AI-based web service composition: a review

TITR_A_1110061.3d (Standard Serif-IETE)

(215.9£279.4mm)

29-10-2015

14:50

PROOF COVER SHEET Author(s): Article title:

 Guillermo Rodrıguez, Alvaro Soria and Marcelo Campo AI-based web service composition: a review

Article no: Enclosures:

1110061 1) Query sheet 2) Article proofs

Dear Author, 1. Please check these proofs carefully. It is the responsibility of the corresponding author to check these and approve or amend them. A second proof is not normally provided. Taylor & Francis cannot be held responsible for uncorrected errors, even if introduced during the production process. Once your corrections have been added to the article, it will be considered ready for publication. Please limit changes at this stage to the correction of errors. You should not make trivial changes, improve prose style, add new material, or delete existing material at this stage. You may be charged if your corrections are excessive (we would not expect corrections to exceed 30 changes). For detailed guidance on how to check your proofs, please paste this address into a new browser window: http://journalauthors.tandf.co.uk/ production/checkingproofs.asp Your PDF proof file has been enabled so that you can comment on the proof directly using Adobe Acrobat. If you wish to do this, please save the file to your hard disk first. For further information on marking corrections using Acrobat, please paste this address into a new browser window: http://journalauthors.tandf.co.uk/production/acrobat.asp

2. Please review the table of contributors below and confirm that the first and last names are structured correctly and that the authors are listed in the correct order of contribution. This check is to ensure that your name will appear correctly online and when the article is indexed. Sequence 1 2 3

Prefix

Given name(s) Guillermo  Alvaro Marcelo

Surname Rodrıguez Soria Campo

Suffix

TITR_A_1110061.3d (Standard Serif-IETE)

(215.9£279.4mm)

29-10-2015

14:50

Queries are marked in the margins of the proofs, and you can also click the hyperlinks below. Content changes made during copy-editing are shown as tracked changes. Inserted text is in red font and revisions have a blue indicator . F n Changes can also be viewed using the list comments function. To correct the proofs, you should insert or delete text following the instructions below, but do not add comments to the existing tracked changes.

AUTHOR QUERIES General points: 1. Permissions: You have warranted that you have secured the necessary written permission from the appropriate copyright owner for the reproduction of any text, illustration, or other material in your article. Please see http://journalauthors.tandf.co.uk/permissions/ usingThirdPartyMaterial.asp. 2. Third-party content: If there is third-party content in your article, please check that the rightsholder details for re-use are shown correctly. 3. Affiliation: The corresponding author is responsible for ensuring that address and email details are correct for all the co-authors. Affiliations given in the article should be the affiliation at the time the research was conducted. Please see http://journalauthors. tandf.co.uk/preparation/writing.asp. 4. Funding: Was your research for this article funded by a funding agency? If so, please insert ‘This work was supported by ’, followed by the grant number in square brackets ‘[grant number xxxx]’. 5. Supplemental data and underlying research materials: Do you wish to include the location of the underlying research materials (e.g. data, samples or models) for your article? If so, please insert this sentence before the reference section: ‘The underlying research materials for this article can be accessed at / description of location [author to complete]’. If your article includes supplemental data, the link will also be provided in this paragraph. See for further explanation of supplemental data and underlying research materials. 6. The CrossRef database (www.crossref.org/) has been used to validate the references. Resulting changes are tracked in red font.

Q1. AU: Please provide department name for the affiliation, if available. Q2. AU: You have not included an Acknowledgement section. Please either supply one with your corrections or confirm that you do not wish to include one. Q3. AU: A disclosure statement has been inserted. Please correct if this is inaccurate. Q4. AU: The CrossRef database (www.crossref.org/) has been used to validate the references. Mismatches between the original manuscript and CrossRef are tracked in red font. Please provide a revision if the change is incorrect. Q5. AU: Please provide missing publisher location for the [5] references list entry. Q6. AU: Please provide missing volume number for the [10] references list entry. Q7. AU: Please provide conference location for the [22] references list entry. Q8. AU: References have been renumbered in the references list and in text citations in order to maintain sequential order. Please check, and provide revisions if necessary. Q9. AU: Please update reference [48].

TITR_A_1110061.3d (Standard Serif-IETE)

(215.9£279.4mm)

