Towards two-stage service representation & reasoning: from lightweight annotations to comprehensive semantics Stefan Dietze, Neil Benn, Hong Roland Siebes Dong Liu, John Domingue Qing Yu, Carlos Pedrinaci Department of Computer Science Knowledge Media Insititute Knowledge Media Insititute Vrije Universiteit The Open University The Open University Amsterdam, The Netherlands Milton Keynes, MK76AA, UK Milton Keynes, MK76AA, UK Telephone number, incl. country code {d.liu, j.b.domingue}@open.ac.uk {s.dietze, n.j.l.benn, h.q.yu, {r.m.siebes}@few.vu.nl c.pedrinaci}@open.ac.uk ABSTRACT targeted WSDL/SOAP-based Web services, which are not Semantics are used to mark up a wide variety of data-centric Web prevalent on the Web. Secondly, due to the inherent complexity resources but are not used to annotate online functionality in required to fully capture computational functionality, creating significant numbers. That is despite considerable research SWS descriptions has represented an important knowledge dedicated to Semantic Web Services (SWS). This has led to the acquisition bottleneck and has required the use of rich knowledge emergence of a new Linked Services approach with simplified and representation languages and complex reasoners. There exists an less costly to produce service models, which targets a wider inherent conflict between the need to capture comprehensive and audience and allows even non-SWS developers to annotate meaningful service semantics – to allow reasoning-based services. However, such models merely aim at enabling semantic automation of any sort – and the requirement to keep the costs for search by humans or automated service clustering rather than providing services descriptions low in order to simplify the automation of service tasks such as discovery or orchestration. modeling process and to ensure that efficient and scalable Thus, more expressive solutions are still required to achieve solutions can be implemented. Hence, despite considerable automated discovery and orchestration of services. In this paper, amount of research dedicated to the SWS visio, so far there has we describe our investigation into combining the strengths of two been little take up of SWS technology within non-academic distinct approaches to modeling semantic Web services – environments. “lightweight” Linked Services and “heavyweight” SWS The prevalent lack of impact of SWS technology is particularly automation - into a coherent SWS framework. In our vision, such concerning since Web services as such are in widespread use integration is achieved by means of model cross-referencing and throughout the Web nowadays, where applications use distributed model transformation and augmentation. HTTP requests via rather lightweight interface technologies such as RESTful services, HTTP GET-style request or XML-feeds. Categories and Subject Descriptors Hence, the SWS challenges are of increasingly crucial importance D.3.3 [Programming Languages]: for today’s highly distributed Web applications. These issues led to the emergence of more simplified SWS approaches to which we shall refer here as “lightweight”, such as WSMO-Lite [9] or the Keywords Micro-WSMO/hRESTs [5] approach which replace Semantic web services, semantic web, WSMO, WSMO-Lite, web “heavyweight” service semantics with less comprehensive and services, minimal service model. less costly to produce service models represented in RDF and hence, complying with the infrastructure of the growing Semantic 1. INTRODUCTION Web. Analogous to the Linked Data term [1], this approach was The past decade has seen a range of research efforts in the area of recently dubbed as the Linked Service approach [7]. Due to the Semantic Web Services (SWS), mainly aiming at the automation fact that such service annotations are much easier to produce and of Web service–related tasks such as discovery, orchestration or can be populated with references to widely established Linked mediation via broker-based approaches. Building on formal Data vocabularies, they address a much wider audience and allow service semantics, several frameworks, such as SAWSDL [8], even non-SWS experts and lay people to describe and annotate OWL-S [6] and WSMO [4], have been proposed which aim at services. However, those models merely aim at enabling formalizing semantic service descriptions, which usually cover structured, semantics-enabled search by humans or automated aspects such as service capabilities, interfaces or non-functional service clustering, and more expressive solutions are required to properties. Besides, a considerable research community evolved achieve greater levels of automation. around these SWS frameworks, providing, for instance, annotation and execution tools based on these formal SWS 2. TWO-STAGE SERVICE ANNOTATION frameworks [3][2]. AND REASONING In the Web context semantics are used to mark up a wide variety In order to tackle the introduced challenges, we aim at combining of data-centric resources but are not used to annotate online the two distinct SWS representation approaches functionality in any form in significant numbers. The reasons for this are two-fold. Firstly, SWS research has for the most part (R1) lightweight Linked Services, and (R2) heavyweight SWS descriptions. application developers benefit from both low cost for providing annotation and a high level of automation. In that, while taking While both approaches partially share common schema entities, advantage of service models produced by a large non-expert e.g. both cover aspects such as interfaces and non-functional audience, both structured search for service instances by humans properties of services, they differ significantly in certain other as well as automation of service tasks is supported. In our vision, aspects, for instance, the way the service models are being integration between lightweight service annotations and produced, the nature of the actual produced models or the kind of comprehensive SWS specifications is achieved by different means reasoning facilitated by each approach. For instance, while (R1) is of (a) model cross-referencing and (b) model transformation and being produced collaboratively as a joint effort by a potentially augmentation. While the current solution provides an overall large group of service providers and consumers, it allows to framework for integrated service models which support different consider a range of perspectives on one particular service and to levels of automation, future work needs to address the gather annotations and RDF-model references to a wide range of investigation of automated model transformation mechanisms in existing RDF vocabularies. Hence, they can be described as multi- order to support the seemless integration of instances across faceted, deliberately incomplete and incoherent. In contrast, the distinct service models schemas. Besides, future work needs to models usually subsumed under (R2), e.g. WSMO-based service investigate the effort required to populate the introduced specifications, reflect the perspective of one particular SWS knowledge bases and the level of automation which is supported. provider and describe a service following a meta-model which aims at exhaustive modeling of a service in terms of its core identifying aspects, such as its capabilities or behavioral 4. REFERENCES characteristics. Here, one strives for a much greater level of [1] Bizer, C., T. Heath, et al. (2009). "Linked data - The Story expressivity and detail and particularly takes into account So Far." Special Issue on Linked data, International Journal execution-related aspects. Therefore, such descriptions could best on Semantic Web and Information Systems (IJSWIS). be described as comprehensive, potentially complex and coherent. [2] Dietze, S., Benn, N., Domingue, J., Conconi, A., and Cattaneo, F. (2009) Two-Fold Semantic Web Service Developers Applications Matchmaking – Applying Ontology Mapping for Service annotate & reuse services request goals Discovery, 4th Asian Semantic Web Conference, Shanghai, (1) referencing (a) Light-weight Service Annotations (b) Semantic Web Service China. Execution Environment (2) transformation [3] Domingue, J., Cabral, L., Galizia, S., Tanasescu, V., Web Service Web Service Web Service Web Service Web Service Gugliotta, A., Norton, B., Pedrinaci, C.: “IRS-III: A broker- based approach to semantic Web services”, Jounal of Web Fig. 1. From lightweight service annotations to heavyweight Semant, pp. 109-132. Elsevier Science Publishers B. V, Semantic Web Services descriptions—the overall approach. 2008. Depending on the quality of the produced service models, the [4] Fensel, D.; Lausen, H.; Polleres, A.; de Bruijn, J.; Stollberg,, representational approach (R2) facilitates reasoning that allows M.; Roman, D.; and Domingue, J. 2007. Enabling Semantic for automation of certain service-related tasks such as discovery Web Services: The Web Service Modeling Ontology. or orchestration but are costly to produce. In contrast, models as Springer. in (R1) are less intricate, but also allow only limited reasoning, [5] Kopecky, J.; Vitvar, T.; and Gomadam, K. 2008. such as clustering of services or structured searches by humans. MicroWSMO. Deliverable, Conceptual Models for Services Working Group, URL: http://cms- While these approaches currently co-exist without a well-defined wg.sti2.org/TR/d12/v0.1/20090310/d12v01_20090310.pdf. relationship, we propose two different bi-directional correlations, [6] Martin, D., Burstein, M., Hobbs, J., Lassila, O., McDermott, which are under investigation: D., McIlraith, S., Narayanan, S., Paolucci, M., Parsia, B., (C1) service model cross-referencing, Payne, T., Sirin, E., Srinivasan, N., and Sycara, K. (2004). (C2) service model transformation and augmentation. OWL-S: Semantic Markup for Web Services. Member submission, W3C. W3C Member Submission 22 November Under (1), we subsume all kinds of references between models 2004. across (a) and (b) as depicted in Fig 1. For instance, a lightweight [7] Pedrinaci, C., Liu, D., Maleshkova, M., Lambert, D., service annotation could point to a heavyweight SWS description Kopecky, J., and Domingue, J. (2010) iServe: a Linked that models the same service more exhaustively or vice versa. Services Publishing Platform, Workshop: Ontology That would allow semantics to be exploited in (a) as well as (b) Repositories and Editors for the Semantic Web at 7th for reasoning of different sorts, for instance, to perform some Extended Semantic Web Conference. clustering based on (a) to reduce the amount of potentially [8] Sheth, A. P., Gomadam, K., and Ranabahu, A. (2008). interesting services for a given query in (b). In addition, (2) Semantics enhanced services: Meteor-s, SAWSDL and SA- considers the transformation between models across (a) and (b), REST. IEEE Data Eng. Bul l., 31(3):8–12. either manually or (semi-)automatically. [9] Vitvar, T.; Kopecky, J.; Viskova, J.; and Fensel, D. 2008. Wsmo-lite annotations for web services. In Hauswirth, M.; 3. CONCLUSION Koubarakis, M.; and Bechhofer, S., eds., Proceedings of the We have described a two-stage approach to semantic service 5th European SemanticWeb Conference, LNCS. Berlin, representation. By integrating collaborative and user-driven Web- Heidelberg: Springer Verlag. scale service annotations with comprehensive SWS specifications,