|10:00 - 19:00|
VENUE: University College LondonChemistry Lecture Theatre
Christopher Ingold Building, Ground Floor
20 Gordon Street
London WC1H 0AJ [Location]
conference Linked Data: The Future of Knowledge
Organization on the Web
was ISKO UK's most successful event yet. The programme of nine speakers
from the UK and Europe attracted 160 registrations with yet more on a
waiting list. On the day, 150 of these attended, all but filling the
lecture theatre to capacity.
Several speakers referred to Tim Berners-Lee's thoughts on Linked Data principles, which he first wrote about in July 2006. The idea is simple: you can add value to your information by linking it to that of others.
The conference set out to examine just how linked datasets can be published and how value can be added by linking them using Semantic Web technologies like URIs, RDF and SKOS supported by traditional KO meta-structures like ontologies and thesauri. The intention was also to present some real-world examples of Linked Data and to discover who are the current leaders in the field.
The conference revealed not only what huge amounts of data remain to be made available, but also some of the barriers to its free publication. Although the technology may not yet be fully mature, it is commercial interests, copyright, and sometimes plain ignorance of its potential value which pose the greatest barriers to progress.
Linked Data: The Future of Knowledge Organization on the Web was organized by ISKO UK in collaboration with the UCL Department for Information Studies and was sponsored by Talis and PoolParty.
We are also very grateful to Conrad Taylor for recording and photographing the event on our behalf.
See also outputs from the earlier related ISKO UK event.
Summaries of the talks and biographies of the speakers are presented below. For a full report on the conference, please refer to the ISKO UK blog.
Please note that speakers' presentations are available on SlideShare. Audio recordings of the talks synchronized to the slides, will be available in due course.
Linked Data: A
Tipping Point for the Semantic Web
Governments around the world collect information. This information is often non-personal public sector information. It relates to transport and education, health and the environment, business and leisure. Publishing and releasing this information in a machine readable way could lead to huge economic and social improvements. It would also make the process of Government more transparent and give rise to new classes of information business.
This talk outlined developments in the UK and US which are showing the potential of building a Web of Open Linked Data using Semantic Web Standards. It discussed the technical as well as social and cultural challenges. It gave examples of the applications that can result and considered the impact this could have across a number of dimensions.
Deputy Head (Research) of the School of Electronics and Computer
Science at the University of Southampton. He is a Director of the Web
Science Trust, and of the Web Foundation – organisations that
seek to advance our understanding of the Web and promote the Web's
positive impact on society.
Isaac (Netherlands): SKOS
and Linked Data |
SKOS (Simple Knowledge Organization Systems) is a vocabulary to express data on knowledge organization systems in RDF. SKOS is meant to be as easy of use as possible, fitting existing practices while keeping general enough to fit a high number of use cases and KOS configurations. In this talk a short introduction to SKOS was given. A specific emphasis was put on how SKOS allows connecting knowledge organization systems together, a key feature for scenarios that aim at leveraging existing vocabularies on the web of data and making them truly linked resources. SKOS can thus be seen as one of the crucial pieces of technology to build a conceptual layer for the linked data cloud--connecting concepts as a "glue" complementary to the identification of "real-world entities" which is one of the current foci of the linked data initiative.
This talk presented how this vision is being adopted in a number of projects, in the Cultural Heritage sector and beyond. The speaker also gave an idea of the main challenges that should be tackled, based on the work of the W3C Library Linked Data incubator group. Among these, two crucial issues are the identification of links between KOSs different organizations and the connection of these KOS to descriptions of objects from different contexts.Antoine Isaac works as scientific coordinator for Europeana and researcher at the Vrije Universiteit Amsterdam. He has been working for a number of years on using Semantic Web technology in the Cultural Heritage environment, focusing on the representation and interoperability of collections and their vocabularies (STITCH, TELplus and EuropeanaConnect projects). He has been a member of the W3C Semantic Web Deployment Working Group, involved in the design of SKOS. He currently serves as co-chair of the Library Linked Data incubator group and SKOS community contact. Read more at http://www.few.vu.nl/~aisaac.
Europeana is a single access point to the digitalised European cultural and scientific heritage. The online portal was launched by the President of the EU Commission in November 2008 and currently provides access to over 9 million materials from the European libraries, museums, archives, galleries, and audiovisual collections. More than 1000 memory institutions contribute cultural content in Europeana in a wide range of formats (images, texts, audio and video) and in all European languages.
