pCaoppeyrrsightby© 2th0e10 fpoarpethrse' inaduivthidoursa.l Copying permitted only for private and aepdcuaibtdloisershm.eidc paunrdposecos.pyTrihgihstedvolubmye itiss
10-Jun-2010: submitted by Christoph Lange 11-Jun-2010: published on CEUR-WS.org
iServe: a Linked Services Publishing Platform Carlos Pedrinaci, Dong Liu, Maria Maleshkova, David Lambert, Jacek Kopecky, and John Domingue Knowledge Media Institute, The Open University
Walton Hall, Milton Keynes, MK7 6AA, UK
c.pedrinaci@open.ac.uk Abstract. Despite the potential of service-orientation and the e orts devoted so far, we are still to witness a signi cant uptake of service technologies outside of enterprise environments. A core reason for this limited uptake is the lack of appropriate publishing platforms able to deal with the existing heterogeneity in the service technologies landscape and able to provide expressive yet simple and e cient discovery mechanisms. In this paper we describe iServe, a novel and open platform for publishing services which aims to better support their discovery and use. It exposes service descriptions as linked data expressed in terms of a simple vocabulary for describing services of di erent kinds with annotations in diverse formalisms. In addition to describing iServe, this paper also highlights the set of principles behind iServe, which we believe are essential for other generic repositories of semantic information notably ontology repositories. 1
Web services are software systems o ered over the Internet via platform and
programming-language independent interfaces de ned on the basis of a set of
open standards such as WSDL, SOAP, and further WS-* speci cations [
One of the main reasons for the paucity of service repositories to date has
been the fact that, although they are relatively complex, they do not support
expressive queries, limiting their usefulness [
Before any signi cant uptake of services can take place on the Web, better mechanisms for creating, publishing and discovering services must be in place. In particular, service publication must be able to deal with service heterogeneity (e.g., dealing with both WSDL services and Web APIs), it must be based on the use of lightweight semantics able to support relatively advanced yet e cient discovery, and it must be combined with an appropriate set of tools able to support users in the annotation and publication of services.
In this paper we describe iServe, a platform for the seamless publication
and discovery of services developed in the context of the EU project SOA4All.
iServe addresses the publication of services from a novel perspective based on
lessons learned from the evolution of the Web of Data [
The remainder of this paper is organised as follows. First we cover the state of the art in services description and publication (Section 2). We then introduce iServe and the core principles that underpin the approach (Section 3). Next, we discuss openness as one of the essential characteristics of the approach, and highlight how existing editors and ontology indexing systems can be connected to iServe to better support the creation and publication of service annotations (Section 4). Finally, we present our main conclusions and introduce lines for future research (Section 5). 2
The Universal Business Registry part of Universal Description Discovery and
Integration (UDDI) [
A major reason for the lack of success of repositories such as UDDI was the
fact that, although these registries are relatively complex, they do not support
expressive queries, limiting their usefulness [
The Web services ecology has recently seen a major evolution with the advent and proliferation of Web APIs and RESTful services [10], and there has not been much progress on, or even concern with, means for describing and discovering these newer kinds of services. Perhaps the most popular directory is ProgrammableWeb.com, which at the time of this writing lists 1,700 APIs and 4,600 mashups. This directory is based on the manual submission of APIs by users and currently provides simple search mechanisms based on keywords, tags, or a simple classi cation, none of which are particularly expressive. APIHut [11] is a platform that claims to increase the accuracy of keyword-based search of APIs compared to ProgrammableWeb.com or plain Google search, although it does not provide richer discovery mechanisms.
SWS [
Enhancing service repositories with semantics and supporting automated discovery has been one of the key topics SWS research has tackled [18{20]. Despite demonstrating advantages of semantic annotations in discovering services, particularly in terms of accuracy and in dealing with heterogeneous data models, SWS work has overlooked the additional complexity involved in creating semantic annotations for services. Consequently, there is no signi cant body of SWS published in a convenient manner on the Web: the largest public SWS repository today is probably OPOSSum, a test collection with less than 3000 service annotations which provides programmatic access to its content solely through direct access to the database management system [21].
