Ontology Evolution in the Corporate Semantic Web

Fatma Chamekh, Guilaine Talens and Danielle Boulanger

Magellan Center- Information System Team, University of Jean Moulin, 6 cours Albert Thomas 69008, Lyon, France

Keywords: Semantic Web, Ontology Evolution, Multi-Agent System, Knowledge Management, Ontology Version.

Abstract: We present in this paper an approach for ontology evolution in the corporate semantic web. We particularly

focus on the ontology and resources evolution which are two important components of the corporate

semantic web. Ontology evolves with added and updated resources. The concerned ontological entities by

the resource modifications are linked to the documents. To manage dynamically changes we use the MAS

paradigm. The evolution process is distributed in the different agents of the system. Each of them has a

particular role.

1 INTRODUCTION

Nowadays, to better capitalize and share their

knowledge, the companies set up knowledge

management systems. They enrich their in-house

knowledge by capturing knowledge extracted from

external information sources. The new generation of

the web technologies as the semantic Web-

authorizes to describe sources by increasing

meaning of documents via metadata.

To represent knowledge of a specific domain, the

ontology concept is a possible approach (Grüber,

1993; Studer et al., 1998).

The integration of semantic Web technologies in

knowledge management systems provides new

perspectives. The coupling between the company’s

communication tools (intranet, intraweb) as well as

semantic web technologies leads to create

company’s memory as a Corporate semantic Web

(Gandon, 2002; Luong et al., 2006). It consists of

resources (documents), ontologies and, semantic

metadata/annotations.

The environment of organisations is widely

heterogeneous, distributed and evolutive. The main

challenge consists of the capture of this changing

environment to satisfy the dynamic experts’needs

and requirements. Indeed, the application of changes

to the one of components (document, and ontology)

involves the evolution of the Corporate Semantic

Web. In the context of intranet, the employees share

the documents and reuse them to write others.

The knowledge extracted from documents allows

the ontology evolution. This evolution is performed

by a MAS (Multi-Agent System). Each agent

manages a step of evolution process: Change

identification, change analysis, change propagation

and version management.

This paper is structured as follows: Section 2

lists related works about ontology evolution and

multi-agent systems. Section 3 describes our

approach and system architecture. In section 4, we

present a case study. Finally, section 5 gives a

conclusion and some perspectives.

2 RELATED WORKS

In this section, we describe related works concerning

ontology evolution and multi-agent systems.

2.1 Ontology Evolution

The management of the ontology evolution is

defined by (Stojanovic et al., 2003) as the “timely

adaptation of ontology to the arisen changes and the

consistent propagation of these changes to

dependent artefacts”.

(Klein and Noy, 2003) define ontology

versioning as “the ability to manage ontology

changes and their effects by creating and

maintaining different variants of the ontologies”.

(Stojanovic, 2004) proposed an approach for the

management of evolution and the maintaining of

consistency for KAON ontologies.

(Klein, 2004) suggested a process of ontology’s

version management.

182

Chamekh F., Talens G. and Boulanger D..

Ontology Evolution in the Corporate Semantic Web.

DOI: 10.5220/0004438401820189

In Proceedings of the 15th International Conference on Enterprise Information Systems (ICEIS-2013), pages 182-189

ISBN: 978-989-8565-60-0

 2013 SCITEPRESS (Science and Technology Publications, Lda.)

Based upon the previous works, several approaches

have been developed:

Onto-Evoal (Ontology Evolution-Evaluation)

(Djedidi et al., 2010) is a system to manage ontology

evolution and evaluation. It is made up of three

levels: Evolution level process, pattern level and

historical level.

CoSWEM (Corporate Semantic Web Evolution

Management) (Luong et al., 2006) designed to

manage Corporate Semantic Web Evolution

especially on the two components ontology and

semantic annotation. This approach focuses on the

propagation of ontology changes to the semantic

annotations.

Rogozan (Rogozan and Paquette, 2005)

developed an approach for ontology evolution

relatively to educational semantic web. The author

presented a process of evolution in two phases: a

phase of ontology evolution and a phase of version

management.

