Web Services Discovery
A Novel Social Networks Approach based on Communities
Abdelmalek Metrouh
1
, Hassina Seridi-Bouchelaghem
2
and Farid Mokhati
3
1
LAMIS Laboratory, Department of Mathematics & Computer Science, University of Tebessa, Tebessa, Algeria
2
LABGED Laboratory, Department of Computer Science, University of Badji Mokhtar, Annaba, Algeria
3
LAMIS Laboratory, Department of Mathematics & Computer Science, University of Tebessa, Tebessa, Algeria
Keywords: Web Services, Social Networks, Discovery, Recommendation, Community.
Abstract: Nowadays Web services have become a new focal point of all the technological market of IT. Their number
has grown rapidly and the task of their discovery resting on standards, UDDI and ebXML becomes more
and more difficult. These standards have their own inherent limitations; they describe only the functional
aspect of Web services and prohibit the possibility of combining them. Most proposed approaches for Web
services discovery focused on the description of Web services themselves and neglect their interaction. In
this paper we propose a novel approach which combines social networking with the principles of
recommender systems for the Web services discovery. We also defined the concept of Web services
community in a social network of Web services that gave us the opportunity to establish an abstraction level
between the client applications and these last ones.
1 INTRODUCTION
Created to facilitate commercial exchanges, Web
services take their roots in distributed computing and
the advent of the Web. The Web services technology
aims to standardize the presentation of services
offered by a company and make their access
transparent to any type of platforms, through a
number of interoperability standards (Yu et al.,
2008) (Margaria, 2007) (Papazoglou et al., 2007).
The current approach of Web services
composition which consists in defining business
processes, i.e. sequences of reusable Web services,
is a static approach (Dustdar and Schreiner, 2005).
The sequences of Web services are defined in
advance. In this case, managing the applications
scalability based on Web services integration is a
difficult task. Composing Web services dynamically
according to their functionalities and constraints that
may occur during the process of composition is an
approach that may be considered. This approach
requires that the discovery process of Web services
should be dynamic. The discovery of a Web service
is a preliminary step to its use or recess in its
selection in a composition process. It is through the
mechanism of discovery that we can locate and carry
Web services.
To further reduce the task of Web services
discovery we propose to create Web services
communities that allow us to reduce the search space
of these latter. To date there is no definition in terms
of consensus of Web services community.
Consequently, several definitions have been
proposed in the literature. Maamar et al. (2007)
considered a community as a means to provide a
common description of a desired functionality
without referring explicitly to a specific Web
service. Benatallah et al. (2003) defined a
community as a collection of Web services with
common features, although these Web services may
have separate non-functional properties, as different
providers and different QoS parameters. In this
paper, we describe informally, the notion of Web
services community as relationship between Web
services without any particular weight (house
construction, hotel reservation). Besides, a Web
service may belong to different communities. More
ample details on the construction of services
communities are given in section 4.
In this paper we propose a new approach for
Web services discovery in the context of social
networks. It essentially allows: (i) defining social
networks in the context of Web services, (ii)
defining Web services community and (iii)
316
Metrouh A., Seridi-Bouchelaghem H. and Mokhati F..
Web Services Discovery - A Novel Social Networks Approach based on Communities.
DOI: 10.5220/0004094103160319
In Proceedings of the 14th International Conference on Enterprise Information Systems (ICEIS-2012), pages 316-319
ISBN: 978-989-8565-11-2
Copyright
c
 2012 SCITEPRESS (Science and Technology Publications, Lda.)

proposing an algorithm for Web services dynamic
discovery in the context of social networks. The rest
of this paper is organized as follows. Motivation and
related work are given in section 2. Section 3 gives
an overview of social networks. Section 4 describes
our approach for Web services discovery in the
context of social networks. Conclusion and future
works are given in section 5.
2 RELATED WORK
Maamar et al. (2009) used recommender systems in
the context of social networks for Web services
discovery. Their works consist of proposing two
scenarios. The first scenario is: the recommendation-
based associations that could enrich the process of
composition with additional Web services and are
not usually requested by the user. The second one is
that recommendation-based associations, with an
emphasis on the robustness that could allow direct
selection (and possibly automatic) of a Web service,
which will replace a failed Web service. Maamar,
Wives, Badr, Elnaffar, Boukadi, and Faci, (2011)
introduced a semantic dimension for calculating the
similarity between Web services in the same spirit as
the work proposed by Maamar et al. (2009). Also
Maamar, dos Santos, Wives, Badr, Faci and de
Oliveira (2011) introduced the concept of Web
services competition in addition of the concepts
discussed by Maamar et al. (2009) to build social
networks for service discovery.