Linked Data Journey
For many dipping their toe in the world of Linked Data, the question of how to approach implementation soon arises. What steps do I need to take? Do I need to be coding everything in RDF? Do I need to be opening all my data to the Web? What does minting URIs mean? Do I need a triple store? What about licensing? — all valid questions that learning from others will help answer.
Linked Data has only been around for a short while, but it has rapidly gained visibility in the wider community. In addition to the obvious enthusiasts, significant organisations are engaging with and implementing Linked Data services. These organisations, which include the BBC, UK Government, and the New York Times, did not introduce these Semantic Web techniques and technologies overnight. By observing the journeys that these organisations have taken, Richard identified the approaches used, lessons learnt, and the resulting successful outcomes.
The journey that these organisations have taken include stops along the way where things did not work out quite how they expected, such as projects that moved on without Linked Data, and releases of data that were criticised by the wider community. Other, more positive, experiences include the improvement in navigation, and search engine optimisation, for Linked Data backed web sites. As well as this overview of how others have introduced and implemented Linked Data in their organisations, Richard highlighted the benefits of doing things in a Linked Data way.
Richard Wallis has been with the UK's leading Library Systems and Semantic Web technology company, Talis, for over twenty years. This coupled with his passion for and involvement with new and emerging technology trends, gives him a unique perspective of the issues challenging Information professionals today. As a Technology Evangelist he is at the forefront in promoting, explaining, and applying new and emerging Web and Semantic Web technologies in the wider information domain. Richard is an active blogger and regular podcaster in the Talking with Talis series.Talis started 40 years ago as a co-operative working on data management and sharing. Based at the University of Birmingham, Talis applied the latest technology to the ongoing task of keeping their libraries in order. 40 years on, Talis has grown to become the UK market leader in library management, and has kept at its core the idea of shared innovation through technology.
Dale (UK): Linked
Data in local government - The Knowledge Hub |
The Knowledge Hub (Khub) is a “Web 2” social media development and builds on the IDeA Community of Practice (CoPs) initiative which, with 60,000 registered users, is already the most advanced online practitioner group in the public sector.
The development has four
He is both an evangelist and practitioner in the use of Web 2.0 technologies and Social Media applications to support personal learning and knowledge sharing.
He was the business lead and information architect for the community of practice platform currently deployed across the UK local government sector, the largest network of its type, and continues to play a key role in the support of virtual communities of practice for value creation in public services.
He is a regular contributor to professional journals and magazines and was recently appointed as Chairman of Incisive Media’s Online Information Executive Committee (and frantically planning for this year's conference on 30th Nov - 2nd December!).
Hepp (Germany): Linked
Data in E-commerce - The GoodRelations ontology
More than 50% of a developed nation's Gross Domestic Product is used for establishing and maintaining the exchange of goods and services, and a large share of that is consumed for the search for potential suppliers and consumers. A key variable that determines that effort is the specificity of the objects being exchanged, which is generally on the rise: We produce and consume much more specific objects than a decade ago.
In this talk, Prof. Hepp outlined how Linked Data can be used to weave a giant graph of information about products, offers, stores, and related facts. This will reduce the effort for business matchmaking on a Web scale. Centerpiece of that graph is the GoodRelations ontology, a global schema for exposing such facts as Linked Data on the Web. GoodRelations is the second most popular conceptual schema on the Web of Data and one of the few examples of academic research in the field that has been adopted by several Fortune 500 companies, like BestBuy or Yahoo.
More information on GoodRelations is at http://purl.org/goodrelations/.
Martin Hepp is a professor of general management and e-business at Universität der Bundeswehr München in Germany, where he heads the e-business and Web Science Research Group. He holds a master’s degree in business management and business information systems and a PhD in business information systems from the University of Würzburg (Germany). His key research interest is in using structured, linked data on a Web scale for e-business, in particular matchmaking and product data reuse. As part of his research, he developed the GoodRelations and eClassOWL ontologies, now widely used by companies like BestBuy and Yahoo for describing offers on the Web.
Also, he was the organizer of more than fifteen workshops and conference tracks on conceptual modeling, Semantic Web topics, and information systems and member of more than sixty conference and workshop program committees, including ASWC, ESWC, IEEE CEC/EEE, and ECIS, and is an associate editor of the International Journal on Semantic Web and Information Systems (IJSWIS).
More information and publications are available at http://www.heppnetz.de/.
