A Semantic Web Model for Ad Hoc Context-aware Virtual Communities

Application to the Smart Place Scenario

Pierre Maret, Fr

erique Laforest and Dimitri Lanquetin

Universit

e Jean Monnet, F-42000 Saint-Etienne, France

Keywords:

Communities, Mobility, Services, User, Semantic Web, RDF, Ontology.

Abstract:

In this paper, we propose a model for an open framework that allows mobile users to create and to participate

to context-aware virtual communities. The model we propose and implement is a generic data model fully

compliant with the semantic web data model RDF. This model is suited to let mobile end-users use, create and

customize virtual communities. We combine fundamentals for a decentralized semantic web social network

with context-aware virtual communities and services. Smart cities scenarios are typically targeted with this

approach. It can be implemented in places like metro stations, museums, squares, cinemas, etc. to provide ad

hoc context-aware information services to mobile users.

1 INTRODUCTION

A virtual community (VC) is a computer sup-

ported space where information can be shared in-

between stakeholders with common interest or pur-

pose (Preece, 2001; Koch et al., 2002). Millions

of VCs exist on the Internet and they form what is

called the Social Web, where people can join, leave

and participate in many VCs (Bouras et al., 2005). In

this paper, we address the topic of context-aware vir-

tual communities and more speciﬁcally we present a

data model and its implementation suited to let mobile

end-users use but also create and customize VCs.

Some community platforms take advantage of the

mobile context of their users (GPS coordinates, IDs

of close devices) to provide a speciﬁc service. Ex-

amples are numerous: Personal Smart Spaces (Gal-

lacher et al., 2012), Group Me (Guo et al., 2012),

Meeting Assistant (Zenker et al., 2012; Al Ridhawi

et al., 2012). They use context information to provide

guidance, proﬁle matchmaking or recommendations.

Each platform provides a speciﬁc service. Users will-

ing to use (not even to combine) various services face

the so called ”silo” problem and they have to reg-

ister to each platform (Yeung et al., 2009). Also,

these platforms don’t let users deﬁne the services they

need in their communities. The platforms propos-

ing user-deﬁned communities require programming

skills. Synthesizing all these remarks, nowadays com-

munity platforms are closed platforms that do not al-

low to take proﬁt to the open world of open data, com-

munities are hardly tunable by their creators without

computing skills, the sharing of proﬁle information or

knowledge between communities and with the open

world is hard or even impossible.

In smart cities, users are supposed to move and use

various services. So, there is need for a model and an

open architecture for users to create the communities

they want, with the services they want, and at the loca-

tion they want (or more generally in the context they

want the services to be used).

In this paper, we propose a model and a platform

architecture for ad hoc context-aware communities:

communities can be created by naive end-users with

a selection of pre-deﬁned semantically described ser-

vices. Users get access to these communities with

a single sign-in. Our model makes use of Seman-

tic Web technologies which allows it to provide self-

describing services and to offer a open door to the

open linked data inand out of the virtual communi-

ties. The semantic web approach promotes open data

model, machine readable data and reasoning, which

let the users be supported by personal software agents

to manage their personal data and social relations in

respect with the chosen policy. Our platforms also

uses the multi-agent paradigm to support users’ in-

teractions. These interactions take advantage of con-

text information to help services tune their offered

functionalities. The multi-agent paradigm empha-

sizes autonomy and cooperation of communicating

entities with goals. Our design typically addresses

smart cities scenarios. It can be applied in places

591

Maret P., Laforest F. and Lanquetin D..

A Semantic Web Model for Ad Hoc Context-aware Virtual Communities - Application to the Smart Place Scenario.

DOI: 10.5220/0004876905910598

In Proceedings of the 16th International Conference on Enterprise Information Systems (ICEIS-2014), pages 591-598

ISBN: 978-989-758-028-4

 2014 SCITEPRESS (Science and Technology Publications, Lda.)

like metro stations, museums, public squares, cine-

mas, park places, etc.

We ﬁrst present a state of the art related to the

main features for communities: contents, user proﬁle

and services. Then we make a tour of existing com-

munity platforms. In section 3 we present our concep-

tual framework along the three main features from the

state of the art. In section 4 we provide a description

of our implementation, and conclude in section 5.

