A Web Service Composition Framework Based on Centrality and Community Structure - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2015

A Web Service Composition Framework Based on Centrality and Community Structure

Résumé

Reusing existing services for developing business process applications helps reducing the development time and cost. Besides, existing atomic services may not be able to answer users' requirements. Therefore, as manual composition takes time and is error prone, automatic service composition methods are needed. Their objective is to find the best composite service with minimum number of services and the best performance in term of Quality of Service (QoS). In this paper, a complex network based service composition framework is proposed. It is based on a network representation of the Web service space, making it possible to exploit efficiently various centrality properties and its community structure. It applies the A* search algorithm by considering different network centralities (degree, betweenness, closeness) together with various static and dynamic QoS attributes (execution time, availability and number of calls, changing values of QoS over different timestamps) in its heuristic. Furthermore, the network community structure is integrated in the process, in order to reduce the execution time and increase the accuracy of composition.
Fichier principal
Vignette du fichier
A Web Service Composition Framework based on Centrality and Community Structure.pdf (872.1 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01534564 , version 1 (07-06-2017)

Identifiants

Citer

Sophea Chhun, Kanokwan Malang, Chantal Cherifi, Néjib Moalla, Yacine Ouzrout. A Web Service Composition Framework Based on Centrality and Community Structure. 11th IEEE International Conference on Signal-Image Technology & Internet-Based Systems (SITIS 2015), Nov 2015, Bangkok, Thailand. ⟨10.1109/SITIS.2015.34⟩. ⟨hal-01534564⟩
97 Consultations
149 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More