Integration via Meaning: Using the Semantic Web to deliver Web Services Brian Matthews, Business & Information Technology Dept, CLRC b. m. matthews@rl. ac. uk
Who am I? • Group Leader, Information Science & Engineering, Business & Information Tech Department – Advanced development of business Information Systems – Developing research programme in information technology and distributed systems • Deputy Manager, W 3 C Office for the UK and Ireland Brian Matthews, CRIS 2002, 30/08/02
What’s in this talk? • A look at some initiatives in the Web world. – Web Services and the GRID – The Semantic Web • How they fit into a bigger picture – Towards a Universal Distributed Information System – GRIDS • The problems along the way – Metadata, Workflow, Trust • The CLRC Information Systems research programme – And how it is working towards the UDIS • Conclusions Brian Matthews, CRIS 2002, 30/08/02
What’s in this talk? • A look at some initiatives in the Web world. – Web Services and the GRID – The Semantic Web • How they fit into a bigger picture – Towards a Universal Distributed Information System – GRIDS • The problems along the way – Metadata, workflow, trust • The CLRC Information Systems research programme – And how it is working towards the UDIS • Conclusions Brian Matthews, CRIS 2002, 30/08/02
W 3 C Initiatives • Two major initiatives of the W 3 C (and others) are creating much interest at the moment: – Web Services – The Semantic Web • Both wish to develop the web into a system of the future which is: – Seamless, interactive and automated. • How do they fit together? Brian Matthews, CRIS 2002, 30/08/02
Web Services • Web Services are perhaps the most heavily promoted information technology at the moment “Web Services are XML representations of programs, objects, messages or documents accessible over the Internet for the purpose of direct application to application interaction” IONA Technologies, http: //www. xmlbus. com/learn/webserviceswp. pdf • Generally the consensus is that Web Services are programs which: – Are accessible over the Web – Expose an XML interface – Can be located through a Web service registry – Communicates using XML messages over standard protocols – Allows application to application – no human, no browser Brian Matthews, CRIS 2002, 30/08/02
Web Service Components • Web-services provide a loosely coupled middleware architecture for a global distributed computation system • Simple Object Access Protocol (SOAP) – SOAP v 1. 2 (IBM & MS) – W 3 C XML Protocol WG • Web Services Description Language (WSDL) – WSDL v 1. 1 (IBM & MS) W 3 C Note Mar. 2001 • Universal Description, Discovery and Integration (UDDI) – UDDI. org v 2. 0 specification June 2001 • Also: WSIL, WSFL, SAML, XACML …. • Backed by Microsoft, IBM, Sun, Apache … Brian Matthews, CRIS 2002, 30/08/02
Web Service Architecture UDDI 2. Queries Directory 1. Registers with Directory Company Y 3. Gets Web Service info 4. Negotiates service agreement Company X WSDL Web Service 5. Access service using SOAP and XML Messages Brian Matthews, CRIS 2002, 30/08/02
Other approaches • Other approaches to producing a global distributed computation system : – – Enterprise Java Beans eb-XML CORBA The GRID • All now converging on Web Services as a basis. A few words on the GRID… Brian Matthews, CRIS 2002, 30/08/02
The GRID • The Grid provides persistent environments that enable software applications to integrate computational and information resources that are managed by diverse organisations in widespread locations. – E-Science means science done through distributed global collaborations, using very large data collections, tera-scale computing resources and high performance visualisation. ” • The GRID was conceived as the successor to the Web – “the Web on steroids” – But the Web has caught up! Brian Matthews, CRIS 2002, 30/08/02
NASA Information Power GRID: Collaborative Engineering SFO Hangars Digital White Board United MOC Line Mechanic Wireless Digital Bridge - Video - Audio - Non- Destructive Imaging - Sketchpad CAD/CAE - Portable Maintenance Aid Printer CAD/CAE Table CAD/CAE Printer Boeing Rapid Response Center Remote Displays Digital White Board Virtual Iron Bird United Maintenance Systems Boeing Maintenance Systems NASA Ames Aviation Extra. Net/ & IPG Brian Matthews, CRIS 2002, 30/08/02 Remote Displays Digital White Board CAD/CAE Table CAD/CAE Printer
Web vs. Grid Services • Web Services are usually characterised by: – Long-life times – once a service has been published it is persistently available. – Stateless interaction – a normal http model. • Grid Service requirements: – Dynamic creation of new services – Long-lasting, state-based interactions. • Reconciliation is necessary – – and Web-Services will evolve this way. Brian Matthews, CRIS 2002, 30/08/02
