Within the numerous and heterogeneous web services offered through different sources, automatic web services composition is the most convenient method for building complex business processes that permit invocation of multiple existing atomic services. The current solutions in functional web services composition lack autonomous queries of semantic matches within the "Ontology-based semantic matchmaking approach" of web services, which are necessary in the composition of large-scale related services.
In this paper, we propose a graph-based Semantic Web Services composition system consisting of two subsystems: The management-time subsystem is responsible for dependency
Ontology-based semantic matchmaking approach preparation in which a dependency graph of related services is generated automatically according to the proposed semantic matchmaking rules. The run-time subsystem is responsible for discovering the potential web services and nonredundant web services composition of a user's query using a graph-based searching algorithm.
The proposed approach was applied to healthcare data integration in different health organizations and was evaluated according to two aspects: Web services WS composition is a method used to combine existing WS from heterogeneous systems to build more complicated business processes that match with user requirements.
WS composition also accommodates the development of systems capable of automatic execution of multiple individual WS simultaneously [
Ontology-based semantic matchmaking approach ]. However, these technologies do not offer well-defined semantic and expressive capability for solving semantic service discrepancies that occur due to disagreement in the meaning, interpretation, or intended use of service information.
In most cases, this situation drives the challenge of creating an automated WS composition system that focuses on solving the problems of WS heterogeneities. These problems necessitate semantic matching of input and output parameters to combine multiple relevant services.
Having a fuck
Richer semantics for WS provide greater automation of selection, composition, and invocation of heterogeneous services. Semantic Web Services SWS [ 45 ] have emerged to facilitate automation and dynamism in WS discovery, selection, composition, and monitoring. In recent decades, many approaches for WS composition have been proposed, and certain approaches, such as the work of Oh et al.
However, most of these proposals suffer from high complexity and time consumption for large-scale WS composition. Other approaches, such as the work of Rodriguez-Mier et al. However, such previous works did not present a method with which to prepare the WS dependency graph and further necessitated the difficult task of manually updating the graph.
Although the work of Yue et al. In this paper, we propose a graph-based SWS composition system and introduce a dependency graph preparation approach that aims to resolve the problem of semantic discrepancies
Ontology-based semantic matchmaking approach the use of semantic matchmaking rules to automatically generate the WS dependency graph.
The proposed approach enables the inference engine to perform flexible semantic matches to create a graph model of the related services. This approach is capable of supporting scalable data within graph model by storing it in a corresponding graph database.
We further propose a nonredundant WS composition approach that can efficiently search the most satisfactory services for customer queries
Ontology-based semantic matchmaking approach a dependency graph search technique. To ensure that the proposed approaches can be applied in practical settings, we have developed web-based applications consisting of the graph management tools and WS composition search engine, which are necessary for discovery and publication of complex services in a healthcare domain.
Additionally, our proposed approaches are evaluated according to two aspects: The remainder of the paper is structured as follows. Section 2 presents a review of the related literature, Section 3 outlines the proposed system architecture, Section 4 offers motivating examples, Section 5 presents the graph-based SWS composition methodology of the proposed system, Section 6 illustrates the system implementation, Section 7 presents a system evaluation and discusses the contributions and makes comparison with other works, and conclusions and recommendation for future work are summarized in Section 8.
Ontology-based semantic matchmaking approach composition enables achievement of particular goals through a process of primitive controls and exchanges. The objective in promoting the SWS is to create a flexible layer for development of an automatic system with dynamic discovery, composition, and execution of WS [ 4 ]. The use of such semantic descriptions enables a more flexible "Ontology-based semantic matchmaking approach" expressive Ontology-based semantic matchmaking approach for discovery, composition, and execution of WS.
Many research works have aimed at techniques of discovering, composing, or developing services as reviewed in Rao and Su [ 26 ], Lemos et al.
The need still exists for automatic WS composition to solve the problems within various domains. Many research efforts have been conducted in automatic WS composition using different techniques.
In the context of the AI planning technique, the work of Hatzi et al.
Ontology-based semantic matchmaking approach. Advances...
The approach is based on transforming the WS composition problem into a planning problem that is encoded in PDDL and solved by external planners. The produced composite services are transformed back to OWL-S. The work of Zou et al. A WS composition planning problem is subsequently fed into an AI planner to automatically find a composition plan corresponding to the given composition request. The work of Puttonen et al. The framework aims to extract the planning actions from the OWL-S service descriptions and create a mapping from each action to convert the acquired solution plans into composite OWL-S processes.
The results are intended to reduce the workload of developing semantic WS descriptions and enable automatic composition and deployment of workflow descriptions. Through logic-based technique and algorithms, the work of Rao et al.
The work of Kwon and Lee [ 30 ] proposed a nonredundant WS composition approach based on a two-phase algorithm capable of efficient searching of the scalable WS data using the relational database indexing technique. In the context of a semantic-based technique, the work of Kona et al. The work of Talantikite et al. Another work was proposed by Bansal et al. Within the graph-based technique, models of WS composition have been proposed in several research studies. The work of Hashemian and Mavaddat [ 35 ] proposed an original approach that used a graph search algorithm for WS composition with functional capability.
The work of Dong-Hoon and Kyong-Ho [ 36 ] proposed an accurate WS composition approach by enhancing the functional semantic consideration in graph searching.
The work of Ukey et al. The work of Wang et al. The work of Rodriguez-Mier et al. The work of Lin et al. This work aims to design a cost-effective WS composition algorithm to obtain multiple service compositions using fewer numbers of WS at low costs and within an acceptable execution time. Finally, the work of Shin et al. The main contributions of this paper and a comparison of our
Ontology-based semantic matchmaking approach composition approach with the other approaches will be discussed in Section 7.
