Eyal Oren

We address the problem of scalable distributed reasoning, proposing a technique for materialising the closure of an RDF graph based on MapReduce. We have implemented our approach on top of Hadoop and deployed it on a compute cluster of up to 64 commodity machines. We show that a naive implementation on top of MapReduce is straightforward but performs badly and we present several non-trivial optimisations. Our algorithm is scalable and allows us to compute the RDFS closure of 865M triples from the Web (producing 30B triples) in less than two hours, faster than any other published approach.

Workflow models have been used and refined for years to execute processes within organisations. To deal with collaborative processes (choreographies) these internal workflow models have to be aligned with the external behaviour advertised through Web service interfaces. However, traditional workflow management systems (WfMS) do not offer this functionality. Simply sharing and merging process models is often not possible, because workflow management lacks a widely accepted standard theory for workflow models. Multiple research and standardisation efforts to integrate different workflow theories have been proposed over the years. XPDL is the most widely used standard for process model interchange and supported by over 80 systems. However, XPDL also lacks the possibility to relate a workflow model to its possible choreography interface abstractions.

To remedy this situation, we propose to abstract the XPDL model to a higher-level model, perform the integration and the compaction algorithms at that level and then ground it back to the desired choreography models. We develop and use an integrated ontology which is based on the XPDL standard for this purpose. To facilitate the abstraction and grounding, we present a mapping procedure to automatically translate XPDL and BPMN workflow models into this ontology. After translation, these models are annotated with a parameterised role model and other collaborative properties. We present a compaction procedure that automatically maps the annotated models into external choreography interfaces that expose only the relevant information for a particular partner collaboration. Our procedure is agnostic with respect to the target choreography model. We demonstrate our approach using WSMO choreographies which enables us to automatically generate interface models from any WfMSs that supports XPDL export.

Web Science requires efficient techniques for analysing large datasets. Many Semantic Web problems are difficult to solve through common divide-and-conquer strategies, since they are hard to partition. We present MaRVIN, a parallel and distributed platform for processing large amounts of RDF data, on a network of loosely-coupled peers. We present our divide-conquer-swap strategy and show that this model converges towards completeness. We evaluate performance, scalability, load balancing and efficiency of our system.

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.

RDF is increasingly being used to represent large amounts of data on the Web. Current query evaluation strategies for RDF are inspired by databases, assuming perfect answers on finite repositories. In this paper, we focus on a query method based on evolutionary computing, which allows us to handle uncertainty, incompleteness and unsatisfiability, and deal with large datasets, all within a single conceptual framework. Our technique supports approximate answers with “anytime” behaviour. We present scalability results and next steps for improvement.

We present a method for predicting affective comprehension of political news texts through frequency-oriented latent semantic analysis (LSA). We predict the affective meaning of words and sections of Dutch political news texts, based on the linguistic intuitions of Dutch thesaurus creators and word similarities according found after LSA. We evaluate our approach through cross-fold validation and by predicting the affective meanings previously attached to phrases by human coders in large-scale content analysis of Dutch elections campaigns.

Developers of Semantic Web applications face a challenge with respect to the decentralised publication model: how and where to find statements about encountered resources. The “linked data” approach mandates that resource URIs should be de-referenced to return resource metadata. But for data discovery linkage itself is not enough, and crawling and indexing of data is necessary. Existing Semantic Web search engines are focused on database-like functionality, compromising on index size, query performance and live updates. We present Sindice, a lookup index over resources crawled on the Semantic Web. Our index allows applications to automatically locate documents containing information about a given resource. In addition, we allow resource retrieval through uniquely identifying inverse-functional properties, offer a full-text search and index SPARQL endpoints. Finally we introduce an extension to the sitemap protocol which allows us to efficiently index large Semantic Web datasets with minimal impact on the data providers.