29-10-2015

14:50

How to make corrections to your proofs using Adobe Acrobat/Reader Taylor & Francis offers you a choice of options to help you make corrections to your proofs. Your PDF proof file has been enabled so that you can edit the proof directly using Adobe Acrobat/Reader. This is the simplest and best way for you to ensure that your corrections will be incorporated. If you wish to do this, please follow these instructions: 1. Save the file to your hard disk. 2. Check which version of Adobe Acrobat/Reader you have on your computer. You can do this by clicking on the “Help” tab, and then “About”. If Adobe Reader is not installed, you can get the latest version free from http://get.adobe.com/reader/. 3. If you have Adobe Acrobat/Reader 10 or a later version, click on the “Comment” link at the right-hand side to view the Comments pane. 4. You can then select any text and mark it up for deletion or replacement, or insert new text as needed. Please note that these will clearly be displayed in the Comments pane and secondary annotation is not needed to draw attention to your corrections. If you need to include new sections of text, it is also possible to add a comment to the proofs. To do this, use the Sticky Note tool in the task bar. Please also see our FAQs here: http://journalauthors.tandf.co.uk/production/index.asp. 5. Make sure that you save the file when you close the document before uploading it to CATS using the “Upload File” button on the online correction form. If you have more than one file, please zip them together and then upload the zip file. If you prefer, you can make your corrections using the CATS online correction form. Troubleshooting Acrobat help: http://helpx.adobe.com/acrobat.html Reader help: http://helpx.adobe.com/reader.html Please note that full user guides for earlier versions of these programs are available from the Adobe Help pages by clicking on the link “Previous versions” under the “Help and tutorials” heading from the relevant link above. Commenting functionality is available from Adobe Reader 8.0 onwards and from Adobe Acrobat 7.0 onwards. Firefox users: Firefox’s inbuilt PDF Viewer is set to the default; please see the following for instructions on how to use this and download the PDF to your hard drive: http://support.mozilla.org/en-US/kb/view-pdf-files-firefox-without-downloading-them#w_using-a-pdfreader-plugin

TITR_A_1110061.3d (Standard Serif-IETE)

(215.9£279.4mm)

29-10-2015

14:50

IETE TECHNICAL REVIEW, 2015 VOL. 0, NO. 0, 1 8 http://dx.doi.org/10.1080/02564602.2015.1110061

AI-based web service composition: a review  Guillermo Rodrıguez, Alvaro Soria and Marcelo Campo Q1

ISISTAN-CONICET Research Institute, Universidad Nacional del Centro, Tandil, Buenos Aires, Argentina

5

10

ABSTRACT

Web Service composition allows developers to create applications by capitalizing on the D2serviceD3oriented D4architecture paradigm. Such applications are rapidly deployable and offer developers reuse opportunities and access to an ample variety of complex systems. However, the challenge when composing services is addressing D5quality-of-D6service (QoS) issues. Thus, we attempt to shed light on the role of D7artificial D8intelligence (AI) in assisting developers to deal with QoS-based web service compositions. This paper characterizes contemporary approaches that use AI to explore alternative solutions. We concluded that AI has aimed at exploiting the semantic resources to produce flexible and adaptive-to-change web service compositions.

1. Introduction 15 The main research purpose of Web services is to

achieve interoperability among distributed and heterogeneous applications to offer value-added services [1]. Therefore, flexible composition of Web services to fulfill the given challenging requirements is one of 20 the most important objectives in this research field. Automatically composing Web services involves two main processes: vertical and horizontal composition. Vertical composition consists in defining an appropriate combination of simple processes to perform a 25 composition task, whereas horizontal composition consists in determining the most appropriate Web service for each component process, from among a set of functionally equivalent ones [2]. A Web service may have numerous implementations, all 30 of which have the same functionality, but may have different quality-of-service (QoS) properties. Thus, a significant research problem in Web service composition is how to select a Web service implementation for each of the Web services so that the composite Web service 35 delivers the best overall performance. There may be incompatibilities between Web services at the time of implementation; these incompatibilities may be due to dependency constraint and/or conflict constraint. The former occurs when the implementation of a certain 40 Web service demands the implementation of another particular Web serviceD9, whereas the latter occurs when the implementation of a certain Web service excludes the possibility of including a set of implementations in the Web service composition [3]. © 2015 IETE

ARTICLE HISTORY

Received 12 May 2015 Accepted 10 October 2015 KEYWORDS

Service composition; web services; artificial intelligence; service-oriented architecture; service-oriented development; quality of service

In this work, we provide a detailed, conceptualized, and 45 synthesized analysis of 38 significant AI research works that have aimed at composing services. We also identify open research issues and challenges in the aforementioned research areas. Our results can be used by both researchers and practitioners. Researchers can analyze 50 current approaches to identify potential niches for further research, while practitioners may analyze strengths and weaknesses of current approaches and select the more suitable ones for their professional contexts. The rest of this paper is organized as follows. Section 2 55 presents related work. Section 3 describes the research methodology. Section 4 describes the classification scheme to organize the research works. Section 5 characterizes current Web service composition approaches that use AI methods. Section 6 outlines conclusions and 60 future lines of work.