Consistology (Jaziri et al., 2010) is a tool to

manage ontology evolution and versioning. The

authors proposed an ontology evolution approach

based on change kits which control the

inconsistencies generated by each type of change.

The consistency of ontology is anticipated by

suggesting all the possible resolutions and their

effects on the ontology according to a set of rules

defined by the system. A new version of ontology

will be created. Each version of ontology is stored in

a log.

Evolva (Zablith et al., 2008) is an ontology

management framework, aiming to reduce user

intervention. It explores background knowledge

sources (Wikipedia, WordNet, online ontologies….).

The system is composed of five components:

information discovery, data validation, ontological

changes, evolution validation and evolution

management.

(Freddo and Tacla, 2009) proposed an

integration approach of social web with semantic

web. This approach is composed of two phases:

 Ontology learning from folksonomies: it

consists of populating the tag ontology (SCOT),

identifying the relation between each pair of tags

(tag1, tag2) (tag1, tag3), and interacts with user in

order to create concepts and instantiate concepts and

relations.

 Ontology evolution from folksonomies: in

this phase the authors propose to update the base

ontology after each change of folksonomies. After

identifying the changes, the system studies the

relation between the extracted tag and the entities in

the “base ontology” using an ontology alignment

method.

Most of the existing systems for the ontology

evolution uses patterns, plugins, and software

modules. Thus, the task of automation system

remains difficult.

The multi-agent system can be a tool to solve the

problem related to the evolution of ontology, among

others resolution of inconsistencies.

2.2 Multi-Agent System (MAS)

A software agent is considered as an entity with

goals, is able to actions endowed with domain

knowledge and situated in an environment (Stone

and Veloso, 2000). MAS is suitable for the domains

that involve interactions between different people or

organizations with different (possibly conflicting)

goals and proprietary information (Jennings, 1995).

An agent evolves in an environment and is able to

perceive what surrounds it, to communicate with the

other agents. It has an autonomous behaviour in the

aim of satisfying its objectives. Moreover, each

agent has knowledge of its environment.

Dynamo (Sellami et al., 2009) is a MAS for

dynamic ontology construction from domain specific

text documents. Architecture is composed of two

modules:

 DYNAMO Corpus analyzer is a twofold

module of textual corpus processing: An extractor

of terms fills up the database with candidate terms,

and an extractor of lexical relations provides triplets

constituted of two candidate terms and the syntactic

relation between them.

 DYNAMO MAS is composed of term agent

and concept agent. It uses as an entry the triplet

provided by the first module. The built ontology is

proposed to the user. He can accept, refuse or

modify it. DYNAMO generates ontology in OWL

file relatively RTO model (Resources Termino-

Ontological).

(Deen et al., 2006) proposed an approach for

dynamic ontology integration and mapping between

global and local ontologies in a multi-agent

environment. Each agent captures knowledge about

its own schema (local knowledge) and also

knowledge about the schemas of its acquaintances

(partial global knowledge). The authors employed a

specific thesaurus to resolve semantic conflicts. The

agent can add new knowledge (concept, property or

instance) but it is not really an ontology evolution.

In ACSIS (Boulanger et al., 2000), a MAS allows to

dynamically solve semantic conflicts for the

cooperation of information systems (Talens and

Boulanger, 2009). The global knowledge base –

OntologyEvolutionintheCorporateSemanticWeb

183

ontology– is distributed inside informational and

wrapper agents involved in negotiation protocols to

solve conflicts and insuring the completeness of the

answer to a global multi-base query.

These meta-data encapsulated inside agents build

several ontologies sharing a common description.

Different agents interact to answer to a user query.

In order to give the better results the agent

ontologies dynamically evolve to deduce

information between the database schemas and the

query.