Powell (Eduserv, UK): Linked
Data - The Long and Winding Road
On June 17-19, 1998 the first meeting of the Dublin Core Datamodel Working Group took place at the offices of OCLC in Dublin, Ohio. We didn’t know it at the time but the meeting represented the first real step on what has since become a long journey undertaken by the Dublin Core Metadata Initiative to align itself with the Semantic Web and, in particular, with the RDF model. That journey was used to highlight and discuss some of the issues now faced by the Linked Data community as it seeks to become the de facto approach to sharing data openly on the Web.
Andy Powell is Research Programme Director at Eduserv where he is responsible for specifying and delivering a programme of internal and external research and standards-making activities. His primary areas of interest include: metadata, the Semantic Web and Linked Data; repositories, research data, resource discovery and scholarly communication; identity and access management.
Andy has been active in the Dublin Core Metadata Initiative since about 1996. He is a member of the DC Advisory Board and was previously a member of the DC Usage Board and chair of the DC Architecture Working Group. Andy jointly authored the DCMI Abstract Model and several other Dublin Core technical specifications. More recently he helped create the RDF-based Scholarly Works Application Profile for the JISC. He has also been involved in a wide range of other digital library projects, services and standards-making activities.
Goodwin (UK): Linking
to Geographic Data
Most data references some kind of location whether it be a place name, address, postcode or some kind of coordinate. It is clear that location provides an important data integration hub on the linked data web. This talk discussed the challenges around constructing linked data for geographic and spatial information with particular emphasis given to the work done at Ordnance Survey. Examples of other geo-spatial linked data in the wild were given along with examples of applications built using them. Some discussion occurred on some of the services being built to get the most from geo-spatial linked data and the challenges we will face in the future.
Prior to joining Ordnance Survey, John spent ten years studying mathematics at university, culminating in a PhD in Relativity. John joined Ordnance Survey's IT department in 1998 where he stayed for two years. After a brief stint in the private sector, John returned to Ordnance Survey in 2001 where he moved to the research department and was tasked with looking at something called the 'semantic web' and determining possible implications it might have on Ordnance Survey as an information provider.
Initial work involved constructing ontologies in the Web Ontology Language (OWL) to describe Ordnance Survey data to aid data integration and reuse. In more recent years, John's focus has shifted to the linked data web. To this date John has been responsible for producing Ordnance Survey's linked data offerings.
(punkt. netServices, Austria):
SKOS Thesaurus Management utilizing
Linked Data |
Building and maintaining thesauri are complex and laborious tasks. PoolParty is a Thesaurus Management Tool (TMT) for the Semantic Web, which aims to support the creation and maintenance of thesauri by utilizing Linked Open Data (LOD), text-analysis and easy-to-use GUIs, so thesauri can be managed and utilized by domain experts without needing knowledge about the semantic web. Some aspects of thesaurus management, like the editing of labels, can be done via a wiki-style interface, allowing for lowest possible access barriers to contribution.
PoolParty can analyse documents in order to glean new concepts for a thesaurus. Additionally a thesaurus can be enriched by retrieving relevant information from Linked Data sources and thesauri can be imported and updated via LOD URIs from external systems and also can be published as new linked data sources on the semantic web.Andreas Blumauer is co-founder of punkt. netServices (established in 1998), a Vienna based provider of knowledge management systems and semantic technologies. He is lecturer at several universities in the area of semantic knowledge management. In 2003 he founded the Semantic Web Company which is a consultancy company for Semantic Web with a focus on technology consulting, media economy and metadata management. He has an academic background in Computer Science and Business Administration.
Andreas edited two compendia on the Semantic Web for Springer which were released as one of the first books on that topic for the German speaking community.Andreas currently works in several consulting projects for customers in the financial sector, media industry and engineering industry.
Vatant (France): Porting
terminologies to the Semantic Web
Terminologies have been developed for years in the closed world of enterprises, targeting the specific technical needs or specific communities of users. Their aim is to ease semantic interoperability across resources and systems dealing with well-defined, vertical domains. On the other hand, Semantic Web technologies and the growing Linked Data Cloud are deploying in a global scope, using a unified system of identifiers (URI), a generic data model (RDF) and the universal HTTP protocol to identify and exchange description of resources.