2. Related work Numerous researchers have surveyed approaches related to automatic service composition. For instance, Dutsdar D10and Schreiner discuss an urgent need for service compo- 65 sition, along with the required technologies to perform service composition, and presentD1 numerous composition strategies based on existing composition platforms and frameworks [4]. Rao D12and Su present an overview of research efforts on automatic Web service composition, 70 both from the workflow and AI planning research community [5]. Peer exclusively addresses the use of AI planning in dynamic and incomplete-information Web

TITR_A_1110061.3d (Standard Serif-IETE)

2

75

80

85

90

95

(215.9£279.4mm)

29-10-2015

14:50

G. RODRIGUEZ ET AL.

service composition contexts [6]. Strunk summarizes, classifies, and evaluates research efforts on QoS-aware service composition. FirstD13, he indicates that the tasks required by composite service, along with their interactions, control, and data flow are identified. SecondD14, an appropriate concrete service is selected and bound to the task [7]. Bartalos D15and Bielikova survey automatic Web service composition approaches regarding not only issuesD16 related to semantics of services, but also formalization of pre/post conditions of the automatic dynamic composition problem [8]. Another related survey has dealt with exploring different type of composition techniques, which rely on machinereadable specification, and comparing them based on some of their properties such as automation, dynamicity, and user involvement, among others [9]. Recently, Garriga et al. have provided an overview of current standardized mechanisms for Web service composition by considering technologies for service composition, automation level, composition time, standards conformance techniques, verification and validation capabilities, service composition specification, QoSD17 awareness, and adaptation level [10].

3. Research methodology The goal of this research is to identify major works on the use of AI in composing Web services, and, thereafter, 100 to classify and integrate these works so as to discover gaps, critical issues, and opportunities for further study and research. The review process addressed in this paper follows the four-step process model adopted in [11] namely material collection, descriptive analysis, category 105 selection, and material evaluation. Our research spans over twelve years (2002D18 2015) and we reviewed a total of 38 approaches for composing services that use AI methods. Figure 1 depicts the distribution of the articles by the artificial intelligence techniques 110 applied in the research works. Out of 38 research works for service composition, we highlight that 14 took place in International IEEE conferences (38.8%), 1 in a ACM Conference (2.77%), 8 in Elsevier Journals (22.22%), 4 were published in Springer Journals (11.11%), and 6 in 115 IEEE Journals (16.67%), among others.

4. Classification schema To evaluate the works reviewed and make a comparative description among them, we have identified common criteria in terms of strengths and weaknesses regarding 120 relevant issues in the field of service discovery, service composition and service implementation:

 AI technique. This feature represents the AI technique or set of AI techniques (i.e. AI D20planning) used by the approach, together with parameters and main operators. This criterion attempts to detect the benefits of the technique, as well as its limitations to deal with certain issues.  Language. This criterion refers to the programming language used to implement the approach, such as Java, or a model for notation such as Meta-object Facility (MOF).  Model language. This feature describes the languages or protocols utilized by the approach to model and specify Web services (i.e. OWL-S, BPEL, among others).  Platform and deployment environment. This criterion deals with the set of tools that are part of the development environment, such as Eclipse or Visual Studio. Furthermore, this criterion addresses deployment environment (i.e. Web or desktop applications), architectural style (i.e. CORBA or ESB), and operating systems (Windows or Linux), amongD21 others.  Exhaustive test. This criterion evaluates whether the approach was exhaustively validated with a considerable number of case studies.  Quality-attribute properties. This criterion describes the QoS and other non-functional features maximized by the approach, such as robustness, performances, and cost-effectiveness, amongD2 others.  Optimization of composition. This criterion deals with the environment in which the optimization takes place, such as design time, D23runtime, compilation time, and network level.  Semantic web. This criterion refers to the use of Semantic Web technologies to represent the user requirements for each service in the service composition process, such as OWL-S.

125

130

135

140

145

150

155

5. Characterization of approaches In this section, 35 significant approaches of AI-based 160 Web service composition are analyzed according to the aforementioned set of features. We review widely used AI techniques namely D24planning, D25genetic D26algorithms (GAs), D27information D28retrieval, and D29collaborative D30filtering, among others. 165 The aim of AI D31planning is to find a series of actions that allow any entity, in this case a service composition problem, to mutate from an initial state towards a final state by achieving a predetermined goal. We think that the widespread use of AI planning may stem from the fact 170 that it is a suitable technique to deal with dynamic

TITR_A_1110061.3d (Standard Serif-IETE)

(215.9£279.4mm)

29-10-2015

14:50

IETE TECHNICAL REVIEW

3

Figure 1. Distribution of AI research works in WS D19composition.

composition in contextD32 with incomplete information; however, the technique may be enhanced with semantic information for approximating the optimal composite 175 services when exact solutions are not found. In order to address the aforementioned issues of AI planning, evolutionary approaches arise as feasible tools to explore. From a computational point of view, the Web service selection problem is a typical constrained combi180 natorial optimization problem. Thus, 3G D AD4s3 have resulted in effective and efficient tools for solving the problem. GAs represent a more scalable option and are more suitable to handle generic QoS attributes than AI planning and 5i3D nteger 6p3D rogramming approaches. Furthermore, evo185 lutionary approaches provide faster composition when replanning takes place, since current QoS deviates from the estimated one, leading to constraint violations; nonetheless, there are issues that jeopardize the application of GA to compose Web services. FirstD7,3 it is not possible to 190 define high-priority goals within the fitness function. SecondD8,3 the size of the input 9h 3D as directD04 impact on the performance of the algorithm. Finally, GA may become instable when the constraints’ density is considerable high.