2 CONTENTS, USER PROFILE

AND SERVICES IN VCS

2.1 Contents

Contents are information pieces exchanged between

users within a community. The underlying data model

impacts the sharing of these contents within the com-

munity platform and out of the platform: inputs,

queries, exports, imports of information can be en-

visioned differently depending on this model. The

model can be either platform dependent or open.

Most social networking websites use their own

data models, speciﬁcally designed for their architec-

ture. Moreover, the data model is rarely publicly

available. Platform-dependent data formats make it

nearly impossible to import/export contents, proﬁles

and social relations (Razmerita and Firantas, 2009). It

leads to the well-known isolation of social networks.

Platform independent data models are public.

Since community platforms run on the Internet and

in a Web environment, norms or standards from these

ﬁelds are the best candidates to share models. SIOC

(Breslin et al., 2005; Bojars et al., 2008) has been

proposed for the representation of social data. It is

founded on the Semantic Web technologies. The Se-

mantic Web approach (Berners-Lee et al., 2001) aims

at representing data in terms of RDF

triples: (term

relation value) or (term1 relation term2). It is a user-

and machine-readable model. Each item of a triplet

can be described anywhere on the Internet (URI, Uni-

form Resource Identiﬁer). This builds the so-called

linked data. Sesame (Broekstra et al., 2002) and Jena

are Java frameworks used to store RDF triples. Rea-

soning processes can be launched to infer facts, es-

pecially in conjunction with formal description of do-

mains (ontologies) which are also written with RDF

(more speciﬁcally using RDF-based languages: OWL

and RDFS). SIOC is written in the RDF/OWL lan-

guage. Its use provides models and techniques for a

platform independent data model.

http://www.w3.org/RDF/.

2.2 User Proﬁle

Some platforms (Gallacher et al., 2012; Ganti et al.,

2012; Guo et al., 2012) propose to enrich dynamically

the user proﬁle with context information: physiologi-

cal parameters, GPS coordinates, nearby devices, etc.

The management of several proﬁles is a complex task

for end-users. Some works have proposed to federate

different proﬁles (Mitchell-Wong et al., 2007). How-

ever, all characteristics cannot be federated and, for

average users, keeping proﬁles safe and up to date

quickly becomes an issue. This is crucial because

user proﬁles contain personal information. A com-

plex management of proﬁles raises privacy issues:

users may not adequately control data sharing.

A unique proﬁle used on several community plat-

forms eases its management and allows the user to

better control the access rights to personal data. The

proﬁle can be stored either on a remote system or on

the user’s side. Several proposals can be found in the

literature (Seong et al., 2010; Koch and Worndl, 2001;

Grzonkowski et al., 2009). The most advanced model

for proﬁle description is FOAF. It is described as an

ontology, also written in the RDF/OWL language.

2.3 Services in VCs

A virtual community offers services to its participants

to communicate i.e. create, modify, read, share infor-

mation. It can take the form of forum, blog, chat,

private messaging, walls... Services are key compo-

nents of communities as they can strongly inﬂuence

participation and adoption. Roles are used to deﬁne

different rights of users: a forum can have a moder-

ator, members, experts, etc. Other types of services

can be found such as services related to the location,

to information retrieval, to recommendation, etc.

The services offered on a platform are determined

by the developers or the administrators. Also services

are by deﬁnition not available out of the platform and

they are not exportable: they are made for a given

custom architecture and data model. Only some VC

platforms allow users to create customized communi-

ties, but they require programming skills (see 2.4).

When the number of services increases, users

should receive recommendations to select the most

relevant ones regarding their preferences, social re-

lations, and context (Schubert and Koch, 2003). In

(Broens et al., 2004; Meyffret et al., 2010), ontolo-

gies for the descriptions of services, for the user’s

preferences and for context information are used to

improve service discovery and selection. In (Sinner

et al., 2004), a framework is used to match semantic

user proﬁles with semantic descriptions of location-

ICEIS2014-16thInternationalConferenceonEnterpriseInformationSystems

592

based services for recommendation.

