OUTLINE CAPITALIZATION OF COLLECTIVE KNOWLEDGE: Knowledge - - PDF document

CAPITALIZATION OF COLLECTIVE KNOWLEDGE: FROM KNOWLEDGE ENGINEERING, MULTI- AGENTS TO CSCW AND SOCIO SEMANTIC WEB

Nada Matta 1 and Davy Monticolo 2 1Tech-CICO-University of Technology of Troyes nada.matta@utt.fr 2 SET-University of Technology of Belfort Montbeliart davy.monticolo@utbm.fr 1

OUTLINE

 Knowledge management and Knowledge

Engineering

 Definitions  Process  Knowledge Capitalization approaches  Cooperative Knowledge  Defintinions  Traceability and capitalization approaches  Socio-semantic Web  Definition  Semantic Web  Examples

2

KNOWLEDGE

 Knowledge is data, information used in a given

context

 Knowledge is [Polyani]  Tacit  Explicit

3

KNOWLEDGE MANAGEMENT AND KNOWLEDGE ENGINENRING

4

OUTLINE

 Knowledge management process  Knowledge Enginnering process  Corporate Memory  Knowledge Engineering approaches  CommonKADS  MASK

5

KNOWLEDGE MANAGEMENT CYCLE

[NONAKA & TAKEUSHI]

Tacit Explicit Explicit Tacit

Externalization Combination Internalization Socialization

Knowledge Engineering ICT Semantic Web Community of Practices

6

KNOWLEDGE ENGINEERING Knowledge engineering is an approach allowing problem solving extracting and modelling [Aussenac,

Bradshow], [Newell]

Documents Experts Conceptual Model KB TextMining Interviews Observation …

• What
• Why
• How

7

KNOWLEDGE MANAGEMENT CYCLE

[NONAKA & TAKEUSHI]

Tacit Explicit Explicit Tacit

Externalization Combination Internalization Socialization

Knowledge Engineering

Corporate Memory

8

« A corporate memory is a persistent and explicit representation of knowledge and information of an organization » [Van Heijst, 96], [Dieng et al, 03] Several memory types: Profession memory, project memory, management memory CORPORATE MEMORY

9

PROFESSION MEMORY

Profession memory is the externalization of the knowledge produced in and for a given domain

 Structure :  Definition of the problem (or the process)  Problem solving methods  Description of manipulated concepts

10

KNOWLEDGE ENGINEERING APPROACHES

 CommonKADS  Generic models  MASK  Process  Models

11

COMMONKADS [BREUKER ET AL,94]

Generic Models Library Analysis Selection MC Adaptation Formalisation, Implementation

12

GENERIC MODELS LIBRARY

 Generic task models  Problem solving methods library [Benjamins].

13

TASK TYPES

Task Types Analysis Modification Synthesis Classify Prediction Repare Change Design Planning Diagnostic Evaluate Monitoring

14

Prediction

Observations

Specify

Behaviour description

Transform

New behaviour System Environment

15

Diagnostic

Observations

Symptom detection

Abnormal Observations Normal Observations Generate hypotheses Hypotheses

Discriminate hypotheses

Additional Observations Diagnostic Heuristics

16

Evaluate

Case Description

Abstract

Elements

Match

Norms

Specify

Measures System Model Decision classes

17

Monitoring

Differences Environment Select Observations Compare Parameters Select Criteria System model Transform Variable Values Instantiate Parameter Values Classify History Difference Classes

18

Design

Intention

Specify

Requirements

Evaluate

Artefact

Build

Composites Rules/Laws Violated requirements

19

Planning

Environment

Identify

Goal list

Assemble

Tasks, Time, ...

Specify

Plan Objective Plan models

Evaluate

No satisfied Objective

20

CML : CONCEPTUAL MODELLING LANGUAGE

Level Entity Problem solving

Task Task, Task structure Inference Inference, Inference structure

Domain

Domain Concept, relation, expression, attribute

21

THE MASK METHOD [ERMINE,

MATTA, CASTILLO] Co-building Consensus Sensibilisation Training

Knowledge Book

22

PROCESS MODEL

Sketch Technician Conceive the shape programming Tools stitch Knowledge Shape Realization

• f the

product Electronic machine 2 “fontures” Knowledge on machines Prototype Verify all the sensitive points Model Knowledge

• styling and model

making

• technique on the

chosen stitches Validated program Technician Technician 23

PB SOLVING MODEL

Verification product Verify the product in flattering position (smooth table with or without light) Verify the dimension in height, width, 1/2 scale Look at the weak places fallen from machine Verify the weak places: Edges Motives Armholes (crossing sleeve / body) Decreases Verify the good goes well with model or by wearing test 24

CONCEPT MODEL

Minimal edition Edition Edition for the structures Edge rib Main Crossing sleeve Necklace Reduction edition Reverse Abolition of the auxiliary edition 25

REFERENCES



Knowledge Acquistion as modelling, M. Ford, J.M. Bradshaw (Eds), 1993



CommonKADS Library for expertise modelling, Reusable problem solving components, J. Breuker and W. Van de Velde (Eds), IOS press, Amsterdam, 1994



• A. Newell, The Knowledge level, Artificial Intelligence Journal, 19(2), 1982



[Van Heijst et al, 97] Van Heijst G., Schreiber A. Wielinga B., Using Explicit Ontologies in KBS

• Development. International Journal of Human Computer Studies, Vol. 46, 1997.



