=Paper=
{{Paper
|id=Vol-464/paper-11
|storemode=property
|title=Metasocial Wiki — Towards an interlinked knowledge in a decentralized social space
|pdfUrl=https://ceur-ws.org/Vol-464/paper-02.pdf
|volume=Vol-464
|dblpUrl=https://dblp.org/rec/conf/semwiki/CanoRC09
}}
==Metasocial Wiki — Towards an interlinked knowledge in a decentralized social space==
https://ceur-ws.org/Vol-464/paper-02.pdf
Metasocial Wiki - Towards an interlinked
knowledge in a decentralized social space
Amparo E. Cano1 , Matthew Rowe2 , Fabio Ciravegna3
Department of Computer Science,
University of Sheffield,
Sheffield, United Kingdom
A.Cano1 ,m.rowe2 , F.Ciravegna3 @dcs.shef.ac.uk
Abstract. This paper introduces a new approach to semantic wikis. In
this approach users coming from different social networks can be merged
into a common space to enable collaboration. This approach makes use of
the user’s identity representation and keeping track of the user’s interests
according to the type of annotations encountered in the content they add.
Keywords: semantic wiki, digital identity, collective intelligence,
social networks.
1 Introduction
According to Metcalfe’s law[1], the larger the network the more valuable it be-
comes. Although individual thinkers invent and discover, it is groups, which
typically refine and extend innovations. Moreover, highly developed ideas rarely
emerge from single and isolated thinkers, they usually come as a result of a pro-
cess of interaction [2]. Even though semantic wikis have proven to be a successful
tool for collaborative working, there is still a long way to go to fully exploit users
collective intelligence. In this paper we describe a new approach to semantic wiki
systems. In this approach the skills from users coming from different social net-
works are merged into a common collaborative space called MetaSocial. The
MetaSocial project (a resarch proposal) introduces this approach and some of
the challenges it presents.
1.1 Social Networks and Semantic Wiki Systems
Work by [2] and [3] report that users prefer to establish friend relationships
with other similar users in a social space, regarding for instance socio-cultural
traits. This has lead to the aggregation of people from different backgrounds
into common spaces where they share similar interests and tastes. Social net-
works (SN) differ in the services they provide, targeting different demographics
with different purposes. All SN are valuable in terms of aggregating people with
distinguishable features. For instance LinkedIn1 aims for professional exchanges
1
http://www.linkedin.com
between individuals, while Orkut targets casual and leisure exchanges between
family and friends. In that sense, SN help to define an individual in terms of a
context that includes the type of people they target. Although research in SN
have demonstrated the great potential that SN present for generating knowl-
edge, less attention has been paid on how to harness users’ knowledge in gener-
ating collaborative content. Some of the current SN offer capabilities for creating
communities within the SN, however they dont provide tools for collaboration.
Moreover, until the appearance of Open Social2 , the communication between
independent SN was not possible.
The collaboration has been set aside from SN. Although collaborative sys-
tems, such as a wiki, provide tools for user participation into common tasks, e.g.,
discussion pages, there is still a lack of ties that prevent users from propagating
information and from promoting their participation in a given task, by bringing
this to the attention of the user’s acquaintances. There is also still a gap between
SN and collaborative systems. One possible solution to this situation would be
to enable the capability of managing relationships within semantic wiki engines;
however that would not be enough, as it would leave aside the potential benefit
that the diversity of SN can bring into a common collaborative space.
1.2 Semantic Wiki and the linked Web
The semantic web is not about putting data together but about making links
between the data [4]. So far, semantic wiki engines have acted as isolated content
stores. Although efforts such as [5] have established the existence of links between
same instances of data from different semantic wikis, there is still a broken link
between the author of the data and the identity of the author. The integration
of digital identities into semantic wiki engines can be perceived as an attempt
to break away from one of the wiki principles regarding minimal access control,
which refers to the capability of contributing in an anonymous way or by using
a registered username, which can hold minimal authoring information.
