it satisfies some desirable logical properties. Second, two algorithms are developed to instantiate the revision operator. Since in general, these two algorithms are computationally too hard, we propose a third algorithm as a more efficient alternative. We implemented the algorithms and provide evaluation results on their efficiency, effectiveness and meaningfulness in the context of two application scenarios: Incremental ontology learning and mapping revision.

We propose a novel method for reasoning in the description logic

$\mathcal{SHIQ}$

. After a satisfiability preserving transformation from

$\mathcal{SHIQ}$

to the description logic

$\mathcal{ALCI}b$

, the obtained

$\mathcal{ALCI}b$

Tbox

$\mathcal{T}$

is converted into an ordered binary decision diagram (OBDD) which represents a canonical model for

$\mathcal{T}$

. This OBDD is turned into a disjunctive datalog program that can be used for Abox reasoning. The algorithm is worst-case optimal w.r.t. data complexity, and admits easy extensions with DL-safe rules and ground conjunctive queries.

We describe RDF123, a highly flexible open-source tool for translating spreadsheet data to RDF. Existing spreadsheet-to-rdf tools typically map only to star-shaped RDF graphs, i.e. each spreadsheet row is an instance, with each column representing a property. RDF123, on the other hand, allows users to define mappings to arbitrary graphs, thus allowing much richer spreadsheet semantics to be expressed. Further, each row in the spreadsheet can be mapped with a fairly different RDF scheme. Two interfaces are available. The first is a graphical application that allows users to create their mapping in an intuitive manner. The second is a Web service that takes as input a URL to a Google spreadsheet or CSV file and an RDF123 map, and provides RDF as output.

The Semantic Desktop is a means to support users in Personal Information Management (PIM). Using the open source software prototype Gnowsis, we evaluated the approach in a two month case study in 2006 with eight participants. Two participants continued using the prototype and were interviewed after two years in 2008 to show their long-term usage patterns. This allows us to analyse how the system was used for PIM. Contextual interviews gave insights on behaviour, while questionnaires and event logging did not. We discovered that in the personal environment, simple has-Part and is-related relations are sufficient for users to file and re-find information, and that the personal semantic wiki was used creatively to note information.

For easing the exchange of news, the International Press Telecommunication Council (IPTC) has developed the NewsML Architecture (NAR), an XML-based model that is specialized into a number of languages such as NewsML G2 and EventsML G2. As part of this architecture, specific controlled vocabularies, such as the IPTC News Codes, are used to categorize news items together with other industry-standard thesauri. While news is still mainly in the form of text-based stories, these are often illustrated with graphics, images and videos. Media-specific metadata formats, such as EXIF, DIG35 and XMP, are used to describe the media. The use of different metadata formats in a single production process leads to interoperability problems within the news production chain itself. It also excludes linking to existing web knowledge resources and impedes the construction of uniform end-user interfaces for searching and browsing news content.

In order to allow these different metadata standards to interoperate within a single information environment, we design an OWL ontology for the IPTC News Architecture, linked with other multimedia metadata standards. We convert the IPTC NewsCodes into a SKOS thesaurus and we demonstrate how the news metadata can then be enriched using natural language processing and multimedia analysis and integrated with existing knowledge already formalized on the Semantic Web. We discuss the method we used for developing the ontology and give rationale for our design decisions. We provide guidelines for re-engineering schemas into ontologies and formalize their implicit semantics. In order to demonstrate the appropriateness of our ontology infrastructure, we present an exploratory environment for searching and browsing news items.

We study the problem of distributed RDFS reasoning and query answering on top of distributed hash tables. Scalable, distributed RDFS reasoning is an essential functionality for providing the scalability and performance that large-scale Semantic Web applications require. Our goal in this paper is to compare and evaluate two well-known approaches to RDFS reasoning, namely backward and forward chaining, on top of distributed hash tables. We show how to implement both algorithms on top of the distributed hash table Bamboo and prove their correctness. We also study the time-space trade-off exhibited by the algorithms analytically, and experimentally by evaluating our algorithms on PlanetLab.

In this paper, we present a framework for developing ontologies in a modular manner, which is based on the notions of interfaces and knowledge encapsulation. Within the context of this framework, an ontology can be defined and developed as a set of ontology modules that can access the knowledge bases of the others through their well-defined interfaces. An important implication of the proposed framework is that ontology modules can be developed completely independent of each others’ signature and language. Such modules are free to only utilize the required knowledge segments of the others. We describe the interface-based modular ontology formalism, which theoretically supports this framework and present its distinctive features compared to the exiting modular ontology formalisms. We also describe the real-world design and implementation of the framework for creating modular ontologies by extending OWL-DL and modifying the Swoop interfaces and reasoners.