Breuker J., Van de VeldeW., Common-KADS Library for expertise modelling Reusable problem solving components, Frontiers in Artificial Intelligence and Applications, J. Breuker and W. Van de Velde (EDS), Amsterdam: IOS.Press 1994.



Benjamins R., Problem solving methods of diagnosis, Rapport de Thèse de l'université d'Amsterdam, ISBN 90-9005877-X, Amsterdam, 1993.



Acquisition et Ingénierie des connaissances, Tendances Actuelles, Coordination: N. Aussenac-Gilles,

• P. Laublet, C. Reynaud, Cepadues Editions, 1996



[Nonaka et al, 95] Nonaka I., Takeuchi H.: The knowledge-Creating Company: How Japanese Companies Create the Dynamics of Innovation. Oxford University Press, 1995



Ermine J.L. – Les systèmes de connaissances, Eds. Hermès 1996 (2nd Edition 2000)



Méthodes et Outils pour la gestion des connaissances, R. Dieng, O. Corby, A. Giboin, J. Golebiowska, N. Matta, M. Ribière, Dunod, 2000



Dieng-Kuntz R., Matta N., Knowledge Management and Organizational Memories, Kluwer Academic Publishers, 2002.



Matta N. Zaher L., Applications of knowledge Engineering for Design, Methods and Tools for effective Knowledge Life-Cycle Management, Bernard A., Tichkiewitch S. (Eds), Springer, 2008.



CASTILLO, O., et MATTA, N. A knowledge acquisition system for the French Textile and Apparel. Knowledge-Based Intelligent Information and Engineering Systems: 9th International Conference, KES 2005, 14-16 September 2005, Melbourne. 6 p.



CASTILLO, O., et MATTA, N. An operational learning system definition. 19th International Joint Conference on Artificial Intelligence IJCAI 2005 – Workshop on knowledge Management and Organizational Memories, 1-5 August 2005, Edimbourg. 12 p.



Polyani M. (1958), Personal knowledge, Chicago : University of Chicago Press, 1958. 26

COOPERATIVE KNOWLEDGE

27

OUTLINE

 Cooperative Knowledge definition  Project memory  Traceability approaches  Traceability and capitalization approache

28

COOPERATIVE KNOWLEDGE

 Knowledge produced in cooperative activity

(projects, cooperative decision making, etc.):

 Organizational dimension: actors, tasks, resources,

constraints,

 Cooperative dimension: negotiation, argumentation,

etc.

29

PROJECT MEMORY

Problème Proposition Proposition Argument Argument Argument Decision

Environment and organization Process

Techniques Tools Methods Directives Procedures Goals Constraints Requirements Competences References

Carter Arbre Ca/ Ar Ar/Ca Ar /Ext Ca/ Ext Guider en rotation l’arbre/carter Carter Arbre Cart/ Arbre Arbre /Cart compresseur Vue Technologue

q3 déc4 obj

4 cc

rhu4 , r

ma4, rin4

q5 ad5 déc

5

rhu5 rma5 rin5

• bj5

co5 cr

5

vd5

Guider en rotation l’arbre/carter

q4 ac4

Relationships

Product representation Design Rationale

Roles

Explicitation of the experience learned during project realization [Matta]

30

Approaches to handle cooperative knowledge

• CSCW-Design rationale:

 IBIS [Coklin, 98], QOC [McLean, 91] , DRAMA [Brice]

(Design rationale tree)

 DIPA (Problem solving model) [Lewkowicz, 99]  DRCS (Graphs : Concepts, Relations) [Klein, 93]

• Project Management:

 DRCS (Graphs : Concepts, Relations) [Klein, 93]

31

TRACEABILY METHODS

IBIS, QOC, DRAMA

Problem Proposition Proposition Argument Argument Argument

Representation guided by the decision-making DIPA, DRCS

Problem Task Proposition Interpretation Artifact Decision Constraint Argument

Representation of the dynamics

• f problems solving

32

TRACEABILY AND CAPITALIZATION METHOD

Decision Making Organization Product DYPKM [Bekhti, Djaiz, Matta]

33

QOC: QUESTION, OPTION, CRITERIA TREE

[MCLEAN, 91]