Semantic Web applications share a large portion of development effort with database-driven Web applications. Existing approaches for development of these database-driven applications cannot be directly applied to Semantic Web data due to differences in the underlying data model. We develop a mapping approach that embeds Semantic Web data into object-oriented languages and thereby enables reuse of existing Web application frameworks. We analyse the relation between the Semantic Web and the Web, and survey the typical data access patterns in Semantic Web applications. We discuss the mismatch between object-oriented programming languages and Semantic Web data, for example in the semantics of class membership, inheritance relations, and object conformance to schemas. We present ActiveRDF, an object-oriented API for managing RDF data that offers full manipulation and querying of RDF data, does not rely on a schema and fully conforms to RDF(S) semantics. ActiveRDF can be used with different RDF data stores: adapters have been implemented to generic SPARQL endpoints, Sesame, Jena, Redland and YARS and new adapters can be added easily. We demonstrate the usage of ActiveRDF and its integration with the popular Ruby on Rails framework which enables rapid development of Semantic Web applications.

RosettaNet is an industry-driven e-business process standard that defines common inter-company public processes and their associated business documents. RosettaNet is based on the Service-Oriented Architecture (SOA) paradigm and all business documents are expressed in DTD or XML Schema. Our “ontologically-enhanced RosettaNet” effort translates RosettaNet business documents into a Web ontology language, allowing business reasoning based on RosettaNet message exchanges. This chapter describes this extension to RosettaNet and shows how it can be used in business integrations for better interoperability. The usage of Web ontology languages in RosettaNet collaborations can help accommodate partner heterogeneity in the setup phase and can ease the back-end integration, enabling for example more competition in the purchasing processes. It provides also a building block for a semantic SOA with discovery, selection and composition capabilities.

Managing and enabling knowledge is a key to success in our economy and society. The problem of knowledge management can generally be tackled from two sides: top-down and bottom-up. Many approaches have been taken from the top down, in which the organisation aimed to better manage their internal knowledge by installing central knowledge repositories. Many of these systems were less accepted than expected. Along with the Web 2.0 notions of user-provided content and collective intelligence, more bottom-up approaches to knowledge management were developed. In this chapter we describe an individual-centric, bottom-up approach to personal knowledge management (PKM). PKM is the individual management of knowledge from a subjective perspective.

The move towards open and interlinked data on the Web and the Semantic Web results in more open systems. In contrast to traditional database-driven applications, open systems liberate the data that they operate on: sources are decentralised, data can be semi-structured with arbitrary vocabulary and contributions can be published anywhere. Opening up existing applications and their data would improve knowledge management but raises challenges: how to programmatically access and manipulate the web of linked data, how to visualise and navigate the information graph, how to converge user-provided content, and how to find relevant data in distributed sources. This thesis offers algorithms and components that simplify and support knowledge management based on Semantic Web technology. We address four areas of Semantic Web application development: programmatic access: how to program against the flexible graph-based model; data navigation: how to navigate arbitrary information spaces; data entry: how to guide users through collaborative recommendation; and data discovery: how to locate relevant data sources. Our hypothesis is that the issues of programmatic access, data navigation, data entry, and data discovery can be addressed, with acceptable results, through the sole introspection of instance data at runtime, without relying on fixed schema structures at design time. In all four areas we devise solutions (an object-oriented data mapping, a generic navigation interface, a collaborative recommendation algorithm and a scalable lookup service) that are domain-independent, rely only on instance data and dynamically adjust to the available data.

Developers of Semantic Web applications face a challenge with respect to the decentralised publication model: where to find statements about encountered resources. The “linked data” approach, which mandates that resource URIs should be de-referenced and yield metadata about the resource, helps but is only a partial solution and not followed widely. We present a lookup index over resources crawled on the Semantic Web. Our index allows applications to automatically retrieve sources with information about a certain resource. In contrast to more feature-rich Semantic Web search engines, our index is purposely limited in scope and functionality to achieve highly scalability and maintainability.

“Mash-ups” are an upcoming paradigm for online applications that combine functionality from various sources. The Semantic Web, an extension of the current Web with more support for data integration and data reuse, significantly eases the development of such “mash-ups”. But existing application frameworks offer only limited support for (integration of) Semantic Web data, such as a social networking site combined with scientific publications. We analyse the additional needs for application development on the Semantic Web, focusing mostly on the semantic differences between the object-oriented and RDF(S) datamodel, and explain why dynamically typed languages are well-suited for this task. We introduce our SWAF framework, discuss its implementation and demonstrate its use for rapid development of an example social networking application.

