Dionisis Margaris, Panayiotis Georgiadis, and Costas Vassilakis
In Proceedings of the IEEE International Conference on Research Challenges in Information Science (RCIS), 2015

Abstract:
Recommender systems aim to propose items that are expected to be of interest to the users. As one of the most successful approaches to building recommender systems, collaborative filtering exploits the known preferences of a group of users to formulate recommendations or predictions of the unknown preferences for other users. In many cases, collaborative filtering algorithms handle complex items, which are described using hierarchical tree structures containing rich semantic information. In order to make accurate recommendations on such items, the related algorithms must examine all aspects of the available semantic information. Thus, when collaborative filtering techniques are employed to adapt the execution of business processes, they must take into account the services’ Quality of Service parameters, so as to generate recommendations tailored to the individual user needs. In this paper, we present a collaborative filtering-based algorithm which takes into account the web services’ QoS parameters in order to tailor the execution of business processes to the preferences of users. An offline clustering technique is also introduced for supporting the efficient and scalable execution of proposed algorithm under the presence of large repositories of sparse data.

Note: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.

Dionisis Margaris, Costas Vassilakis, and Panayiotis Georgiadis
In Proceedings of the ACM Symposium on Applied Computing (SAC), 2015

Abstract:
In this paper, we present a framework which provides runtime adaptation for BPEL scenarios. The adaptation is based on (a) quality of service parameters of available web services (b) quality of service policies specified by users (c) collaborative filtering techniques, allowing clients to further refine the adaptation process by considering service selections made by other clients, (d) monitoring, in order to follow the variations of QoS attribute values and (e) on users’ opinions services they have used.

Note: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.

Costas Vassilakis, Kareliotis Christos
Proceedings of MCIS 2009

Abstract:
In the context of service-oriented computing, the introduction of the Quality-of-Service (QoS) aspect leads to the need to adapt the execution of programs to the QoS requirements of the particular execution. This is typically achieved by finding alternate services that are functionally equivalent to the ones originally specified in the program and whose QoS characteristics closely match the requirements, and invoking the alternate services instead of the originally specified ones; the same approach can also be employed for tackling exceptions. The techniques proposed insofar, however, cannot be applied in a secure context, where data are encrypted and signed for the originally intended recipient. In this paper, we introduce a framework for facilitating adaptation in the context of secure SOA.

Note: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.

Dionisis Margaris, Panayiotis Georgiadis and Costas Vassilakis
Proceedings of IEEE RCIS 2013

Abstract:
WS-BPEL has been adopted as the predominant method for composing individual web services into higher-level business processes. The designers of WS-BPEL scenarios define at development time the specific web services that will be invoked in the context of the business process they model; in the context however of the current web, where each functionality is offered by multiple service providers, under different quality of service parameters, using a fixed BPEL scenario has been recognized to be inadequate for servicing the diverse needs of business processes clients. To this end, WS-BPEL scenario execution adaptation has been proposed, mainly allowing clients to specify quality of service policies, which drive the dynamic selection of the services that will be invoked. In this paper, we present a framework extending the quality of service-based adaptation mechanisms with collaborative filtering techniques, allowing clients to further refine the adaptation process by considering service selections made by other clients, in the context of the same business processes.

Note: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.

George Lepouras, Costas Vassilakis, Anya Sotiropoulou, Dimitrios Theotokis, Akrivi Katifori
International Journal of e-Government

Abstract:
Organisations nowadays are in the process of developing network-enabled systems through which they deliver electronic services to citizens, customers and enterprises. Often, such services need to be combined in order to cover all aspects of a service consumer¢s life event. The composition of different services though is usually left to the service consumer, who needs to manually locate the individual services and drive the process of obtaining results from some services and feeding them as input to subsequent ones until all relevant services have been executed. Although it would be possible for organisations to improve their level of service through provision of composite services, realizing thus the concept of one-stop government, i.e. by making available mechanisms that would undertake the task of input collection, invocation and execution synchronisation of individual services and delivery of the final result as a reply, such facilities have not been made yet widely available. This shortage stems partly from financial considerations, since the frequent changes in the regulatory framework of both the individual services and in their interoperation requirements or in the technical aspects of the service implementation render the development and maintenance of composite services inexpedient and partly from technical issues, since format or representational incompatibilities in parameters and results hinders automation developments. In this paper we present an active blackboard architecture, which automates the task of service composition based on the semantics of individual services and the data dependencies between them. The blackboard incorporates registries, which can be employed for facilitating service discovery and an execution engine that arranges for dynamic service composition and execution.

