Joyce El Haddad

CPN-TWS: a coloured petri-net approach for transactional-QoS driven Web Service composition

In this paper, we address the issue of selecting and composing Web Services (WSs) considering functional and Quality of Service (QoS) requirements combined with transactional properties. We formalise the WS functional, QoS and transactional properties as well as the WS Composition problem. We have extended Coloured Petri Net (CPN) formalism to incorporate transactional WSs properties in the composition process. We present and prove our CPN Transactional WS (CPN-TWS) selection algorithm that satisfies the user query functional conditions expressed as input and output attributes, QoS requirements represented by weights over criteria, and transactional properties expressed as a risk level. The result of our algorithm is a CPN corresponding to a Transactional Composite WS whose WS components locally optimise the QoS. We present experimental results to study the performance and the efficiency of our selection algorithm compared with an exhaustive one.

Web Service Selection for Transactional Composition

In this work we address the issue of selecting and composing Web Services (WSs) considering functional requirements and transactional properties. We formalize the WS composition problem using the user query, the transactional properties of WSs and the composite WSs definition. We extends Colored Petri Net (CPN) formalism to incorporate transactional WSs properties. We present a CPN-WS selection algorithm that satisfies the user query functional conditions expressed as input and output attributes, and transactional properties expressed as a risk level. The result of our algorithm is a Colored Petri-Net allowing to execute a transactional composite WS.

SPL-TQSSS: A Software Product Line Approach for Stateful Service Selection

An important problem in Web services composition process is optimal selection of services meeting the user functional requirements (tasks of a workflow) and ensuring a reliable execution of the composition. Therefore, non-functional properties of services such as their transactional behavior as well as their Quality of Service (QoS) must be considered. In this context, a challenging objective is to assist users in integrating on the fly the operations of services to realize their required tasks by further meeting their transactional and QoS preferences. Towards this purpose, we present SPLTQSSS, a Software Product Line based approach for Stateful (conversation-based) Service Selection problem with Transactional and QoS support. SPL-TQSSS considers the set of functionally-equivalent services as part of a service family by modeling their internal operations using Feature Models. Then, SPL-TQSSS chooses the best services, from the service families matching with every task of the workflow, which fit with the user transactional preference and satisfy QoS constraints.

Trust-Based service discovery in multi-relation social networks

With the increasing number of services, the need to locate relevant services remains essential. To satisfy the query of a service requester, available service providers has first to be discovered. This task has been heavily investigated from both industrial and academic perspectives based essentially on registers. However, they completely ignore the contribution of the social dimension. When integrating social trust dimension to service discovery, this task will gain wider credibility and acceptance. If a service requester knows that discovered services are offered by trustworthy providers, he will be more confident. In this paper, we present a new discovery technique based on a social trust measure that ranks service providers belonging to the service requesters multi-relation social network. The proposed measure is an aggregation of three measures: the social position, the social proximity and the social similarity. To compute these measures, we take into account both semantic and structural knowledge extracted from the multi-relation social network. Semantic information includes service requestor and provider profiles and their interactions. Structural information includes among other the position of service providers in the multi-relation social network graph.

A transactional-qos driven approach for web service composition

Existing Web Service architectures provide the basis for publishing applications as Web Services (WSs), and for composing existing WSs to provide new functionalities. To fully meet user requests when WSs are composed, functional characteristics of the WSs as well as Quality of Service (QoS ) parameters and transactional capabilities of their executions, need to be simultaneously considered. QoS parameters describe WSs in terms of their behavior; transactional capabilities state whether a service is reliable during execution time if unpredictable failures occur. We formalize this WS composition problem as an optimization problem that considers at the same time functional, QoS , and transactional requirements. We also define a utility function that combines functionality, QoS , and transactional WS properties, to guide the service compositor into the space of compositions that best meet the QoS and transactional criteria. In addition, we propose a service compositor, named PT-SAM-Transac, which adapts a Petri-Net unfolding algorithm and efficiently traverses the space of optimal compositions. Our experiments show that PT-SAM-Transac outperforms a state-of-the-art solution (called SAM) by identifying compositions that better meet the QoS and transactional criteria, while the composition time of both approaches are in the same order of magnitude.