2.3 Discussion

From the analysis of the related works (see table1),

two types of works exist:

 The user can add or modify directly the

ontology. The evolution of ontology is based on the

process described in (Stojanovic, 2004). Moreover,

(Klein, 2004; Rogozan and Paquette, 2005) and

(Luong et al., 2006) propose approaches based on

the correction of inconsistencies after their

production. In (Jaziri et al., 2010) an anticipatory

approach is proposed to prevent and anticipate

inconsistencies. (Rogozan and Paquette, 2005) and

(Luong et al., 2006) describe Semantic web

evolution approach that allows changing two

components: ontology/semantic annotation or

ontology/resources.

 (Zablith et al., 2008) and (Freddo and Tacla,

2009) focuse on identifying new information added

to the ontology. The second type performs either

through the analysis of the Folksonomy created by

open and uncontrolled systems (social tagging

system), or by using of the background knowledge

sources.

Other approaches with MAS only propose design

or integration of ontologies with evolution.

However, Dynamo allows ontology modification but

there is no process of ontology evolution.

In ACSIS (Talens and Boulanger, 2010), the

ontologies inside agents dynamically evolve

relatively the negotiation protocols. The ontology

evolution is performed by versioning. The MAS role

is to build the semantic interoperability between

databases. In our knowledge, none system performs

the ontology evolution through a MAS. We propose

an approach of ontology evolution based on a MAS.

Each agent realises a part of the evolution process

dynamically evolves by adding of new documents.

Indeed concepts, properties and instances extracted

from documents imply ontology evolution. The

ontological entities may index the corresponding

documents.

3 PROPOSED APPROACH

3.1 Context

The organisations share knowledge among their

members, create and collect new knowledge.

Sharing the activity in a group also implies to share

the knowledge involved in the activity. Inside an

enterprise, different services exist. Each of them

produces documents; they can be consulted by the

users. The goal of our proposition is to classify the

documents thanks to ontologies. Furthermore,

ontology servers help an organization to keep track

of all concepts and notions used in its documents

and to clarify the importance of transactions and

business processes. The ontology is defined in

(Grüber, 1993) as “an explicit specification of a

conceptualization “. The adding and the updating of

documents perform the updating of ontologies.

Therefore, the documents evolve; the ontologies

must follow the evolution. But, the modification of

the ontology can involve a new classification of the

documents.

Table 1: Comparative analysis.

Stojanovic

Klein

and al

Onto-

Evoal

Coswem Rogozan Consistology Evolva ACSIS

Freddo

and al

Ontology

evolution

process

Yes No Yes Yes Yes Yes Yes No No

System

Semi-

automatic

Semi-

automatic

Semi-

automatic

Automatic Automatic

Semi-

automatic

utomatic

utomatic Automatic

Ressources

Internal Internal Internal Internal Internal Internal External Internal Floksonomies

Ontology

Consistency

Yes No Yes Yes Yes Yes Yes No No

User

intervention

Yes Yes Yes Yes Yes Yes No No No

ICEIS2013-15thInternationalConferenceonEnterpriseInformationSystems

184

At the beginning, experts of each domain must build

the domain ontology to describe the vocabulary

related to a specific domain. For example,

marketing, human resources, financial ontologies are

designed. Our framework allows the user to evolve

the domain ontology when he drops off a new

document on the intranet. After, when a user

searches a document, key words are asked to find it.

The system proposes all the documents containing

these terms in all the ontologies. To realize these

changes we must follow a multi-step process

presented in the next section.

3.2 Evolution Process

Our approach follows an evolution process based on

Stojanovic and Klein’s process. It includes four

phases:

 Identification of changes enables to represent

the transformations at the request of the user (or

expert) and to clarify them formally. We distinguish

the changes that can make by the expert when he:

-Adds a new document includes add concept,

modify concept, add property, modify property, add

instance, modify instance.

-Updates a new document includes add concept,

modify concept, add property, modify property, add

instance, modify instance.

-Removes a document includes delete concept,

delete property, delete instance.

An expert can directly add, update or delete a

concept, property and instance if he considers

necessary to handle ontology. He can consult the

changes history stored in the log of changes.

 The analysis of the changes consists in

studying the effect of the changes on the consistence

of ontology and the consistency between ontology

and documents. To resolve this problem, we defined

consistence rules. Examples of consistence rules:

-Ontology should not have isolated concepts

-Ontology should not contain two occurrences for a

same concept

-Each ontological entity should be connected to

concept, property or instance.

