Semantic Web services (SWS) aims at extending traditional Web services
with machine-readable semantic descriptions of their functionality and
interfaces in order to increase the degree of automation for
service-based applications, e.g., by allowing the discovery, binding
and composition of services to be performed automatically.
This talk will provide a quick introduction to Semantic Web Services,
will discuss what have been the past achievements in this research area. The talk will also try to
analyze what are the problems that are hindering semantic web services to be largely adopted and how
future work in the area can contribute to solve such issue.
2. 2 Federico M. Facca federico.facca@sti2.at RetoKrummenacher reto.krummenacher@sti2.at http://www.sti-innsbruck.at
3. Semantic Technology Institute Innsbruck Institute at the University of Innsbruck (est. 1669) which is currently the largest education facility in Austria. Founded as a research group under the guidance of Prof. Dieter Fensel in 2003. Status of a research institute at the University of Innsbruck since January of 2006. Main research areas: Semantic Web, Semantic Web Services, Service-Oriented Architectures. 3
4. Projects Currently involved in a number of FP6 and FP7 EU projects related to the Semantic Web and Semantic Web Services such as 4
6. Making this real…STI International The mission of Semantic Technology Institute Internationalis to establish semantics as a core pillar of modern computer science. STI is organized as an association of jointly interested academic, industrial and governmental parties. It provides services to facilitate research, education, and commercialization activities around semantic technologies and the service web beyond the boundaries of individual projects or initiatives.
9. 9 Overview Background and Motivation Service Web Semantic Web Services SOA4All: A Global Service Delivery Platform Highly Flexible Service Offer for the Future Internet Conclusion
11. The Rise of the Service Economy [IBM Survey on national labor data, 2004] 11
12. Background Computer science is entering a new generation The previous generation was based on abstracting from hardware The emerging generation comes from abstracting from software and sees all resources as services in aservice-oriented architecture(SOA) In a world of services, it is the service that counts for a customer and not the software or hardware components that implement the service Service-oriented architectures are rapidly becoming the dominant computing paradigm 12
13. From SaaS to XaaS In a service-oriented world everything is a service Programs are services Devices are services Different types of media (audio, video, text) are integrated Environments are dynamic and open Mobility; Ubiquity; RFID Service orientation needs to scale up to open and dynamic environments of billionsof services 13
14. XaaS: Amazon – S3 & EC2 “Infrastructure as a service” Amazon Simple Storage Service (S3) Write and read objects up to 5GB 15 cents GB / month to store 20 cents GB / month to transfer Amazon Elastic Compute Cloud (EC2) allows customers to rent computers on which to run their own computer applications virtual server technology 10 cents / hour 14
15. State of affairs Current SOA solutions are however still restricted in their application context to companies’ intranets A ‘Service Web’ with billions of services depends on resolving fundamental challenges that SOA does not address currently Currently there exists only around 30000 Web services on the Web Number of Web services found during the past 26 months [seekda.com, August 2009] 15
17. Requirements for Service Web A Service Web with billions of services can be realized only if SOA can deal with Openness– everybody can act as a provider or consumer of services Heterogeneity– services are created in isolation from one another thus interoperability is an issue Distributedness– there is no central control of services. Services can appear, change or disappear at any time in an uncontrolled fashion Scalability– with so many services available on the Service Web the Human may become the bottleneck 17
18. How to enable Service Web? A Web-scale service delivery platform Any time and anywhere service consumption Heterogonous execution platforms New paradigms to engineer, integrate, deploy services Flexibility Customization Semantics as scalability enabler Service customization Service federations 18 [Prof. dr. Lutz Heuser, SAP: “Towards afuture Internet of Services”]
19. Semantics and Service Web Semantics is a required key enabler for automation of the service life-cycle at Web-scale, but if misused it may become a bottle-neck It does not make sense to describe (or assume that) Amazon services in a complete way (i.e. using ~30 billions RDF triples!) In a world of billions of services it may cost too much to find the “optimal” service in relation to the reward of having actually found the optimal solution Pragmatic approaches in service discovery will focus on utility, i.e., stop the search process when a service is found that is “good” enough to fulfill a request Also, it is unrealistic to assume that semantic descriptions of services are correct and complete, i.e., duplicate the functionality of a service at the description level 19
21. Semantic Web and Web Services 21 It’s all about automation! Web Services UDDI, WSDL, SOAP Semantic Web Services Dynamic WWW URI, HTML, HTTP Semantic Web RDF, RDF(S), OWL, etc. Static
24. Interfaces (usage) Provide the formally specified terminology of the information used by all other components Connectors between components with mediation facilities for handling heterogeneities
29. Used mediators OO Mediators (ontology import with terminology mismatch handling)Ontology Elements: Concepts set of entities that exists in the world / domain Attributes set of attributes that belong to a concept Relations define interrelations between several concepts Functions special type of relation (unary range = return value) Instances set of instances that belong to the represented ontology Axioms axiomatic expressions in ontology (logical statement) 25
31. The Web Service Element WSMO Web service descriptions consist of non-functional, functional, and the behavioral aspects of a Web service A Web service is a computational entity which is able (by invocation) to achieve a users goal. A service in contrast is the actual value provided by this invocation 27
35. Example: Web Service Discovery 34 Web service: sells train tickets for trips within Europe Goal: buy a travel ticket from Vienna to Berlin Reasoning Travel Ticket Europe Train Ticket Match! Vienna & Berlin
37. Mediators Mediation Data Level - mediate heterogeneous Data Sources Protocol Level - mediate heterogeneous Communication Patterns Process Level - mediate heterogeneous Business Processes 36