Our approach is based on the interaction
between Web services in the context of Web 2.0. We
introduce recommendation-based techniques by
focusing on social networks to solve Web services
discovery problem by allowing these Web services
taking advantage of the previous composition
scenarios in which they participated. We plan to
focus on collaboration-based associations. A
scenario that allows us to really introduce the
problem of Web services discovery in the context of
social networks. It is neither substitution, nor the
recommendation of additional Web services
proposed by Maamar et al. (2009). Compared to the
others approaches, our approach proposes two new
kinds of links or associations between Web services
in a social network: The collaboration-based
associations without any particular weight and
recommendation-based associations moderate and
expressed by specific weights adjusted dynamically,
in the context of Web 2.0, without using semantic
techniques. Semantic approaches proposed by
Maamar, Wives, Badr, Elnaffar, Boukadi, and Faci
(2011) and that of Maamar, dos Santos, Wives,
Badr, Faci and de Oliveira (2011) can not compete
with the approaches in the context of Web 2.0 in
terms of execution time, which remains a sore point
for applications in the world of e-commerce.
3 OVERVIEW OF SOCIAL
NETWORKS
A social network is a dynamic structure modelled by
nodes and edges. Nodes usually refer people and / or
organizations and are connected together by social
interactions. In recent years, social networks have
become very popular. Their fields of applications are
varied and are growing day by day. We can quote
the e-commerce, the artificial intelligence and the
social and political sciences.
The works that introduce social networking in
the field of the discovery and the composition of
Web services are relatively recent. The social
networks of Web services differ from conventional
social networks. These last ones are based on the
absolute cooperation and mutual assistance between
their members (i.e., no competition). By cons, Web
services in social networks are especially
competitive (Maamar et al., 2009) (Werthner et al.,
2007). In a social network of Web services, Web
services are in permanent interaction. New links can
be formed and existing ones may disappear or be
changed.
4 SOCIAL NETWORK-BASED
APPROACH FOR WEB
SERVICES DISCOVERY
The proposed discovery process of Web services
rests on the creation of a social network of Web
services for one or more registers UDDI. This
process is developed as an algorithm for Web
services discovery, presented farther in this section.
At first we describe how to build a social network of
Web services.
4.1 Building Social Networks of Web
Services
Web services are the only constituents of the social
network and designate the nodes. We propose two
kinds of links or associations between Web services:
Collaboration-based associations (C) and
Recommendation-based associations (R).
WebServicesDiscovery-ANovelSocialNetworksApproachbasedonCommunities
317

Mason‐WS
Constructionprocedure
Constructionprocedure
Constructionprocedure
Constructionprocedure
Constructionprocedure
Constructionprocedure
Home'sConstruction
Home'sConstruction
Home'sConstruction
Carpenter‐WS
Electrician‐WS
Home'sConstruction
Plumber‐WS
Architect‐WS
Paintre‐WS
Building_Premit‐WS
Figure 1: Example of Web services communities
definition in a social networks of Web services.
Collaboration-based Association (C): The
associations of collaboration are defined upon Web
services communities. Each community is defined
by a tree of Web services; community “Home's
Construction” in Figure 1, for example. Add a Web
service to a particular community means inserting it
at the end of the corresponding tree. A Web service
may belong to one or more communities; in Figure1,
the Web services Painter-WS, Mason-WS,
Electrician-WS, Plumber-WS and Carpenter-WS are
belonging simultaneously to both communities
“Home's construction” and “Construction
procedure”.
Recommendation-based Association (R): in the
same Web services community, a Web service could
propose that new Web services resulting from
recommendation-based associations should be part
of a composition. Web services resulting from
recommendation-based associations may be required
to satisfy a user query. For example, under a
subcontract, a prime contractor wanting to build a
house could call an electrician, a painter or a
carpenter. This last one could or not express
explicitly this interest. But if a contracting authority
wishes to manage himself building construction and
thus assumes the role of prime contractor. He will
need to involve all stakeholders.
Example: WS
i
will recommend that WS
j
could be
part of the composition scenario with a weight wR
ij
.
WS
j
could also recommend WS
i
with a weight wR
ji
.
The weight of the recommendation-based
association wR
ij
(or wR
ji
) is a calculated numerical
value between 0 and 1 and is given by the following
equation:
ionparticipatWS
selectionWS
WSWSwR
i
J
jiij
=),(
(1)
|WS
i
participation| and |WS
j
selection| represent the
number of times that WS
i
participated in scenarios
composition and number of times that WS
j
has been
appointed by WS
i
to participate in these scenarios
composition. wR
ji
is calculated by the same
equation.
The equation (1) is inspired by the works of
Maamar et al. (2009) and Maamar, Hacid and Huhns
(2011).
4.2 Algorithm for Web Services
Discovery
The proposed algorithm is based on the definition of
three subgraphs of the graph G associated with the
social network of Web services; a subgraph of
collaboration cG that defines communities
associated with Web services whose associations are
characterized by no weight; a subgraph of
recommendation rG that defines associations of
recommendation characterized by weight between 0
and 1 given by the equation (1) and an undirected
graph rG' which is a graph associated to the graph of
recommendation rG whose weights of associations
are defined by the equation (2).