2.4 Contents, User Proﬁle and Services

in Existing VCs Systems

Popular social networking websites such as Facebook

use proprietary models. User proﬁles are stored on

the proprietary platform which is based on a cen-

tralized architecture. Different approaches have been

proposed

. Diaspora is a software environment for

distributing the components of the social network at

the users’ sides. Evri has platform-dependent pro-

ﬁles and services, however it uses ontologies for the

semantic annotation of contents. dgFOAF (Schwa-

gereit et al., 2010) proposes users to be registered in

a unique community and to launch services from dif-

ferent platforms. Elgg and Liferay allow the develop-

ment of community platforms. Services can be added

as plug-ins (or portlets), making community creation

and upgrading customizable. However, they require

programming skills.

The multi-agent approach has been also used for

VCs (Gunasekera et al., 2012; Aguero et al., 2012).

Koch (Koch and Lacher, 2000) introduces agencies

to separate user proﬁle and community participation

(i.e. use of community services). Guidelines are made

available for people with programming skills wanting

to develop a community platform, with no restriction

on contents or services representation. (Fahad et al.,

2012) is a community platform based on the multi-

agent framework Cartago

. End-users can create and

join communities, and they can delegate these tasks to

autonomous agents. The communities are limited to

four services: mailbox, forum, information dispatcher

and personal box.

Properties of community platforms are summa-

rized in table 1. None of these let end-users create

and customize their VCs in selecting services and as-

sociating them to context information. Few platforms

use open (semantic web) models.

3 A DATA MODEL FOR AD-HOC

CONTEXT-AWARE VIRTUAL

COMMUNITIES

In this section, we propose our data model to develop

the framework for ad hoc context-aware communi-

ties. The model relies on RDF, and consists of three

main elements: (i) A user proﬁle based on the FOAF

http://diasporaproject.org/, http://www.evri.com/,

http://elgg.org/, http://liferay.com/.

http://cartago.sourceforge.net.

model and encompassing also user’s context descrip-

tion, (ii) A community model that extends SIOC with

communities-related properties, and (iii) A service de-

scription model that allows the use of on-the-shelf ser-

vices.

Table 1: Properties of community platforms. (1)has an open

semantic data model; (2)allows multi-platform; (3) imple-

mentation of services.

Platforms (1) (2) (3)

Facebook, Twitter no no developer’s choice

Diaspora no yes developer’s choice

Evri yes no developer’s choice

dgFOAF yes yes developer’s choice

Elgg, Liferay no no developers’ plug-ins

Fahad et al. no yes predeﬁned services

Koch et al. no yes developer’s choice

3.1 User Proﬁle

As we have seen in section 2.2, a unique proﬁle stored

at user’s side is desirable. We use and extend the

FOAF ontology. The model (Fig. 1(a)) has the fol-

lowing properties: foaf:interest (a topic about which

the actor wants to share, obtain, create information),

foaf:friend (another actor known by the user), hasA-

vatar (a delegate of the actor in the systemhosted by

an electronic device of the user), hasGoal (goal to be

achieved by the avatar, see 3.4). Properties foaf:friend

and foaf:interest are taken from the FOAF ontology.

The Actor class refers to both the foaf:Person and the

UserAccount class in SIOC. The SIOC:Avatar class

has no relationship with our Avatar class.

Figure 1: Actor model (a), Avatar model (b).

In addition, avatars have the property hasContext,

which is the capability of the device hosting the avatar

to deliver context data (Fig. 1(b)). We deﬁne context

data as a simple set of pairs (descriptor, value). These

pairs are used to represent current information sensed,

ASemanticWebModelforAdHocContext-awareVirtualCommunities-ApplicationtotheSmartPlaceScenario

593

Figure 3: Content (a) and Community (b) models.

Figure 2: Context model.

calculated or known by the device hosting an avatar.

Context data may relate to location but also to battery

state, network connections, user’s behavior (through

an accelerometer), etc. The property hasContextData

is the piece of information describing the state of the

device hosting the avatar or a status of the actor (Fig.

2). It contains hasDescriptor (a string describing the

information) and hasValue (context data value).