The Semantic Web is a distributed environment for knowledge representation and reasoning. The distributed nature brings with it failing data sources and inconsistencies between autonomous knowledge bases. To reduce problems resulting from unavailable sources and to improve performance, caching can be used. Caches, however, raise new problems of imprecise or outdated information. We propose to distinguish between certain and cached information when reasoning on the semantic web, by extending the well known

$\mathcal{FOUR}$

bilattice of truth and knowledge orders to

$\mathcal{FOUR-C}$

, taking into account cached information. We discuss how users can be offered additional information about the

reliability

of inferred information, based on the availability of the corresponding information sources. We then extend the framework towards

$\mathcal{FOUR-T}$

, allowing for multiple

levels of trust

on data sources. In this extended setting, knowledge about trust in information sources can be used to compute, how well an inferred statement can be trusted and to resolve inconsistencies arising from connecting multiple data sources. We redefine the stable model and well founded semantics on the basis of

$\mathcal{FOUR-T}$

, and reformalize the Web Ontology Language OWL2 based on logical bilattices, to augment OWL knowledge bases with trust based reasoning.

The emerging paradigm of service-oriented computing requires novel techniques for various service-related tasks. Along with automated support for service discovery, selection, negotiation, and composition, support for automated service contracting and enactment is crucial for any large scale service environment, where large numbers of clients and service providers interact. Many problems in this area involve reasoning, and a number of logic-based methods to handle these problems have emerged in the field of Semantic Web Services. In this paper, we build upon our previous work where we used Concurrent Transaction Logic (CTR) to model and reason about service contracts. We significantly extend the modeling power of the previous work by allowing iterative processes in the specification of service contracts, and we extend the proof theory of CTR to enable reasoning about such contracts. With this extension, our logic-based approach is capable of modeling general services represented using languages such as WS-BPEL.

Integration of heterogeneous services is often hard-wired in service or workflow implementations. In this paper we define an execution model operating on semantic descriptions of services allowing flexible integration of services with solving data and process conflicts where necessary. We implement the model using our WSMO technology and a case scenario from the B2B domain of the SWS Challenge.

Efficient and scalable discovery mechanisms are critical for enabling service-oriented architectures on the Semantic Web. The majority of currently existing approaches focuses on centralized architectures, and deals with efficiency typically by pre-computing and storing the results of the semantic matcher for all possible query concepts. Such approaches, however, fail to scale with respect to the number of service advertisements and the size of the ontologies involved. On the other hand, this paper presents an efficient and scalable index-based method for Semantic Web service discovery that allows for fast selection of services at query time and is suitable for both centralized and P2P environments. We employ a novel encoding of the service descriptions, allowing the match between a request and an advertisement to be evaluated in constant time, and we index these representations to prune the search space, reducing the number of comparisons required. Given a desired ranking function, the search algorithm can retrieve the top-

k

matches progressively, i.e., better matches are computed and returned first, thereby further reducing the search engine’s response time. We also show how this search can be performed efficiently in a suitable structured P2P overlay network. The benefits of the proposed method are demonstrated through experimental evaluation on both real and synthetic data.

We present Redgraph, the first generic virtual reality visualization program for Semantic Web data. Redgraph is capable of handling large data-sets, as we demonstrate on social network data from the U.S. Patent Trade Office. We develop a Semantic Web vocabulary of virtual reality terms compatible with GraphXML to map graph visualization into the Semantic Web itself. Our approach to visualizing Semantic Web data takes advantage of user-interaction in an immersive environment to bypass a number of difficult issues in 3-dimensional graph visualization layout by relying on users themselves to interactively extrude the nodes and links of a 2-dimensional graph into the third dimension. When users touch nodes in the virtual reality environment, they retrieve data formatted according to the data’s schema or ontology. We applied Redgraph to social network data constructed from patents, inventors, and institutions from the United States Patent and Trademark Office in order to explore networks of innovation in computing. Using this data-set, results of a user study comparing extrusion (3-D) vs. no-extrusion (2-D) are presented. The study showed the use of a 3-D interface by subjects led to significant improvement on answering of fine-grained questions about the data-set, but no significant difference was found for broad questions about the overall structure of the data. Furthermore, inference can be used to improve the visualization, as demonstrated with a data-set of biotechnology patents and researchers.