GRID Architectures • Leading GRID tools developed by the Globus project – www. globus. org • Standardisation through the Global Grid Forum • Open Grid Service Architecture is based on Web Services – Convergence predicted Brian Matthews, CRIS 2002, 30/08/02
The Semantic Web • The Web is one HUGE data structure • However, when you build a data structure, you know meaning behind it • The Web is chaotic - why are resources are linked? – Imagine a library where all the books have the same text on the cover, and the only catalogues are compiled by photocopying the books, cutting up the copies, and arranging the words in the order of frequency. Johan Hjelm • Thus Google is great at returning all the pages on the web that mention "Tim Berners-Lee“ – – But what about returning those pages written by Tim Berners -Lee? • The Semantic Web adds well-defined meaning to describe the Web (Metadata). Brian Matthews, CRIS 2002, 30/08/02
Add Meaning to Resources Brian Matthews, CRIS 2002, 30/08/02
A Layered Architecture Brian Matthews, CRIS 2002, 30/08/02
Annotation: a Semantic Web Application • Allows user to add comments to other web sites • And make comments on the comments • Uses RDF Metadata Brian Matthews, CRIS 2002, 30/08/02
Semantic Web – current status • The Semantic Web has been around several years: – – Base technologies well-established Gone through several iterations Lots of academic interest Convincing applications are still missing • However, many demonstrators and interesting applications – Dublin Core, Thesauri, Annotations. – CC/PP, P 3 P, PICS, – Need to demonstrate the benefit of a common framework. Brian Matthews, CRIS 2002, 30/08/02
What’s in this talk? • A look at some initiatives in the Web world. – Web Services and the GRID – The Semantic Web • How they fit into a bigger picture – Towards a Universal Distributed Information System – GRIDS • The problems along the way – Metadata, workflow, trust • The CLRC Information Systems research programme – And how it is working towards the UDIS • Conclusions Brian Matthews, CRIS 2002, 30/08/02
Standing Back: Towards a UDIS • We have two approaches to the evolution of the Web: – Web Services – access to computation as well as documents – Semantic Web – adds information to the structure of the Web. • How might we combine these to provide a Universal Distributed Information System ? • A starting point: Keith Jeffery has proposed a three level architecture… Brian Matthews, CRIS 2002, 30/08/02
GRIDS The Knowledge Grid INDIRECTION DATA TO KNOWLEDGE The Information Grid The Data & Computation Grid Brian Matthews, CRIS 2002, 30/08/02 CONTROL DELEGATION
Data/Computation Grid • Base methods for accessing, invoking and scheduling data, network and computation resources. • User has to : – – – know where resources are located, understand what services the resources offer negotiate access directly with the resource owner access and invoke them directly control the order of interaction • Characteristic tool: the Browser • Where the Web (and the Grid) is now – Web services offer access to computation resources – Not very intelligent! Hard to be universal! Brian Matthews, CRIS 2002, 30/08/02
Information Grid • Resources available via descriptions of those resources (Metadata). • Discovery and negotiation of resources within domains of practice • Metadata processed via simple querying • Meaning of the metadata is agreed within a particular community. – Agreed DTDs, XML Schemas etc Brian Matthews, CRIS 2002, 30/08/02
Portals • Characteristic tool: the Portal • Users have to : – – Know where portals are located, But do not need to know where resources are located Understand the resources supplied via the agreed semantics Can negotiate access indirectly through the portal, or be passed off. • Thus there is a: – Decrease in prior knowledge of where resources are – Increase in delegation of functions to the portal. • Some intelligence has been capture in the domain schema – but the user has to do a lot! Brian Matthews, CRIS 2002, 30/08/02
Knowledge Grid • Discovery and access to resources is negotiated through the semantics of metadata. • Contextualised access to information, utilising knowledge of semantics and reasoning. • Classifying information against published ontologies, – interoperability between different semantics, • Reasoning for negotiation • Reasoning to discovery new information from old. Brian Matthews, CRIS 2002, 30/08/02
Agents • Characteristic tool: the Agent • Users have to : – Instruct their agents with their requirements for resources – Set conditions on which they will be willing to access resources (quality of service requirements, privacy, cost). – But do not need to know the location of the resources the agents use. – Do not need to understand resources in their own terms • Thus there is again a: – Decrease in prior knowledge of where resources are – Increase in delegation of functions to the agents. • Much intelligence been captured in the agent system. Brian Matthews, CRIS 2002, 30/08/02