3.2 Content and Communities

Contents consist of pieces of information coded in

RDF format with the properties hasAuthor, hasTopic,

type and value which associates a value of a RDF

class deﬁned as a URI (Fig. 3(a)). The sharing of con-

tents happens in the frame of communities. We deﬁne

the Community ontology as an extension of the SIOC

ontology (Fig. 3(b)). It is composed of hasOwner,

hasTopic (contents used as subject of the community),

hasMember (actors), hasService (see 3.3), hasRole

(see 3.4) and hasContent (contents shared within the

community). SIOC also provides the notion of topic

Figure 4: Service model.

with a dedicated property that usually refer to a se-

mantic web resource.

3.3 Services

We deﬁne a service as a set of operations. For in-

stance, a blackboard is a service, with the operations

write, read and delete. Services in our framework are

related to the management of contents. They are im-

plemented to create and get contents from communi-

ties (hasContent property). As all communities are

described using the data model deﬁned above, any

service can be used in communities. Services are de-

scribed using the ontology given in Fig. 4. It contains

the name of the service, its description and operations.

The Service class of SIOC has been adapted to deﬁne

our Service class.

Properties of an operation are described in Fig.

5(a). Notice that the hasRequiredContext property in-

troduces the context of the avatars (see 3.1). This is

used to introduce the context-aware property of op-

erations. The property hasCommand relates to the

exchanges of information: Get, Put, Delete or Post;

it uses hasInputParameter and hasOutputParameter

A parameter (Fig. 5(b)) has a description, a type

(RDF class) and constraints applied on its value (RDF

triple). The hasMultiProperty is used to indicate mul-

tivalued properties.

ICEIS2014-16thInternationalConferenceonEnterpriseInformationSystems

594

Figure 5: Operation model (a) and Operation parameter model (b).

3.4 Goals and Roles

Goals are deﬁned by actors. They are part of their per-

sonal data and consist of sets of operations that avatars

will execute (Fig. 6(a)). The goal model uses part of

the operation’s model deﬁned above to ease service

discovery and selection. Roles deﬁne the rights to ex-

ecute operations in the frame of a community. A role

is composed of a set of actors and a set of operations

(Fig. 6(b)). The notion of role also exists in the SIOC

ontology with the class Role that can be referenced by

a UserAccount.

Figure 6: Goal model (a) and Role model (b).

4 FRAMEWORK

IMPLEMENTATION

In the previous section, we have presented the data

model of our framework. It allows users, commu-

nities and services modeling, and more speciﬁcally

context-aware operations within communities. It is

based on RDF principles and can be easily extended

or adjusted. In this section, we present the imple-

mentation of this model and we show how it allows

end-users to easily create their customized communi-

ties with different services and for dedicated contexts.

The compliance with the RDF principles is kept with

the framework implementation.

4.1 Main Implementation Choices

Our implementation combines the semantic web tech-

nologies with the multi-agent approach. Indeed,

agents (users, avatars) as well as information and ser-

vices are distributed in different concrete or virtual

(i.e. computer supported) locations. Furthermore

agents interact with others and share information in

the frame of the communities to achieve user’s goals.

We use the multi-agent platform Cartago to im-

plement our system. In Cartago, agents interact with

other agents but also with the environment through

the artifact abstraction. We implement data store,

avatars, communities and services with this abstrac-

tion. Through artifacts, agents can write/read data,

get context values, create and participate to communi-

ties and execute operations. Cartago proposes also the

concept of workspace to group artifacts and agents.

We use this concept to group communities based on

ASemanticWebModelforAdHocContext-awareVirtualCommunities-ApplicationtotheSmartPlaceScenario

595

context values (for instance the location). We call

Space of Communities (SoC) a set of communities

hosted in a same workspace. Distributed communities

can also be realized since communication in-between

workspaces is possible.

As previously mentioned, we use RDF triples to

model contents within communities, users’ proﬁles,

and services. This RDF implementation conforms to

open linked data models. We use Sesame repositories

(Broekstra et al., 2002) within artifacts to store and

access these triples. Services are built on top of Web

services. Details of the framework architecture are

described in the next subsections and summarized in

ﬁgure 7.