exploration evidenced how AI techniques have contributed to facilitate the service composition task. As shown in Table 1, planning techniques (37.5% of the research 205 works) and evolutionary algorithms (20.5% of the research works) have been widely used to build executable workflows of composite Web services that satisfy users’ requirements, especially non-functional requirements such as availability, robustness, performance, and 210 adaptability, among others (column 6 of Table 1). Most of the approaches described in Table 1 are modelled by means of BPEL (column 7), since they are in charge of building workflows, and use Java-based tools and inference engines (column 4). Furthermore, seman- 215 tic tools are utilized through ontologies modelled using OWL-S. Thus, as expected, most approaches address runtime composition optimizations. A limitation widely detected in the aforementioned approaches is the lack of independence of programming platforms and vendors. 220 Moreover, few research works are focused on robustness and observability of devices and services (column 6). Consequently, we consider that further research to develop reliable and robust Web service compositions is needed. 225

Aiming to deal with the issues mentioned above, other 195 AI techniques have been applied to compose Web serv-

ices, such as Horn rules, Petri D41nets, functional clustering, D42multi-D43agent D4systems, Markov chains, D45swarm D46intelligence, and D47reinforcement D48learning.

5.1. Findings 200 We have reviewed numerous approaches and frame-

works that have been developed in order to provide widely usable Web service composition platforms. The

5.2. Advantages and disadvantages of AI methods 5.2.1. AI planning This method is suitable for dynamic Web service composition with incomplete information. Along this line, con- 230 straint-optimization techniques are useful for dynamic, distributed, and uncertain environments, especially when user intervention at runtime is required to find

N/A Semantic Web Rule Language, Golog, Java Java Java Visual C# N/A Java N/A

Planning HTN-D6planning

Planning

Goal D71decomposition and RPG Genetic D76algorithmD7 Genetic D80algorithm C CBR Genetic D82algorithmD83

Genetic D84algorithm C D85particle D86swarm D87optimization Genetic D8algorithm C D89ant D90colony

[19] [20]

[21]

[22] [23] [24] [25]

[26]

Visual C# Java Java Java Java, Prolog Java, OWL OWL-S OWL-S OWL-S Java OWL-S Visual Basic

Genetic D95algorithm C D96constraint D97optimization

Genetic D10algorithmD10 Genetic D102algorithm Multi-D106agent D107system (MAS)

Propositional D109logic

Propositional D1logic

Clustering C D14planning

Constraint D18optimization Constraint D12optimization

Markov C Bayesian D125learning

MDP C HTN-D128planning

Swarm D134intelligence C D135genetic D136algorithm

[28]

[29] [30] [31]

[32]

[33]

[34]

[35] [2]

[36]

[37]

[38]

.Net (Windows Vista)

BPELWS4J API, IBM Websphere N/A

Keikaku algorithm tool, IBM UDDI LINDO Solver, METEOR-S N/A

JENA

JESS

Eclipse (Windows) EclipseD103 Jadex platform

.Net

N/A

Eclipse, WSIF, WSDL2JavaD72 .Net N/A Eclipse (Windows), OWL API N/A

Eclipse

JBOSS SHOP2, Eclipse

Eclipse Eclipse, Prolog

N/A Eclipse, OWLS2PDDL

Yes

No

Yes

No No

Yes

No

No

Yes Yes No

Yes

Yes

No

No Yes No Yes

Yes

No No

Yes Yes

No No

No Yes

Yes

Exhaustive test

Flexibility, D129usability, D130availability, D13reliability, D132performanceD13 Scalability, D137performance

Performance, D10expressiveness Performance, D12costeffectiveness Reliability, D15flexibility, D16costeffectiveness Performance, D19precision Performance, D12interoperability, D123usability Robustness, D126adaptability

Scalability Reliability, D104availability Usability, D108adaptability

Reliability, D91availability, D92costeffectiveness, D93performance Reliability, D98precision

Performance, D64availability Performance, expressiveness Performance Performance, D67trustworthiness Scalability, D68interoperability, D69cost-effectiveness Responsiveness, D73scalabilityD74 Scalability, D78extensibility Feasibility, D81flexibility Availability, performance, cost-effectiveness Feasibility, performance

Cost-effectiveness Effectiveness, D57scalabilty, D58robustness Usability Availability

Reachability

Quality-attribute D53properties

workflow

BPEL4WS, MDP model OWL-S, MDP models

BPEL4WS OWL-S

Horn model, Petri D13nets, SAWSDL OWL-S

Custom model BPEL Belief-DesireIntention Horn model

Custom model

BPEL

none

PDDL PDDXML YAWL WS-BPEL WS-BPEL

OWLS-XPLAN

BPEL OWL, SHOP2

OWL DL

XML OWL

Custom model WSDL

DAML-S, OWL-S

Model D54language

D138Runtime

none

none

(continued)

none

D127Runtime Design time

ontology ontology

Semantic annotations ontology

none

Knowledge-based crossover operator none none ontology

none

none

D120Runtime D124Runtime

D17Runtime

Design time

Design time

Design time D105Runtime Design time

D9Runtime

D94Runtime

Design time

OWL-S none ontology OWL-S

OWL-S

D70Runtime D75Runtime D79Runtime Design time Design time

OWL-S OWL-S

OWL-S ALE ontology

none OWL-S

D59Runtime Design time D65Runtime Runtime, design time Design time Design time

none none

none

Semantic D56web

Compilation and run time N/A Network level

Optimization of D5composition

29-10-2015

N/A

PDDL, Java Golog

PDDL PDDL, CCC, Java

UNPOP Planner Eclipse

N/A

Platform and D51deployment D52environment

(215.9£279.4mm)