-Each ontological entity should be linked to

document

The rules must manage the effects of each

modification in the ontology. We present the

authorized changes in table 2.

To treat inconsistencies when they occur, we

defined additional operations. To identify the

adequate corrective operations related to each type

of change, it is necessary to determine the types of

changes and inconsistencies. If different possibilities

exist, i.e., different additional operations can be

applied with different effects, the users have to

choose the appropriate additional changes to

implement. The various operations and their impact

to the consistence of ontology are displayed to the

users in order to assist them.

 The propagation of the changes consists in

checking the consistence of the dependent artefacts

(ontology, document) after each change. When

ontology is modified, the different mappings must

be checked between the ontologies.

 The management of the versions (Vn and

Vn+1) consists of validation of the changes enables

to create the Vn+1 version and keeping record of the

ontology library and annotating the whole of the

changes. The annotations are recorded in log of

changes. The classified documents can be differently

referenced. It concerns changes in ontology, in

document but also the classification of documents.

Finally, the document is referenced in different

concepts, attributes and instances. The URI is stored.

The linked data are used to better manage and make

easier information retrieval. The key concept of

Table 2: Effects of Changes

Changes

Ontological

entity position

Consequences

Add Concept

Leaf concept None

Non leaf

concept

Consequences on the sub

concepts and instances

Delete

Concept

Leaf concept Consequences instances

Non leaf

concept

Consequences on the sub

concepts, upper concepts, and

instances

Rename

Concept

Leaf concept

Consequence on the concepts

and upper concepts

Creation semantic link

Non leaf

concept

Consequence on the concepts

and upper concepts

Creation semantic link

Add

property

Leaf

concept

Consequence on the upper

concepts

Non leaf

concept

Consequence on the upper

concepts and sub concepts

Delete

property

Leaf concept None

Non leaf

concept

Consequence on the sub

concepts and instances

Rename

property

Leaf concept

Creation semantic link

Consequence on the concepts

and upper concepts

Non leaf

concept

Creation semantic link

Consequence on the upper

concepts and sub concepts

OntologyEvolutionintheCorporateSemanticWeb

185

Linked Data (D’Aquin, 2010; Health and Bizier,

2011; Health et al., 2012) is based on the idea that

the mechanisms used to share and interlink

documents on the Web can be applied to share and

interlink data and metadata about these documents,

as well as the concepts they relate to.

3.3 Architecture of the System

The architecture of our system (see figure 1) is

founded on a multi-agent system. These agents

interact together to meet the change needs for the

user while respecting the structural and semantic

constraints. The user starts the evolution process by

carrying out a request for change.

The agents interact to run different processing steps.

In order to provide the consistence of ontology, our

system guides the user by suggesting many choices

of change operations. The system architecture (see

Figure1) is made up of four agents. Firstly we

assigned to each agent a process step.

 User agent detects the change to be realized

by analyzing the user request. If the user

adds/modifies a new document the user agent throws

a term extractor module. This one backs the entities,

the agent informs the ontology management agent of

a type of changes and concerned entities. If the user

deletes a document, the agent gives the document

URI to the ontology management agent.

 Ontology management agent analyses a change.

It searches similar ontological entity on the domain

ontology. We use existing similarity measures like

Jaccard (Jaccard, 1901), Levenstein (Levenshtein,

1966) and n-grammes (Damerau et al., 1971). It

informs the inconsistency management agent about

the changes that can be held.

 Inconsistency management agent checks the

effects of changes on the ontology consistence. After

each request for change, it receives the consistence

type from the ontology management agent. So, it

must propose additional changes in order to guide

the user to work out the operation of change. It is a

BDI agent (Beliefs, Desires, and Intentions). It is

composed of three layers: The beliefs (knowledge)

are the whole of the consistence rules as formal

concepts. The desires are the whole of the changes

rules. The intentions are the additional changes

corresponding to the user’s changes.

 Version management agent: it generates a new

version of ontology after validation of the changes.