This section presents an overview of the graph-based SWS composition system, which is divided into two subsystems of management time and run time, as illustrated in Figure 1. The details of the processes in management-time and run-time subsystems are described in the dependency graph preparation and WS composition subsections of Section 5respectively. This section illustrates
Ontology-based semantic matchmaking approach example of WS as shown in Table 1.
These services extended from our previous work [ 4041 ] and were developed for retrieval of healthcare data from heterogeneous Electronic Health Record EHR systems of different health organizations. The example consists of eight operations of WS i. Given a query q1"Ontology-based semantic matchmaking approach" requested input of q1 is
Ontology-based semantic matchmaking approach, and the requested outputs of q1 are health-numberorganization-namedistrict-nameand zip-code.
Certain requested outputs of q1such as the organization-namedistrict-nameand zip-codeare satisfied in the operations OIDIand ZIrespectively. Although the semantic matchmaking techniques of WS composition are limited to small-scale WS, this research proposes graph-based search algorithms to efficiently find a nonredundant composition of WS with a large-scale WS. The research also creates a systematic method of WS dependency graph preparation to enable the inference engine to perform semantic matching between the input and output parameters of WS.
Ontology-based semantic matchmaking approach section presents the definition of semantic parameter matching and the components of a graph-based WS composition ontology, which are used in two main processes of the proposed system, that is, dependency graph preparation and WS composition. Alfredo Cuzzocrea, Marco Fisichella, A...
These processes are described as follows. According to the semantic matchmaking technique of WS capabilities designed "Ontology-based semantic matchmaking approach" on the ontology concepts, the matching types can be classified into several levels, such as Exact, Plug-in, Subsumes, and Fail [ 42 ].
This research considers only three matching types of Exact, Subsumes, and Fail, as defined in Table 2 for use in parameter matching of dependency graph preparation in the next section. In this section, we propose the graph-based SWS composition ontology, which is used to represent the components of WS in the dependency graph preparation. A portion of the graph-based ontology, which is also expressed in OWL language, is illustrated in Figure 2.
The proposed ontology consists of three main classes: GraphElementProcessand SemanticMatching.
The GraphElement class, which consists of the Arc and Vertex as subclasses, is derived from the directed graph theory and is used to describe the dependency of services. The Arc class represents redundant operations that have the same functions through the property hasOperation and describes the input and output through the tail and head properties, respectively.
The Vertex class is defined to represent input or output parameters of operations represented in Arc. The Process class, which is derived from the OWL-S process model specification [ 43 ], describes 1 the characteristics of the AtomicProcess and the CompositeProcess classes through hasInput
Ontology-based semantic matchmaking approach, hasOutputhasPreconditionand hasEffect properties and 2 the control constructs Perform and Sequence of the CompositeProcess class.
The SemanticMatching class describes the semantic similarity of parameters through the sourceParameter and targetParameter properties, which are classified into two types of Exact and Subsumeas defined in Table 2. This graph-based SWS composition ontology is used in the dependency graph preparation in the next section.
The dependency graph preparation is the process of constructing the relationships of the input and output parameters of the atomic process as a graph. This process consists of three interrelated processes of parameter preparation, parameter matching, and graph generation processes, as described in the following subsections.
Parameter preparation is a process used to divide the parameter names of the atomic processes into meaningful keywords. For this process, the basic components of the graph-based SWS composition ontology, such as AtomicProcessParameterKeywordand Contextare defined as sets of atomic processes of WS,
Ontology-based semantic matchmaking approach and output parameters of atomic processes, keywords of parameters, and contexts of parameters, respectively.
Each basic component is used in the following subprocesses. The execution results of Rules 1—3 for Example 1 are further presented in a portion of the graph-based SWS composition ontology, as illustrated in Figure 3. The results of execution of Rules for Example 2 are illustrated in Figure 4.
Parameter matching is the process of locating the semantic similarity between pairs of input and output parameters in WS operations. The matching instances are subsequently created as outputs. The parameter matching processes consists of keyword matching and matching filtering processes, as described below. Let K cm and K kn be sets of keywords of parameter p cm
Ontology-based semantic matchmaking approach p knrespectively. To calculate the similarity value between keywords, we apply the equation proposed by Wu and Palmer wup [
Ontology-based semantic matchmaking approach ], which is defined in the following function.
Otherwise, the subsume match can be determined through the adjustable threshold of semantic similarity degree. In this paper, the threshold of subsume match is set as 0.
The semantic matching between parameters can be performed using the following rules:. This instance has the district-code and CodeOfTerritory as the source and target parameters, respectively, and has two pairs of equivalent source-target keywords: An example of this exact matching instance generated according to Rule 6 is illustrated in Figure 5.
Abstract: We propose a formal...
An example of
Ontology-based semantic matchmaking approach matching instances generated according to Rules 6 and 7 is illustrated in Figure 6. Although this example contains both subsume and exact instances, one of these matching instances is eliminated through the matching filtering process described in the next section. The coefficient of matching between p cm and p kndenoted as Co jac p cmp knis calculated according to Jaccard's coefficient [ 46 ], as shown in the following equation:.
The return value of 1 is lowest if p is equal to 0 and is the highest if q and r are both equal to 0. This semantic approach allows a more flexible and dynamic matching mechanism based on semantic descriptions stored in ontologies.
Index Terms– Ontology. set of algorithms that solve the semantic matchmaking problem. A formal approach to ontology-based semantic match of skills descriptions.
In  a similarity-based approach for
Ontology-based semantic matchmaking approach service matchmaking is based on the Ontology Language (OWL) for semantic web services which supports the.
MORE: Indiatimes matchmaking