Online identity can be defined as the representation of one’s persona in a
digital context[6]. It is worth noting that in the case of online identity the user has
the ability to define a personal facade that represents him in a particular context.
People in social environments tend to present just a facet of their identity for
others to perceive. The integration of digital identities with wiki engines would
allow the users to select their preferred persona to represent his authorship on
a given wiki, while the wiki engine could still allow the anonymous authoring.
The integration of digital identities into wiki engines open a promising field
in which the contextualization of a persona can be developed according to the
activities carried out by this persona on the Web. Services integrating digi-
tal identities would not only pull information from the user’s identity service
provider, but also push information about the task carried out by the user that
would help to better define the user’s persona’s interests and knowledge. The
aggregation of information containing linked data coming from different data
2
OpenSocial, http://www.opensocial.org
silos could help in building up a decentralized digital identity [7]. The openID
version 2.0 OpenID3 protocol for digital identity providers already defines a at-
tribute exchange protocol however it doesn’t provide a standard way of defining
and developing users’ personas’ contexts.
Until now, just a few semantic wiki engines have integrated the openID stan-
dard into their engines however they use it just as an authentication service. On
the other hand, none of them have introduced the use of the social intelligence
included in the users’ FOAF4 files.
Incorporating that information would enable the merging of users’ SN infor-
mation (including profile, social graph and interests for instance). This informa-
tion can help to correlate users coming from different SN. Consider a user who
is part of an alternative rock community within meta-social and is interested in
releasing his first album. He creates a new project with that subject. People with
similar interests can be advised to join the project. The user should be able to
try to establish a relationship with other users, not necessarily a friendship but
a colleague relationship, advertising his project. In this way he could establish
relationships with people from Myspace interested in music, which could advise
him on the design of the album’s cover, or with people from LinkedIn who could
help him to better position his album on the market.
Integrating the user’s identity and social information into semantic wiki en-
gines would enable linkage between user’s identity with the type of contents he
adds (presumably the type of topics he is interested in) and the relationships he
establishes. Keeping track of this information will enable better tracking of users’
interests and would facilitate the improvement of suggestion engines. Moreover
making this information available to the user’s identity service provider, would
help to decentralize the information given by the user, enabling it to be reused
in any other system linked to the user’s identity.
2 Proposed Approach
The proposed functionalities for Metasocial are contained within two separate
parts. Each part depends on semantic technologies to enhance existing services.
A semantic wiki plays a crucial role in bringing the below functionalities together
as we now explain.
2.1 Social Functionalities
Metasocial will allow users to collaborate from multiple social networks, therefore
managing user accounts will be addressed to map individual accounts in different
social spaces. As mentioned, OpenID will be used as a single user identity URI
for each user of Metasocial. Social graphs using both the FOAF and SIOC5
specifications will be imported into Metasocial from multiple distributed social
3
OpenID, http://openid.net
4
FOAF, http://www.foaf.org
5
SIOC, http://www.sioc.org
web platforms and linked together. The intuition behind this functionality is
to enhance the information attributed to each user thereby offering intelligent
suggestions based on their prior knowledge. Such social graph interlinking also
contributes to current initiatives to address identity fragmentation and data
portability 6 .
Status information of each social network member will be described using
the Online Presence Ontology 7 , and with current initiatives such as Smesher 8
it is possible to convert such updates to semantic representations usable by the
system. For example, if a person’s status describes how they are busy working on
a given project task, then Metasocial would not suggest additional work. Possible
collaborations with work colleagues are suggested based on the imported social
graphs, each social network member also has a list of interests extracted from the
hosting service. Suggestions are then made based on such interests for specific
projects. The converse is also true in that users are suggested projects and work
based on their interests described in multiple social graphs. Combining such
identity fragments we build a more complete profile of the user allowing projects
spanning both the Semantic Web and Social Web to be suggested. We use this
example as a very simple indicator of a trivial suggestion task 9 .
