A Web Services Matchmaking Engine for Web Services
نویسندگان
چکیده
This paper concentrates on the issue of matchmaking in the context of web services. It provides a brief review of the difference between directory services and matchmaking facilities and explains why directories such as UDDI are important but insufficient for web services and need to be complemented with advanced matchmaking facilities. It discusses the requirements that web services place on matchmaking, namely symmetry of information exchange, the ability of each party to specify requirements of the other party, rich languages to describe services and their consumers as well as their demands, and the ability to dynamically update and configure what is being offered. These requirements are addressed by the Web Services Matchmaking Engine (WSME) a powerful matchmaking engine capable of matching complex entities, and a Data Dictionary Tool for defining the language of the corresponding matchmaking process. The WSME matchmaking process and property and rules languages are described. An example of how a dynamic market for selling and buying Capacitors can be created with WSME is given. Finally, conclusions and possible future avenues of work are presented. 1 Finding Compatible Web Services and Consumers Ensuring that web service providers and consumers are compatible business partners will become increasingly important as web service technology and its exploitation for serious business purposes become more pervasive. In this context, several issues are of paramount importance: • Facilitating the specification of complex information models by developing appropriate languages for describing services as well as consumers. • Reaching agreements in specific domains on service and consumer definitions that can be used to create internal or external markets. • Building matchmaking facilities that can deal with the complexity of service and consumer descriptions and the two-way relationship between them. EC-Web, DEXA 2003 conference, Prague 1-5 September 2003 This paper concentrates on the issue of matchmaking in the context of web services. It provides a brief review of the difference between directory services and matchmaking facilities and explains why directories such as UDDI are important but insufficient for web services and need to be complemented with advanced matchmaking facilities. It discusses the requirements that web services place on matchmaking, namely symmetry of information exchange, the ability of each party to specify requirements of the other party, rich languages to describe services and their consumers as well as their demands, and the ability to dynamically update and configure what is being offered. These requirements are addressed by the Web Services Matchmaking Engine (WSME) a powerful Matchmaking Engine capable of matching complex entities, and a Data Dictionary Tool for defining the language of the corresponding matchmaking process. The WSME matchmaking process and property and rules languages are described. An example of how a dynamic market for selling and buying Capacitors can be created with WSME is given. Finally, conclusions and possible future avenues of work are presented. 2 Directories and Matchmaking Facilities Most early directories provided a mapping from a name to an address. With work on distributed system in the 80’s, directories changed to a more advanced form where the search is not carried out for a specific entry by name but rather based on its attributes. In other words, it is possible to search for something of a certain ‘type’ with certain attributes, rather than having to know its name. Along with the notions of searching by attribute rather than by name, came the idea of qualifying the search with an expression, referring to the description of the service and acting as a requirement. This was the basis of the ANSAware Trader [ANSAware 93] and the CORBA/ODP Trading Service [OMG 96, ODP 95]. Such services can be regarded as an advanced form of a directory, a ’Yellow Pages’ in an electronic form. One shortcoming of those directories is that they provide an asymmetric form of selection, where the selection criteria are provided entirely by the potential consumer. Work on advanced Matchmaking Engines (MMEs) extended these ideas by introducing symmetry into the selection process [HOFFNER 00]. Here, the potential consumer is able to provide a description of themselves and what they offer as consumers. This allows the service provider to specify their demands of the potential consumer, thereby being able to select them, just as the consumer selects the service. This advanced form of matchmaking opened up ways of dynamically updating the description of the service at matchmaking rather than advertising time. By coupling this with the notion of symmetry, directories can be more dynamic and facilitate the configuration of what is being offered to suit the needs of the specific consumer, thus introducing dynamic customisation. To summarise it is possible to discern a spectrum ranging from the simplest form of a directory, implemented as a simple query to a database, and the matchmaking process with its multi-dimensions: symmetry, rich description and requirement languages, dynamic configuration, etc. [FIELD 