4.2 Implementation of Users Proﬁles

The user proﬁles are implemented as described in the

data model (see 3.1) under the form of RDF triples.

Each user’s avatar is composed of an artifact (Pro-

ﬁle artifact) and 3 agents (Proﬁlor, Space Detector,

Participator). The Proﬁle artifact is used for access-

ing a Sesame repository which contains triples about

context information, information topics and friends as

deﬁned in 3.1. The Proﬁlor agent lets users manage

their proﬁles. The Space Detector agent is a contex-

tual detector of SoCs. The Participator agent is cre-

ated when a SoC is detected. It allows the user to join

or create communities: the agent ﬁlters the informa-

tion received from the SoCs, depending on the user’s

topics of interest, context and goals and notiﬁes the

user. This process is privacy-aware since no personal

data is transmitted to the SoC. Notice that as the user

proﬁle is unique, avatars of a single user interact to-

gether to merge data and keep up to date.

4.3 Implementation of Communities

and their Services

As stated before, Cartago artifacts are used to imple-

ment communities. Once a community is created,

a community artifact allows the access to a Sesame

repository. It stores the data of this community:

shared contents, roles, services (see 3.2). A dedicated

SoC artifact manages the list of communities and the

list of services the SoC holds.

Services interfaces are also implemented as arti-

facts. They provide a set of methods for querying (us-

ing SPARQL), adding, deleting triples in the commu-

nity repository. The availability of a given service in

a community is decided by the community creator.

4.4 Smart City Scenario and Discussion

Our system can be used for the following scenarios.

A city has implemented in the most populated quar-

ters the ad hoc context-aware community platform

and has settled several spaces of communities with

different web services. A ride sharing service and

a question/answer service are among these. Opera-

tions of the ride sharing service are: declare an offer,

search for daily rides, and book for a ride. The ques-

tion/answer service has two main operations: search

a query and get answers; and read a query and give

an answer. This second service is implemented at the

train station.

As a committed citizen, Alice wants to develop

daily ride sharing in her close area. She has learned

the existence of the community platform and decides

to register and to create her avatar (stored on her de-

vice). She creates a community with ”ride sharing”

as topic, the street name as the context for launching

the services, and she selects operations from the ride

sharing service. John is searching for ”car ride” on

the platform. Since he leaves close to Alice (same

context), and thanks to the semantic analysis (same

content), the platform proposes him to join Alice’s

community.

Bob and Colin have joined the platform. They of-

ten wait in the train station and have learnt that the

train company has created a question/answer com-

munity. Bob loves winning points every day in an-

swering questions asked by other participants. Colin

loves asking questions about subjects he wants to

learn about.

The scenarios illustrate the kind of implementa-

tions our data model can support for smart city ap-

plications. Users can manage their proﬁles on their

mobile devices, their avatars can collect context data,

and pieces of information can be gathered or deliv-

ered automatically following their goals. Spaces of

Communities (with their artifacts, data stores and ser-

vices) can be hosted by servers in speciﬁc places such

as museums, cinemas, stores, campus, points of in-

terest... Communities with ad hoc services can be

created by users with no speciﬁc knowledge in com-

puter science, thanks to the use of the semantic web

technologies. Users can join communities and thus

receive or deliver information.

Compared to other approaches, our approach pro-

poses a user proﬁle which is unique and which con-

tains context values from the user’s avatar. It can

be used in all Spaces of Communities. The models

for user proﬁle, context and data are open and com-

pliant with the Semantic Web technologies. Knowl-

edge items are RDF triples contained into URIs. This

ICEIS2014-16thInternationalConferenceonEnterpriseInformationSystems

596

Figure 7: Framework architecture.

opens our system to an unlimited amount of knowl-

edge, as it follows the open linked data principles and

technologies. Following such standards also allows

to include other value added services ; for example

reasoning on data can improve the user’s experience

with better recommendations. Finally, services can be

added into Spaces of Communities and they can be

easily included by end-users in the communities they

create. No programming skill is required : the use

of standards for the semantic description of services

allows their inclusion with a high level of interaction

with end users. Different types of communities can

be created, using different services according to the

communities needs. The semantic description of the

avatar context also allows the use of such information

by services ; they can adapt their behavior to the end-

user context of usage.

