Using Social Networks based on Data Mining to Promote
New Technologies for Women from Emerging Countries
Syrine Ben Meskina
, Valérie Monfort
and Achraf Ben Miled
SOIE, 41, Rue de la Liberté, Cité Bouchoucha 2000, Le Bardo, Tunis, Tunisia
Université de Paris1, Panthéon Sorbonne, France
Abstract. Nowadays Social Network Websites become increasingly numerous
and millions of people are registered. They become famous and useful in sever-
al domains, in particular for sales and marketing activities all over the world.
Their success is gradually growing in industrialized countries. In our paper, we
firstly show how this technology is used to promote and enlarge sales networks.
Secondly, we investigate the ontologies dedicated to represent the Social Net-
work Semantics and their limits. We propose to use this approach to a research
project which aims at addressing women from both rural and urban areas of
emerging low income countries, who are individually or within group involved
with sustainable hand craft and/or fair trade business leading to a varying de-
gree of economic mobility. We finally conclude by proposing a solution that
aims to promote Sales Social Networks based on Data Mining techniques.
1 Introduction
Nowadays, it seems easy and usual to utilize new embedded applications on PDA and
Social networks. Past months showed impacts of these technologies during social
movements. Moreover, economical analysis showed companies aim to keep their
clients and to promote one to one relationship to increase their sells and their busi-
ness. They use social networks, Web sites, mails, and SMS, to promote and improve
their products.
However, what is obvious for industrialized countries is not so easy for emergent
countries. Costs of the devices such as laptops, PDA, software developments to de-
velop Web sites and simple mobile phones are heavy for emerging countries. Moreo-
ver, technologies are not disseminated and technical infrastructures as networks and
relay antennas do not cover the whole country. Sometimes it is not possible for elec-
tronic transaction to pay in Euro or Dollars, or by credit cards.
Based on a research project, we address women of emergent low incomes coun-
tries. These women are involved and are developing sustainable, hand craft and/or fair
trade business. The project is structured into several main steps such as:
- Elicitation: understanding the manner women use new technologies and social
- Defining and improving behavioral patterns to adopt new technologies
- Storing and analyzing behaviors
Ben Meskina S., Monfort V. and Ben Miled A..
Using Social Networks based on Data Mining to Promote New Technologies for Women from Emerging Countries.
DOI: 10.5220/0004106901060117
In Proceedings of the 10th International Workshop on Modelling, Simulation, Verification and Validation of Enterprise Information Systems and 1st
International Workshop on Web Intelligence (WEBI-2012), pages 106-117
ISBN: 978-989-8565-14-3
2012 SCITEPRESS (Science and Technology Publications, Lda.)
- Proposing new behavior, technical solutions, and training
These steps are the cornerstone of a process to improve iteratively. The aims are to
improve business of these women by offering them true technological means adapted
to their habits and to the technical and legal contexts of their countries. Moreover, it is
a way to think about how to propose to the greatest part of the humanity new cheaper
technologies. So, to develop a specific affordable business for emerging countries.
We focus on the ontologies used to characterize users in emerging countries. Moreo-
ver we propose a manner to analyze information to understand how new technologies
are used by women in emerging countries and if they have an added value for busi-
ness. We aim to measure the impacts and the benefits of the new technologies and the
usage of the propositions we can do to improve business.
Thus, we briefly describe the project, secondly, we introduce and define the basic
relative concepts; Social Networks and the ontologies dedicated to them as well as the
relative social and semantic problems. In the fourth section, we will briefly cite the
main limitations of the evoked ontolgies and then mention, as a solution, some tech-
niques of ontologies composition. Section 6 shows related works. Then, in the re-
maining sections, we will present our proposed work. Section 7 concludes the paper
and outlines the future work.
2 Project Presentation
Present status of Web 2.0 is far advanced than that of the basic DiNuccian ideology in
this concern, and the modern Web 2.0 is smartly admixed with the facilitation of
participatory information sharing, interoperability and user centric design where the
Social Networks are considered as the cornerstone of this system. So, the develop-
ment of Business Intelligence techniques with Web 2.0 and Internet offer an access to
less expensive and/or free resources and provide opportunities to economic and social
actors. The essential of these opportunities and resources aims to transform time into
information to enable a better decision making. Enterprises and e-commerce are the
most impacted areas by this evolution.