The SIOC Ontology can be used to express information from the online community sites in a machine-readable form using RDF. This rich data structure allows us to easily analyse and extract social relations from these community sites. We use SIOC information to analyse the social relations between users through the content that they create. We introduce metrics for social neighbourhood and social reputation, formally expressed as SPARQL queries over the SIOC data. Finally, we demonstrate these algorithms in our Social SIOC Explorer prototype.

We extend the Ruby on Rails framework towards a more complete Semantic Web application framework. Our SWORD plugin provides developers with a set of tools and libraries for managing Semantic Web data and rapid Semantic Web Application development. We describe the functionality of our SWORD plugin and demonstrate its use for rapid development of a social networking application.

When creating Semantic Web data, users have to make a critical choice for a vocabulary: only through shared vocabularies can meaning be established. A centralised policy prevents terminology divergence but would restrict users needlessly. As seen in collaborative tagging environments, suggestion mechanisms help terminology convergerce without forcing users. We introduce two domain-independent algorithms for recommending predicates (RDF statements) about resources, based on statistical dataset analysis. The first algorithm is based on similarity between resources, the second one is based on co-occurrence of predicates. Experimental evaluation shows very promising results: a high precision with relatively high recall in linear runtime performance.

Developers of Semantic Web applications face a challenge with respect to the decentralised publication model: where to find statements about encountered resources. The “linked data” approach, which mandates that resource URIs should be de-referenced and yield metadata about the resource, helps but is only a partial solution and not followed widely. We present a simple lookup index that crawls and indexes resources on the Semantic Web. Our index allows applications to automatically retrieve sources with information about certain resource. Our index is, in contrast to more feature-rich Semantic Web search engines, limited in scope and functionality and therefore simple, small, and scalable.

Since online Semantic Web applications are based on existing Web infrastructure, developing these applications could leverage experiences with and infrastructure of existing frameworks. These frameworks need to be extended to deal with the different nature of Semantic Web data. We introduce several extensions to the Ruby on Rails Web development framework to support Semantic Web application development, demonstrated in the development of a SIOC explorer. This online application integrates information from heterogeneous communities, allowing users to explore this information and find relevant posts and topics across these sites without the need to manually visit the different sites.

Object-oriented programming is the current mainstream programming paradigm but existing RDF APIs are mostly triple-oriented. Traditional techniques for bridging a similar gap between relational databases and object-oriented programs cannot be applied directly, given the different nature of Semantic Web data, as can for example be seen in the semantics of class membership, inheritance relations, and object conformance to schemas. We present ActiveRDF, an object-oriented API for managing RDF data that offers full manipulation and querying of RDF data, does not rely on a schema and fully conforms to RDF(S) semantics. ActiveRDF can be used with different RDF data stores, adapters have been implemented to generic SPARQL endpoints, Sesame, Jena, Redland and YARS and new adapters can be added easily. In addition, integration with the popular Ruby on Rails framework enables fast development of Semantic Web applications.

We present a semantic B2B gateway based on the WSMX semantic Service-Oriented Architecture to tackle heterogeneities in RosettaNet messages. We develop a rich RosettaNet ontology and use the axiomatised knowledge and rules to resolve data heterogeneities and to unify unit conversions. We use adaptive executable choreography definitions to easily integrate new sellers into existing RosettaNet collaborations.

Several recent initiatives aim to provide a Memex-inspired semantic desktop that would integrate with or replace our current physical and electronic desktop. For these semantic desktop initiatives to succeed, we need to consider how people organise their work and use their desktop. If we do not consider this existing work, the semantic desktop might very well suffer from the low adoption rate that is visible in other personal information management solutions. The contribution of this paper is not technical, but presents an overview of relevant semantic desktop literature from the personal information management and human-computer interaction domains. We extract six practical lessons: focus on the individual, forget rigid classifications, follow the links, remember the context, value the power of paper, and keep it simple.

Data on the Semantic Web is semi-structured and does not follow one fixed schema. Faceted browsing is a natural technique for navigating such data, partitioning the information space into orthogonal conceptual dimensions. Current faceted interfaces are manually constructed and have limited query expressiveness. We develop an expressive faceted interface for semi-structured data and formally show the improvement over existing interfaces. Secondly, we develop metrics for automatic ranking of facet quality, bypassing the need for manual construction of the interface. We develop a prototype for faceted navigation of arbitrary RDF data. Experimental evaluation shows improved usability over current interfaces.