GRIDS Architecture • Three agent types: – Resource – User – Broker • The Semantic Web provides the means of controlling agents • We thus have resource integration via meaning Brian Matthews, CRIS 2002, 30/08/02
What’s in this talk? • A look at some initiatives in the Web world. – Web Services and the GRID – The Semantic Web • How they fit into a bigger picture – Towards a Universal Distributed Information System – GRIDS • The problems along the way – Metadata, Workflow, Trust • The CLRC Information Systems research programme – And how it is working towards the UDIS Brian Matthews, CRIS 2002, 30/08/02
What is needed for GRIDS to become UDIS? • Take a user’s point of view – what do you need? – Some data to analyse – Some programmes to build and run on the data – A resource access and usage framework • In a UDIS, these are much magnified. We need: – Metadata – Workflow – Trust Brian Matthews, CRIS 2002, 30/08/02
Metadata • Information describing resources – Resources are anything “web-accessible” • Data, services, networks, facilities, people, …. • Allows actions concerning a resource without accessing it – Resource discovery (Provenance and description) – Resource negotiation (Access control and policies) – Resource usage (Type information) • Need better metadata – In wider domains – Covering aspects other than data description – Deeper content – Ontologies and more complex constraints • Need to be able to interoperate between metadata formats – Reasoning over and between metadata descriptions. Brian Matthews, CRIS 2002, 30/08/02
Workflow • How to “program” across a Web Service Infrastructure – – Combination of Web services Scheduling and integration Reconciling datatypes in APIs Monitoring and auditing • GRID requirement that this combination should be dynamic • Relating Business Process Rules to Web Service infrastructure. – High-level business rules – Lower-level XML level interfaces. – Again relating semantics Brian Matthews, CRIS 2002, 30/08/02
Trust • Establishing that the interactions between actors are trustworthy – Security: access control, authentication and authorisation and TRUST is not only security policies – Reliability and dependability – Quality ratings – Personalisation: Privacy, user preferences, accessibility TRUST is not only security • Building dynamic virtual organisations – Transferring trust from third parties – Establishing service-level agreements which can be relied upon • Establishing trust between agents that have no prior knowledge of each other – could prevent the establishment of a UDIS Brian Matthews, CRIS 2002, 30/08/02
Trust in the Semantic Web bargain rates value Ford Bank Which £ 10000 offers £ 20000 limit guarantees 935783468 Access Broker card number price Privacy http: //www. rustycars. com vendor Brian Matthews, CRIS 2002, 30/08/02 buyer
What’s in this talk? • A look at some initiatives in the Web world. – Web Services and the GRID – The Semantic Web • How they fit into a bigger picture – Towards a Universal Distributed Information System – GRIDS • The problems along the way – Metadata, workflow, trust • The CLRC Information Systems research programme – And how it is working towards the UDIS • Conclusions Brian Matthews, CRIS 2002, 30/08/02
The BITD Research Programme • CLRC’s Business & Information Tech. Department has an R&D programme developing such a vision – Establishing portal technology as common base: • using metadata and web services. – Taking Grid technology from Science into other domains: • education, commerce, government – Developing core technologies • agents, semantic web, knowledge management, Grid architectures – and their combination – Investigating key components • Metadata, workflow, trust – Disseminating results across communities • W 3 C Office, ERCIM, GGF … http: //www. bitd. clrc. ac. uk Brian Matthews, CRIS 2002, 30/08/02
Portals • Science Data Portal – Developing the “Information Grid” – General Science Metadata + Web Service based architecture – Talk on Saturday • Information Portal – Bring some of these ideas into a business context – CERIF based • Library Portal – Access to publications – Links to the above portals. Brian Matthews, CRIS 2002, 30/08/02
GRASP • Grid Based Application Service Provision • Introducing GRID technology to business applications – Combine the scalability of the GRID. . . –. . . with the usability of ASP. . . –. . . to realize powerful, low cost Application Service Provision • Key issues: – Integrating ASP and Grid technology – Resource discovery and management – Trust (security, reliability, quality of service) • Partners: – Logic. DIS, CLRC, CS-Systèmes d’Information CSSI, Consorzio CRMPA (I), Schlumberger. Sema, University of Stuttgart Brian Matthews, CRIS 2002, 30/08/02
