Adaptive Search and Learning-Based Approaches for Automatic Web Service Composition

2011 ◽  
pp. 135-188
Author(s):  
Nikola Milanovic ◽  
Miroslaw Malek
Keyword(s):  
Web Service ◽  
Service Composition ◽  
Composition Operators ◽  
Web Service Composition ◽  
Search Problem ◽  
Service Architecture ◽  
Abstract Machines ◽  
Automatic Composition ◽  
Automatic Service Composition ◽  
The Impact

We investigate architectural properties required for supporting automatic service composition. First, composable service architecture will be described, based on modeling Web services as abstract machines supported by formally defined composition operators. Based on the proposed infrastructure, we introduce and analyze several options for achieving automatic service composition by treating it as a search problem. Namely, basic heuristic search, probabilistic, learning-based, decomposition, and bidirectional automatic composition mechanisms will be presented and compared. Finally, we discuss the impact and outlook for automatic composition.

scholarly journals Vartotojo valdomos internetinių paslaugų automatinės ir pusiau automatinės kompozicijos atlikimo galimybės

2013 ◽  
Vol 65 ◽  
pp. 34-44
Author(s):  
Dalė Dzemydienė ◽  
Arūnas Miliauskas
Keyword(s):  
Web Service ◽  
Service Composition ◽  
Research Work ◽  
Web Service Composition ◽  
Ai Planning ◽  
End User ◽  
Automatic Methods ◽  
Automatic Composition ◽  
Composition Process

Šio mokslinio tyrimo tematika nagrinėja internetinių paslaugų kompozicijos atlikimo priemones ir būdus. Straipsnyje aprašomi vartotojo poreikiams pritaikytų internetinių paslaugų kompozicijos automatiniai ir pusiau automatiniai kūrimo būdai ir metodai, plačiau nagrinėjamos šių metodų galimybės ir apribojimai. Analizuojami moksliniuose straipsniuose pateikiami internetinių paslaugų kompozicijos atlikimo metodai ir išskiriami trys pagrindiniai šių paslaugų kompozicijos būdai: darbų srautų modeliais grindžiamas, paslaugų šablonais grindžiamas ir automatinis paslaugų kompozicijos metodas. Tyrimo tikslas – pateikti pasiūlymą, kuris leistų automatiniu ar pusiau automatiniu būdu kurti internetinių paslaugų kompozicijas pagal vartotojų poreikius. Pusiau automatinio internetinių paslaugų komponavimo uždavinio sprendimo būdą iliustruoja sudėtingas kelionės planavimo pavyzdys. Internetinių paslaugų kompozicijai atlikti siūloma taikyti dirbtinio intelekto planavimo metodus. Pateikiama tokio uždavinio sprendimo koncepcija, kuri grindžiama fragmentiniais kituose projektuose gautais paslaugų komponavimo rezultatais ir bando sujungti visą internetinių paslaugų kompozicijos procesą: nuo vartotojo keliamų kompozicijos reikalavimų įvedimo iki tinkamo paslaugų kompozicijos rezultatų pateikimo. Possibilities of automatic and semi-automatic end-user driven web service compositionDalė Dzemydienė, Arūnas Miliauskas  SummaryOur research work relates to the main principles, means and current limitations of the end-user driven automatic and semi-automatic web service composition. It analyses automatic and semi-automatic composition approaches found in literature and classifi es them as workfl ow-based, template-based and automatic methods. The aim of this research is to provide a proposal how to construct semiautomatic or automatic end-user driven web service composition. An approach is illustrated by the multi-complexity of service composition in travel domain. We analyze a conceptual solution that covers the whole composition process: from an end-user submitting composition requirements until the presentation of the a composition execution results. Some methods of an artifi cial intelligence (AI) planning research fi eld were used in proposed web service composition approach.

scholarly journals Distributed Data-intensive Web Service Composition

2021 ◽  
Author(s):  
Soheila Sadeghiram
Keyword(s):  
Web Services ◽  
Web Service ◽  
Service Composition ◽  
Problem Formulation ◽  
Web Service Composition ◽  
User Preferences ◽  
Distributed Data ◽  
Multi Objective ◽  
Data Intensive ◽  
The Impact