Wikis are becoming popular knowledge management tools. Analysing knowledge management requirements, we observe that wikis do not fully support structured search and knowledge reuse. We show how Semantic wikis address the requirements and present a general architecture. We introduce our SemperWiki prototype which offers advanced information access and knowledge reuse.

Currently external business process descriptions (choreographies) are disconnected from the internal processes (workflows), leading to several problems. Directly mapping internal to external processes requires a quadratic amount of mappings; an intermediate ontology reduces the amount of necessary mappings but is not trivial to construct, due to the variety in workflow metamodels. In this paper we introduce our multi metamodel process ontology (m3po), which is based on various existing reference models and languages from the workflow and choreography domain. This ontology allows the extraction of arbitrary choreography interface descriptions from arbitrary internal workflow models. We also report on an initial validation: we translate an IBM Websphere MQ Workflow model into the m3po ontology and then extract an Abstract BPEL model from the ontology.

Wikis tend to be used more and more in world-wide, intranet and increasingly even in personal settings. Current wikis are data islands. They are open for everyone to contribute, but closed for machines and automation. In this paper we define a wiki interchange format (WIF) that allows data exchange between wikis and related tools. Different from other approaches, we also tackle the problem of page content and annotations. The linking from formal annotations to parts of a structured text is analysed and described.

Although most developers are object-oriented, programming RDF is triple-oriented. Bridging this gap, by developing a truly object-oriented API that uses domain terminology, is not straightforward, because of the dynamic and semi-structured nature of RDF and the open-world semantics of RDF Schema. We present ActiveRDF, our object-oriented library for accessing RDF data. ActiveRDF is completely dynamic, offers full manipulation and querying of RDF data, does not rely on a schema and can be used against different data-stores. In addition, the integration with the popular Rails framework enables very easy development of Semantic Web applications.

Semantic Wikis allow users to semantically annotate their Wiki content. The particular annotations can differ in expressive power, simplicity, and meaning. We present an elaborate conceptual model for semantic annotations, introduce a unique and rich Wiki syntax for these annotations, and discuss how to best formally represent the augmented Wiki content. We improve existing navigation techniques to automatically construct faceted browsing for semistructured data. By utilising the Wiki annotations we provide greatly enhanced information retrieval. Further we report on our ongoing development of these techniques in our prototype SemperWiki.

Cross-organisational interoperability is a key issue for success in B2B e-commerce applications. To achieve this interoperability, choreography descriptions are necessary that describe how the business partners can cooperate. In existing approaches, these choreography descriptions are independent of the internal workflows of the partners. We present a framework for extracting choreography interface descriptions from internal workflow models. The framework helps companies to reuse information from their internal process models and enables a collaborative integration scenario. Our approach consists of two steps: first we translate internal workflow models into a common formal model, then we generate arbitrary choreography interfaces from it. Our ontology (m3pl) is based upon the Process Specification Language (PSL) as a starting point for modelling internal workflows and reuses extensions defined in the First Order Ontology for Web Services (FLOWS).

In this paper we introduce our multi metamodel process ontology (m3po), which is based on various existing reference models and languages from the workflow and choreography domain. This ontology allows the extraction of arbitrary choreography interface descriptions from arbitrary internal workflow models. We also report on an initial validation: we translate an IBM Websphere MQ Workflow model into the m3po ontology and then extract an Abstract BPEL model from the ontology.

Wikis are popular collaborative hypertext authoring environments, but they neither support structured access nor information reuse. Adding semantic annotations helps to address these limitations. We present an architecture for Semantic Wikis and discuss design decisions including structured access, views, and annotation language. We present our prototype SemperWiki that implements this architecture.

Wikis are collaborative authoring environments, and are very popular. The original concept has recently been extended in two directions: semantic Wikis and personal Wikis. Semantic Wikis focus on better retrieval and querying facilities, by using semantic annotations of pages. Personal Wikis focus on improving usability and on providing an easy-to-use personal information space. We combine these two developments and present a semantic personal Wiki. Our application SemperWiki offers the usability of personal Wikis and the improved retrieval and querying of semantic Wikis. Users can annotate pages with RDF together with their normal text.The system is extremely easy-to-use, provides intelligent navigation based on semantic annotations, and responds instantly to all changes.