Connexion type ? Numerical Hybrid Analogical Flux Performance Cost Support Objection

34

QOC: OPTION LINKS

Connexion type ? Numerical Hybrid Analogical Flux Performance Cost Optical connexion Performance Installation 35

DIPA [LEWKOWICZ, 00]

 Cooperative problem solving model  Structure based on principal concepts

36

PROBLEME Data description abstraction INTERPRETATION PROPOSITION AGREEMENT implementation selection

• pposition /

precision

• pposition /

precision Abstract CONSTRAINT evaluation evaluation Concrete CONSTRAINT 37

DIPA DIPA synthesis DIPA analysis Problem Goal Malfunction Fact Requirement Symptom Interpretation Functionality Cause Abstract constraint Constraint Constraint Proposition Means Corrective action Concrete constraint Constraint Constraint Agreement Choice Choice Tool: MemoNet: a structured discussion forum

38

Semi-structured notes Structuring report Validation

DYPKM: DYNAMIC PROCESS OF KM

[BEKHTI, DJAIZ, MATTA]

39

MEETING NOTES

Question: Autonomy  Part1: How competences needed to do risk

evaluation ?

 Part3: Can company ask for a supplier to do

risk evaluation

 Part5: The Risk evaluation principles must

consider the particularity of each company

 Part1: It is important to have some guaranty

rules.

40

STRUCTURED FORM

Criteria: Required resources List competences that the company must to have in order to do evaluation (Participant 3) Criteria: flexibility Defining contextual procedures (Participant 1) Criteria: Control The evaluation can be done by sub-contracting, but the company has to learn how to evaluate risk (Participant 3) Criteria: flexibility It is not the same than autonomy but, we can add adaptability principle (Participant 1) Criteria: Control Company must to nave the control

• f the evaluation (Participant 2)

Problem Element: Allow the autonomy of the company Arguments Suggestions Decision: - Modify the autonomy principle

• Create another principle about adaptability

Criteria: Required resources List competences that the company must to have in order to do evaluation (Participant 3) Criteria: flexibility Defining contextual procedures (Participant 1) Criteria: Control The evaluation can be done by sub-contracting, but the company has to learn how to evaluate risk (Participant 3) Criteria: flexibility It is not the same than autonomy but, we can add adaptability principle (Participant 1) Criteria: Control Company must to nave the control

• f the evaluation (Participant 2)

Criteria: Required resources List competences that the company must to have in order to do evaluation (Participant 3) Criteria: flexibility Defining contextual procedures (Participant 1) Criteria: Control The evaluation can be done by sub-contracting, but the company has to learn how to evaluate risk (Participant 3) Criteria: flexibility It is not the same than autonomy but, we can add adaptability principle (Participant 1) Criteria: Control Company must to nave the control

• f the evaluation (Participant 2)

Problem Element: Allow the autonomy of the company Arguments Suggestions Decision: - Modify the autonomy principle

• Create another principle about adaptability

41

CRITERIA TREE : DESIGN PROBLEMS

[MATTA] Proposition Strategy Understanding Acceptability Conditions Resources Requirements Preferences Execution Consequences Finality Coordination Terminology Interpretation Formulation Redundancy Clarity Quality Universality Structure Characteristics Conformity Product Pertinence Validity Complete Constraints Interaction Extension Responsibility Cooperation Flexibility Motivation Engagement Method Assistance Feasibility Behavior Safety Rules Elasticity Applicability 42 Participation Validity Conformity Redondance Terminology Resources

• f ERP

Understanding The last session is not coherent with participating There the same element in finality principle Reformulate the last session

• Do not explicit the principle,

but incite to participating

• Dot not focus on the

animator of the evaluation,

• Mention work medecine
• Replace « mainly » .
• Delete the brakets in order to Put

the element in thé same level. It is necesary to define all employee types Delete the sentence

• « Shared results » is not clear
• « The external participants » is not clear
• Replace « shared

« by « validate by all the participants »

• Replace « the » by

« different »

ARGUMENTATION VIEW

43

Focus ERP Pb Terminology P1 Consulting Exhaustivness P4 Validity P3 Psychology Biochemistry Flexibility P2 Ergonomics P3 Psychology

COMPETENCES VIEW

44

COMPETENCES/DECISION MAKING

45 46

RÉFÉRENCES



Brice A. – Design Rationale Management (DRAMA), http://www.quantisci.co.uk/DRAMA/



Karsenty L. – An Empirical Evaluation of Design Rationale Documents, Proceedings of CHI, R. Bilger, S. Guest, and M. J. Tauber (Eds), 1996



[Klein, 93] Klein M., Capturing Design Rationale in Concurrent Engineering Teams, IEEE, Computer Support for Concurrent Engineering, January 1993.