The current state of the art evidences limited use of service technologies on a Web
scale, and existing technologies for publishing and discovering services remain
rather simple, providing limited support and usually based on keyword-based
search. This type of mechanism has proven to be insu cient for the needs of
software developers, hampering uptake [
The Web of Data is a relatively recent e ort, derived from research on the
Semantic Web, whose main objective is to generate a Web exposing and
interlinking data previously isolated in silos. The Web of Data is based upon four
simple principles, known as the linked data principles, which essentially dictate
that every piece of data should be given an HTTP URI which, when looked
up, should o er useful information using standards like RDF and SPARQL [
iServe is a novel and open platform for publishing semantic annotations of services based on a direct application of linked data principles to publish service annotations expressed in terms of a simple vocabulary for describing services of di erent kinds (e.g., WSDL and Web APIs) with annotations in diverse formalisms (e.g., OWL-S, WSMO-Lite). More concretely, iServe is driven by the following conclusions drawn from previous research on service repositories and the progress made by the Web of Data: { Semantics are essential to reach a minimum level of automation during the life-cycle of services; { The annotation of services should be simpli ed in as much as possible; { On the Web, lightweight ontologies together with the possibility to provide custom extensions prevail against more complex models; { Any solution to deploying services that aspires to be widely adopted should build upon the various approaches and standards used on the Web, including Web APIs, RDF, and SPARQL; { Linked data principles are an appropriate means for publishing large amounts of semantic data, both for human and machine consumption; { Links between publicly available datasets are essential for the scalability and the value of the data exposed.
In the remainder of this section we describe iServe in detail, focusing rst on the overall approach and architecture, then describing the ontology it uses, and nally presenting the import mechanisms that make it largely compatible with a wide range of service annotations. 3.1
Overall Approach iServe supports publishing Linked Services expressed in terms of a simple conceptual model that is suitable for use by both humans and machines, and which abstracts away the existing heterogeneity around service kinds and annotation formalisms. In particular iServe: { supports importing service annotations in a range of formalisms (e.g., SAWSDL,
WSMO-Lite, MicroWSMO) that cover both WSDL services and Web APIs; { provides means for publishing semantic annotations of services which are automatically assigned a resolvable HTTP URI; { includes support for content negotiation so that service annotations can be returned in HTML for human users, or in RDF for machine interpretation; { provides a SPARQL endpoint allowing advanced querying over the services annotations; { o ers a read/write REST API so that services can easily be retrieved and published from remote applications; and { automatically generates links between the published service annotations and additional documents on the Web such as the original service description or documentation so that users and machines can easily discover more information.
The architecture of iServe, depicted in Figure 1, comprises a crawler, a RESTful API, a set of import mechanisms, and an RDF store. The crawler collects existing annotations from the Web in order to publish them in iServe. Given that not many annotations are published on the Web, the crawler currently deals only with known sets of service annotations such as the SAWSDL retrieval test collection (SAWSDL-TC)2 and the OWL-S retrieval test collection (OWLS-TC)3. The RESTful API, implemented using Restlet4, provides operations for accessing service annotations and service documentation, allowing remote applications to publish and discover services. The import mechanisms provide support for importing annotations in diverse formalisms by transforming them into the Minimal Service Model, explained in more detail below, which is used by iServe for publishing them as linked data. This service model provides a common vocabulary for service annotations, smoothing away the heterogeneity of di erent formalisms such as OWL-S, SAWSDL and WSMO, allowing humans and machines to discover service annotations originally described using heterogeneous conceptual models through a single vocabulary. Finally, iServe captures the service annotations together with some provenance information including the annotation author and the creation or modi cation date in an RDF store. RDF storage and querying support is provided by Swift OWLIM,5 although it is accessed via RDF2Go6 to maintain independence with respect to the concrete store used. The RDF 2 See http://www.semwebcentral.org/projects/sawsdl-tc/ 3 See http://www.semwebcentral.org/projects/owls-tc/ 4 See http://www.restlet.org/ 5 See http://www.ontotext.com/owlim/ 6 See http://semanticweb.org/wiki/RDF2Go store provides us with a SPARQL endpoint that is made available to external applications in order to interact with iServe for retrieving services. Section 2 covered the state of the art around the publication of services and semantic annotations, highlighting the existing heterogeneity in terms of languages and formalisms used, the main approaches adopted thus far for publishing and discovering services, and the need of nding a trade-o between expressivity of service descriptions and their complexity. In building a system like iServe it is necessary to provide a common vocabulary, able to describe services in a way that allows machines to automatically locate and lter services according to their functionality or the data they handle, independent of the formalism originally used to describing them.