[27]

[17] [18]

Planning Fast-D60forward D61planning and HTN-D62planning PlanningD63 Planning

[15] [16]

PDDL, Prolog PDDL, Java

Estimated-regression planning Planning

[13] [14]

WS-BPEL

Language

Planning

AI D50technique

4

[12]

Research D49work

Table 1. Web service composition approaches that use AI methods.

TITR_A_1110061.3d (Standard Serif-IETE) 14:50

G. RODRIGUEZ ET AL.

TITR_A_1110061.3d (Standard Serif-IETE)

(215.9£279.4mm)

29-10-2015

14:50

D156Runtime

D158Runtime

D160Runtime

SAWSDL

SAWSDL, SWRL

OWL-S

OWL-S

D154Runtime OWL-S

User’s preferences none

D147Runtime D149Design time D15Runtime UML BPEL, Petri nets WSDL, OWL

Scalability Reachability Decentralization, concurrency, D150contingency Cost-effectiveness, D153reliability Personalized customization of services Execution time, availability, D157reputation Management of uncertainty Yes No Yes NASC System N/A Intel Pentium dual core processor 2.4 GHz

5

optimal solutions. Moreover, OWL-S enriches planning approaches since it facilitates modelling and specifying 235 Web service composition problems. Additionally, the combination with Situation Calculus allows for reducing the search space and number of alternative plans, but increasing performance. As disadvantages, a major issue in planning is the scal- 240 ability. Furthermore, planning fails to be suitable for a choreography-based Web service composition, since it involves decentralized control, concurrent workflows, and contingency.

5.2.2. Evolutionary methods

245

These methods result promising when high scalability and management of generic quality attributes are required. In fact, D16GAs are effective and efficient for solving the Web service selection problem. However, the performance of the D162GAs is sensitive to the increase of search 250 space.

Yes

No

Yes

Yes

Cloud Environment

Cloud Environment

N/A

Intel Core i3 CPU 3.2GHz, Windows 8, 4GB RAMD159

MatLab

Web Service Composition technologies N/A

Ant D152colony

AI D15planning C user/domain preferences Graph search-based algorithm

Petri nets have been proposed for service composition verification and validation, such as deadlock freedom and safety properties. Moreover, in environments where 255 both input/output compatibility and exceptional paths are considered, Petri D163nets are highly recommended. However, Petri nets are unable to map workflow patterns, which usually involve multiple instances, complex synchronizations, or non-local withdrawals onto high- 260 level Petri nets.

Java

5.2.4. Markov chains This technique is suitable for environments in which designers have to address non-deterministic behaviour of Web services along the composition process. 265

5.2.5. Functional clustering Graph-plan

HTN-D146planning Propositional D148logic Planning

PSL N/A

N/A

Reinforcement D139learning C MDP Propositional D142logic Propositional D14logic

OWL-S, Java N/A PDDL, Java

5.2.3. Petri nets

Language

In combination with planning, this technique is suggested for environments in which reliability, QoSD164 awareness, and flexibility are highly required.

[48]

[47]

[46]

[45]

[42] [43] [44]

[40] [41]

5.2.6. Reinforcement learning [39]

AI D50technique Research D49work

Table 1.(Continued )

none

none Distributed Operational semantics none none none D143Runtime D145Runtime Petri nets, BPEL DAML-S, DAMLCOIL Yes No N/A KarmaSIM

Reliability Reachability

none D14Runtime MDP model Yes Q-Learning algorithm

Adaptability, D140performance

Exhaustive test Platform and D51deployment D52environment

Quality-attribute D53properties

Model D54language

Optimization of D5composition

Semantic D56web

IETE TECHNICAL REVIEW

This technique, together with Markov chains, is suitable to achieve optimal Web Service composition at runtime.

270

TITR_A_1110061.3d (Standard Serif-IETE)

6

(215.9£279.4mm)

29-10-2015

14:50

G. RODRIGUEZ ET AL.

5.2.7. Swarm intelligence By using multi-objective optimal-path selection model275 ling for Web Service composition, this technique has yielded favourable results in efficiency, performance, and scalability.

5.3. Future directions Dynamic Web services present several challenges due to 280 the need to discover and handle repositories with numerous Web services, achieving high performance and reliability. In this context, AI methods have aimed to automate the dynamic composition tasks by maximizing QoS-based properties. However, a considerable issue to 285 overcome is maintaining the existing running Web service composition solutions because of changes of business requirements, deployment environment, and other dynamic factors. Current efforts automate only a component of the entire problem. Therefore, a real automation 290 process that addresses all composition concerns, without human intervention, has not been developed yet.