Lewkowicz M., Zacklad M., MEMO-net, un collecticiel utilisant la méthode de résolution de problème DIPA pour la capitalisation et la gestion des connaissances dans les projets de conception, IC’99, Palaiseau, 14-16 juin 1999, p.119-128.



Bekhti S., Matta N., A Formal Approach to Model and Reuse the Project Memory, Journal of Universal Computer Science (J.UCS). Springer, 2003, http://www.jucs.org/



DJAIZ, C., et MATTA, N. Traceability and capitalization of project memory. Journal of Universal Computer Science (J.UCS), 2007.



MacLean A., Young R.M., Bellotti V.M.E., Moran T.P., Questions, Options, and Criteria: Elements of Design Space Analysis, Human-Computer Interaction, Vol.6, 1991.



Conklin J.E. et Begeman M.L. – gIBIS: A Hypertext Tool for exploratory Policy Discussion, ACM Transactions on Office Informations Systems, 6,303-331, 1998.



Matta N., Corby O., Prasad B.. A Generic Library of Knowledge Components to Manage Conflicts in CE tasks, Concurrent Engineering Research and Applications (CERA) Journal, Volume 6, Number 4, December 1998.



Huyet Anne-Lise, Matta N., Traçabilité, La conception industrielle de produits, Management des hommes, des projets et des informations, Volume 1, (B. Yannou, M. Bigand, Th. Gidel, Ch. Merlo, J.-P. Vaudelin (Eds). Hermes, 2008.



Matta N., Ribière M., Corby O., Lewkowicz M., Zacklad M., Project Memory in Design, Industrial Knowledge Management - A Micro Level Approach, Rajkumar Roy (Eds), Springer-Verlag, 2000



Matta N., Corby O., Conflict Management in Concurrent Engineering: Modelling Guides, Computational Conflicts, Conflict Modeling for Distributed Intelligent Systems, H.J. Muller and R. Dieng (Eds), Springer, 2000 47

SOCIO-SEMANTICWEB [ZACKLAD, CAHIER, BENEL]

48

OUTLINE

 Socio-Semantic Web definition  Semantic Web  Example

49

SOCIO-SEMANTIC WEB

 The Social Semantic Web can be seen as a Web of

collective knowledge systems, which are able to provide useful information based on human contributions and which get better as more people participate, [Zacklad], [Cahier, Benel, Zaher].

 Social Semantic Web combines technologies,

strategies and methodologies from the Semantic Web, Social Software and the Web 2.0.

50

SEMANTIC WEB [BERNERS-LEE]

 The Semantic Web is an evolving extension of

the World Wide Web in which the semantics of information and services on the web is defined, making it possible for the web to understand and satisfy the requests of people and machines to use the web content.

51 52

SOCIO-SEMANTICWEB PRINCIPLES

[ZACKLAD, CAHIER]

• To co-build maps by users themselves
• Let the community imagine its own

architecture of cooperation and its socio- semantic activity Use CSCW and social approaches to handle Socio-SemanticWeb

53

Document

description, revision, signature

Interpretation

heuristic modelling

Intersubjectivity

viewpoints comparison

[Benel, Cahier]

54

AGORAE METHOD [CAHIER] 1) STARTING FROM DOCUMENTARY RESOURCES

URLs / DOCUMENTAY RESOURCES Adhikari, T. B., R. C. Basnyat, et al. 1999. "Virulence of Xanthomonas oryzae pv.

• ryzae on rice lines containing single

resistance genes and gene combinations." Plant disease 83(1): 46-50. 55

2) AGREEMENT IN THE GROUP ON DOCUMENTS CONSIDERED

• NAMING OF EACH CONSIDERED DOCUMENT
• STANDARD ATTRIBUTES (AUTHOR…)

URLs / RESOURCES FRAGMENTS DOCUMENTAIRES « Documents » Text 3 Title, authors… Text 4 Title, authors… Text 1 Title, autthors… Text 2 Title, authors…

56

URLs / RESSOURCES FRAGMENTS DOCUMENTAIRES

« Documents »

Text 3

Titre, auteurs…

Texte 4 Titre, auteurs… Texte 2

Titre, auteurs…

• Mainstream
• Peripheral
• Emergeant
• Out of scope

(centrality)

Text 2

Titre, auteurs…

Text 1

Titre, auteurs…

3) BUT POSSIBLE CONFLICTS WITHIN THE GROUP ON THE « CENTRALITY » OF THESE CONSIDERED DOCUMENTS Topics = « heuristic » attributes

57 SYNTHESIS MAP

Topic-1 By pdvX Conception / développement URLs / DOCUMENTARY RESOURCES

« Documents »

Topic-2

Point of view actor 3 Point of view actor 2

By pdvY By pdvZ By pdvX

Topic-1

N1 N2(centralité) N3 N>3

Point of view actor 1

Topic-4

Text 3

Titre, auteurs…

Text 1

Titre, auteurs…

Text 2

Titre, auturs…

Text 4 Titre, auteurs…

Topics = « concepts », « heuristic » attributes » (coming for example from a text-mining tool)

