THEMATIC OASES

Towards a Social Semantic Web

Maria Teresa Pazienza, Savino Sguera and Armando Stellato

DISP, University of Tor Vergata, Via del Politecnico 1, Rome, Italy

Keywords: Semantic Web, Social

Web, Social Frameworks.

Abstract: A few years have passed from the first alarming yells about the unmanageable growth of data which is

literally exploding from the Web. While Web2.0 technologies, born and grown from the crowd of the web

community, are reaching their full maturity, and with W3C eventually managing to give concreteness to

Berners-Lee Semantic Web Vision through a plethora of new languages and protocols, the same problem is

still a living matter. Lots of vendors and providers offer social services with more-than-overlapping aspects,

but with no intentions of sharing their data. RSS aggregators as well as blog and mailing list scrapers, erupt

tons of data which are irritatingly replicated by search engines indexes. At the same time, most of the

information services (wikis, blogs, mailing lists, forums, newsgroups etc...) still maintain their traditional

functionalities and move no step forward reaching any kind of interoperability. In this paper we analyze the

current scenario and propose our personal view on how new Semantic Web technologies could be employed

to give life to a new generation of social, heterogeneous and coordinated informative services: Thematic

Oases.

1 INTRODUCTION

The original vision of the Web, as Tim Berners-Lee

exposed it in his book: “Weaving the Web”

(Berners-Lee, 2000), seems now very close to its

realization: a web equally participated by major

publishing entities as well as from ordinary people

willing to make public their thoughts, artworks,

opinions and ideas. Today technologies and

approaches coming from new Web2.0 paradigms

have helped this dream come true: now we have

millions of people blogging, contributing to social

networks by producing and publishing huge amounts

of digital stuff in several multimedia formats, and

sharing their knowledge through Wikipedia or in

domain specific wikis. Yet the several possibilities

that new Web era is giving to the mass, are

producing a lot more information than before, but

are not sensibly improving the way we access it: the

first alarming yells about the crescent unmanageable

grow of data which were characterizing the growth

of the traditional Web are still an important warning

to take into in account. Semantic Web technologies

and standards fostered by the W3C are trying to

address this issue, by providing vocabularies and

methodologies for organize web data in a

decentralized and neutral way. The process of

realization of the Semantic Web layer cake of

language and protocols is near to the end, though,

paradoxically, the adoption of these standards is

hampered by those which autonomously generated

from the Web2.0 stream of innovation. It appears

evident as these new standardized and open

technologies should make their way through the

resistance of existing service providers, possibly

beneficiating from the success of new open semantic

applications and frameworks.

In this paper we

analyze the current scenario, considering state-of-

the-art on social/semantic organization of data, and

propose our personal view on how Semantic Web

technologies could be employed to give life to a new

generation of social, heterogeneous and coordinated

informative services: Thematic Oases

2 SCENARIO AND RELATED

WORKS

Most recent works and discussions on the web made

clear that the need for actual data portability and

shareability has become the top priority for web and

Software as a Service (SaaS) applications providers.

There is an amazingly increasing number of

fferently flavoured social web applications, which

351

Teresa Pazienza M., Sguera S. and Stellato A. (2008).

THEMATIC OASES - Towards a Social Semantic Web.

In Proceedings of the Fourth International Conference on Web Information Systems and Technologies, pages 351-354

DOI: 10.5220/0001532603510354

 SciTePress

inherently lead to data replication all over the web:

every time a user joins a new social service she

probably has to sign up, invite friends, add/remove

friends, generally ask for email addresses too,

requiring people to send out address verification

emails, not even citing the tedious “lost

email/password” issues. Negative implications of

having users’ data tied to a proprietary platform –

worthwhile to name Facebook (Facebook, ©) here,

given the exponential growth in terms of interest and

users it has been experiencing (100.000 new users

per-day, mostly in the golden over-25-years-old

market share) – with its own markup language and

its own set of API, are evident: user’s experience is

based upon a specific framework and set of enabling

technologies, while data portability is – often

deliberately – not granted. On the other side,

developers are forced to master the n-th set of REST

API, to write applications that are not even close to

the “write once, run everywhere” paradigm which

underlies enterprise software engineering principles,

sparkling programmers’ and managers’ interests.

A solution to this problem has been proposed by

Google with the Open Social (Google ©) API set,

which is a specification for widgets and applications

deployable on social networks. Open Social defines

three broad areas of specification:

Widget/Application, Friends, Activity.