<p>Service-oriented architecture (SOA) encourages the creation of modular applications involving Web services as the reusable components. Data-intensive Web services have emerged to manipulate and deal with the massive data emerged from technological advances and their various applications. Distributed Data-intensive Web Service Composition (DWSC) is a core of SOA, which includes the selection of data-intensive Web services from diverse locations on the network and composes them to accomplish a complicated task. As a fundamental challenge for service developers, service compositions must fulfil functional requirements and optimise Quality of Service (QoS), simultaneously. The QoS of a distributed DWSC is not only impacted by the QoS of component services and how the compositions are generated, but also by the locations of services and data transformation between services. However, existing works often neglect the impact of locations and data on service composition. The distributed DWSC has not been sufficiently studied in the literature. In this thesis, we first define the single-objective distributed DWSC that includes communication (e.g. bandwidth), Web service (execution time) and data (data cost) attributes. To this aim, we consider bandwidth information of communication links obtained using the location information of services. Based on the problem formulation, we then address the distributed DWSC problem by developing EC-based approaches. Those EC-based approaches are designed to incorporate domain-knowledge for effectively solving the distributed DWSC problem. Afterwards, we study the multi-objective distributed DWSC to satisfy different QoS requirements. In particular, the QoS-constrained distributed DWSC problem and user preferences are considered. For finding trade-off solutions for those problems, new Multi-objective Evolutionary Algorithms (MOEAs) are proposed based on the current Non-dominated Sorting Genetic Algorithm-II (NSGA-II). Furthermore, a new problem formulation for the dynamic distributed DWSC (D2−DWSC) problem with bandwidth fluctuations is proposed. An EC-based approach is developed to solve the D2-DWSC. Finally, extensive empirical evaluations are conducted that demonstrate the high performance of our proposed methods in finding composite services with good QoS.</p>

A Formal Verification for Web Service Composition Based on CCS

2008 ◽  
Vol 392-394 ◽  
pp. 330-334 ◽  
Author(s):  
B.S. Yun ◽  
J.W. Yan ◽  
M. Liu
Keyword(s):  
Formal Verification ◽  
Web Service ◽  
Service Composition ◽  
Composition Operators ◽  
Web Service Composition ◽  
Composition Algebra ◽  
Dynamic Behaviors ◽  
Formal Tool ◽  
Transformation Mechanisms

Web service composition is a complex and error-prone process. To guarantee its correctness, CCS is exploited as a formal tool to model Web service composition. Composition algebra is defined to specify the rules through which a compositive service can be generated. The transformation mechanisms from service compositions into CCS formalisms according to their corresponding composition operators are gone into details. Then, the automatic reasoner CWB is used to validate the constructed model, by which the dynamic behaviors of the model can be verified and such composition errors as deadlocks can be detected in advance, thus enhancing the composition reliability and avoiding runtime failure. Finally, an example is given to illustrate the effectiveness of this approach.

Adaptive Web Service Composition

2014 ◽  
pp. 278-298
Author(s):  
Areeg Samir
Keyword(s):  
Web Services ◽  
Web Service ◽  
Service Composition ◽  
Business Processes ◽  
Web Service Composition ◽  
Business Services ◽  
Service Architecture ◽  
Crosscutting Concerns ◽  
Dynamic Composition ◽  
Functional Attributes

Service-Oriented Architecture supports sharing resources and transforming business services into a set of linked Web services. Web services rely on non-functional attributes managed through Web Service standards (WS-*) and Quality of Service (QoS) specifications. However, traditionally, the functionality related to QoS and WS-* specifications is scattered and tangled all over the main service code, making the maintenance of these Web services expensive and complex. Aspect-Oriented Programming (AOP) provides abstraction techniques and language constructs to manage and separate these crosscutting concerns from other parts of the system. This chapter focuses on explaining the concepts of dynamic and adaptive Web service composition and proposes an adaptive Web service architecture to enhance reusability of services using the Aspect-Oriented approach. This approach enables separating crosscutting concerns such as QoS and WS-* specifications in aspect Web services and integrating them with the base Web services on the fly. This architecture is based on AO4BPEL, an aspect-oriented extension to BPEL, which reduces the complexity in dynamic selection and reuse of non-functional attributes. This methodology can facilitate dynamic composition of services and business processes in on-premise and Cloud computing environments.