38. Mediators Four different types of mediators in WSMO ggMediators: mediators that link two goals. This link represents the refinement of the source goal into the target goal or state equivalence if both goals are substitutable ooMediators: mediators that import ontologies and resolve possible representation mismatches between ontologies wgMediators: mediators that link Web services to goals, meaning that the Web service (totally or partially) fulfills the goal to which it is linked. wgMediators may explicitly state the difference between the two entities and map different vocabularies (through the use of ooMediators) wwMediators: mediators linking two Web services 37
39. The WSMO Framework 38 Conceptual Model for SWS Execution Environment for SWS Formal Language for WSMO Ontology & Rule Language for the Semantic Web
41. 40 Motivation The Web currently contains 30 billion Web pages Children can create Web pages BUT the Web contains only ~28,000 ‘true’ Web services (seekda.com) Only technologically experienced people can create and work with Web services
42. 41 Two Core Objectives “Billion of Services”: SOA4All will transform the Web into a domain where billions of parties are exposing and consuming services in a seamless and transparent fashion. “4 All”: SOA4All will integrate the service world of large enterprises, SMEs, and end-users enabling them to engage as peers within a network of equals. http://www. .eu
43. 42 Approach Context: user profiles, execution monitoring, service data, social context Web: openness, decentralization, n:m relations, statelessness Semantics: formal models, service and goal descriptions, processes Web2.0: content prosumers, service prosumers, communities
44. 43 SOA4All Architecture ‘semantic service descriptions’ ‘semantic process descriptions’ ‘semantic goal descriptions’
45. 44 Semantic Spaces Use of semantics in SOA4All requires a scalable and distributed data management infrastructure for: Repository for service annotations in RDF Infrastructure for sharing monitoring and execution data Process repository of composition information User profile management infrastructure Semantic Spaces provide: Web-style publish and read operations (persistent storage) Shared data management Interaction mechanism for collaborative activities Event-based notification services
46. 45 Annotation of Services Representation Languages WSML Reasoners Annotation Mechanisms WSMO-Lite, MicroWSMO
48. 47 Lightweight Service Modelling A common service model is expressed in RDF Schema, using only the WSMO features motivated by SAWSDL references WS-* Stack services attached to lightweight semantic descriptions via SAWSDL RESTful services attached to lightweight semantic descriptions via microformants
49. WSDL Simplified 48 Web service input Operation 1 output input Operation 2 . output . . input Operation N output
50. 49 Semantics in Service Model F N B I SAWSDL modelReference Web service input Operation 1 output input Operation 2 . output . . input Operation N output Functional, Non-Functional, Behavioural, Information model
51.
52. hRESTS allows aspects of the service description to be annotated
53.
54. Goal Formalisation Semantic goal descriptions match the WSMO-Lite service annotations. SPARQL can be used as simplest discovery algorithm by matching operations, input, and output messages. More sophisticated matching (based on conditions, effects or NFPs) requires axiomatic reasoning (e.g. WSML). 52
55. 53 Example: Service Discovery SOA4All Studio: Consumption Platform ervice G SOA4All Runtime: DSB & Platform Services S Ranking & Selection Discovery Q O ntology Crawler Reasoner O S Communication via DSB S Service Registry uery O Q S Semantic Space oal G
56. 54 Example: Service Discovery SOA4All Studio: Consumption Platform S ervice G SOA4All Runtime: DSB & Platform Services S Ranking & Selection Discovery O Q ntology Crawler Reasoner O S Communication via DSB S Service Registry uery O Q S Semantic Space oal G
57. Service Composition SOA4All Studio: Provisioning Platform oal G P SOA4All Runtime: DSB & Platform Services Design-Time Composer Template Generator ntology Reasoner Composition Optimizer Neglected is the monitoring data that is provided by the DSB to the Composition and Execution. O O Q Execution Engine O uery „DISCOVERY“ Q G Semantic Space rocess P P Communication via DSB
58. Service Composition SOA4All Studio: Provisioning Platform oal G P SOA4All Runtime: DSB & Platform Services Design-Time Composer Template Generator ntology Reasoner Composition Optimizer Neglected is the monitoring data that is provided by the DSB to the Composition and Execution. O O Execution Engine O uery „DISCOVERY“ Q G Semantic Space rocess P P Communication via DSB
60. From one that fits allto personalized software Traditional software engineering and provisioning solutions suffer from lack of flexibility A software to fits all type of customers Modern trends in products variability showed how customization increase revenues Web scale delivery of customized software How can we achieve mass customized software as with traditional products? Economy showed that the only way to enable small competitors to stay on the market is by federating and providing high-added-value service bundles Dynamically created federations of services to better match user’s demand How can we enable providers to federate together at web-scale with a high degree of automation? 58
61. Parametric Services and Semantics High level service customization can be achieved by making services parametric Automatic deploy-time and run-time customization of parametric services requires proper languages and methods Semantics enable description of such aspects and automatic reasoning over them through application of problem solving methods and parametric design 59
62. Service Federations and Semantics Global scale delivery of services, including services provided by small providers can be achieved by automated federation of services Requires tools and languages for enabling negotiation among services and service providers Semantics is the means to enable negotiation among providers, supporting heterogeneity resolution and making possible optimization of the federation via reasoning techniques and problem solving methods 60
64. Conclusion Future Internet requires: Platforms and languages for Service Web Methods and languages for mass customization of services Semantic Web techniques can be used to provide approximate descriptions of services … … however not as a replacement of service technology. 62
65. Summary 63 Global service delivery Web services stagnate Semantic Web services SOA4All
Notes de l'éditeur
Human becomes the bottleneck. We need something to support humans in the service related tasks.
Representation languages are used to describe ontologies for a particular domainA service description is annotated with elements from these ontologiesA reasoner can answer questions about these service descriptions.The instances are services that are associated with the financial service category