4) « DESIGN MAPS » FROM EACH ACTOR, BUT POSSIBLE CONFLICTS WITHIN THE GROUP ABOUT THE « CONCEPTS » USED TO GROUP THE LOW-LEVEL TOPICS

58 59

1-Technology

telecoms Terminals

3-Services & products 2-Application and usages

Health IPV6 commerce Work

4-delive-

rables

Car

URL DOCUMENTARY RESOURCES & FRAGMENTS

ILLUST RATIO N DKN Cristal Archipel Puma2

Entities

PR09 #8

Relations:

r4 r2 r3 r7

Keys (HyperTopic basic constructs): Point of View : Topic: Entity: Resource:

r1

Hypertopic meta-semiotic

ADSL

HYPERTOPIC

is the knowledge representation (a « metasemiotic ») that the actors need to use, in order to co-construct , to discuss, etc., the collective map

Item: Items

PDAs

60

Hypertopic model (UML representation)

item Value Attribute Topic Point of view

• Doc. Res.

1

* * * * * * * * * *

• Hypertopic is focused excusively on a very few basic constructs

(inspired by the Topis Map)

• It Gives to many end-users the ability to edit the map (items, topics)

without any particular training

• It makes easier to deploy the co-building within large communities

www.hypertopic.org

61

HYPERTOPI C

Point of View: concurrent caracterisations of the item

2

Item: identifier of the situation /

• f the artefact object of the inquiry

1

62

Topics: heuristic thematization

• f the item

Point of View: concurrent caracterisations of the item

2 3

Item: identifier of the situation /

• f the artefact object of the inquiry

1

HYPERTOPI C

63

Correlation A Standard attributes: referential specification

• f the item

Topics: heuristic thematization

• f the item

Point of View: concurrent caracterisations of the item

2 3 4

Item: identifier of the situation /

• f the artefact object of the inquiry

1

HYPERTOPI C

64

5

Correlation B Correlation A Correlation C Standard attributes: referential specification

• f the item

Resources: Documentation of the item Topics: heuristic thematization

• f the item

Point of View: concurrent caracterisations of the item

2 3 4

Item: identifier of the situation /

• f the artefact object of the inquiry

1

HYPERTOPI C

65

HOW TO COLLECTIVELY CONSTRUCT AND MAINTAIN AN HYPERTOPIC MAP

It is useful to distinguish:

 - a « bootstrapping » phase  to define the item, to define the first set of « points of view»  based (eventually) on folksonomies or on the confrontation of

actors’ personal « design maps »

 leading (eventually) to a « synthesis map » usable by the

group

 - a phase of maintenance / evolution of the map

66

HOW TO ARTICULATE THE MODELS REQUIRED FOR CO-BUILDING HYPERTOPIC MAPS?

HYPERTOPIC

Model for knowledge representation within the Socio Semantic Web

+ instrumented activity model ? Basic roles to edit the map, e.g.:

• Tagger (propose tags, indexes items)
• Contributor (edits/indexes items
• Semantic editor (edits/ associates topics

+ informal

roles: discussion, annotations…

for each Hypertopic map node

“co-building” participative method ? + knowledge representa tion

Map

• f the domain

for the Actor1

+

Map

• f the domain

for the Actor 2 Hypertopic Map for the community in the domain

“co-building” method ?

+…=

The « socio-semantic activity » methodological challenge

67

AGORÆ EXAMPLE: « E-CATALOG » OF PROJECTS AND INITIATIVES IN THE FIELD OF SUSTAINABLE DEVELOPPEMENT

68

69 70 71 72

Application Application

uses all Hypertopic concepts: points of view,

73

Application Application

uses all Hypertopic concepts: points of view, topics

74

Application Application

uses all Hypertopic concepts: points of view, topics , items

75

Application Application

uses all Hypertopic concepts: points of view, topics , items , attributes

76

Application Application

uses all Hypertopic concepts: points of view, topics , items , attributes , resources

77

Application Application

uses all Hypertopic concepts: points of view, topics , items , attributes , resources , and possibility to build the map within the inquiry activity

78

Application Application

uses all Hypertopic concepts: points of view, topics , items , attributes , resources , and possibility to build the map within the inquiry activity