Collaborative tagging systems have nowadays become important data sources for populating semantic web applications. For tasks like synonym detection and discovery of concept hierarchies, many researchers introduced measures of tag similarity. Even though most of these measures appear very natural, their design often seems to be rather ad hoc, and the underlying assumptions on the notion of similarity are not made explicit. A more systematic characterization and validation of tag similarity in terms of formal representations of knowledge is still lacking. Here we address this issue and analyze several measures of tag similarity: Each measure is computed on data from the social bookmarking system del.icio.us and a semantic grounding is provided by mapping pairs of similar tags in the folksonomy to pairs of synsets in Wordnet, where we use validated measures of semantic distance to characterize the semantic relation between the mapped tags. This exposes important features of the investigated similarity measures and indicates which ones are better suited in the context of a given semantic application.

The continued increase in Web usage, in particular participation in folksonomies, reveals a trend towards a more dynamic and interactive Web where individuals can organise and share resources. Tagging has emerged as the de-facto standard for the organisation of such resources, providing a versatile and reactive knowledge management mechanism that users find easy to use and understand. It is common nowadays for users to have multiple profiles in various folksonomies, thus distributing their tagging activities. In this paper, we present a method for the automatic consolidation of user profiles across two popular social networking sites, and subsequent semantic modelling of their interests utilising Wikipedia as a multi-domain model. We evaluate how much can be learned from such sites, and in which domains the knowledge acquired is focussed. Results show that far richer interest profiles can be generated for users when multiple tag-clouds are combined.

We introduce

$\text{\sf{ELP}}$

as a decidable fragment of the Semantic Web Rule Language (SWRL) that admits reasoning in polynomial time.

$\text{\sf{ELP}}$

is based on the tractable description logic

$\mathcal{EL}^{\mathord{+}\mathord{+}}$

, and encompasses an extended notion of the recently proposed

DL rules

for that logic. Thus

$\text{\sf{ELP}}$

extends

$\mathcal{EL}^{\mathord{+}\mathord{+}}$

with a number of features introduced by the forthcoming OWL 2, such as disjoint roles, local reflexivity, certain range restrictions, and the universal role. We present a reasoning algorithm based on a translation of

$\text{\sf{ELP}}$

to Datalog, and this translation also enables the seamless integration of DL-safe rules into

$\text{\sf{ELP}}$

. While reasoning with DL-safe rules as such is already highly intractable, we show that DL-safe rules based on the Description Logic Programming (DLP) fragment of OWL 2 can be admitted in

$\text{\sf{ELP}}$

without losing tractability.

A large amount of terms (classes and properties) have been published on the Semantic Web by various parties, to be shared for describing resources. Terms are defined based on other terms, and thus a directed dependence relation is formed. The study of term dependence is a foundation work and is important for many other tasks, such as ontology maintenance, integration, and distributed reasoning on the Web scale. In this paper, we analyze the complex network characteristics of the term dependence graph and the induced vocabulary dependence graph. The graphs analyzed in the experiments are constructed from a large data set that contains 1,278,233 terms in 3,039 vocabularies. The results characterize the current status of schemas on the Semantic Web in many aspects, including degree distributions, reachability, and connectivity.

This paper presents a new semantic relatedness measure on ontologies which considers especially the object properties between the concepts. Our approach relies on two hypotheses. Firstly, using only concept hierarchy and object properties, only a few paths can be considered as “semantically corrects” and these paths obey to a given set of rules. Secondly, following a given edge in a path has a cost (represented as a weight), which depends on its type (

$is\mbox{-}a$

,

$part\mbox{-}of$

, etc.), its context in the ontology and its position in this path. We propose an evaluation of our measure on the lexical base WordNet using

$part\mbox{-}of$

relation with two different benchmarks. We show that, in this context, our measure outperforms the classical semantic measures.

In cultural heritage, large virtual collections are coming into existence. Such collections contain heterogeneous sets of metadata and vocabulary concepts, originating from multiple sources. In the context of the E-Culture demonstrator we have shown earlier that such virtual collections can be effectively explored with keyword search and semantic clustering. In this paper we describe the design rationale of ClioPatria, an open-source system which provides APIs for scalable semantic graph search. The use of ClioPatria’s search strategies is illustrated with a realistic use case: searching for ”Picasso”. We discuss details of scalable graph search, the required OWL reasoning functionalities and show why SPARQL queries are insufficient for solving the search problem.

Modern businesses operate in a rapidly changing environment. Continuous learning is an essential ingredient in order to stay competitive in such environments. The APOSDLE system utilizes semantic web technologies to create a generic system for supporting knowledge workers in different domains to learn work. Since APOSDLE relies on three interconnected semantic models to achieve this goal, the question on how to efficiently create high-quality semantic models has become one of the major research challenges. On the basis of two concrete examples-namely deployment of such a learning system at EADS, a large corporation, and deployment at ISN, a network of SMEs-we report in detail the issues a company has to face, when it wants to deploy a modern learning environment relying on semantic web technology.