The Vn+1 version are stored in an ontology library.

However, the Vn management is stored in the log of

ontology versions to memorize change traces.

The architecture of system contains: Log

changes for storing all annotations of changes; Log

of versions contains all versions of ontology.

Ontology Library is contains all ontologies of

domain. A learning module is implemented to

extract news terms from the document. We use

existing tools text2onto (Maedche and Staab, 2000),

Gate (Cunningham, 2001), RapidMiner (Hunyadi,

2010)…).

4 CASE STUDY

We explain the ontology evolution process through

an example (see Figure1).

Figure 1: Architecture of the system.

ICEIS2013-15thInternationalConferenceonEnterpriseInformationSystems

186

4.1 Add New Document

Our system offers a guide to an expert in the process

of the corporate semantic web evolution. We present

it through an example. We suppose the expert want

to add a new document, he sends a request through

the research interface (1). The user agent treats the

request and activates the learning module (2). This

one extracts terms from the document. Different

tools are currently in study (text2onto, GATE…).

The user agent passes on the list of generated terms

in the XML file (see Figure 2) (3). The extracted

XML file contains two concepts “mobile commerce”

and “electronic transfer”. The user agent informs the

ontology management agent about the changes

which are Add

document and transmits the list of

extracted terms (4). This one searches in the domain

ontology (see Figure 3) similar concepts (5). In this

step, different existing similar measures (Jaccard,

Levenstein …) are used to find concepts, properties

and instances. The similar concepts are “commerce”,

“electronic commerce” and “E-commerce” to

“mobile commerce” and “electronic commerce” and

“Electronic data exchange” to “electronic transfer”.

The ontology management agent suggests that:

 “mobile commerce” is a subclass of

“commerce”

 “mobile commerce” is an upper class of

“commerce”

 “mobile commerce” is a sub class of

“electronic commerce”

 “mobile commerce” is an upper class of

“electronic commerce”

 “mobile commerce” is a sub class of “E-

commerce”

 “mobile commerce” is an upper class of “E-

commerce”

 “electronic transfer” is sub class of “electronic

commerce”

 “electronic transfer” is an upper class of

“electronic commerce”

 “electronic transfer” is sub class of “electronic

data exchange”

 “electronic transfer” is sub class of “electronic

data exchange”

Figure 2: Extracted XML file.

The ontology management agent transmits the

details of changes to the inconsistency management

agent (6). This one studies the impact of changes on

the domain ontology (7). For this, it verifies the

consistence rules:

 “mobile commerce” and “electronic transfer”

are concepts; they should be connected to a

concept, property or instance.

 “mobile commerce” and “electronic transfer”

does not exist in the domain ontology. They are

not redundant concepts.

 “electronic data exchange” is a leaf concept.

If we add “electronic transfer” there is none

consequence

 “commerce”, “electronic commerce” and “e-

commerce” are non leaf concepts. The

ontology management agent studies the impact

of changes to the properties of “commerce”

“electronic commerce”, “e-commerce” and

“electronic data exchange”. The inconsistency

management agent proposes the additional

changes (8). When the expert chooses one of

these: We suppose that he validates the

“mobile commerce” subclass of “commerce”

and “electronic transfer” is a sub class of

“electronic commerce”. The user agent

transmits this information to the version

management agent. This later saves the Vn

version in the log of version and Vn+1 version

in the ontology library. It saves the

modifications in the log of changes. The

document URI is stored in each ontological

entity.

4.2 Delete Document

We suppose that the expert wants to delete a

document, he sends request through the research

interface. The user agent treats the request. It

informs the ontology management agent about the

changes and transmits the URI of the document. The

ontology management agent researches the

instances, properties and concepts related to this

document. This one transmits the details of changes.

The inconsistency management agent studies the

impact of the changes and proposes additional

changes. When the expert chooses one of these, the

user agent communicates the information to the

version management agent. This one saves the Vn

version in the log of version and Vn+1 version in the

ontology library. It saves the modifications in the log

of changes.