Web Service Automatic Composition Based on Extended Token of Petri Net

2010 ◽  
Vol 20-23 ◽  
pp. 220-225 ◽  
Author(s):  
Bo Yang ◽  
Shun Lin Song
Keyword(s):  
Web Services ◽  
Web Service ◽  
Petri Net ◽  
Service Composition ◽  
Web Service Composition ◽  
Composition Method ◽  
Automatic Composition ◽  
Service Modeling ◽  
Composition Process ◽  
The Relationship

By analyzing the existing Web service composition，a automatic composition method of Web Service based on Petri net was been proposed. This paper primarily extended the Petri net’s tokens, which been expressed as the parameters of atomic services (the services can not be decomposed), as a model for Web services. Then the improved Web service composition method was used to compose the atomic services, making them flowing in the Petri net in accordance with the relationship between the composition of atomic services, and verify the correctness of the service. Finally a specific example was used to analyze and realize the service modeling and automatic composition process of Web service.

scholarly journals Distributed Data-intensive Web Service Composition

2021 ◽  
Author(s):  
Soheila Sadeghiram
Keyword(s):  
Web Services ◽  
Web Service ◽  
Service Composition ◽  
Problem Formulation ◽  
Web Service Composition ◽  
User Preferences ◽  
Distributed Data ◽  
Multi Objective ◽  
Data Intensive ◽  
The Impact

<p>Service-oriented architecture (SOA) encourages the creation of modular applications involving Web services as the reusable components. Data-intensive Web services have emerged to manipulate and deal with the massive data emerged from technological advances and their various applications. Distributed Data-intensive Web Service Composition (DWSC) is a core of SOA, which includes the selection of data-intensive Web services from diverse locations on the network and composes them to accomplish a complicated task. As a fundamental challenge for service developers, service compositions must fulfil functional requirements and optimise Quality of Service (QoS), simultaneously. The QoS of a distributed DWSC is not only impacted by the QoS of component services and how the compositions are generated, but also by the locations of services and data transformation between services. However, existing works often neglect the impact of locations and data on service composition. The distributed DWSC has not been sufficiently studied in the literature. In this thesis, we first define the single-objective distributed DWSC that includes communication (e.g. bandwidth), Web service (execution time) and data (data cost) attributes. To this aim, we consider bandwidth information of communication links obtained using the location information of services. Based on the problem formulation, we then address the distributed DWSC problem by developing EC-based approaches. Those EC-based approaches are designed to incorporate domain-knowledge for effectively solving the distributed DWSC problem. Afterwards, we study the multi-objective distributed DWSC to satisfy different QoS requirements. In particular, the QoS-constrained distributed DWSC problem and user preferences are considered. For finding trade-off solutions for those problems, new Multi-objective Evolutionary Algorithms (MOEAs) are proposed based on the current Non-dominated Sorting Genetic Algorithm-II (NSGA-II). Furthermore, a new problem formulation for the dynamic distributed DWSC (D2−DWSC) problem with bandwidth fluctuations is proposed. An EC-based approach is developed to solve the D2-DWSC. Finally, extensive empirical evaluations are conducted that demonstrate the high performance of our proposed methods in finding composite services with good QoS.</p>

Adaptive Web Service Composition

2016 ◽  
pp. 2139-2158
Author(s):  
Areeg Samir
Keyword(s):  
Web Services ◽  
Web Service ◽  
Service Composition ◽  
Business Processes ◽  
Web Service Composition ◽  
Business Services ◽  
Service Architecture ◽  
Crosscutting Concerns ◽  
Dynamic Composition ◽  
Functional Attributes

Service-Oriented Architecture supports sharing resources and transforming business services into a set of linked Web services. Web services rely on non-functional attributes managed through Web Service standards (WS-*) and Quality of Service (QoS) specifications. However, traditionally, the functionality related to QoS and WS-* specifications is scattered and tangled all over the main service code, making the maintenance of these Web services expensive and complex. Aspect-Oriented Programming (AOP) provides abstraction techniques and language constructs to manage and separate these crosscutting concerns from other parts of the system. This chapter focuses on explaining the concepts of dynamic and adaptive Web service composition and proposes an adaptive Web service architecture to enhance reusability of services using the Aspect-Oriented approach. This approach enables separating crosscutting concerns such as QoS and WS-* specifications in aspect Web services and integrating them with the base Web services on the fly. This architecture is based on AO4BPEL, an aspect-oriented extension to BPEL, which reduces the complexity in dynamic selection and reuse of non-functional attributes. This methodology can facilitate dynamic composition of services and business processes in on-premise and Cloud computing environments.

Sign in / Sign up

Export Citation Format

Share Document