The glue between the terminology world and the Semantic Web will be ensured by nothing but vocabularies published in RDF, and currently SKOS is the favourite language for such publication. But SKOS has been built on a concept-centric model, leveraging mainly the thesaurus world, standards and best practices. In SKOS, concepts and their semantic relationships are defined independently of the terms used to name them (labels). SKOS is quickly becoming the lingua franca to migrate legacy vocabularies to the Semantic Web across the librarian community.
The SKOS-XL extension makes provision for description of terms themselves, considering them as first-order citizens, allowing the description of more specific relationships to concepts than just “preferred” and “alternative”, as well as other fine-grained information such as context of use, translation, acronyms, lexical variants. But does this (non-normative) extension meet the requirements of terminologists? Does it meet the requirements of terminology standards such as developed by ISO/TC 37? How is the terminology community involved in this process?
As approaches to this issue, the speaker first quickly presented the model underlying the new management system for EUROVOC, a vocabulary presenting itself as a thesaurus, but with extensions of expressivity at the terminological level. Bernard also looked at the lexvo.org initiative, which proposes a semiotic approach to terminology in the Semantic Web framework.Bernard Vatant is a former maths teacher graduated from the Ecole Normale de l’Enseignement Technique (ENSET) in 1975. Bernard Vatant integrated the Mondeca team as a consultant in the domain of ontologies and knowledge representation languages in 2000. His experience in modelling and data migration has built up thanks to the diversity of Mondeca’s customers and projects, in domains as various as scientific and medical terminologies, tourism and local government, or legal publication.
has been the representative for Mondeca in several working groups or
standard bodies such as ISO (ISO 13250 Topic Maps standard,
ISO 25964 standard for Thesauri), OASIS (former chair of the
Technical Committee on Published Subjects), or W3C (OWL, SKOS,
currently Invited Expert in the Library Linked Data Incubator Group).
|Panel discussion: The Semantic Web: Threat or Opportunity for Conventional KO?|
Poster DisplaysChristophe Dupriez (Belgium): Integrating applications and projects: Dynamic and repeatable transformation of existing Thesauri and Authority lists into SKOS + Cross-tabulation of Concepts Linked Data
The Belgium Poison Centre is moving from one strongly integrated database to different open source applications, each focussed on a domain. The users are continuously asking for "What is linked to XYZ?": we have to provide an ergonomic and efficient way to retrieve concepts somehow related to "XYZ" and present those concepts with links toward their different uses in one or another application.
ASKOSI (Application Services for Knowledge Organisation and System Integration) has been built to collect existing Thesauri or Authority lists downloaded from an external URL or dynamically viewed within the relational database of an independent application or read from a local file (Excel, XML, RDF, SKOS, OWL). The different applications can then register in ASKOSI how and how much they are using each concept of the Thesauri or Authority lists.
The users can then search, browse the concepts, sort their usages by the different applications and branch to the linked data. Especial care has been taken to support the volume of data managed within the response time constraints of a Poison Centre. We present our current results, our current challenges and how the ASKOSI project is open to sharing and external collaboration.
Christophe Dupriez is a computer scientist working with libraries, poison centres, universities and many other knowledge based organisations since 1979. Current projects are WindMusic.org (Music Library of the CDMC Haute Alsace) and with the Belgium Poison Centre, using DSpace and other open source applications integrated under the umbrella of ASKOSI.org.
Barry Norton (Germany):Linked Open Services
Poster not supplied
Linked Open Services are a radical re-evaluation of the means to produce, describe and compose services on the Semantic Web. They harness the technologies favoured by Linked Data - RDF, SPARQL and HTTP - rather than imposing rule languages, SOAP and WSDL. Like Linked Data, and unlike traditional Semantic Web Services, they aim towards contribution to the knowledge on the Semantic Web. They are open in the sense that they are openly exposed, using only the Web as a platform, not hidden away inside closed services infrastructures. To these ends Linked Open Services are based, like Linked Data, on a simple set of principles to be explored, refined and supported in an open, community-driven manner.
Barry Norton is a postdoc in the AIFB institute and in KSRI at KIT. He is a work package leader in the SOA4All EU FP7 Integrated Project, as he was previously at the University of Innsbruck, and in the completed IPs SUPER and DIP in Framework 6 at the Open University in the UK. He holds a PhD in Computer Science from the University of Sheffield, where he was in the Lirics, Dot.Kom and AKT projects. He is co-chair of the OASIS Technical Committee on Semantic Execution Environments, and editor of the Semantic SOA Reference Ontology.