Formally, we define the graph G as a triple
(S,C,R), where S is the set of vertices, C is the set of
edges of Collaboration and R is the set of arcs of
Recommendation. Furthermore, the graphs formed
by S and C, on the one hand, and S and R, on the
other hand, are respectively called cG (Graph of
collaboration) and rG (Graph of recommendation)
of G. We also define an undirected graph rG'=(S,R')
corresponding to the directed graph rG=(S,R) whose
edges R' are valued by the equation (2). WS
i
and WS
j
are two Web services of the graph rG'. The weight
wR(WS
i
,WS
j
) of the edge (WS
i
,WS
j
) is given by the
following equation:
2
),(),(
),(
ijjijiij
ji
WSWSwRWSWSwR
WSWSwR
+
=
(2)
Web services communities that form the social
network are identified and indexed at the time of
their inscription. In fact, the interface is performed
in order to exploit these communities directly.
Semantic techniques can be used for extracting the
community associated with the user query. It is also
supposed that it concerns only one community.
ICEIS2012-14thInternationalConferenceonEnterpriseInformationSystems
318

Figure 2: Algorithm for Web services discovery process
based on social networks.
5 CONCLUSIONS AND FUTURE
WORKS
In this paper, we focused on Web services discovery
in the context of social networks. We have defined
two types of links in a social network of Web
services. Links or collaboration-based associations
formed from Web services communities, and links
or recommendation-based associations within these
same communities. We defined a community as the
report between Web services correlated (home
construction, organization of a trip, hotel
reservation) without any particular weight. The
combined exploitation of these two types of
associations as part of an algorithm for Web services
discovery, allowed us to reduce considerably this
task.
Our implementation is a work in progress. We
are presently implementing the proposed algorithm
where we considered a graph G associated with the
Web services social network of 2100 nodes (Web
services) and 5000 edges.
The current work could be extended by
introducing the notion of community multi-criteria
in the formation of the collaboration-based
associations. We could also review the
recommendation-based associations and suggest that
the users take part more in the process of
recommendation of these last ones.
REFERENCES
Benatallah, B., Sheng, Q. Z., Dumas, M., (2003). The
Self-Serv Environment for Web Services
Composition. IEEE Internet Computing, 7(1), 40–48.
Dustdar S. and Schreiner, W., (2005). A Survey on Web
services Composition. International Journal of Web
and Grid Services, 1(1), 1–30.
Maamar, Z., dos Santos, P. B.,Wives, L. K., Badr, Y.,
Faci, N. and de Oliveira, J. P. M., (2011). Using Social
Networks for Web Services Discovery. IEEE Internet
Computing, 15(4), 48–54.
Maamar, Z., Hacid, H. and Huhns, M. N., (2011). Why
Web Services Need Social Networks. IEEE Internet
Computing, 15(2), 90–94.
Maamar, Z., Lahkima, M., Benslimane, D., Thirand, P.,
Sattanathan, S., (2007). Web Services Communities :
Concepts & Operations. In: Proceding of the 3rd
International Conference on Web Information Systems
and Technologies. Mars, 3–6, 2007. Bacelona, Spain.
Maamar, Z., Wives, L. K., Badr, Y. and Elnaffar, S.,
(2009). Even Web Services Can Socialize: A New
Service-Oriented Social Networking Model. In: the
Proceedings of the International Conference on
Intelligent Networking and Collaborative Systems
(INCoS). 4–6, November, 2009. Barcelona, Spain.
Maamar, Z., Wives, L. K., Badr, Y., Elnaffar, S., Boukadi,
K. and Faci, N., (2011). LinkedWS: A novel Web
services discovery model based on the Metaphor of
social networks. Simulation Modelling Practice and
Theory, 19(1), 121–132.
Margaria, T., (2007). Service is in the Eyes of the
Beholder. IEEE Computer, 40(11), 33–37.
Papazoglou, M., Traverso, P., Dustdar, S. and Leymann,
F., (2007). Service-Oriented Computing: State of the
Art and Research Challenges. IEEE Computer, 40(11),
38–45.
Yu, Q., Bouguettaya, A. and Medjahed, B., (2008).
Deploying and Managing Web Services: Issues,
Solutions and Directions. The VLDB Journal, 17(3),
537–572.
Werthner, H., Hansen, H. R. and Ricci, F., (2007).
Recommender systems. In: Proceding of the 40th
Annual Hawaii International Conference on System
Sciences. 3–6, January, 2007. Waikoloa, Big Island,
HI, USA.
WebServicesDiscovery-ANovelSocialNetworksApproachbasedonCommunities
319