LinkedWS: A Novel Web Services Discovery Model Based on the Metaphor of Social Networks
Web services are increasingly becoming the de facto implementation for the service-oriented architecture paradigm for enterprises due to their ease of use. Nevertheless, discovering these Web services is still hindered by many challenges that are partially attributed to shortcomings found in the discovery registry models (e.g., UDDI) used nowadays. These registries do not, for example, capture the rich information resulting from the various types of interactions between Web services. To address these shortcomings, and inspired by the conventional human social networks on the net, like Facebook and Twitter, we develop LinkedWS, a social networks discovery model to capture the different interactions that occur between Web services. Based on these interactions, specialized relationships are spawned and discerned. Examples of these relationships are collaboration and substitution. This paper describes LinkedWS and its potential, and reports on its implementation status.
Available from: Zakaria Maamar
- "These experiences identify (i) the peers that Web services have collaborated with, (ii) the peers that have replaced Web services when they failed, and (iii) the criteria that were used to select Web services over other peers . In our previous work for instance  , the contributions were on engineering social Web services and making them sign up in social networks, respectively. 1 Specific contributions for the current work include 1. elaborating a new taxonomy for social networks of Web services at the intra-and inter-community levels. "
[Show abstract] [Hide abstract]
ABSTRACT: This paper discusses a framework to manage Web services using the concept of community and the metaphor of social networking. On the one hand, a community gathers Web services that offer similar functionalities together. These Web services are referred to as either master or slave. On the other hand, social networking captures all interactions that occur between Web services located in the same or separate communities. Five interactions are identified and referred to as supervision, substitution, competition, collaboration, and recommendation. The mining exercise over the social networks that capture these interactions results in assigning social qualities to Web services, similar to those found in people’s daily life such as selfishness, fairness, and trustworthiness. Experiments showing the mining exercise are also reported in this paper.
Available from: mars.ing.unimo.it
- "In the area of service-oriented computing, it is getting recognized that social networking at the level of services can notably facilitate service discovery and composition , . For instance, in the LinkedWS proposal , a social network of services is dynamically built by analyzing the patterns of co-invocation and similarity, and defining the social relationships between services accordingly. Our proposal commits to this idea, but instantiates it to the specific scenario of pervasive computing services and devices. "
[Show abstract] [Hide abstract]
ABSTRACT: Middleware infrastructures for pervasive computing, in order to be able to support services and users activities, have to deal with both spatially-situated and socially-situated interactions. In this paper we present the solution adopted in the SAPERE middleware that exploits the graph of a social networks, and combines it with relations deriving from spatial proximity, to drive the topology of interactions among users, devices and services. This results in a middleware that facilitates the development and management of services that are adaptive to both spatial and social concerns, and can support effective service discovery and orchestration, and naturally tackles privacy issues.
Available from: Marco Mamei
- "Also in the area of service-oriented computing it is getting recognized that social networking can notably facilitate, or make more reliable and trustable, service discovery and composition , , . For instance, in the LinkedWS proposal , a social network of services is dynamically built by analyzing the patterns of co-invocation and similarity, and defining the social relationships between services accordingly to those patterns. Then, the discovery of Web services can be notably facilitated by " navigating " the resulting social network of services to, e.g., discover the most suitable partners for a composition or recommend alternatives between equivalent services. "
[Show abstract] [Hide abstract]
ABSTRACT: Any middleware for pervasive computing services has to e ectively support both spatially-situated activities and social models of interactions. In this paper, we present the solution integrated in the tuple-based SAPERE middleware to tackle this problem. The idea is to exploit the graph of a social network along with relations deriving from spatial proximity to rule the actual topology of interactions among devices, users and services. The proposed approach can facilitate the autonomous and adaptive activities of pervasive services while accounting for both social and spatial issues, can support e ective service discovery and orchestration, and can enable tackling critical privacy issues. I. Introduction The spread of pervasive computing technologies, smart phones above all (1), is leading to the emergence of an inte- grated and very dense socio-technical infrastructure for the provisioning of innovative general-purpose digital services (2), (3). That infrastructure will be used to ubiquitously access services for better interacting with the surrounding physical world and with the social activities occurring in it. Also, the infrastructure will be very open, enabling users to deploy customized services and to make available own whose components may have to interact based on physical relations; and (ii) supporting the need for users to interact, and based on their social relations and in respect of privacy issues, facilitating services/devices interaction and composi- tion based on such social relations, other than simply based on spatial proximity.
Data provided are for informational purposes only. Although carefully collected, accuracy cannot be guaranteed. The impact factor represents a rough estimation of the journal's impact factor and does not reflect the actual current impact factor. Publisher conditions are provided by RoMEO. Differing provisions from the publisher's actual policy or licence agreement may be applicable.