However, if this artifact seems coherent and promising for companies, its imple-
mentation remains limited because it is faced to several constraints such as the una-
vailability of fitted infrastructure as well on the technological aspect, a range of re-
strictions by laws and hindrances from a degree of regional / global politics. These
conditions are applied in developing countries; the situation remains so complex that
researchers use the expression: new technologies excluded or digital divide, by show-
ing the possible relations with social and economic or banking exclusion.
To face to this new kind of gap between societies, several research works have
been realized to find the adequation between new technologies and usage capacities
of the wider part of population targeted. Most of the peoples of these countries have
weak access to the benefits of new technologies and e-commerce.
During last past years, several works have been aiming to offer technological
products and services in adequacy with means, needs and social habits of populations
in emerging countries. These domains concern with well education, health, agricul-
ture, telecommunications, e-commerce, etc. Companies are also trying to enter these
markets by using sells strategies along with market structure and its needs (such as
rural phone).
One of the starting points of these works is the incorporation of the concept of
“Base of the Pyramid”. This concept is based on a simple idea, i.e. 4 milliard people
who live with less than few dollars per day represent a set of opportunities in terms of
demand of goods and services. These set of opportunities is linked to the “non inclu-
sive” strategies deployed by economic actors, private or public, particularly for
emerging countries.
Till the publication of first research works about the concept, and critics, several
projects and programs were launched. Some of these works show the importance of
the new technologies in the socio-economic development of this population. The new
technologies improve the socio economic situation of this population (Base of Pyra-
mid) by bringing them information and expertise for consumers and managers. How-
ever the results of these works remain limited and a global approach should allow
combining a local understanding of habits and needs of this population with a tech-
nical expertise of the product or service, and a sectoral expertise. The ‘applied’ part of
the global technological innovations in this particular concern is required to be flexi-
ble enough that it can be necessarily molded as per the physico-socio-politico-
economic characteristics of regional domain of application.
Firstly, this project is based on a global approach based on synergies between spe-
cialists in different domains in particular: sociology, economics, computer science,
and marketing. It aims to study the impacts of new technologies on populations of
both countries, Algeria and Tunisia. Focus to be given on two actors categories: very
small companies and members of formal or non formal social networks. Principal aim
is to address these two categories and to study the impact of technologies on business
activities proposing new habits to use new technologies. This approach is fully itera-
tive and incremental. Knowing current habits and usage pattern by auditing and elicit-
ing users allows noticing lacks and proposing new approaches. The different pro-
posals and its effects are audited and improved.
Secondly, the result of this study can imply the requirements of new kinds of af-
fordable technologies fitted to emerging countries. Moreover, these new technologies
have to be adapted to the profile of the users, for instance, the users can be analpha-
bets and the media have to support text to speech services. The definition of new
paradigms for a great part of humanity has the consequences to create start up to
commercialize new packaged products and allied services. On top of all, generative
programming technologies are an issue to automatically define and generate applica-
tions without requiring costly investments in developing e-commerce or M-commerce
Thirdly, this project requires to be based on a performing technical infrastructure
to analyze information usage and to help users to learn new technologies usage.
Moreover, this infrastructure is based on: Web services providing service based ap-
proach and interoperability between applications, social networks, mobility, etc. So,
proposal may be given to develop and stabilize this architecture to support marketing
and social research works, and to promote a true course of actions to help and train
women to new technologies and future fitted measures for improving business.
The proposed research will involve triangulation among a variety of different
sources of data. Hence, we intend to conduct both formal and informal on- and off-
site interviews with actors. We will also analyze archival materials such as internal
documents as well as articles and evaluate existing case studies and other relevant
Case studies may be needed to deduce a common knowledge and identify stages
and obstacles impeding progress from one stage to the next, as well as characteristics
of projects that are believed to successfully navigate through the processes. Depend-
ing on the measured intensity of factors, it will be possible to classify different tech-
nological orientations and to predict on which development stage a sector or a country
is situated.