295

300

305

310

Furthermore, the advent of the decentralized paradigm results in more challenges to service composition. FirstD165, it is still complicated and timeD16-consuming to implement, test, and debug composition of low-coupled services in service-oriented systems with current strategies and tools. SecondD167, interoperability mechanisms are highly necessary due to the heterogeneity of devices in a decentralized environment; in this context, the exploration of ontologies seems to be a promising line of research to achieve portability and interoperability. ThirdD168, selfadapting service composition must also deal with mobile devices with limited resources and computational capabilities; thus, it is necessary to explore strategies adaptable to topology changes within the environment to coordinate service composition by considering mobility patterns, platform battery lifetime, fault tolerance, and reliability. FourthD169, current service composition approaches need to be aware of the numerous devices and their failures; then, security, privacy, and trust are still open issues [49].

6. Conclusion Planning and D170GAs have been the most widely used AI techniques to automate the Web service composition 315 process in dynamic environments; however, issues related to Web service compatibilities and portability demand research to efficiently compose Web services in heterogeneous and decentralized environments.

To sum up, the main contributions of our work are the updating of the SOA research agenda and identification 320 of open issues and future lines of work, by pointing out that the future Internet has to face existing challenges in terms of scalability, mobility, security, trust, awareness, and adaptability. As future work, we are planning to extend our study by including cloud-related criteria, 325 since Software as a D17service is taking an important role in Q2 the Internet application paradigm.

Disclosure statement No potential conflict of interest was reported by the authors.

REFERENCES 1. M. Singh, “Being interactive: Physics of service composition,” IEEE Internet Comput., Vol. 5, no. 3, pp. 6 7, 2001. 2. A. B. Hassine, S. Matsubara, and T. Ishida, “A constraintbased approach to horizontal web service composition,” The Semantic Web-ISWC, Lect. Notes Comput. Sci., Vol. 4273, pp. 130 43, 2006. 3. M. Tang, and L. Ai, “A hybrid genetic algorithm for the optimal constrained web service selection problem in web service composition,” in Proceedings of the IEEE Congress on Evolutionary Computation, Barcelona, Jul. 18 23, 2010, pp. 1 8. 4. S. Dustdar, and W. Schreiner, “A survey on web services composition,” Int. J. Web grid Serv., Vol. 1, no. 1, pp. 1 30, 2005. 5. J. Rao, and X. Su, “A survey of automated web service composition methods,” Semantic Web Services and Web Process Composition, Lect. Notes Comput. Sci., Vol. 3387, 2005, pp. 43 54. 6. J. Peer, Web Service Composition as AI Planning: A Survey. University of St. Gallen, 2005. 7. A. Strunk, “Qos-aware service composition: A survey,” in Proceedings of the IEEE 8th European Conference on Web Services, Ayia Napa, Dec. 1 3, 2010, pp. 67 74. 8. P. Bartalos, and M. Bielikova, “Automatic dynamic web service composition: A survey and problem formalization,” Comput. Informatics, Vol. 30, no. 4, pp. 793 827, 2011. 9. S. Kumar, and R. Mishra, “Semantic web services composition,” IETE Tech. Rev., Vol. 25, no. 3, pp. 105 21, 2008. 10. M. Garriga, A. Flores, A. Cechich, and A. Zunino, “Web service composition mechanisms: A review,” IETE Tech. Rev., pp. 1 8, 2015. 11. S. Zhang, C. Lee, H. Chan, K. Choy, and Z. Wu, “Swarm intelligence applied in green logistics: A literature review,” Eng. App. Artif. Intell., Vol. 37, pp. 154 69, 2015. 12. P. Bertoli, M. Pistore, and P. Traverso, “Automated composition of web services via planning in asynchronous domains,” Artif. Intell., Vol. 174, no. 3 4, pp. 316 61, 2010. 13. D. McDermott, “Estimated-regression planning for interactions with web services,” AIPS 2, Vol. 2, pp. 204 11, 2002.

Q3

330 Q4

335

340

345

Q5

350

355

360 Q6

365

370

TITR_A_1110061.3d (Standard Serif-IETE)

(215.9£279.4mm)