Modern knowledge management is based on the orchestration of dynamic communities that acquire and share knowledge according to customized schemas. However, while independence of ontological views is favoured, these communities must also be able to share their knowledge with the rest of the organization. In this paper we introduce K-Forms and K-Search, a suite of Semantic Web tools for supporting distributed and networked knowledge acquisition, capturing, retrieval and sharing. They enable communities of users to define their own domain views in an intuitive way (automatically translated into formal ontologies) and capture and share knowledge according to them. The tools favour reuse of existing ontologies; reuse creates as side effect a network of (partially) interconnected ontologies that form the basis for knowledge exchange among communities. The suite is under release to support knowledge capture, retrieval and sharing in a large jet engine company.

Medical ontologies have become the standard means of recording and accessing conceptualized biological and medical knowledge. The expressivity of these ontologies goes from simple concept lists through taxonomies to formal logical theories. In the context of patient information, their application is primarily annotation of medical (instance) data. To exploit higher expressivity, we propose an architecture which allows for reasoning on patient data using OWL DL ontologies. The implementation is carried out as part of the Health-e-Child platform prototype. We discuss the use case where ontologies establish a hierarchical classification of patients which in turn is used to aid the visualization of patient data. We briefly discuss the treemap-based patient viewer which has been evaluated in the Health-e-Child project.

We present a tool, called the

Requirements Analysis Tool

that performs a wide range of best practice analyses on software requirements documents. The novelty of our approach is the use of user-defined glossaries to extract structured content, and thus support a broad range of syntactic and semantic analyses, while allowing users to write requirements in the stylized natural language advocated by expert requirements writers. Semantic Web technologies are then leveraged for deeper semantic analysis of the extracted structured content to find various kinds of problems in requirements documents.

The documentation of Enterprise Research Planning (ERP) systems is usually (1) extremely large and (2) combines various views from the business and the technical implementation perspective. Also, a very specific vocabulary has evolved, in particular in the SAP domain (e.g. SAP Solution Maps or SAP software module names). This vocabulary is not clearly mapped to business management terminology and concepts. It is a well-known problem in practice that searching in SAP ERP documentation is difficult, because it requires in-depth knowledge of a large and proprietary terminology. We propose to use ontologies and automatic annotation of such large HTML software documentation in order to improve the usability and accessibility, namely of ERP help files. In order to achieve that, we have developed an ontology and prototype for SAP ERP 6.0. Our approach integrates concepts and lexical resources from (1) business management terminology, (2) SAP business terminology, (3) SAP system terminology, and (4) Wordnet synsets. We use standard GATE/KIM technology to annotate SAP help documentation with respective references to our ontology. Eventually, our approach consolidates the knowledge contained in the SAP help functionality at a conceptual level. This allows users to express their queries using a terminology they are familiar with, e.g. referring to general management terms. Despite a widely automated ontology construction process and a simplistic annotation strategy with minimal human intervention, we experienced convincing results. For an average query linked to an action and a topic, our technology returns more than 3 relevant resources, while a naïve term-based search returns on average only about 0.2 relevant resources.

Market Blended Insight (MBI) is a project with a clear objective of making a significant performance improvement in UK business to business (B2B) marketing activities in the 5-7 year timeframe. The web has created a rapid expansion of content that can be harnessed by recent advances in Semantic Web technologies and applied to both Media industry provision and company utilization of exploitable business data and content. The project plans to aggregate a broad range of business information, providing unparalleled insight into UK business activity and develop rich semantic search and navigation tools to allow any business to ’place their sales proposition in front of a prospective buyer’ confident of the fact that the recipient has a propensity to buy.

Home automation has recently gained a new momentum thanks to the ever-increasing commercial availability of domotic components. In this context, researchers are working to provide interoperation mechanisms and to add intelligence on top of them. For supporting intelligent behaviors, house modeling is an essential requirement to understand current and future house states and to possibly drive more complex actions. In this paper we propose a new house modeling ontology designed to fit real world domotic system capabilities and to support interoperation between currently available and future solutions. Taking advantage of technologies developed in the context of the Semantic Web, the DogOnt ontology supports device/network independent description of houses, including both “controllable” and architectural elements. States and functionalities are automatically associated to the modeled elements through proper inheritance mechanisms and by means of properly defined SWRL auto-completion rules which ease the modeling process, while automatic device recognition is achieved through classification reasoning.