Pragmatically, the project will be divided into four main steps as follows:
- Firstly, the behavior of Tunisian and Algerian professionals is observed and the
way to use of new technologies as well as social and commercial spheres will be stud-
ied. These behaviors are based on formal and informal knowledge to collect and ana-
lyze qualitative and quantitative data with rules. These techniques were already used
to collect qualitative and quantitative data. Knowledge on Algerian and Tunisian
professionals will be modeled, structured and validated according to business experts,
computer scientists, marketing experts and sociologists. Sociologists, marketing ex-
perts and computer scientists provide a behavioral and habits patterns repository pro-
posing and improving fitted solutions to facilitate the integration of new technologies
and social networks during commercial transactions initiated by professionals. These
knowledge are parameterized, aggregative, changeable, modifiable and flexible, ac-
cording to functional, non functional, and technical needs, but also according to ambi-
ent context, economical and social environment.
- Secondly, the study will link the current behavior of professionals faced to new
technologies and social networks to offer solutions or improvements.
- The third step proposes to dynamically updated users’ profiles and behavioral
patterns through the data from data mining analysis, data warehouse, marketing anal-
ysis and sociometry. This approach is iterative and incremental according to the man-
ner users appropriate new technologies and social networks usage. New manners are
proposed with feed backs. In the same time innovative approaches and/or technolo-
gies is to be framed.
- The fourth step proposes to use innovative solution defined in the third step, to
check the way it is appropriated and to train people to improve their business.
3 Basic Concept
3.1 Social Networks
A Social Network is a dynamic social structure composed of nodes and canals or arcs
such that the firsts represent individuals or organizations and the lasts represent their
According to [11], a Social Network is, generally, defined as a relationship set of a
specific type (for example: collaboration, support, control …) between a set of actors.
This definition calls the Barnes’s notion of mesh in relationships as well as the traffic
notion to transmit data between individuals over the mesh.
In Sociology, the Social Networking definition is different from Social Network
one. In fact, it references the employed means to inter-connect individuals. Social
Networking refers to multimedia social structure which essentially facilitates commu-
nication between individuals, groups and organizations. Since the apparition of the
Web 2.0 having a user-centered design, the Internet users become actors and the net-
working concept has been extended to sites building social networks on the Web.
Social Networks are based on connections between virtual identities which are
created on the Net and which may be veritable or not. So, the profiling concept was
included by the majority of the Social Network definitions. The main characteristics
of the classical ones are turning around that profiling notion and are summarized on;
create profile which represents a virtual identity necessary for connecting to some
members of the Social Network, look for friends with whom the relation-
ship is desired and finally draw links with other profiles which is the aim of the Social
Network The most commonly recognized social networking Web sites include Face-
book (specially in Tunisia), LinkedIn and MySpace… They are used for professional,
cultural, sport activities and specially to do marketing and for sales. For the last usage
domain, we call them Sales Social Networks.
In addition, to social Network sites, there are other ways to accomplish social me-
dia marketing through blogging, instant messaging, widgets … which become in-
creasingly popular internationally. Using Social Networking as a form of social media
marketing is the new fad.
Facebook, blog platforms and other Social Networks have proven to be essential
for sales people as well (For example: in Tunisia, all sales persons or people have at
least one page or profile on the social Web networking. It allows them sharing ideas
and digital photos with their friends and publishing real as well as virtual events to
add value…). Social Networks present the fastest, cheapest and easiest way to know
more new customers, new ideas improving and enlarging their business. Technically,
Social Networks are based on specific meta models called ontologies.
3.2 Social Networks and Ontologies
Web becomes “social”. Hence, it is necessary to use a common model to represent the
semantic of the majority of these Websites. Ontologies, mainly are defined by classes
and properties, and provide common semantics for resources on the Web semantic
according to [6]; [9]; [1];[4].
To represent social data, some common representation models are dedicated to the
social Web semantic adopting the idea that semantics can help social Websites and
vice versa. In fact, social media sites can interoperate by calling common semantics,
defined using agreed-upon semantic formats, to describe people, content objects and
the connections that link them all together. Current data on the Web own different
formats such as: relational Data Base, XML, CSV… Hence, the integration and the
link of these data become complex and infeasible. For instance, RDF presents a
standardized way to publish data on the Web thanks to its triplets represented by la-
beled graphs.
(Berners-Lee, 1998) has introduced the idea of Linked Data which are based on
URIs for naming, identifying resources, providing information about them and includ-
ing links to other ones in the aim to discovering more things. Within the framework of
LOD Project (Linked Open Data), the existing datasets are translated into RDF and
linked to each other (for example DBpedia and Geonames, Freebase, BBCprograms
As users may dispose of different accounts on the same or different social Web-
sites on the Web, it is interesting to establish links between them.