29-10-2015

14:50

IETE TECHNICAL REVIEW

375

380

385

390

395

400

405 Q7

410

415

420

425

430

14. S. Oh, D. Lee, and S. Kumara, “Effective web service composition in diverse and large-scale service networks,” IEEE Trans. Serv. Comput., Vol. 1, no. 1, pp. 15 32, 2008. 15 M. El Falou, M. Bouzid, A. Mouaddib, and T. Vidal, “Automated web service composition using extended representation of planning domain,” in Proceedings of the IEEE International Conference on Web Services, Beijing, Sep. 23 26, 2008, pp. 762 3. 16. M. Klusch, A. Gerber, and M. Schmidt, “Semantic web service composition planning with owls-xplan,” in Proceedings of the AAAI Fall Symposium on Semantic Web and Agents, USA. AAAI Press, 2005. 17. O. Hatzi, D. Vrakas, M. Nikolaidou, N. Bassiliades, D. Anagnostopoulos, and L. Vlahavas, “An integrated approach to automated semantic web service composition through planning”, IEEE Trans. Serv. Comput., Vol. 5, no. 3, pp. 319 32, 2012. 18. F. Lecue, A. Leger, and A. Delteil, “DL reasoning and AI planning for Web service composition,” in Proceedings of the 2008 IEEE/WIC/ACM International Conference on Web Intelligence and Intelligent Agent Technology, Sydney, Dec. 9 12, 2008, pp. 445 53. 19. X. Li, X. Tang, Z. Song, X. Yuan, and D. Chen, “A flow: An automated web services composition system based on the AI planning and workflow,” in Proceedings of the IEEE International Conference on Progress in Informatics and Computing, Shangai, Dec. 10 12, 2010, pp. 1067 71. 20. E. Sirin, B. Parsia, D. Wu, J. Hendler, and D. Nau, “HTN planning for web service composition using SHOP2,” Web Semantics: Sci. Serv. Agents World Wide Web, Vol. 1, no. 4, pp. 377 96, 2004. 21. G. Zou, Y. Chen, Y. Xu, R. Huang, and Y. Xiang, “Towards automated choreographing of web services using planning,” in Proceedings of the AAAI Conference on Artificial Intelligence, AAAI, 2012. 22. M. Kuzu, and N. Cicekli, “Dynamic planning approach to automated web service composition,” Appl. Intell., Vol. 36, no. 1, pp. 1 28, 2012. 23. L. Ai, and M. Tang, “A penalty-based genetic algorithm for QoS-aware web service composition with inter-service dependencies and conflicts,” in Proceedings of the International Conference on Computational Intelligence for Modelling Control & Automation, Vienna, Dec. 10 12, 2008, pp. 738 43. 24. Y. Fanjiang, Y. Syu, C. Wu, J. Kuo, and S. Ma, “Genetic algorithm for QoS-aware dynamic web services composition,” in Proceeding of the International Conference on Machine Learning and Cybernetics, Qingdao, Jul. 11 14, 2010, pp. 3246 51. 25. Y. Fanjiang, and Y. Syu, “Semantic-based automatic service composition with functional and non-functional requirements in design time: A genetic algorithm approach,” Inf. Softw. Technol., Vol. 56, no. 3, pp. 352 73, 2014. 26. J. Liu, J. Li, K. Liu, and W. Wei, “A hybrid genetic and particle swarm algorithm for service composition,” in Proceedings of the 6th International Conference on Advanced Language Processing and Web Information Technology, Luoyang, Aug. 22 24, 2007, pp. 564 7. 27. H. Liu, F. Zhong, B. Ouyang, and J. Wu, “An approach for QoS-aware web service composition based on improved genetic algorithm,” in Proceedings of the International

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

7

Conference on Web Information Systems and Mining, Sanya, Oct. 23 24, 2010, pp. 123 8. M. Tang, and L. Ai, “A hybrid genetic algorithm for the optimal constrained web service selection problem in web service composition,” in Proceedings of the IEEE Congress on Evolutionary Computation, Barcelona, Jul. 18 23, 2010, pp. 1 8. H. Jiang, X. Yang, K. Yin, S. Zhang, and J. Cristoforo, “Multi-path QoS-aware web service composition using variable length chromosome genetic algorithm,” Inf. Technol. J., Vol. 10, no. 1, pp. 113 9, 2011. G. Canfora, M. Di Penta, R. Esposito, and M. Villani, “An approach for QoS-aware service composition based on genetic algorithms,” in Proceedings of the ACM Conference on Genetic and Evolutionary Computation, New York, 2005, pp. 1069 75. M. Santofimia, F. Moya, F. Villanueva, D. Villa, and J. Lopez, “An agent-based approach towards automatic service composition in ambient intelligence”, Artif. Intell. Rev., Vol. 29, no. 3 4, pp. 265 76, 2008. S. Ponnekanti, and A. Fox, “SWORD: A developer toolkit for web service composition,” in Proceedings of the 11th World Wide Web Conference, Vol. 45, Honolulu, May 2002. X. Tang, C. Jiang, and M. Zhou, “Automatic web service composition based on Horn clauses and Petri nets,” Expert Syst. Appl., Vol. 38, no. 10, pp. 13024 31, 2011. F. Wagner, F. Ishikawa, and S. Honiden, “Qos-aware automatic service composition by applying functional clustering,” in Proceedings of the IEEE International Conference on Web Services, Washington, Jul. 4 9, 2011, pp. 89 96. R. Aggarwal, K. Verma, J. Miller, and W. Milnor, “Constraint driven web service composition in METEOR-S,” in Proceeding of the IEEE International Conference on Services Computing, Hong Kong, Sep. 15 18, 2004, pp. 23 30. P. Doshi, R. Goodwin, R. Akkiraju, and K. Verma, “Dynamic workflow composition using Markov decision processes,” in Proceedings of the 2004 International Conference on Web Services, California, Jul. 6 9, 2004, pp. 576 82. C. Kun, J. Xu, and S. Reiff-Marganiec, “Markov-HTN planning approach to enhance flexibility of automatic web service composition,” in Proceedings of the IEEE International Conference on Web Services, Los Angeles, Jul. 6 10, 2009, pp. 9 16. W. Zhang, C. Chang, T. Feng, and H. Jiang, “QoS-based dynamic web service composition with ant colony optimization,” in Proceedings of the IEEE 34th Annual Computer Software and Applications Conference, Seoul, Jul. 19 23, 2010, pp. 493 502. Y. Wang, and E. Stroulia, “Flexible interface matching for web-service discovery,” in Proceedings of 4th International Conference on Web Information Systems Engineering, Rome, Dec. 10 12, 2003, pp. 147 56. P. Chan, and M. Lyu, “Dynamic web service composition: A new approach in building reliable web service,” in Proceedings of the IEEE International Conference on Advanced Information Networking and Applications, Okinawa, Mar. 25 28, 2008, pp. 20 5. S. Narayanan, and S. McIlraith, “Simulation, verification and automated composition of web services,” in