CLRC in GRASP • Lead architects – Monitor existing GRID technologies – Ensure that GRASP complies with the emerging standards – Define all necessary middleware services (security service, information service, Qo. S. . . ) – Design the infrastructure of the ASP-GRID model • Host of the e-learning test-bed – Personalized learning environment using knowledge and goals of the user Brian Matthews, CRIS 2002, 30/08/02
Pellucid • To assist organisationally mobile workers in public sector organisations – Benefits from leveraging their own knowledge/experience and that of others who have previously held the post • Pellucid will offer active support. – Via an agent-based knowledge management system • Partners: – CCLRC, Sadiel, Softeco, Cd. G, Jd. A, MMBG, Cyronet, II-SAS Brian Matthews, CRIS 2002, 30/08/02
The Pellucid Platform • Agent-based architecture customisable to public administrations – supporting and guiding workers – managing workflows – information search and retrieval • An organisational memory – Abstract, record and store knowledge and experience. – Bring under Knowledge Management • Formalised knowledge and experience, integrated into the agents. – Agent based reasoning Brian Matthews, CRIS 2002, 30/08/02
SWAD-Europe • Semantic Web Advanced Development in Europe • Purpose is to encourage the use of Semantic Web tools and techniques now: – By an outreach programme – By developing practical demonstrators – By providing tools and standards • Many application and demonstration areas – Annotations, querying, scaleability, knowledge management, accessibility, web service integration • Partners: – Univ. of Bristol, W 3 C-INRIA, CCLRC, HP Labs, Stilo Brian Matthews, CRIS 2002, 30/08/02
CLRC in SWAD-Europe • Three major areas – Developing XML Schemas from the Semantic Web – Developing tools and techniques for representing thesauri in the Semantic Web • Especially Multilingual Thesauri – Developing tools and techniques for representing and processing Trust relationships in the Semantic Web. Brian Matthews, CRIS 2002, 30/08/02
Further application areas • i. Trust – Working group considering Trust issues • Le. GE-WG – Working group considering Grid applications in the e. Learning field • Marine XML – Developing metadata and infrastructure for distributed Marine Information Systems • CORAS – Looking at design issues around Risk Management Brian Matthews, CRIS 2002, 30/08/02
What’s in this talk? • A look at some initiatives in the Web world. – Web Services and the GRID – The Semantic Web • How they fit into a bigger picture – Towards a Universal Distributed Information System – GRIDS • The problems along the way – Metadata, workflow, trust • The CLRC Information Systems research programme – And how it is working towards the UDIS • Conclusions Brian Matthews, CRIS 2002, 30/08/02
Summary • The traditional Web infrastructure is on the verge of a step change: – Web Services making processing resources accessible – Semantic Web making resource processing automatable • Their combination allows a further step change towards a universal distributed information system – Autonomous intelligent agents guided by semantics – Reasoning for resource discovery and negotiation • Not the only one to have noted this: Semantic GRID, DAML, SWWS project, Social Science Dream Machine …. Brian Matthews, CRIS 2002, 30/08/02
Key Areas • Key areas of research to realise this: – Metadata: to provide the resource description for discovery and negotiation – Workflow: to describe and combine services to provide the required functions – Trust: to provide an assured framework for the user. Brian Matthews, CRIS 2002, 30/08/02
Will it happen? • Is this just another ambitious grand scheme? – – Yes – but it just might work! Vendor support strong Funding agencies support strong The fundamental infrastructure has changed: • Faster networks and computers, bigger storage capacity • Everyone is now connected - and used to it! • Uses Web Virtues: openness, decentralisation, standards. • Need to controlling complexity and making it easy and robust to use – It’s how the existing Web took off! – Semantic web controlled agents should enable this Brian Matthews, CRIS 2002, 30/08/02
Finally • Trust is the key – people need to have faith in a complex and impenetrable system • Realising these goals should be an exciting prospect. The Web and the Grid: from e-Science to e-Business , Euro. Web 2002 Conference – Oxford, 17 -18 December 2002 http: //www. w 3 c. rl. ac. uk/Euro. Web/ Brian Matthews, CRIS 2002, 30/08/02
Скачать презентацию Integration via Meaning Using the Semantic Web to
69b3ba2c2fbe7f00e63316c40b06bbb1.ppt