The best-known approaches to annotating services semantically are OWL-S, WSMO, SAWSDL, and WSMO-Lite for WSDL services, and MicroWSMO, and SA-REST for Web APIs. To cater for interoperability, iServe uses what can essentially be considered the largest common denominator between these formalisms which we refer to as the Minimal Service Model (MSM). The MSM, rst introduced together with WSMO-Lite [15], is a simple RDF(S) ontology able to capture the core semantics of both Web services and Web APIs in a common model supporting the common publishing and search of services, yet still permitting framework-speci c extensions to remain attached and thereby bene tting those clients able to comprehend those formalisms.
The MSM, denoted by the msm namespace in Figure 2, de nes Services which have a number of Operations. Operations in turn have input and output MessageContent descriptions, and Faults. A MessageContent may be composed of MessageParts which can be mandatory or optional. The addition of message parts extends the earlier de nition of the MSM as described in [15]. The intent of the message part mechanism is to support ner-grain discovery based on message parts, mirroring the granularity of SAWSDL and allowing to distinguish between mandatory and optional parts.
iServe also uses the SAWSDL, WSMO-Lite and hRESTS vocabularies, depicted in the gure with the sawsdl, wl, and rest namespaces respectively. The SAWSDL vocabulary captures in RDF the three kinds of annotations over WSDL and XML Schema, namely modelReference, liftingSchemaMapping and loweringSchemaMapping that SAWSDL supports. SAWSDL supports the annotation of WSDL and XML Schema syntactic service descriptions with semantic concepts, but does not specify a particular representation language nor does it provide any speci c vocabulary that users should adopt. The modelReference property links syntactic service elements to semantic models via URIs, while the schema mapping properties indicate data transformations from a syntactic representation to its semantic counterpart and vice versa.
WSMO-Lite builds upon SAWSDL by extending it with a model specifying the semantics of the particular service annotations. It provides a simple RDFS ontology together with a methodology for expressing functional and non-functional semantics, and an information model for WSDL services based on SAWSDLs modelReference hooks. In particular, WSMO-Lite makes explicit the intended meaning for modelReference annotations without modifying SAWSDL. Instead, it provides a vocabulary to annotate the URIs pointed to by SAWSDL annotations. The WSMO-Lite vocabulary includes the classes NonFunctionalParameter, FunctionalClassificationRoot, Condition, Effect and Ontology. With these, an annotator can type SAWSDL annotation references without adding information directly to the WSDL description.
The hRESTS vocabulary [22] extends the MSM with speci c attributes for operations to model information particular to Web APIs, such as a URITemplate to describe the URI for invocation, and method to indicate the HTTP method used for the invocation. For methods, iServe uses the draft W3C HTTP in RDF vocabulary [23]. 4
The fundamental objective pursued by iServe is to provide a platform able to publish service annotations in a way that would allow people to achieve a certain level of expressivity and re nement in discovering services, while remaining simple and convenient both for human and machine use. The simple conceptual model explained earlier is a principal building block to support this as a general model able to abstract away the existing conceptual heterogeneity among service annotation approaches without introducing considerable complexity from a knowledge acquisition and computation perspectives. Thanks to its simplicity, the MSM captures the essence of services in a way that can support service matchmaking and invocation and still remains largely compatible with the RDF mapping of WSDL [24], with WSMO-based descriptions of Web services, with OWL-S services, and with services annotated according to WSMO-Lite and MicroWSMO. Although providing a formal mapping for each of these languages is out of the scope of this paper, we note that the elements captured in the MSM are largely common to existing models. The mapping is not lossless, but appropriate use of rdfs:isDefinedBy, covered next, can help circumvent this limitation and still provide a common ground for publishing Linked Services in the Web of Data in a way that is amenable to automated processing and where more expressive de nitions can be linked if needed.