FOAF (Friend of a Friend) presents an ontology designed by Tim Berners-Lee for
describing people and the existing relationships between them using the RDF standard
vocabulary. It mainly investigates the identity notion, personal profiles and social
networks and can be integrated with other Web Service vocabularies. It aims to repre-
sent the concept of agent and the different related subclasses (persons, agent groups,
organizations as well as some properties related to these concepts such as names,
knowledge, membership…) based on the management of the distributed identities
over the Web. Consequently, it inter-relates different processes as interlinking identi-
ties and social networks to associate individuals to all their online accounts over the
social Web by exporting data from different sources. It is used for FriendFeed,
LiveJournal, MyBlogLog…
SIOC (Semantically-Interlinked Online Communities) ontology is defined to de-
scribe the online activities of communities on the Web such as: blogs, forums, wikis,
etc. In fact, it fully describes the content and the structure of the social Website by
representing the socio-cultural data; communities and their activities, produced doc-
uments as well as their organizations and inter-connection. It creates a kind of bridge
to link several Web tools to collect the different elements of communities from sepa-
rate places on the Web. It is composed of a core of eleven classes concerning the
social aspect (user accounts) as well as the structural aspects (contents and containers)
and properties describing their relations. It includes two main modules such as: Ser-
vices and Types concerning the social Web content. SIOC uses the RDF vocabulary
to describe the previous referenced elements. It also uses objects defined by other
ontologies such as: FOAF, SKOS, DublinCore, RSS etc. The goal of SIOC ontology
is to address interoperability issues on the social Web when it is combined to FOAF
(Bojares and al., 2008).
The tagging concept consists of associating some information to resource. It al-
lows sharing, exchanging and reusing of the data that was tagged over the social Web.
The TagOntology proposed by (NewMan, 2005) is based on the Gruber’s tag which is
composed of three parts such as: i) the tag representing the word or the sentence rec-
ognized by humans and machines, ii) the annotated resource representing the part to
be tagged and that must be identified using an URI or another similar labeling service,
and finally iii) the tagger representing the user which indicates the tag. Tags should be
named to differentiate them to create them such as resources which can be uniquely
identified. The contribution of the TagOntology consists on adding a tagging visibility
SCOT (Social Semantic Cloud of Tag) ontology is designed for representing the
structure and the semantic of a tag set. More exactly, this standard model describes
tag clouds (tags and co-occurrence), allows the portability to move from its own tag
cloud to another one. It, also, provides the possibility of developing the global share
(sharing tag clouds between services and between users). An important contribution
consists on supporting the social interoperability betwen heterogeneous sources.
MOAT (Meaning of a Tag) ontology is a standardized model designed to define
tags meanings. It provides a Web Service structure which supports the publishing of
semantically enriched content (free tagging). In fact, it allows users associating mean-
ings to tags via URIs on the social Web. It is a collaborative approach as it shares
meanings in a community and brings the tagged content to the linked data Web.
CommonTag is close to MOAT and allows associating tags to meaningful re-
sources. NiceTag Ontology appeared for which tagging meets speech act theory and
investigates on the link between a tag and a tagged item. It is, also, possible to go in
the opposite way by extracting ontologies from tags thanks to FolksOntology which is
a semi-assisted of relationships between tags.
It is also interesting to localize entities. In fact, it is relevant to geographically
identify the social network as well as its members and partners. It links, for example,
two instances to the same subject using the GeoVocabulary. It is also possible to use
Geonames to provide geographical data.
Moreover, geospatial ontologies localize data according to space and time.
3.3 Problems Met
There are many problems relative to Sales Social Networks. They can be categorized
on two groups.
The first type is social. In fact, to preserve the profitability of its business, it is im-
portant to point that the adoption mobile technologies are primordial (being always
connected and reachable using the smart phones and laptops, to be up to date…)
The emergence of smart phones and laptops with small screens seems to be more
attractive especially for emerging markets. But the prohibitive costs of high-end smart
phones, high-speed Internet, additional memory and new technologies, in general,
decelerate the development of mobile social networking. On the other hand, emerging
markets are behind on mature markets of payment, remote buying, using credit cards,
no delivery… this is due to the non-adaptation of the infrastructure.