All of these still have a long way to evolve but,

yet being not standard at all, they bring powerful

concepts of openness and interoperability into the

social network marketplace. Personal data, however,

are not limited in any way in scope and practice to

the usual profile-related information: depending on

the service being used, personal information span

from pictures to videos, from wikis to blog posts,

from forums to discussion groups; the list would go

a long way. Heterogeneous information sources

continuously change in nature and content, moving

around highly dynamic centroids, topics, which

attract people sharing interests or just the desire of

publishing something: personal data, pictures,

artworks etc... To name a few, Facebook, Myspace

(MySpace.com, ©) and Flickr groups (Yahoo, ©),

YouTube channels (YouTube, LLC ©) and LinkedIn

(LinkedIn Corporation ©) or web sites aggregating

similar feeds from different sources. It is also the

case of newsgroups, or wikis. None of the above,

however, gives the user a thorough understanding

nor a total access to topic-related information.

Thus, building virtual communities of people

sharing the same areas of interest, and moving onto

topic-driven web surfing and information sharing is

a key aspect in (re)organizing world’s information.

This is what Radar Networks promises with the

forthcoming Twine (Radar Networks ©) which

promises to set as the first mainstream Semantic

Web application. Twine will construct a RDF graph

mapping relationships among people and topics as

well, giving the user full control over information

organization, providing a mean to share knowledge

with like-minded people. Twine follows successful

past experiences from both the industrial (see the

examples so far) and research worlds: consider past

Semantic Browsers emerged from the research

community, geared towards personal semantic

bookmarking, like Semantic Turkey (Griesi,

Pazienza, & Stellato, 2007), social semantic

annotation, as for Piggie Bank (Huynh, Mazzocchi,

& Karger, November, 2005), or Web Services

composition (Dzbor, Motta, & Domingue, 2004).

3 THEMATIC OASES

Industrial and research worlds are thus sharing the

same views and aims, colliding towards a new web

vision, where “knowledge” is no more a huge

amount of (semi)structured text but, turning into a

cloud of overlapping concepts, topics and domains.

At the same time, it emerges the need for a better

organization of the huge amount of information

erupting from the stream of available technologies.

To find the way through the plethora of

information sources, differentiating in content,

presentation and accessibility, a viable approach

would be to make these layers explicit, be able to

recognize them as several possible manifestations of

the same knowledge, and organize them

accordingly.

It makes sense, then, to make the jump from

single (and in-interoperable) specific services (blogs,

wikis, forums, discussion groups and so on…) to

huge collectors of information on a open and global

scale, which we dare to call Thematic Oases.

Thematic Oases (TOs, from now on) should provide

the main intellectual stream of interests around

which knowledge should be organized (and different

services be offered). By adopting Semantic Web

standards, TOs would be developed around

ontological repositories of conceptual knowledge,

which will be used as reference vocabularies for

accessing contents of federated (or simply annexed)

services and (socially) bookmarked web pages.

In this scenario, traditional services will be still

reusable and will coexist with their new semantic

counterparts, with the former being semantically

annotated with respect to the ontologies adopted in

given Thematic Oases, and the latter natively

WEBIST 2008 - International Conference on Web Information Systems and Technologies

352

Figure 1: Conceptual Architecture and Use Case of a Repository of Thematic Oases.

supporting a semantic organization of their content.

The main principles of TOs should be:

– Affordable setup: no more heavy bulked Social

Networks held by major company titans. Much

the same way a normal web user can now start a

forum or a blog using third party (often free)

software, Thematic Oases should be at the hand

of any user with the availability of a web host or

of an hosting service

– Accessible by (Semantic?) Search Engines: In

our vision, this is surely something related to the

open nature of TOs, but would beneficiate at the

same time of some commitment from search

engines, which will be able to improve quality

of searches through proper indexing of semantic

annotations publicly exposed by the oases

– Scalable open architecture: a given service may

be explicitly built upon a TO, committing to its

ontologies and content organization. Vice-versa,

in an even more open view, independent

services may be linked by a given TO. This

would allow users to tag the content of these

services according to the oasis’ reference

ontologies, thus easily putting traditional (non

semantic-driven) services immediately into

play. The same would hold for standard web

pages. People could write web pages directly

connected to a TO making explicit reference to

its vocabulary, as embedded RDFa (Adida &

Birbeck, 2007), or could semantically bookmark

an external web page (or annotate part of its

content) against that same vocabulary.

The above principles should promote a new

interpretation of today social networks, where

people gets back the ownership of their own data,

being able to publish them autonomously and freely

move them according to their hosting possibilities.

At the same time, these services could be completely

defined by the users, according to their specific

interests and exigencies, addressing important

themes and coordinating different services around

their explication, whereas current social networks

offer nothing more than well-cooked showcases for

exhibiting our personal data, multimedia, and social

contacts. Thematic Oases could become just mere

aggregators of already existing services, by

providing the possibility of storing semantic

annotations in their internal repositories.