<?xml version='1.0'encoding='ISO-8859-1?>

<concept1>mobile commerce</concept1>

<concept2>electronic transfer</concept2>

</entit

OntologyEvolutionintheCorporateSemanticWeb

187

Figure 3: Commerce ontology.

5 CONCLUSIONS

Our research work tries to bring up two crucial

points about corporate semantic web evolution and

ontology evolution. The first point is how to change

each component and the second point is how to

apply this change to the ontology. We proposed an

approach to manage the corporate semantic web

evolution founded on multi-agent system. The

agents interact between them to manage evolution

process. The ontology evolves dynamically when a

new document is added or modified by the expert.

The evolution history is stored to keep track of

changes. Currently we experiment and compare

different tools to extract terms or concepts. Then our

perspective is to link all the documents of a domain

by the linked data technologies.

REFERENCES

Boulanger, D., Disson, E., Dubois, G., 2000. Object

Oriented Metadata for Secured Cooperation of Legacy

Information Systems. In International Workshop on

Model Engineering IWME'00, 14th European

Conference on Object Oriented Programming,Sophia-

Antipolis and Cannes, France, p 75-82.

Cunningham, H., Maynard, D., and Bontcheva, K., 2001.

Text processing with Gate.

Damerau, F., Hirano, M., Oka, K., and Tagawa, Y., 1971

Markov Models & Linguistic Theory, The Hague

Mouton.

D’Aquin, M., Putting linked data to use in a large higher-

education organisation, 2010. In Workshop on

Interacting with Linked Data (ILD) at the Extended

Semantic Web Conference (ESWC).

Djedidi, R., Aufaure, M. A., Onto-Evo an Ontology

Evolution Approach Guided by Pattern Modelling and

Quality Evaluation, 2010. In the 6th International

Symposium on Foundations of Information and

Knowledge Systems, LNCS: Vol. 5956, G. Berlin,

Germany: Springer. Sofia, Bulgaria, p. 268-305.

Deen, S. M., Ponnamperuma, K., Dynamic ontology

integration in a multi-agent environment, 2006. In

20th International Conference on Advanced

Information Networking and Application AINA’06.

Freddo, A. R., Tacla, C. A., Integrating social web with

semantic web-Ontology learning and ontology

evolution from folksonomies, 2009. In International

Conference on Knowledge Engineering and Ontology

Development, Portugal, p. 247-253.

Gandon, F., 2002, Distributed Artificial Intelligence and

Knowledge Management: Ontologies and multi-agent

systems for a corporate semantic web, PhD Thesis

INRIA.

Grüber, T. R., A translation approach to portable

ontologies, 1993. In Knowledge acquisition, Vol. 5,

n°2, p. 199-220.

Hass, P., Stojanovic, L., Consistent evolution of OWL

ontologies, 2005. In Proceedings of the 2nd European

Semantic WebConference (ESWC), LNCS 3532, p.

182-197.

Hunyadi, D., Rapid miner e-commerce, 2010. In

Proceedings of the 12th WSEAS international

conference on Automatic control, modelling &

simulation, p. 316-321.

Health, T., Bizier, C., Linked Data: Evolving the Web into

a Global Data Space, 2011. Morgan & Claypool.

Health, T., Singer, R., Sahbir, N., Clark, C., and Leaves-

ley, J., Assembling and applying an education graph

based on learning resources in universities, 2012. In

Linked Learning (LILE) Workshop at the World Wide

(WWW) conference.

Jaziri, W., Sassi, N., Gargouri, F., Approach and tool to

evolve ontology and maintain its coherence, 2010.

Metadata Semantic and Ontologies, vol. 5, p. 151-166.

Jennings, N., Controlling cooperative problem solving in

industrial multi-agent systems using joint intentions,

1995. Artificial Intelligence.

Jaccard, P., Étude comparative de la distribution florale

dans une portion des Alpes et des Jura, 1901. Bulletin

de la Société Vaudoise des Sciences Naturelles, vol.

37, p. 547-579.

Klein, M., Noy, M., A component based framework for

ontology evolution, 2003. In Workshop on Ontologies

and distributed System, IJCAI’03, Mexico.