Note: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.

George Lepouras, Costas Vassilakis, Anya Sotiropoulou, Dimitrios Theotokis, Akrivi Katifori
Proceedings o f the Second IEEE Conference on Service Systems and Service Management, 2005.

Abstract:
Web services are functional, independent components that can be called over the web to perform a task. Web services are provided by organizations to enable others to perform tasks the organization offers online. However, with an ever increasing number of web services, finding the web service that performs a certain task is not always easy. Furthermore, adopting an end-user point of view what is needed is the actual result and not the service per se. It is often the case that more than one service have to be combined to produce the anticipated outcome, e.g. in the case of life-events. To this end, we propose an active, ontology-based blackboard architecture that aims at tackling the problems inherent in dynamic synthesis of composite web services and at facilitating user interaction with complex e-government transactions.

Note: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.

Dionisis Margaris, Costas Vassilakis, and Panayiotis Georgiadis
Proceedings of ACM SAC 2013

Abstract:
In this paper, we present a framework which incorporates runtime quality of service-based adaptation for BPEL scenarios, allowing for tailoring their execution to the diverse needs of individual users. The proposed framework also caters for automatically resolving system-level exceptions, such as machine outages or network partitionings, while both scenario execution adaptation and exception resolution maintain the transactional semantics that invocations to multiple services offered by the same provider may bear.

Note: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.

Dionisis Margaris, Costas Vassilakis, and Panayiotis Georgiadis
Science of Computer Programming, Elsevier, vol. 98, pp. 707–734, 2015

Abstract:
In this paper, we present a framework which incorporates runtime adaptation for BPEL scenarios. The adaptation is based on (a) the quality of service parameters of available services, allowing for tailoring their execution to the diverse needs of individual users and (b) collaborative filtering techniques, allowing clients to further refine the adaptation process by considering service selections made by other clients, in the context of the same business processes. The proposed framework also caters maintaining the transactional semantics that invocations to multiple services offered by the same provider may bear.

Note: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder. The paper is available through ScienceDirect

Dionisis Margaris, Panayiotis Georgiadis and Costas Vassilakis
Technical report SDBS-TR-14002

Abstract:
In this technical report we give examples on how Quality of Service-based and Collaborating filtering-based techniques can be combined to drive the adaptation of WS-BPEL scenario execution.

Chris Kareliotis, Costas Vassilakis, Panayiotis Georgiadis
Proceedings o f the IEEE 2007 International Conference on Web Services (ICWS).

Abstract:
Web services have become the key technology in business processes management. Business processes can be self-contained or be composed from sub-processes; the latter category is typically specified using the Web Services Business Process Execution Language (WS-BPEL) and executed by a Web Services Orchestrator (WSO). During the execution however of such a composite service, a number of faults stemming from the distributed nature of the SOA architecture, e.g. network or server failures may occur. WS-BPEL includes provisions for exception handling, which can be exploited for detecting such failures; once detected, a failure can be resolved by invoking alternate web service implementations that perform the same business task as the failed one. However, the inclusion of such provisions is a tedious assignment for the business process designer, while additional effort would be required to maintain the BPEL scenarios in cases that some alternate WS implementations cease to exist or new ones are introduced. In our research we are developing a framework for automating handling of that kind of exceptions. The proposed solution employs a pre-processor that enhances BPEL scenarios with code that detects failures, discovers alternate WS implementations and invokes them, fully thus resolving the exception. Alternate WS implementation discovery is based on service relevance, which takes into account both functional and qualitative properties of web services.

Note: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.

Christos Kareliotis, Costas Vassilakis, Efstathios Rouvas and Panayiotis Georgiadis
Proceedings of the tenth International Conference on Information Integration and Web-based Applications & Services (iiWAS2008)

Abstract:
WS-BPEL has become the predominant technology for specifying and executing composite business processes within the Service Oriented Architecture. During the execution however of such a composite business process, a number of faults stemming from the distributed nature of the SOA architecture (e.g. network or server failures) may occur. To this end, the WS-BPEL scenario designer must exploit the provisions offered by WS-BPEL to catch exceptions owing to system failures and resolve them, typically by invoking some alternate equivalent web service that is expected to be reachable and available. The task of system fault handler specification is though an additional burden for the WS-BPEL scenario designer and the presence of such handlers within the WS-BPEL scenario necessitates additional maintenance activities, as new alternate services become available or some of the specified ones are withdrawn. In this paper, we propose a middleware-based framework for system exception resolution, which undertakes the tasks of failure interception, discovery of alternate services and their invocation. The middleware is deployed and maintained independently of the WS-BPEL scenarios, removing thus the need for specifying and maintaining system faults within the scenarios. We also present performance measures, establishing that the overhead imposed by the addition of the proposed middleware layer is minimal.