Let’s think about a user willing to buy a new

monitor for his Pc. He would access a TO about

computer hardware, then browse the hardware-

ontology looking for video peripherals and getting

the pointer to the monitor concept. He could then

decide to learn a bit more about monitors before

deciding to buy a new one, thus accessing to the

wiki page associated to the concept. Thus he learns

interesting features which he uses to filter out a few

possible models he is interested in. He finally uses

the search comparisons functionality, pointing to

discussions in forums, reports in webpages/wikis,

RSS fed discussions which have been tagged by the

community as comparisons and instantiated wrt the

specific models selected by him.

THEMATIC OASES - Towards a Social Semantic Web

353

3.1 The Ontology Commitment Issue

One of the biggest (and most discussed) problems in

the realization of the Semantic Web (Dijck, 2003),

was about the difficulty in establishing consensus on

domain representations. The major concern was

about the natural resistance of companies and

providers to commit to any kind of knowledge

organization which could not reflect their inner

nature and characteristics or simply properly address

their specific information needs. Also, the different

languages and cultures which must be considered on

a World Wide scale needed to be kept into account,

posing another obstacle towards ontologies seen as

acceptable “shared reference vocabularies” in the

Semantic Web. These fears, appear, at least in part,

as dictated by and old fashioned way of thinking

about knowledge organization. Most widely adopted

ontologies contain now very simple descriptions of

very specific aspects of a domain (or of reality, in

general). The FOAF ontology (Brickley & Miller,

2007) contains just very simple data for describing

people personal information and for establishing

connections between people. The Basic RDF Geo

Vocabulary (Brickley, 2007) is composed of just

three attributive properties for describing WGS84

standard latitude, longitude and altitude, plus a

generic Point concept for addressing points in the

space using these properties. These ontologies can

be easily imported in any more complex knowledge

organization system, with no fear of generating

unsolvable inconsistencies, while leaving the

possibility of providing ad-hoc domain descriptions

for addressing specific needs, by adding arbitrary

concepts and relationships to the imported ones.

This approach guarantees a desirable degree of

shareability of the collected data (at least, on its

higher level descriptive units) while preserving the

intellectual independence in modeling specific

scenarios and domains. Following this approach,

TOs, while allowing for ad-hoc ontologies

developed for their specific needs, should foster

reuse of “standard ontologies”, thus opening to

external linkable services which have been

developed independently from their suggested

vocabulary, as well as enabling peer-to-peering

among different Thematic Oases.

4 CONCLUSIONS AND FUTURE

WORK

In this paper we have presented our vision about a

possible concrete application of Semantic Web

principles to the social paradigms which are

characterizing today Web2.0 era. We firmly believe

that the Web community is ready for embracing new

air breezing from the so called Web3.0 semantic

trend, but is currently blocked by the unavailability

of solid and impacting killer-applications giving a

rightful reason for learning the next step of Web

evolution.

REFERENCES

Adida, B., & Birbeck, M. (2007, October 26). RDFa

Primer. Retrieved from W3C: http://www.w3.org/TR/

xhtml-rdfa-primer/

Berners-Lee, T. (2000). Weaving the Web. New York:

HarperBusiness.

Brickley, D. (2007). Basic RDF Geo Vocabulary.

Retrieved December 10, 2007, from http://

www.w3.org/2003/01/geo/

Brickley, D., & Miller, L. (2007, November 2). FOAF

Vocabulary Specification 0.91, Namespace Document

2 November 2007 - OpenID Edition. Retrieved from

xmlns.com:http://xmlns.com/foaf/spec/20071002.html

Dijck, P. V. (2003, November 15). Themes and metaphors

in the semantic web discussion. Retrieved from

http://poorbuthappy.com/ease/semantic/

Dzbor, M., Motta, E., & Domingue, J. B. (2004). Opening

Up Magpie via Semantic Services. 3rd Intl. Semantic

Web Conference (ISWC04). Hiroshima, Japan:

November.

Facebook | Welcome to Facebook!: http://

www.facebook.com/

December 10, 2007, from Google Code:

http://code.google.com/apis/opensocial/

Griesi, D., Pazienza, M. T., & Stellato, A. (2007).

Semantic Turkey - a Semantic Bookmarking tool

(System Description). 4th European Semantic Web

Conference (ESWC 2007). Innsbruck, Austria.

Huynh, D., Mazzocchi, S., & Karger, D. (November,

2005). Piggy Bank: Experience the Semantic Web

Inside Your Web Browser. Fourth International

Semantic Web Conference (ISWC05), (p. 413-430).

Galway, Ireland.

2007, from LinkedIn: Relationships Matter: http://

www.linkedin.com/

from MySpace: http://www.myspace.com/

from Twine: http://www.twine.com/

Flickr: Groups: http://www.flickr.com/groups/

from YouTube - Broadcast Yourself: http://

www.youtube.com/

WEBIST 2008 - International Conference on Web Information Systems and Technologies

354