Moreover, for instance, there is a very low Internet penetration in many parts of
the world especially for emerging countries seen its high cost. It is, also, useful to take
into account that social conditions and NTIC usage vary widely between emerging
countries capital and regions.
The second type concerns technical problems. According to the Semantic Web
Layer Cake designed by (Berners-Lee, 2001), each layer is composed of one or more
languages and standards dedicated to the ontology manipulation. We will present their
most important limitations. RDF language was appeared in 1999 for describing
knowledge using triplet of the form (subject, predicate, object). It provides only very
simple and elementary RDF descriptions and dispose of a reduced number of prede-
fined constructs.
RDFS language extends RDF and resolves partially the problems relative to ma-
nipulating non accurate data. Using RDF and RDFS present numerous advantages
such as associating namespaces to schemas, supporting multiple heritages as well as
obtaining some descriptions for which several classes are instantiated at a time.
3.4 Data Warehouse and Data Mining
Social Networks Websites dispose of huge databases leading to data storage prob-
lems. Even if such Web sites can be represented through ontologies. These data may
be vital for marketers to be analyzed to promote sales and marketing activities. Thus,
the corresponding studies start from the past data to the actual data going to preview
the future in order to describe the customer behavior, from which comes the necessity
of storing all information about customer.
It is important to recall that user data are usually confidential, thus, marketers aims
to consider even the smallest details about customers as they may be useful which
increases the amount of data.
Data warehouse and data mining techniques can be helpful to identify the perti-
nent data to be considered and to identify and extract the most important rules and
properties in the aim to refine analyses.
4 Towards an Ontology based Solution
4.1 Current Ontologies Selection and Limitations
In the Semantic Web, knowledge and data are generally represented through ontolo-
gies supported by the tagging notion (section 3.2) for the social Semantic Web.
In this issue, several standard ontologies were developed during the last decade.
The well known and most used ones were exposed in section 2.2. However, the ma-
jority of them, when they are used separately, seem to be limited and insufficient to
overcome some problems mainly summarized on distributed sources, heterogeneous
data, tags ambiguity, semantic links between tags... seen that no ontology works under
all conditions. Hence, it is interesting to consider a set of them at a time.
The combination of ontologies is formed using composition techniques which will
be exposed on the following section.
One such combination example can be compound of TagOntology, SCOT, SIOC
and MOAT. This model supports tags, simple and tripartite tagging, agent modeling,
definition of cloud of tags and associating meaning to tags at a time.
4.2 Ontologies Composition
The representation of some systems requires the intervention of more than one onto-
logies which can be heterogeneous and distributed. So, the cooperation of ontologies
is recommended as they may solve such problems (Euzenat and al., 2004).
The approaches dedicated in this purpose use the mapping and alignment tech-
niques. They, respectively, consist of identifying similar entities from different ontol-
ogies and establishing mapping links between distinct representations.
The manipulation of ontologies includes comparison and inter-operation.
The first axis, also called “alignment” processes by mapping equivalent entities
from different ontologies using the adequate metric (Johnson and al., 2006).
The second axis describes, in general, the inter-operation between ontologies by
both preserving and extending them. It is a sort of combining heterogeneous or dis-
tributed ones based on mapping schemas without being modified (copy, align-
ment…). Generally, it is useful to combine the ontologies that describe complemen-
tary domains. The fusion technique allows the creation of only one global and coher-
ent ontology. Thus, the resulting ontology must unify knowledge described by the yet
existing ones (importing the whole original in another one).
The integration technique provides an ontology created by considering only some
proportions from the other original ones (it is different from the complete fusion).
The coordination technique consists of using, at same time, the knowledge de-
scribed by only some parts of the original ontologies in an independent manner.
5 Towards a Data Mining based Solution
The problems of promoting Sales Social Networks depend on the collected data about
costumers. The data may be analyzed for useful correlations between interests and
also between interests and demographic categories.
The reasoning activities over social networks are usually based on the proposed
ontologies to represent these Web sites (for example each Website is represented by
ontology). Then, to model the users of Social Network Websites, we can collect all
information concerning them from other social sites by interlinking several social
Websites in order to better describe the different customer classifications.
This process requires the use of ontology composition techniques. We can use the
combination model evoked in section 4.1 to represent user profiles. In a second step,
it is interesting to look for forming a global ontology on which we can effectuate
some reasoning and inferences in order to preview new marketing activities based on
NTIC use.