435

440

445

450

455

460

465

470

475

480

485

490

495

TITR_A_1110061.3d (Standard Serif-IETE)

8

500

42.

43. 505

44.

510

45.

(215.9£279.4mm)

29-10-2015

14:50

G. RODRIGUEZ ET AL.

Proceedings of the 11th ACM International Conference on World Wide Web, New York, 2002, pp. 77 88. I. Paik, W. Chen, and M. Huhns, “A scalable architecture for automatic service composition,” IEEE Trans. Serv. Comput., Vol. 7, no. 1, pp. 82 95, 2012. W. Tan, Y. Fan, and M. Zhou, “A petri net-based method for compatibility analysis and composition of web services in business process execution language,” IEEE Trans. Aut. Sci. Eng., Vol. 6, no.1, pp. 94 106, 2009. G. Zou, Y. Chen, Y. Yang, R. Huang, and Y. Xu, “AI planning and combinatorial optimization for web service composition in cloud computing,” in Proceedings of the International Conference on Cloud Computing and Virtualization, Singapore, May 17 18, 2010, pp. 1 8. Q. Yu, L. Chen, and B. Li, “Ant colony optimization applied to web service compositions in cloud computing,” Comput. Electr. Eng., Vol. 41, pp. 18 27, 2015.

46. P. W. Wang, Z. J. Ding, C. J. Jiang, and M. C. Zhou, “Design and implementation of a web service based public-oriented personalized health care platform,” IEEE Trans. Syst. Man Cybern., Vol. 43, no. 4, pp. 941 57, 2013. 47. P. W. Wang, Z. J. Ding, C. J. Jiang, and M. C. Zhou, “Constraint-aware approach to web service composition,” IEEE Trans. Syst. Man Cybern., Vol. 44, no. 6, pp. 770 84. 2014. 48. P. W. Wang, Z. J. Ding, C. J. Jiang, M. C. Zhou, and Y. W. Zheng, “Automatic web service composition based on uncertainty execution effects,” IEEE Trans. Serv. Comput., 2015. doi:10.1109/TSC.2015.2412943 49. V. Issarny, N. Georgantas, S. Hachem, A. Zarras, P. Vassiliadist, M. Autili, and A. Hamida, “Service-oriented middleware for the future Internet: State of the art and research directions,” J. Internet Serv. Appl., Vol. 2, no. 1, pp. 23 45, 2011.

515 Q8

520

Q9

525

530

Authors Guillermo Rodrıguez received the Computer Engineer degree from Universidad Nacional del Centro de la Provincia de 535 Buenos Aires (UNICEN), Tandil, Argentina, in 2001, and the Ph.D degree in Computer Science at the same university in 2014. Since 2008, he has been part of ISISTAN Research Institute (CONICET 540 UNICEN). His research interests include Software Architectures, Quality-driven Design and Architecture Materialization. Email: [email protected] 545

550

 Alvaro Soria received the Computer Engineer degree from Universidad Nacional del Centro de la Provincia de Buenos Aires (UNICEN), Tandil, Argentina, in 2001, and the Ph.D degree in Computer Science at the same university in 2009. Since 2001, he has been part of ISISTAN Research Institute (CONICET UNICEN). His research interests

include Software Architectures, Quality-driven Design, Objectoriented Frameworks and Fault Localization. Email: [email protected]

555

Marcelo Campo received the Computer Engineer degree from Universidad Nacional del Centro de la Provincia de Buenos Aires (UNICEN), Tandil, Argentina, in 1988, and the Ph.D degree in 560 Computer Science from Instituto de Informatica de la Universidad Federal de Rio Grande do Sul (UFRGS), Brazil, in 1997. He is currently an Associate Professor at Computer Science Department and Director of the ISI- 565 STAN Research Institute (CONICET UNICEN). His research interests include Intelligent Aided Software Engineering, Software Architecture and Frameworks, Agent Technology and Software Visualization. Email: [email protected]

570