5 CONCLUSION

The design of ad hoc context-aware communities typ-

ically addresses smart cities scenarios. We have pro-

posed a model and an implementation which rely on

the Semantic Web approach. Our approach can be

useful in many different places and for many scenar-

ios. Our data model can be easily extended to suit spe-

ciﬁc needs, and various services can be used by end-

users. Cities, shops, malls, museum, cinemas should

be interested by the approach. In the near future, our

next steps will consist of building a more appropriate

and illustrative user interface for end-user, and of de-

veloping new services which will illustrate different

kinds of communities that could be initiated by users

in different places.

REFERENCES

Aguero, J., Rebollo, M., Carrascosa, C., and Julin, V.

(2012). Developing pervasive systems as service-

oriented multi-agent systems. In Snac, P., Ott, M.,

and Seneviratne, A., editors, Mobile and Ubiquitous

Systems: Computing, Networking, and Services, vol-

ume 73 of Lecture Notes of the Institute for Computer

Sciences, Social Informatics and Telecommunications

Engineering, pages 78–89. Springer.

Al Ridhawi, Y., Al Ridhawi, I., Bruno, L., and Karmouch,

A. (2012). Policy-based personalized context dissem-

ination for location-aware services. In Senac, P., Ott,

M., and Seneviratne, A., editors, Mobile and Ubiqui-

tous Systems: Computing, Networking, and Services,

volume 73 of Lecture Notes of the Inst. for Comp.

Sc., Social Informatics and Telecommunications Eng.,

pages 366–371. Springer.

Berners-Lee, T., Hendler, J., and Lassila, O. (2001). The

Semantic Web. Scientiﬁc American, 284(5):34–43.

Bojars, U., Passant, A., Cyganiak, R., and Breslin, J.

(2008). Weaving SIOC into the Web of Linked Data.

In Linked Data on the Web (LDOW) workshop, in con-

junction with WWW conference.

Bouras, C., Igglesis, V., Kapoulas, V., and Tsiatsos, T.

(2005). A web-based virtual community. Int. J. of

Web Based Communities, 1(2):127.

Breslin, J., Harth, A., Bojars, U., and Decker, S. (2005). To-

wards Semantically-Interlinked Online Communities.

In 2nd Eur. Semantic Web Conference.

Broekstra, J., Kampman, A., and Harmelen, F. V. (2002).

Sesame: A Generic Architecture for Storing and

Querying RDF and RDF Schema. In International Se-

mantic Web Conference - ISWC, pages 54–68.

ASemanticWebModelforAdHocContext-awareVirtualCommunities-ApplicationtotheSmartPlaceScenario

597

Broens, T., Pokraev, S., Sinderen, M. V., Koolwaaij, J., and

P (2004). Context-aware, ontology-based service dis-

covery. Ambient Intelligence, pages 72–83.

Fahad, M., Boissier, O., Maret, P., and Gravier, C. (2012).

Smart places: multi-agent based virtual community

management system. In Proc. of the 4th Int. Workshop

on Web Intelligence and Communities, New York,

USA. ACM.

Gallacher, S., Papadopoulou, E., Taylor, N., Williams, M.,

and Blackmun, F. (2012). Linking between personal

smart spaces. In Senac, P., Ott, M., and Seneviratne,

A., editors, Mobile and Ubiquitous Systems: Com-

puting, Networking, and Services, volume 73 of Lec-

ture Notes of the Inst. for Comp. Sc., Social Infor-

matics and Telecommunications Eng., pages 401–408.

Springer.

Ganti, R., Mohomed, I., Raghavendra, R., and Ran-

ganathan, A. (2012). Analysis of data from a taxi cab

participatory sensor network. In Puiatti, A. and Gu,

T., editors, Mobile and Ubiquitous Systems: Comput-

ing, Networking, and Services, volume 104 of Lec-

ture Notes of the Institute for Computer Sciences, So-

cial Informatics and Telecommunications Engineer-

ing, pages 197–208. Springer.