Article available through the ACM Author-izer service:

Note: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.

Dionisis Margaris, Costas Vassilakis, and Panayiotis Georgiadis
In Proceedings of the ACM Symposium On Applied Computing (SAC), 2016

Abstract:
WS-BPEL scenario execution adaptation has been proposed by numerous researchers as a response to the need of users to tailor the WS-BPEL scenario execution to their individual preferences; these preferences are typically expressed through Quality of Service (QoS) policies, which the adaptation mechanism considers in order to select the services that will ultimately be invoked to realize the desired business process. In this paper, we study the potential to parallelize the execution of the WS-BPEL scenario in order to minimize its response time and/or achieving higher scores in the other qualitative dimensions, such as cost, reliability, etc., at the same time. We also describe, develop and validate a parallelization algorithm for realizing the proposed enhancements.

Note: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.

Dionisis Margaris, Panayiotis Georgiadis, and Costas Vassilakis
In Proceedings of the 8th IEEE International Conference on Service Oriented Computing & Applications, 2015

Abstract:
WS-BPEL scenario execution adaptation has been proposed by numerous researchers as a response to the need of users to tailor the WS-BPEL scenario execution to their individual preferences; these preferences are typically expressed through Quality of Service (QoS) policies, which the adaptation mechanism considers in order to select the services that will ultimately be invoked to realize the desired business process. In this paper, we consider a number of issues related to WS-BPEL scenario adaptation, aiming to enhance adaptation quality and improve the QoS offered to end users. More specifically, with the goal of broadening the service selection pool we (a) discuss the identification of potential services that can be used to realize a functionality used in the WS-BPEL scenario and (b) elaborate on transactional semantics that invocations to multiple services offered by the same provider may bear. We also describe and validate an architecture for realizing the proposed enhancements.

Note: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.

Kareliotis, Costas Vassilakis, Efstathios Rouvas, Panayiotis Georgiadis
SDBS Lab technical report, August 2012

Abstract:
In this technical report, we describe and exemplify the transformations applied by the WS-BPEL preprocessor, in order to produce a BPEL scenario that can be adapted according to QoS specifications, in the architecture described in [1]. [1] adopts a greedy algorithm for performing adaptation, i.e. it uses only the QoS specifications pertaining to the first invoke activity IA to a specific service provider S, so as to decide the service provider to which both IA and further invocations to operations provided by S will be directed. The greedy algorithm may result in suboptimal bindings, while in some cases it may even lead to situations where the middleware is unable to find any appropriate service selection for fully servicing the BPEL scenario, albeit such a path does exist. To this end, a service provider-level adaptation strategy can be employed: the transformed scenario may communicate to the ASOB middleware [1] the information concerning all operation invocations to a specific partner link before the first invocation an operation provided by the specific partner link is executed, and therefore the ASOB middleware can exploit this information to remedy the problems stemming from the greedy nature of the adaptation method specified in [1].

Dionisis Margaris and Costas Vassilakis
In Proceedings of the IEEE Symposium Series on Computational Intelligence, 2016

Abstract:
In this paper, we introduce algorithms for pruning and aging user ratings in collaborative filtering systems, based on their oldness, under the rationale that aged user ratings may not accurately reflect the current state of users regarding their preferences. The aging algorithm reduces the importance of aged ratings, while the pruning algorithm removes them from the database. The algorithms are evaluated against various types of datasets. The pruning algorithm has been found to present a number of advantages, namely (1) reducing the rating database size, (2) achieving better prediction generation times and (3) improving prediction quality by cutting off predictions with high error. The algorithm can be used in all rating databases that include a timestamp and has been proved to be effective in any type of dataset, from movies and music, to videogames and books.

Note: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.