79

THE DOMAIN OF OUR KBM-EXAMPLE : IN THE TRAINING FIELD

Multi-sellers catalogue, in the field of

training for (French-speaking) computer people

500 training themes in a 4-stages hierarchy 1200 training organisms of all sizes 18 000 registered training modules plurality of naming practices, importance of

the original names

30% new themes / year

80

Java Programmation

 IIN05 - 5 Jours à Paris : 28 janvier au 1er février 11 au 15 février 11 au 15 mars 22 au 26 avril 3 au 7 juin 8 au 12 juillet 26 au 30 août 30 septembre au 4 octobre 18 au 22 novembre 9 au 13 décembre à Lyon : 28 janvier au 1er février 11 au 15 février 11 au 15 mars 22 au 26 avril 3 au 7 juin 8 au 12 juillet 26 au 30 août 30 septembre au 4 octobre 18 au 22 novembre 9 au 13 décembre Profils stagiaires : Développeurs, concepteurs et chefs de projet confrontés à l'arrivée massive de l'Objet et des réseaux Internet et Intranet dans les systèmes informatiques Prérequis : Connaissance de la programmation Objet, la connaissance d'Internet (HTML) est un plus Avant ce stage, vous pouvez suivre : ITM04, IXU20 Après ce stage, vous pouvez suivre : IIN15, IIN45 Animateur(s) : Consultant Informaticien Spécialiste Prix HT : 1635 euros (10 724,90 FF)

Imprimer la fiche

S'inscrire

Conditions générales de vente

Envoyer à un collègue


 Objectif : apprendre la programmation en Java et particulièrement la mise en oeuvre des concepts Objet du langage pour développer des serveurs Web et plus généralement des applications sur Internet et Intranet PROGRAMME

Introduction au langage Java, Java et le Web Les origines, le kit de DéveloppementJava ( JDK et les navigateurs La situation de Java parmi les langages orientés objets Eléments et concepts Objets de Java : héritage, encapsulation et polymorphisme Construction de sa première application Java Les éléments de base du langage Java, vue complète des Objets et des classes Définition d'une classe et des membres (données et méthodes) Les constructeurs : les membres statiques, la création et la manipulation d'Objets Encapsulation des membres : publics, privés et protégés Java et l'héritage Classe dérivée, classe de base et référence aux Objets, surcharge des méthodes Interface : définition et utilisation, l'héritage multiple et l'utilisation avancée des interfaces Package Définition et utilisation des packages, organisation physique d'une application avec les packages Classes utilitaires de Java.lang et Java.util Gestion des exceptions Les instructions throw, try, catch et les exceptions Programmation Java Introduction à Abstract Window Toolkit (AWT) et SWING . création d'interfaces utilisateurs . gestion des événements Applet . définition, exemples, les Applets et les Browsers . applet Java et langage HTML Entrée/Sortie . lecture et écriture de fichiers en mode texte, binaire et objet Réseau . utilisation des classes URLConnection, Socket et ServerSocket Connexion aux bases de données Java Data Base Connectivity (JDBC) . architecture et utilisation pour la sélection et la mise à jour des données Méthode pédagogique : Alternance de théorie avec de nombreux travaux pratiques

« TRAINING CATALOGU E »:

MANY « HEURISTICS ATTRIBUTES » THAT AN EXPERT CAN HIGHLIGHT

Profile Prequisite Animator 81

…COMPLETING « STANDARD ATTRIBUTES »

82

KBM FORMATION: MULTIPLE POINTS OF VIEW

Example of associations :

• « prerequisite » between training thèmes

Concept: Learning M odule Point of view by thema Point of view calendar Point of view by cost Point of view quality/reputation Point of view by profession Point of view by learned notion Point of view by project’s goal Point of view by tool

83

KBM-FORMATION IHM-1

84

IHM- 2 DE LA KBM

COOP 2002

85

86

«Enterprise Java Beans » « Java programming » « Java for the Web» Web Projet « Certification SUN » GUI Project

Man Machine Interface Interface Homme Machine Interface Personne Machine

«Certification FAFIEC»

Développer en Java Pour le Web Enterprise Java Beans Formation Programme de Certification Guide des stages Agréé FAFIEC www.orgaform.com www.orgaform.com www.sun.fr/formation/certifications/index.html www.orgaform.com/demo

Project Learning module Label

86

87

Développer en Java Pour le Web Enterprise Java Beans

• ccurrence

Page Web Site Web Formation, programme de certification Guide des stages Agréé FAFIEC www.orgaform.com www.orgaform.com www.sun.fr/formation/certifications/index.html www.orgaform.com/demo

• ccurrence

«Enterprise Java Beans » « Programmer en Java » « développer en java pour le Web» Projet WEB « Certification SUN » Projet IHM «Certification FAFIEC»

• ccurrence

Document audio

• ccurrence

Catalogue papier

Site Web Document Audio Catalogue Site Web

TYPES DES OCCURRENCES

87

Thèmes des stages Séminaires décideurs Conception et développement Technologies

• bjet

C++ Java Développer des EJB Fiche d’attributs : 109.html Fiche descriptive : 109.pdf Labels Labels de stage Labels d’organismes Label SUN … Labellisation De stage Subject Indicator

http:// java.sun.com/products/ejb/docs.html

Prérequis Stage Label

EXAMPLE OF ASSOCIATIONS WITH TM

88

«Enterprise Java Beans »