Klein, M., 2004. Change Management for Distributed

Ontologies. PhD thesis,University of Amsterdam .

Levenshtein, V., Binary codes capable of correcting

deletion, insertions, and reversals, 1966. Soviet

Physics Doklady, vol. 10, n° 8, p. 707-710.

Luong, P-H., Dieng-Kuntz, R. et Boucher, A. Managing

semantic annotations evolution in the CoSWEM

system.2006. In Proceedings of the Third National

Symposium on Research, Development and

Application of Information and Communication

is a

Date Sale

Transaction date

amount

Payement system

Electronic ticket

Whole

E-commerce

Internet marketing

Electronic commerce

Commerce

Electronic data exchange

ICEIS2013-15thInternationalConferenceonEnterpriseInformationSystems

188

Technology (ICT.rda’06), Hanoi (Vietnam), May

2006, p 215 –223.

Maedche, S., Staab, S., 2000. The text-to-onto ontology

learning environment”, In Software Demonstration at

ICCS-2000-Eight International Conference on

Conceptual Structures.

Noy, N. F., Musen, M., PROMPTDIFF: A fixed-point

algorithm for comparing ontology versions, 2002. In

Proceedings of the 18th National Conference on

Artificial Intelligence and Fourteenth Conference on

Innovative Applications of Artificial Intelligence,

Canada.

Rogozan, D., Paquette, G., 2005. Managing ontology

changes in the semantic web, In the IEEE/WIC/ACM

International Conference on Web Intelligence,

Compiegne, France.

Stojanovic, L., Maedche, M., Stojanovic, N., Studer, R.,

Ontology evolution as reconfiguration design problem

solving, 2003. In Proceedings of the 2nd International

Conference on Knowledge CAPture K-CAP’03, p.

162-171.

Stojanovic, L., 2004. Methods and Tools for Ontology

Evolution, PhD thesis, University Karlsruhe.

Stojanovic, L., Maedche, A., Motik, B., Stojanovic , N.,

User-driven ontology evolution management,2002.

In Proceedings of the European Conference on

Knowledge Engineering and Management, p. 285-300.

Stone, P., Veloso, M., Multiagent Systems: Survey from a

Machine Learning Perspective, 2000, In Autonomous

Robotics, vol. 8, p345-383.

Studer, R., Benjamins, R., Fensel, D., Knowledge

engineering: Principles and methods, 1998. Data &

Knowledge Engineering, Volume 25, Issues 1-2, p.

161-19.

Sellami, Z., Aussenac-Gilles, N., Gleizes, M.P., Vers un

outil de construction d’ontologies à partir de textes à

l’aided’un système multi-agent, 2009. Actes de JFO.

Talens, G., Boulanger, D., Séguran, M., Domain

Ontologies Evolutions to Solve Semantic Conflicts,

2007. In Proccedings of Ontologies-Based Databases

and Information Systems, vol. 4623/2007, LNCS

Springer, Berlin /Heidelberg, ISBN 978-3-540-75474-

9_4, p. 51-67.

Talens, G., Boulanger, D., Domain evolution ontology by

versioning, 2009. In Proceedings of the 5th

International Conference on Agents and Artificial

Intelligence – ICAART’09, Porto, Portugal, p. 185-

190.

Talens, G., Boulanger, D., Evolutive ontology by

versioning, 2010. In Proceedings of the 4th

International Conference Research Challenges in

Information RCIS’10, Nice, France, p. 157-168.

Zablith, F., Sabou, M., d’Aquin M., and Motta, E., Using

background knowledge for Ontology evolution, 2008.

In International Workshop on Ontology Dynamics

(IWOD) at the International Semantic Web

Conference (ISWC), Karlsruhe, German.

Zablith, F., D’Aquin, M., Brown, S., and Green-Hughes,

L., Consuming linked data within a large educational

organization, 2011. In Second International Workshop

on Consuming Linked Data (COLD) at International

Semantic Web Conference (ISWC).

OntologyEvolutionintheCorporateSemanticWeb

189