Web-based social networking applications gained considerable popularity in the recent years. However, most social networking platforms su er from the limits of the centralized paradigm they adopted, limiting the mobility of their users, and their chances to bene t from diverse services. In this paper we want to illustrate how these limitations can be overcome by enabling a global naming service for the 'social semantic web'. To do so, we will describe a web tool, called foaf-Omatic, for creating FOAF pro les which is enabled to interact with a global naming system called Entity Name System (ENS). Through a well-known mashup system called Sigma, we will show how the use of globally shared IDs in the de nition of user pro les enables frictionless integrations mechanism supporting web-based application interoperability in dealing with users information.
A social network is a structure composed by individuals or organizations, that
are connected by one or more relationships, such as interests, professional
community, values, ideas, friendship and so on [
In this paper, we focus on the bene ts which may derive from the adoption of a uniform naming system and shared ontologies to create a fertile ground for the de nition of frictionless, entity-centric, generic integration mechanism. We use social network pro les as a typical example of user information that can be de ned across di erent applications and according to di erent standard, causing well known portability issues.
In section 2 we introduce Friend-Of-A-Friend (FOAF) as shared vocabulary, and the Entity Name System (ENS) as a service which allows uniform identi cation of entities across social pro les. In section 3 we describe the entity-centric integration of social network pro les described according to the FOAF ontology and we brie y present a web-based tool, called foaf-O-matic1, for the editing these pro les; we also show how these bene ts may be directly tested through a general purpose mashup system for the Web of Data like Sigma2. In section 4 some related works are presented. Finally, in section 5 we present some conclusion. 2
Friend-Of-A-Friend FOAF is a machine-readable ontology describing persons,
their activities and their relations to other people and objects. The FOAF
project, which de ned and maintains the FOAF ontology [
Entity Name System The ENS [
The founding pillars of a frictionless social network pro le integrations are: (1) Use of shared ontology : all the social network pro les must be structured according to formal ontologies; (2)Uniform identi cation system : people must be identi ed uniformly across di erent social network. pro les.
The combination of these two enables automated and smooth social networks
integration. It is important to notice that ideally, social networking applications
should endorse the same ontology, but when this is not feasible it possible to rely
on techniques for automatic ontology alignment [
To prove that smooth social network user pro le integration is feasible, we have implemented a tool enabling users to manage personal social network proles de ned using the well known FOAF ontology and integrating globally shared identi ers provided through interaction with the ENS. This tool, named foaf-Omatic, is brie y described in section 3.1. For the sake of clarity, we de ned two simple couples of social network pro les. The rst couple was de ned using our tool foaf-O-matic, where users can search the ENS for the identi er of the people they want to talk about and integrate their identi er in the social network pro le (see Figure 2). The second couple using foaf-a-matic, standard tool linked in the foaf project web site, where the integration process is managed by means of equivalence of values of speci c properties (see Figure 1).
As you can see in the Figure 2, the two social network de ned independently according to the FOAF ontology are integrable at zero cost. Indeed, it is possible to merge the two social network by simply putting together the description of the person involved.
In order to evaluate the advantages of the adoption of FOAF and uniform
identi cation system to de ne social networks, we rely on Sigma: a generic web
based tool for (semantic) information mashup. The tool collects information from
di erent sources available on the web, and produces an entity-centric mashup of
the retrieved information. The fact that social network pro les were described
according to the same ontology, and presented the same identi ers for people,
allows the mashup tool to automatically integrate the retrieved information. A
view of Sigma information mashup based on the entity identi er of one of the
author of this article is presented in Figure 3.
Foaf-O-matic, available at http://www.foaf-o-matic.org, is a Rich Internet
Application developed using the Faceltes and Icesoft Icefaces 1.8.2 J2EE AJAX
framework, that supports users in managing their social network pro les. A user
can create new, load and edit the personal social network pro le, and integrate
globally unique identi ers retrieved through guided interaction with an Entity
Name System. Foaf-O-matic allows also to import of list of friends description
contained in FOAF pro les and hCard ( 4, and to save the RDF FOAF
representation of the network on local or remote machine. It is important to notice
that Foaf-O-matic does not force the usage of the identi er retrieved through
the ENS, but also supports the integration of other identi ers by means of OWL
sameAs 5 statements being compatible with LinkedData principles [
Another description of a potential solution for social network portability based
on FOAF is presented in [
5 http://www.w3.org/TR/owl-ref/ behind the FOAF project, providing a more detailed description about how a social network could be described using the FOAF and SIOC ontologies6.
A decentralization of social network system based on FOAF is proposed by
[
A FOAF based peer-to-peer social network system is proposed in [
In this paper we outlined the advantages given by the adoption of shared ontologies and uniform identi cation system when integrating social networks pro le. We presented an example of zero-cost social network pro le integration based on FOAF ontology and relying on the Entity Name System for the identi cation. While doing that, we also brie y presented an existing web application supporting users in creating automatically integrable social network pro les.
It is important to notice that in this context we used social network pro le as known type of user pro le to show the advantages of semantic web technologies combined with the adoption of a uniform identi cation system. Indeed, the
solution we applied to social networks can, in principle, be replicated in any other application environment, creating necessary condition for a frictionless entity-centric information integration, enabling de-facto web application interoperability.