We recall that (Klein, 2001) defined mapping by finding equivalent elements ac-
cording to a similarity measure even if they come from different data representation
(ontology, data base schema). We mention also that Social Networks have huge data
bases containing information about their users. It is possible to proceed by mapping
the real stored data bases and the proposed ontology to resolve our problematic.
In the other hand, we plan to implement some data mining techniques over the
considered data bases which consist of identifying the pertinent information by using
data reduction in a first step. And, then in a second step, effectuate a mapping process
between the ontology and the kept data.
6 Related Works
In the literature review, we focused on several research works as our project handle
different domains; marketing, Semantic Web, ontology manipulation, Web Mining...
Bouzeghoub and Kostodinov [3] studied a personalization of profiles in order to
facilitate the need expression for a particular user or community and to obtain relevant
information from an information system. They defined a set of criteria and prefer-
ences specific to each user or community of users and proposed a generic model of
profile allowing classification tasks over profile’s contents. Slimani [16] defined some
approaches for evaluating semantic associations on domain ontologies in order to
discover heavily linked data collected from separated sources. Jarrar [10] presented
and used ontology for customer complaint management which has been developed in
the CCFORM project. As customers can register all their complaint against any party
about any problem, it can be useful to explore this data in the aim to promote Sales
Social Networks. In order to facilitate the re-use of ontologies, Maedech [15] pro-
posed a ontology learning framework enhancing typical ontologies engineering envi-
ronments for importing, extracting, pruning and evaluating ontologies. Berendt et al.
[2] discussed the combination of Semantic Web and Data Mining, the use of Web
structures to do Web Mining and vice versa as well as the resulting profitability of
integrating closer data. Stumme et al. [17] proposed an automated schemes for learn-
ing the relevant information to automatically enhance several resources as it becomes
impossible to do it manually seen the enormous size of data on the Web. (Gruber,
2007) discussed the combination of the best ideas from the Semantic Web and the
Social Web which depends of individual user contributions. He proposed collective
knowledge systems which unlock the “collective intelligence” of the Social Web with
knowledge representation and reasoning techniques of the Semantic Web. (Li, 2003)
presented an approach for automating the discovering of ontologies from data sets in
the goal of building complete concept models for Web user information needs. The
proposed approach overcomes the problem of obtaining information from user Web
profiles. Li [14] evaluated the previous model to be able of capturing the evolving
patterns to refine the discovered ontologies and assesses their relevance. Li [13] pre-
sented the relationship between Web mining and linked data such user profiles and
proposed an abstract Web mining model for extracting approximate concepts hidden
in user profiles on the Semantic Web based on ontologies. The developed approach
includes an efficient filtering algorithm to filter out most non-relevant inputs.
We found that ontologies handling the context and the interest creiteria can be
useful to resolve our problematic as we look for adapting a user profile to the women
previously described.
There exist some ontologies which investigate the context of Web pages repre-
sented by documents. The Foafnaut ontology, proposed as an extension of the FOAF
ontology, describes one’s contacts details based on the associated context. That family
of ontologies enables us to measure the relationship between persons by calculating
scores using keywords and related contexts.
Contexts depend of a periods of time, location, documents and Web pages as well
people’s preferences and interests. As contexts change interests may change. The
Weighted Interests Vocabulary, specified by Bob Ferris in 2009, attributes weights to
people interests in dynamic manner. In fact, it includes both of E-FOAF: interest
vocabulary which extends the FOAF vocabularies, in 2010, to describe user interests
and preferences (interested by, disinterested by as well as preference degrees…) and
the Interest Mining Ontology.
The Geo ontology is also a FOAF extension proposed by Dan Brickley in 2009 to
specify the coordinates; latitude, longitude and altitude of objects in the Web.
As research works treat separately each one of the domains on which we are fo-
cusing, it is interesting to work under the combination of all of them.
7 Conclusions
We propose the first step of a research project based on new technologies such as:
social networks, mobility, e-commerce. We offer a possibility to proposed fitted solu-
tion to communicate about business such as aircraft. As social networks are based on
Web 2. And ontologies we proposed an overview of current existing ontologies based
on social networks. We have to use and/or to modify existing ontologies or proposing
new one. We also have to link these ontologies with data warehouse and data mining
by proposing fitted algorithms to manage data and their instances. We aim to improve
this approach and to propose concrete results after interviews, structuring and collect-
ing data.