We present IYOUIT, a prototype service to pioneer a context-aware mobile digital lifestyle and its reflection on the Web. The application is based on a distributed infrastructure that incorporates Semantic Web technologies in several places to derive qualitative interpretations of a user’s digital traces in the real world. Networked components map quantitative sensor data to qualitative abstractions represented in formal ontologies. Subsequent classification processes combine these with formalized domain knowledge to derive meaningful interpretations and to recognize exceptional events in context histories. The application is made available on Nokia Series-60 phones and designed to seamlessly run 24/7.

The combination of Semantic Web and Grid technologies and architectures eases the development of applications that share heterogeneous resources (data and computing elements) that belong to several organisations. The Aerospace domain has an extensive and heterogeneous network of facilities and institutions, with a strong need to share both data and computational resources for complex processing tasks. One such task is monitoring and data analysis for Satellite Missions. This paper presents a Semantic Data Grid for satellite missions, where flexibility, scalability, interoperability, extensibility and efficient development have been considered the key issues to be addressed.

As AI developers increasingly look to workflow technologies to perform complex integrations of individual software components, there is a growing need for the workflow systems to have expressive descriptions of those components. They must know more than just the types of a component’s inputs and outputs; instead, they need detailed characterizations that allow them to make fine-grained distinctions between candidate components and between candidate workflows. This paper describes

ProCat

, an implemented ontology-based catalog for components, conceptualized as

processes

, that captures and communicates this detailed information.

ProCat

is built on a layered representation that allows reasoning about processes at varying levels of abstraction, from qualitative constraints reflecting preconditions and effects, to quantitative predictions about output data and performance.

ProCat

employs Semantic Web technologies RDF, OWL, and SPARQL, and builds on Semantic Web services research. We describe

ProCat’s

approach to representing and answering queries about processes, discuss some early experiments evaluating the quantitative predictions, and report on our experience using

ProCat

in a system producing workflows for intelligence analysis.

The Inference Web infrastructure for web explanations together with its underlying Proof Markup Language (PML) for encoding justification and provenance information has been used in multiple projects varying from explaining the behavior of cognitive agents to explaining how knowledge is extracted from multiple sources of information in natural language. The PML specification has increased significantly since its inception in 2002 in order to accommodate a rich set of requirements derived from multiple projects, including the ones mentioned above. In this paper, we have a very different goal than the other PML documents: to demonstrate that PML may be effectively used by simple systems (as well as complex systems) and to describe lightweight use of language and its associated Inference Web tools. We show how an exemplar scientific application can use lightweight PML descriptions within the context of an NSF-funded cyberinfrastructure project. The scientific application is used throughout the paper as a use case for the lightweight use of PML and the Inference Web and is meant to be an operational prototype for a class of cyberinfrastructure applications.

Metadata management is an important aspect of today’s enterprise information systems. Metadata management systems are growing from toolspecific repositories to enterprise-wide metadata repositories. In this context, one challenge is the management of the evolving metadata whose schema or meta-model itself may evolve, e.g., dynamically-added properties, which are often hard to predict upfront at the initial meta-model design time; another challenge is to organize the metadata by semantically-rich classification schemes. In this paper, we present a practical system which provides support for users to dynamically manage semantically-rich properties and classifications in the IBM WebSphere Metadata Server (MDS) by integrating an OWL ontology repository. To enable the smooth acceptance of Semantic Web technologies for developers of commercial software which must run 24 hours/day, 7 days/week, the system is designed to consist of integrated modeling paradigms, with an integrated query language and runtime repository. Specifically, we propose the modeling of dynamic properties on structured metadata as OWL properties and the modeling of classification schemes as OWL ontologies for metadata classification. We present a natural extension to OQL (Object Query Language)-like query language to embrace dynamic properties and metadata classification. We also observe that hybrid storage, i.e., horizontal tables for structured metadata and vertical triple tables for dynamic properties and classification, is suitable for the storage and query processing of co-existing structured metadata and semantic metadata. We believe that our study and experience are not specific to MDS, but are valuable for the community trying to apply Semantic Web technologies to the structured data management area.

A Multimedia Content Marketplace can support innovative business models in the telecommunication sector. This marketplace has a strong need for semantics, co-ordination and a service-oriented architecture. Triple Space Computing is an emerging semantic co-ordination paradigm for Web services, for which the marketplace is an ideal implementation scenario. This paper introduces the developed Triple Space platform and our planned evaluation of its value to our telecommunication scenario.