iServe provides a set of import mechanisms that can take service annotations in several formats, and generate the appropriate RDF in terms of iServe's conceptual model. In particular, the current version can import SAWSDL, WSMO-Lite, MicroWSMO, and part of OWL-S descriptions. The import process generates rdf:isDefinedBy links from the service annotation to the original description le (e.g., the WSDL including the annotations), rdf:seeAlso to any documentation about the service (e.g., the Web API description page itself), and owl:sameAs relations to the published version of the same service annotation in other systems. Currently, the only owl:sameAs links that are generated link to the RDF mapping of WSDL [24], ensuring compatibility with tools using it, but in the future, other links could be generated to connect services stored in other repositories adopting principles similar to iServe's.
In addition to the import mechanisms that provide compatibility and abstract away the heterogeneity in service description formalisms, iServe contributes to publishing services on the Web by automatically hosting service descriptions at a publicly accessible URI, and by o ering means for users or machines to upload service annotations through its RESTful API7.
iServe is integrated with two service annotation editors being developed within the SOA4All project: SWEET [25] (SemanticWeb sErvices Editing Tool) and SOWER (SWEET is nOt a Wsdl EditoR), which support users in annotating Web APIs and WSDL services respectively.8 Both editors are Web applications that can directly be used with a Web browser and provide support for browsing service descriptions and annotating them through a simple interface. Behind the scenes, both editors take care of serialising the descriptions according to MicroWSMO (SWEET) and WSMO-Lite (SOWER) speci cations respectively. Both editors strive to assist users in creating service annotations through two main means: the use of the Web as background knowledge and a direct connection with iServe for the seamless persistence and publication of annotations.
A fundamental part of service annotation concerns the linking of (parts of) service descriptions to ontologies capturing semantically the data model they handle, certain non-functional properties such as the quality of service and price, and their categorisation with respect to reference taxonomies and classi cations of services. Thus far, the authoring of service annotations has essentially been based on the manual creation of both the services' structure and the actual ontologies used in their annotation. Consequently, creating service annotations has been a particularly tedious task which rather than better supporting the integration of services has led instead to additional heterogeneity at the semantic level. Further limitations have also traditionally been brought by the fact that the service annotations and the ontologies used were most often not published publicly which impedes the interpretation and use of services by third-parties.
SWEET and SOWER support the integrated search of suitable domain
ontologies by means of Watson [
Despite the potential of service technologies and the e orts devoted so far, we have yet to witness a signi cant uptake of service technologies outside enterprise environments. There is a tension between the rich descriptions necessary to automate much of the use of services, and the reticence on the part of developers to invest that e ort. A similar situation exists in Semantic Web Services, where the expressive models able to capture the semantics of services stand in contrast to the lightweight tagging approaches that appear more acceptable to users.
In this paper we described iServe, a novel and open platform for publishing services that aims to better support their discovery and use by exposing them as linked data expressed in terms of a simple vocabulary for services. This vocabulary provides a common ground for descriptions of di erent avours of Web services (SOAP services described in WSDL, and RESTful services described in hRESTS), and several kinds of semantic formalisms (including OWL-S, WSMO, and WSMO-Lite). iServe is underpinned by principles we believe are applicable to generic repositories of semantic information. These principles include the application of linked data principles to publish metadata about the content, the automated publication of the actual content at a public URI if necessary, the adoption of a simple vocabulary able to abstract away the heterogeneous vocabularies or formalisms used by other repositories within the domain, and the provisioning of an open API allowing applications to automatically retrieve and publish information.
Additionally in this paper we have shown how the advent of engines like Watson that support searching for ontological entities over the Web has enabled a new way for authoring semantic models which as we have seen can also contribute to authoring service annotations by reducing the amount of e ort required and potentially contributing to reducing integration issues through ontology reuse.
Future work on iServe will focus on the development and re nement of import mechanisms for WSMO and OWL-S, the creation of advanced indexing mechanisms, and capturing non-functional information about services gleaned at runtime from the monitoring infrastructure. Future work on the editor side will focus on better assisting users in the creation of annotations.