Chris Kareliotis, Costas Vassilakis, Efstathios Rouvas, Panayiotis Georgiadis
Proceedings of IEEE International Conference on Web Services, 2009

Abstract:
BPEL/WSBPEL is the predominant approach for combining individual web services into integrated business processes, allowing for the specification of their sequence, control flow and data exchanges. BPEL however does not include mechanisms for considering the invoked services¢ Quality of Service (QoS) parameters and thus BPEL scenarios can neither tailor their execution to the individual user¢s needs or adapt to the highly dynamic environment of the WEB, where new services may be deployed, old ones withdrawn or existing ones changing their QoS parameters. Moreover, infrastructure failures in the distributed environment of the web introduce an additional source of failures that must be considered in the context of QoS-aware service execution. In this work we propose a framework for addressing the issues identified above; the framework allows the users to specify the QoS parameters that they require and it undertakes the task of locating and invoking suitable services. Finally, the proposed framework intercepts and resolves faults occurring during service invocation, respecting the QoS restrictions specified by the consumer.

Note: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.

Kareliotis Christos, Costas Vassilakis, Efstathios Rouvas, Panayiotis Georgiadis
International Journal on Web and Grid Services (IJWGS), 2009 - Vol. 5, No.3 pp. 284 - 320

Abstract:
WS-BPEL is widely used nowadays for specifying and executing composite business processes within the Service Oriented Architecture (SOA). During the execution however, of such business processes, a number of faults stemming from the nature of SOA (e.g. network or server failures) may occur. The WS-BPEL scenario designer must therefore use the provisions offered by WS-BPEL to catch these exceptions and resolve them, usually by invoking some equivalent web service that is expected to be reachable and available. System fault handler specification is though an additional task for the WS scenario designer, while the presence of such handlers within the scenario necessitates extra maintenance activities, as new alternate services emerge or some of the specified ones are withdrawn. In this paper, we propose a middleware-based framework for system exception resolution, which undertakes the tasks of failure interception, discovery of alternate services and their invocation. The process of selecting the alternate services to be invoked can be driven by process consumer-specified QoS policy, specifying lower and upper bounds for each QoS attribute as well as the importance of each QoS parameter. Moreover, the middleware arranges for bridging syntactic differences between the originally invoked services and functionally equivalent replacements to it, by employing XSLT-based transformations. The middleware is deployed and maintained independently of the WSBPEL scenarios, removing thus the need for specifying and maintaining system fault handlers within the scenarios. We also present performance measures, establishing that the overhead imposed by the addition of the proposed middleware layer is minimal.

Note: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.

Kareliotis Christos, Vassilakis Costas, Georgiadis Panagiotis
Proceedings of OOPSLA 2006, Fourth International Workshop on SOA & Web Services Best Practices

Abstract:
Web services have become the leading technology for application-to-application (A2A) communication over distributed and heterogeneous environments. Both academia and industry have strived to enable useful service collaborations among distributed systems without any human intervention. Web service composition can be used to this end, to achieve business automation within one company or realize business-to-business (B2B) integration of heterogeneous software and cross-organizational computing systems. Service composition pro-vides added value, when a web service composition itself becomes a higher level composite web service. However, as business processes are long-lasting transactions, exceptions may often occur, necessitating the replacement of a service component which has been made unavailable, hindering the completion of some business process. In this paper we present an exception resolving approach based on discovering replacement components that are functionally equivalent, taking also into account criteria for qualitative substitutability. The proposed solution introduces the Service Relevance and Replacement Framework (SRRF) which undertakes exception handling.

Note: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.

Costas Vassilakis, George Lepouras, Akrivi Katifori
Proceedings of the the Third IEEE Conference on Service Systems and Service Management - ICSSSM 06, 2006.

Abstract:
Web services are functional, independent components that can be called over the web to perform a task. Besides being used individually to deliver some well-specified functionality, web services may be used as building blocks that can be combined to implement a more complex function. In such compositions, typically some web services produce results that are used as input for web services that will be subsequently invoked. In the execution schemes currently employed, web services producing intermediate results deliver them to some "coordinating entity", which arranges the forwarding of these intermediate results to web services that require them as input. In this paper we present an execution scheme that employs direct communication between producers and consumers of intermediate results. Besides performance improvement stemming from reduction of network communication, this scheme permits consumer web services to employ simpler authenticity and integrity verification algorithms on incoming parameters, when the producer web service is considered trustworthy.

Note: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.