Web Service Composition Based on Petri Nets: Review and Contribution

Web Services (WSs) are the most used implementation of service-oriented architectures allowing the construction and the sharing of independent and autonomous software. WS composition consists in combining several WSs into a Composite one, which becomes a value-added service, in order to satisfy complex users queries. Thus, the WS composition process may imply several phases to identify how and which WSs will conform the Composite WS, including specification, verification, evaluation, WSs selection, and execution. As it is known, Petri Nets are the main formal models used to describe static vision of a system and dynamic behavior of processes. Then, Petri Nets are well suited to model internal operations of WSs and interactions among them as well as to model the processes in all phases of the WS composition process. In this article we present a review of approaches using Petri Nets for WS composition. Moreover, we describe our experiences in this field: a transactional-QoS-driven WS selection approach and a framework for reliable execution of Composite WSs based on Colored Petri Nets.

A Distributed Decision Making and Propagation Approach for Trust-Based Service Discovery in Social Networks

With the emergence of social networks, users show the willingness to use them to find and offer services. A problem arises when the number of available services is increasing and no means to distinguish between two or many providers offering the same service. To overcome this issue, we propose a trust measure defined as the combination of two dimensions namely sociability and expertise. This measure allows to discover trustworthy providers with good services satisfying the requester’s needs. The problem increases when no central control can be fixed due to the distributed nature of social networks. To address that, our work advocates a distributed agent-based service discovery approach where each user is represented by an agent that acts on behalf of him to achieve the service discovery task. The propagation process within the social network is ensured by means of a referral system wherein agents communicate and evaluate referrals based on a distributed knowledge and a decentralized decision-making.

Optimization techniques for qos-aware workflow realization in web services context

With the evolution of Web services technologies, a lot has been done to answer users needs throughout service composition. Service selection is an important step of the service composition process. Multiple services functionally equivalent might be offered by different providers but characterized by different Quality of Service (QoS) values. Since the QoS of the selected services has an impact on the QoS of the produced composite service, the best set of services to be selected is the set that maximize the QoS of the composite service. In the literature, many approaches have been proposed for the QoS-aware service selection problem which has been formalized as an optimization problem. This paper is devoted to the presentation of some optimization techniques and their application to the service selection problem.

Fuzzy ACID properties for self-adaptive composite cloud services execution

The ACID transaction model has played a cornerstone role in service composition to guarantee that composite services (CSs) have transactional support and consistent outcomes even in presence of failures. However, the classical ACID properties are too restrictive for independent and multi‐proprietors services running on distributed environments, such clouds. Transactional properties allow a relaxed atomicity and isolation, providing an “all‐or‐(almost)nothing” model. In previous works, we proposed a model to relax atomicity, called fuzzy atomicity, according user requirements (acceptable fuzzy atomicity). In this article, we extend that model and propose an approach to measure the fuzzy atomicity. Our model allows to self‐adapt the CS execution, taking into account the state of the CS execution and user preferences (ie, the acceptable fuzzy atomicity expressed in the user requirements). This fuzzy atomicity measure is applied in a self‐adaptive CS execution model based on the transactional properties (pivot, compensatable, and retriable) of its component services and on a checkpointing mechanism. It is suitable to cloud computing, which offers cloud services on demand and in which CS execution should preserve the self‐organizing and self‐adaptivity properties of such environment, specially in the presence of failures. We show how it is possible to relax the retriable property in CSs execution, based on our fuzzy atomicity model. Additionally, in this work, we present a comparative analysis of the most recent works in the context of ACID properties relaxation for CS.

Managing the Variability in the Transactional Services Selection

Web service composition is the capability to recursively construct a value added service by means of picking up existing services. An important step in the composition process is the selection step, which includes choosing services located in repositories. The selection approaches of Web services need to consider their specifics which raises important challenges as the management of the inherent service variability in functionality and implementation and ensuring correct execution termination between others. To realize reliable service compositions, transactional properties of services must be considered during the selection step. We argue that the transactional properties should be considered at the operation level of each service to be composed. However, modelling transactional services composition at the operation level drastically increment the complexity of service selection. In order to overcome this difficulty, in this paper we report on our research in progress on transactional service selection, which follows a Software Product Line approach considering the set of services that provide the same functionality as part of a service family. We model the variable operations of the service families using Feature Models. In this way, the selection process consists of selecting each service from a service family such that the aggregated transactional property satisfies the user preference.