Acknowledgements This work was partly funded by the EU project SOA4All (FP7-215219). We thank all the members of the SOA4All project and the Conceptual Models for Services Working Group of STI International. We also thank Pierre Grenon for his insightful comments and Alex Simov for the development of SOWER. 10. Richardson, L., Ruby, S.: RESTful Web Services. O'Reilly Media, Inc. (May 2007) 11. Gomadam, K., Ranabahu, A., Nagarajan, M., Sheth, A.P., Verma, K.: A faceted classi cation based approach to search and rank Web APIs. In: ICWS '08: Proceedings of the 2008 IEEE International Conference on Web Services, Washington, DC, USA, IEEE Computer Society (2008) 177{184 12. Fensel, D., Lausen, H., Polleres, A., de Bruijn, J., Stollberg, M., Roman, D., Domingue, J.: Enabling Semantic Web Services: The Web Service Modeling Ontology. Springer (2007) 13. Martin, D., Burstein, M., J., H., Lassila, O., McDermott, D., McIlraith, S., Paolucci, M., Parsia, B., Payne, T., Sirin, E., Srinivasan, N., Sycara, K.: OWL-S: Semantic Markup for Web Services. http://www.daml.org/services/owl-s/1.0/owl-s.pdf (2004) 14. Farrell, J., Lausen, H.: Semantic Annotations for WSDL and XML Schema. http://www.w3.org/TR/sawsdl/ (January 2007) W3C Candidate Recommendation 26 January 2007. 15. Vitvar, T., Kopecky, J., Viskova, J., Fensel, D.: WSMO-Lite annotations for web services. In Hauswirth, M., Koubarakis, M., Bechhofer, S., eds.: Proceedings of the 5th European Semantic Web Conference. LNCS, Berlin, Heidelberg, Springer Verlag (June 2008) 16. Sheth, A.P., Gomadam, K., Lathem, J.: SA-REST: Semantically interoperable and easier-to-use services and mashups. IEEE Internet Computing 11(6) (2007) 91{94 17. Maleshkova, M., Kopecky, J., Pedrinaci, C.: Adapting SAWSDL for semantic annotations of restful services. In: Workshop: Beyond SAWSDL at OnTheMove Federated Conferences & Workshops. (2009) 18. Sycara, K., Paolucci, M., Ankolekar, A., Srinivasan, N.: Automated discovery, interaction and composition of semantic web services. Web Semantics: Science, Services and Agents on the World Wide Web 1(1) (2003) 27 { 46 19. Klusch, M., Fries, B., Sycara, K.: Automated semantic web service discovery with OWLS-MX. In: AAMAS '06: Proceedings of the fth international joint conference on Autonomous agents and multiagent systems, New York, NY, USA, ACM (2006) 915{922 20. Stollberg, M., Hepp, M., Ho mann, J.: A caching mechanism for semantic web service discovery. In: 6th International and 2nd Asian Semantic Web Conference (ISWC2007+ASWC2007). (November 2007) 477{490 21. Kuster, U., Konig-Ries, B.: Towards standard test collections for the empirical evaluation of semantic web service approaches. Int. J. Semantic Computing 2(3) (2008) 381{402 22. Kopecky, J., Gomadam, K., Vitvar, T.: hRESTS: an HTML Microformat for Describing RESTful Web Services. In: The 2008 IEEE/WIC/ACM International Conference on Web Intelligence (WI2008), Sydney, Australia, IEEE CS Press (November 2008) 23. Koch, J., Velasco, C.A.: HTTP vocabulary in RDF 1.0. Working draft, W3C (October 2009) 24. Kopecky, J.: Web services description language (WSDL) version 2.0: RDF mapping.
Working group note, W3C (June 2007) 25. Maleshkova, M., Pedrinaci, C., Domingue, J.: Supporting the creation of semantic RESTful service descriptions. In: Workshop: Service Matchmaking and Resource Retrieval in the Semantic Web (SMR2) at 8th International Semantic Web Conference. (2009)