Web Services offer an interoperability model that abstracts from the idiosyncrasies of specific implementations; they were introduced to address the increasing need for seamless interoperability between systems in the Business-to-Business domain. We analyse the requirements from this domain and show that to fully address interoperability demands we need to make use of semantic descriptions of Web Services. We therefore introduce the Web Service Execution Environment (WSMX), a software system that enables the creation and execution of Semantic Web Services based on the Web Service Modelling Ontology. Providers can use it to register and offer their services and requesters can use it to dynamically discover and invoke relevant services. WSMX allows a requester to discover, mediate and invoke Web Services in order to carry out its tasks, based on services available on the Internet.

This document provides an overview of the results achieved by the Web Service Execution Environment (WSMX) working group. The mission and goal of the WSMX working group is to define a Semantic Web services (SWSs) architecture and to provide a complete implementation based on the conceptual model of Web Service Modeling Ontology (WSMO). Through this document we advocate that to address Semantic Web services architectural requirements as defined by [SWSA, 2004], the architecture, system components, their interfaces and system execution semantics must be standardized. WSMX is an execution environment that enables discovery, selection, mediation, invocation and interoperation of SWSs. The development process for WSMX includes establishing a conceptual model, defining its execution semantics, developing the architecture of the system, designing the software and building a working implementation of the system. The research results for WSMX provide guidelines and justification for a general SWS architecture. Deliverables of the WSMX working group define the structure of the SWS architecture, its execution semantics, mediation, discovery and invocation mechanisms, and any other functionality needed in the context of executing Semantic Web services. Work carried out by the WSMX working group, apart from its research focus, also provides a reference implementation of WSMO.

The potential to achieve dynamic, scalable and cost-effective infrastructure for electronic transactions in business and public administration has driven recent research efforts towards so-called Semantic Web services, that is enriching Web services with machine-processable semantics. Supporting this goal, the Web Service Modeling Ontology (WSMO) provides a conceptual framework and a formal language for semantically describing all relevant aspects of Web services in order to facilitate the automation of discovering, combining and invoking electronic services over the Web. This document describes the overall structure of WSMO by its four main elements: ontologies, which provide the terminology used by other WSMO elements, Web service descriptions, which describe the functional and behavioral aspects of a Web service, goals that represent user desires, and mediators, which aim at automatically handling interoperability problems between different WSMO elements. Along with introducing the main elements of WSMO, the syntax of the formal logic language used in WSMO is provided. The semantics and computationally tractable subsets of this logical language are defined and discussed in a separate document of the submission, the Web Service Modeling Language (WSML) document.

WSMX is an execution environment for dynamic discovery, selection, mediation and invocation of web services. WSMX builds on WSMO, a conceptual framework for semantically describing web services, goals, ontologies and mediators. The design process of WSMX included formally specifying the operational behaviour of the system. In general, the reasons to formally model system behaviour in a design process are: enlarging developers’ understanding of the system, proving several properties of the (model of the) future system and enabling model-driven execution of components. We present the execution semantics of WSMX and describe whether our approach addressed these requirements.

Current web service technologies lack semantic descriptions of functionality and requirements; semantic markup of web services would allow interoperability and dynamic discovery of services. The Web Service Modelling Ontology (WSMO) provides a framework for semantically describing web services, ontologies, goals and mediators. WSMX is an execution environment for WSMO allowing discovery, mediation and invocation of semantically described services. We give an overview of the current state of WSMX and demonstrate how WSMX can be used in ordering a broadband Internet line. We note two additional requirements for a web service execution environment: that it should be possible to partially defer web service descriptions until runtime and that it should be possible to execute complex goals. We describe how we augmented the software to support these requirements.

Workflow management (WFM) systems are software systems that support the enactment and management of operational business processes. Most WFM systems use an activity-based modeling methodology, which ignores the difference between coordination and production acts. DEMO is a methodology that does adhere to this notion; in addition, the use of DEMO also leads to a more precise definition of certain concepts in WFM theory. To demonstrate that it is feasible to design a WFM system based on DEMO, the design of such a system is introduced.