1. Bachimont, B., 2000. L'intelligence artificielle comme écriture dynamique: de la raison
graphique à la raison computationnelle. In J. Petitot & P. Fabbri (Eds). p 290-319.
2. Berendt, B., Hotho, A. and Stumme, G., 2002. Towards Semantic Web Mining. Lecture
Notes in Computer Science, THE SEMANTIC WEB — ISWC, Volume 2342/2002, p. 264-
3. Bouzeghoub, M., Kostadinov, D., 2005. Personnalisation de l'information: aperçu de l'état
de l'art et définition d'un modèle flexible de profils. Actes de la seconde édition de la Con-
férence en Recherche d'Infomations et Applications (CORIA). Grenoble, France.
4. Gandon, F., 2002. Distributed Artificial Intelligence and Knowledge Management: Ontolo-
gies and Mutli-Agent Systems for a Corporate Semantic Web. Scientific philosopher doc-
torate thesis in informatics, INRIA and University of Nice - Sophia Antipolis - Doctoral
School of Sciences and Technologies of Information and Communication.
5. Gómez-Pérez, A., Fernández-López, M., and Corcho-Garcia O., 2003. Ontological Engi-
neering with examples from the areas of Knowledge Management, e-Commerce and the
Semantic Web. Springer-Verlag New York, Inc., Secaucus, NJ, USA.
6. Gruber, T. R., 1993. A Translation Approach to Portable Ontology Specifications.
Knowledge Acquisition, 5(2), p. 199-220.
7. Gruber, T., 2008. Semantic Grid -The Convergence of Technologies. In ESELVIER Web
Semantics: Science, Services and Agents on the World Wide Web, Volume 6, Issue 1, p. 4–
8. Guarino, N., 1998. Some Ontological Principles for Designing Upper Level Lexical Re-
sources. In Proceedings of the First International Conference on Lexical Resources and
9. Guriano, N., and Giaretta, P., 1995. Ontologies and knowledge bases: Towards a termino-
logical clarification. In Mars, N., editor, Towards Very Large Knowledge Bases:
Knowledge Building and Knowledge Sharing, p. 25-32, Amsterdam, NL. IOS Press.
10. Jarrar, M., 2008. Towards Effectiveness and Transparency in e-Business Transactions, An
Ontology for Customer Complaint Management. A book chapter in "Semantic Web Meth-
odologies for E-Business Applications". Idea Group Inc.
11. Lazega E. Analyse de reseaux et sociologie des organisations. Revue française de sociolo-
gie 35, 2 (Apr. - Jun 1994), 293.
12. Li, Y., 2003. Ontology-based Web mining model: representations of user profiles. In Web
Intelligence Proceedings, IEEE/WIC, p.96 – 103.
13. Li, Y., 2004. Webmining model and its applications for information gathering. In
ESElVIER Knowledge-based Systems, Volume 17, Issues 5–6, p. 207–217.
14. Li, Y., 2006. Mining ontology for automatically acquiring Web user information needs. In:
Knowledge and Data Engineering, IEEE Transactions, Volume: 18, Issue: 4. P. 554-568.
15. Maedche, A., K., 2001. Towards Effectiveness and Transparency in e-Business Transac-
tions, An Ontology for Customer Complaint Management. In Intelligent Systems, IEEE.
16. Slimani, T., Boutheina, B., Y., and Mellouli, K, 2009. Evaluation d'associations séman-
tiques dans une ontologie de domaine. IC 2009: Actes des 20èmes Journées Francophones
d'Ingénierie des Connaissances. Hammamet, Tunisia.
17. Stumme, G., Hotho, A. and Berendt, B., 2006. Semantic Web Mining: State of the art and
future directions. In ESELVIER, Web Semantics: Science, Services and Agents on the
World Wide Web, Volume 4, Issue 2, p. 124–143.
18. Uschold, M., and Gruninger, M., 1996. Ontologies: Principles, Methods and Applications.
AIAI-TR-191, Knowledge Engineering Review.
19. Ushold, M., Jasper, R., and Clark, P., 1999. Three approaches for knowledge sharing: A
comparative study. In KAW - 99.