« Java program ming» « java for the Web» WEB Project « Certification SUN »

Project GUI « Certification FAFIEC »

Project Training module Label

Certified cursus with prerequisite

More advanced module Prerequisit e (module) Certificat of the module

Goal of the training

EXAMPLE OF ASSOCIATION: « CERTIFICATION» OF A TRAINING MODULE

89

«Enterprise Java Beans »

« Programmer en Java »

« développer en java pour le Web»

« Certificati

• n SUN »

«Certification FAFIEC»

Développer en Java Pour le Web

Enterprise Java Beans www.sun.fr/formation/c ertifications/index.html www.sun.fr/formation/c ertifications/index.htm

Stage Label Organisme

« Orgaform »

Certification d’organisme

Qualification de personnel scope2 scope1

Cursus certifié avec pré-requis Certificat du stage

agrément

Certification SUN (scope1) Nom : Certification développeur Occurrences : Rôles : certificat du stage

www.sun.fr/formation/certifications/index.html

• ccurrence
• ccurrence

Nom : Certification d’organisme Occurrences : Rôles : agrément

www.sun.fr/partenaires/certifications/index.html

Certification SUN (scope2)

Certifie

SCOPE MECHANISM

90

COG-DOC: LAB DOCUMENTS

91 92

93

REFERENCES

• Berners-Lee, Tim; James Hendler and Ora Lassila (May 17, 2001). "The Semantic Web". Scientific American
• Magazine. http://www.sciam.com/article.cfm?id=the-semantic-web&print=true. Retrieved on 2008-03-26.
• Manuel Zacklad, Aurélien Bénel, L'Hédi Zaher, Christophe Lejeune, Jean-Pierre Cahier, Chao Zhou. Hypertopic:

une métasémiotique et un protocole pour le Web socio-sémantique, Actes des 18eme journées francophones d'ingénierie des connaissances (IC2007). Francky Trichet (Eds.). pp. 217-228. Cépaduès. 2007. ISBN 978-2-85428- 790-9.

• Zacklad, M. (2003) Communities of Action: a Cognitive and Social Approach to the Design of CSCW Systems, in

Proceedings of GROUP'2003, pp. 190-197, Sanibel Island, Florida, USA.

• Zaher, L. H., J.-P. Cahier, W. A. Turner, et M. Zacklad (2006a). A conflictual co-building method with Agoræ. In

Workshop on Knowledge Sharing in Organizations, (COOP 2006).

• Cahier J.-P., Zaher L'H., Zacklad M., « Cooperative building of multi-points of view topic maps

using Hypertopic and socio-technical approaches”, 3rd International Conference on Topic Maps Research and Applications (TMRA’07 Leipzig,Germany), “Scaling Topic Maps” 11-12 oct. 2007

• Cahier J.-P., Zacklad M., "Towards a Knowledge-Based Marketplace model (KBM) for cooperation between

agents", Actes conference COOP'2002, St Raphael, 4-7june 2002, IOS Press

• Cahier J.-P., Zacklad M., (2004) “Socio-Semantic Web applications: towards a methodology based on the the

Communities of Action”, COOP'04 Workshop on Knowledge Interaction ans Knowledge Management

• Cahier J.-P. , Zaher L'H., Leboeuf, J.P., Pétard X., Guittard, C. Experimentation of a socially constructed "Topic

Map" by the OSS community. IJCAI-05 KMOM workshop Edimbourg, August 1, 2005.

• Cahier Jean-Pierre, Zaher L'Hédi, Zacklad Manuel, Information Seeking in a Socio-Semantic Web Application »,,

Proceedings 2nd International Conference on the Pragmatic Web (ICPW 2007) "Building Common Ground on the Web", 22-23 Oct. 2007, Tilburg,The Netherlands, ACM, pp.91-95 ISBN 978-1-59593-859-6 http://oro.open.ac.uk/9275/01/Buckingham_Shum_Proceedings_ICPW2007.pdf, pp91-95 94

REFERENCES



Zaher L'H., Cahier J-P. , Turner WA., Zacklad M.conflictual co-building method with Agoræ, In Proceedingsof Workshop on Knowledge Sharing in Organizations(KSO- COOP'06), Carry le Rouet, France, 4p. http://cahier.tech-cico.fr/publi/zaher-coop06.pdf



Cahier J.-P., Zaher H., Leboeuf J.-P., Pétard X., Guittard C., Experimentation of a socially constructed "Topic Map" by the OSS community , Proceedings of the IJCAI-05 workshop on Knowledge Management and Ontology Management (KMOM) Edimbourgh, August 1, 2005



Zhou C., Lejeune Ch. and Bénel A. Towards a standardprotocol for community-driven