A semantic wiki provides a useful means to control semantic graphs at-
tributed to individual users, and make inferences based on the type of annotation
the users have used when adding content. Metasocial will maintain a knowledge
base capable of offering a useful collation of knowledge statements expressed
within the wiki. One of the attractions of using a semantic wiki is the ability to
effectively infer suitable projects suggestions and colleagues based on available
semantic information.
2.2 Knowledge Functionalities
Within Metasocial collaborative environment projects will be described using the
DOAP ontology10 including extensions to this specification to capture knowledge
describing more generic projects (at present DOAP is tailored more towards
software based development projects). Looking for a project will be controlled
using a semantic search mechanism by aligning the semantic concepts the user
has expressed an interest in with similar projects. This allows more general and
specialised projects to be returned if the user’s original criteria are not explicitly
matched.
Project management will require tools such as time planners, task managers
and role assigners all usable on the semantic wiki. Social information assigned
to users within the collaborative environment allows suitable roles to be sug-
gested based on the semantics of the user details, i.e., sioc:Role. Managing
6
Date Portability Group. http://www.dataportability.org
7
OPO, www.milanstankovic.org/opo/ontology.html
8
Smesher, http://smesher.com
9
Twine platform already leverages users interests, and track of user’s searches for
suggesting content. See http://www.twine.com
10
Doap, http://usefulinc.com/ns/doap
project milestones would be enhanced through interactions with semantically
linked calendars; allowing project members to receive updates and reminders
about upcoming milestones. Reminders would be controlled automatically by
the semantic preferences stipulated by the project member.
Project work will be labeled using free text tagging, which is in turn aligned
with concepts from a knowledge base, thus controlling term ambiguity and cor-
rect co-referencing. External knowledge sources can also be used 11 for greater
availability of concept definitions. The internal knowledge base would allow dis-
course to be developed specific to that project so that knowledge generated as
a result of this process could then be shared with additional projects. Natural
language style queries could be asked, either returning any derived answers or
relevant knowledge from the knowledge base, or allowing project members to
answer the questions themselves. Completed projects tasks would become less
visible to the user on the semantic wiki based on the task being semantically
defined as complete. This would offer the functionality to display all completed
projects of a specific type, enhancing the knowledge management functionality
on the wiki and encouraging reusability of the existing projects.
3 Conclusions
Metasocial introduces various challenges, most of them concerning the represen-
tation of the user context in an standard format, the extraction of information
(interests in particular) from user added data and the inference of information
from the user’s social graph. Offering a platform where people from different
social networks can not only communicate, but also collaborate easily in the
development of semantic-aware projects according to their interests will help in
allocating the right user with the right skills in the right projects.
References
1. Reed, D. P.: That sneaky exponential beyond metcalfes law to the power of com-
munity building. Context Magazine (1999)
2. Rohilla Shalizi, C.: Social Media as Windows on the Social Life of the Mind CoRR
abs/0710.4911: (2007)
3. Newman, M. E. J.: Mixing patterns in networks. Phys. Rev. E. Vol. 67, (2003)
4. Tim Berners-Lee:Linked Data. http://www.w3.org/DesignIssues/LinkedData.html
(2008)
5. Passant, A., Laublet P.: Towards an Interlinked Semantic Wiki Farm. CEUR Work-
shop Proceedings. SemWiki 2008.
6. Russell, Terrell and Stutzman, Frederic :Proceedings of the American Society for
Information Science and Technology. American Society for Information Science and
Technology . Self-representation of online identity in collected hyperlinks. Vol.44,
2007.
7. Rowe, Matthew.: Proceedings of Linked Data on the Web Workshop 2009, Madrid,
Spain. World Wide Web Conference. Interlinking Distributed Social Graphs. 2009.
11
Such as DBPedia, http://dpedia.org and Freebase, http://www.freebase.com