02]. EC-Web, DEXA 2003 conference, Prague 1-5 September 2003 A service often put forward as a directory for web services, is the Universal Description, Discovery and Integration or UDDI [UDDI 2002] [Graham 01a, b]. UDDI is an example of a directory aimed at to publishing and discovering information about Web services and serves as a repository for web service descriptions with a limited search capability. UDDI is an ‘open’ directory, in the sense that the contents are available to everybody who has access to the directory. Restrictions can be placed on the availability of advertisements, but only on the basis of the identity of the consumer, not by using any attributes that may describe them. The search is thus asymmetric only consumers have the ability to express their requirements of the service and its provider, but not vice versa. UDDI is a static directory its content is specified at advertising time and can only be updated if an advertisement is replaced by a new one. We claim that if web services are to be used as the building blocks for creating internal as well as cross organisational business processes, then UDDI may serve as a basic introduction service but lacks the matchmaking capability essential for selecting the right web services and consumers. 3 Requirements of Web Service Matchmaking 3.1 Symmetry of Information Exchange and Selection The process of finding the right service for a given service consumer is not necessarily a one-way process of having the consumer state their requirements and select a winner from the matching services. Service providers may wish to receive information from the consumer before deciding to make a particular service available to that consumer. The input to the matchmaking process (Figure 1) therefore needs to take account of the demands of both service consumers and providers, relating these demands to information provided by both parties resulting in a symmetric exchange by service consumers and providers of both information and demands. P ro v id er P ro v ider req u ire m e n ts o f co n su m er (R u les) S erv ice d esc r ip tion an d d elivery g u aran te es (P rop erties) C o n su m er d esc r ip tion an d g u ar an te es (P rop erties) M a tch m a k in g E n g in e (M M E ) C o n su m er C on su m er req u ire m e n ts o f p r ov id er (R u les) M a tch m a k in g M a tch m a k in g R
منابع مشابه
S-cbr: Semantic Case Based Reasoner for Web Services Discovery and Matchmaking
Web services are being increasingly adopted as the computing engine of choice for today’s Internet-driven applications. The composition of Web services further advances their advantages by allowing different services to integrate for providing new, value-added services. The automated discovery of adequate Web services is the pre-requisite and core feature for achieving dynamic Web services comp...
متن کاملSemantic Matchmaking via Non-Monotonic Reasoning: the MaMaS-tng Matchmaking Engine
In a semantic Web service (SWS) matchmaking process, given a request, it is obvious that not only a list of services should be returned, but also a ranking of compatible SWSs should be provided. Obviously having semantically annotated services, the ranking should be based evaluating semantic similarity between descriptions. Furthermore, the availability of such descriptions makes explanation of...
متن کاملIntelligent Web Services System Based on Matchmaking Algorithm
Web services, the key technology indispensable for e-business, is a comprehensible service providing the desired information or service in the same environment regardless of time and place by integrating current application systems within a single business or between multiple businesses with standardized technologies through the open network using the Internet. However the current web services ...
متن کاملUtilisation of Case-Based Reasoning for Semantic Web Services Composition
With the rapid proliferation of Web services as the medium of choice to securely publish application services beyond the firewall, the importance of accurate, yet flexible matchmaking of similar services gains importance both for the human user and for dynamic composition engines. In this article, we present a novel approach that utilizes the case based reasoning methodology for modelling dynam...
متن کاملArchitectural Plan for Constructing Fault Tolerable Workflow Engines Based on Grid Service
In this paper the design and implementation of fault tolerable architecture for scientific workflow engines is presented. The engines are assumed to be implemented as composite web services. Current architectures for workflow engines do not make any considerations for substituting faulty web services with correct ones at run time. The difficulty is to rollback the execution state of the workflo...
متن کاملIntelligent Web Services System for automatic framework
Recently Web services have become a key technology which is indispensable for e-business due to its ability to provide the desired information or service regardless of time and place by integrating current application systems within a single business or between multiple businesses with standardized technologies using the open network and Internet. However, the current Web Services Retrieval Sys...
متن کامل