Grzonkowski, S., Gzella, A., Kruk, S. R., Breslin, J. G.,

Woroniecki, T., and Dobrzanski, J. (2009). Sharing

information across community portals with FOAF-

Realm. Int. J. of Web Based Communities, 5(3):351.

Gunasekera, K., Krishnaswamy, S., Loke, S., and Za-

slavsky, A. (2012). Adaptation support for agent

based pervasive systems. In Snac, P., Ott, M., and

Seneviratne, A., editors, Mobile and Ubiquitous Sys-

tems: Computing, Networking, and Services, vol-

ume 73 of Lecture Notes of the Institute for Computer

Sciences, Social Informatics and Telecommunications

Engineering, pages 90–101. Springer.

Guo, B., He, H., Yu, Z., Zhang, D., and Zhou, X. (2012).

GroupMe : supporting group formation with mo-

bile sensing and social graph mining. In 9th Inter-

national Conference on Mobile and Ubiquitous Sys-

tems : Computing, Networking and Services, Beijing,

Chine.

Koch, M., Groh, G., and Hillebrand, C. (2002). Mobile

Communities Extending Online Communities into

the Real World. In Wirtschaftsinformatik, pages 7–17.

Koch, M. and Lacher, M. (2000). Integrating community

services-a common infrastructure proposal. In 4th Int.

Conf. on Knowledge-Based Intelligent Eng. Syst. and

Allied Techn. Proceedings, volume 1, pages 56–59.

IEEE.

Koch, M. and Worndl, W. (2001). Community Support and

Identity Management. In European Conf. on Comp.

Supported Cooperative Work, pages 319–338.

Meyffret, S., Laforest, F., and M

edini, L. (2010). Seman-

tic, Social & Context-Aware Registry for Services. In

Ubimob.

Mitchell-Wong, J., Kowalczyk, R., Roshelova, A., Joy, B.,

and Tsai, H. (2007). Opensocial: From social net-

works to social ecosystem. In Digital EcoSystems and

Technologies Conference, 2007. DEST ’07. Inaugural

IEEE-IES, pages 361–366.

Preece, J. (2001). Sociability and usability in online com-

munities: Determining and measuring success. Be-

haviour & Information Technology, 20(5):347–356.

Razmerita, L. and Firantas, R. (2009). Towards a New Gen-

eration of Social Networks: Merging Social Web with

Semantic Web. In Proceedings of I-KNOW 09 and I-

SEMANTICS 09, number September, pages 412–423.

Schubert, P. and Koch, M. (2003). Collaboration plat-

forms for virtual student communities. In 36th Annual

Hawaii Int. Conf. on Syst. Sc., page 10 pp. IEEE.

Schwagereit, F., Scherp, A., and Staab, S. (2010). Repre-

senting Distributed Groups with dgFOAF. In Euro-

pean Semantic Web Symposium, pages 181–195.

Seong, S.-W., Seo, J., Nasielski, M., Sengupta, D., Hangal,

S., Teh, S. K., Chu, R., Dodson, B., and Lam, M. S.

(2010). PrPl: A Decentralized Social Networking In-

frastructure. In Proc. 1st ACM Workshop on Mobile

Cloud Comp. & Services Social Networks and Beyond

- MCS ’10, pages 1–8, New York. ACM Press.

Sinner, A., Kleemann, T., and von Hessling, A. (2004). Se-

mantic user proﬁles and their applications in a mobile

environment. Art. Intell. in Mobile Systems.

Yeung, C.-m. A., Liccardi, I., Lu, K., Seneviratne, O., and

Berners-lee, T. (2009). Decentralization: The future

of online social networking. In W3C Workshop on the

Future of Social Networking Position Papers.

Zenker, B., Pyatkovka, V., and Drabek, C. (2012). To-

wards meeting assistance: An integrated appointment

assistance and group pedestrian navigation system. In

Puiatti, A. and Gu, T., editors, Mobile and Ubiqui-

tous Systems: Computing, Networking, and Services,

volume 104 of Lecture Notes of the Institute for Com-

puter Sciences, Social Informatics and Telecommuni-

cations Engineering, pages 338–349. Springer.

ICEIS2014-16thInternationalConferenceonEnterpriseInformationSystems

598