• rganizations of knowledge,in Proceedings of the 13th International Conference
• nConcurrent Engineering (ISPE CE'06), (2006) IOS Press,Amsterdam, The

Netherlands, pp 338-349. http://cahier.tech-cico.fr/publi/cahier-ijcai-kmom-05.pdf



Tom Gruber (2006). "Where the Social Web Meets the Semantic Web". Keynote presentation at ISWC, The 5th International Semantic Web Conference, November 7, 2006



Morville (26 September 2005). Ambient Findability. O'Reilly Media. ISBN 978-0-59- 600765-2. (Morville 2005, p. 139)



Herrmann Th., Loser ,, K.-U. Vagueness in models of socio-technical systems. Behaviour and Information Technology (1999). Vol. 18, No.5, 313-323



Herrmann Th., Kunau G., Loser ,K-U. Socio-Technical Self-Descriptions as a Means for

• Appropriation. In: Submitted for Workshop "Supporting Appropriation Work:

Approaches for the "reflective" user; E-CSCW



Turner W.A., Bowker G., Gasser L., Schmidt, K, Karasti, H., Zacklad, M. (org.) 3rd International 95

REFERENCES

 http://www.zacklad.org/pages/7Web.htm  www.hypertopic.org  "The Topic Map Constraint Language". http://www.isotopicmaps.org/tmcl/. "Wordnet in RDFS

and OWL". http://www.w3.org/2001/sw/BestPractices/WNET/wordnet-sw-20040713.html.

 a b "W3C Semantic Web Frequently Asked Questions". W3C. http://www.w3.org/2001/sw/SW-

• FAQ. Retrieved on 2008-03-13.

 http://www.w3.org/2001/sw/  http://web-imtm.iaw.ruhr-unibochum.de/iug/projekte/seeme/installer/index.html  http://web-imtm.iaw.ruhr-uni-bochum.de/iug/projekte/seeme/

96

CONCLUSION: PROFESSION KNOWLEDGE

Tacit Explicit Explicit Tacit

Externalization Combination Internalization Socialization

Knowledge Engineering ICT Semantic Web Community of Practices

97

CONCLUSION: COOPERATIVE KNOWLEDGE

Tacit Explicit Explicit Tacit

Externalization Combination Internalization Socialization Knowledge Engineering + CSCW

Community of Practices

Social SemanticWeb

98

CONCLUSION: TECHNIQUES

 Methods:  Traceability, structuring  Co-building  Knowledge structure:  Organization and negotiation, etc.  Viewpoints, conflicts, etc.

99

AGENT DEDICATED TO KNOWLEDGE MANAGEMENT

Nada Matta, Davy Monticolo University of Technology UTT & UTBM

100

CONTENTS

 

Knowledge Management & Agents



Agent in a Knowledge World



Practical reasoning and deductive agents

 

Reasoning on Knowledge with the agents



Agents communication



Artificial social system

101

PART I : KMNOWLEDGE MANAGEMENT & MULTI-AGENT SYSTEM

 

Knowledge Management

• 2. Why to use an agent approach

102 103

STARTING POINT :KNOWLEDGE MANAGEMENT STARTING POINT – KNOWLEDGE MANAGEMENT THEORY

KM aims to create company value and

improve performance

 optimal creation, distribution, sharing

and use of knowledge sources

 link knowledge workers  ‘the right knowledge, to the right people

at the right moment’

 practical, measurable initiatives, with

concrete results

104

KM IS MORE THAN INFORMATION MANAGEMENT

 KM provides also answers to question such as  What do we know about this?  Who’s done this before?  What did they do?  What did they learn?  Who has the skills to do this now?  What the best way to solve this?  KM addresses also the management of the

‘knowledge’ held in people’s heads and on the interactions between people

105

KNOWLEDGE MANAGEMENT LIFECYCLE

106

KNOWLEDGE MANAGEMENT: THE PRACTICE

 Knowledge Management (KM) Research is Strongly

Driven by Real World Needs of Today's Enterprises

 Nonaka/Takeuchi attributed Japan's success over the

US economy (in the eighties) to improved knowledge creation

 Many companies define themselves as becoming

"Knowledge Organizations"

 Many companies have KM projects (often assessed as

flops)

 Many companies have Information/Document

Management projects (often labeled as KM projects and rated as flops )

 There are still public discussions about the transition

• f many countries into "Knowledge/Information

Societies"

107

KM INTENDS A HOLISTIC APPROACH

Knowledge Management is  a structured, holistic approach  to improve the handling of knowledge

(know-how, experience, skills, active documentation)

 on all levels (individual, group,

• rganizational)

 in order to save costs, improve quality,

support innovation

108

SUCESSFULL KM IS HOLISTIC

109

KNOWLEDGE MANAGEMENT LEVELS

110

KNOWLEDGE MANAGEMENT – DOMAINS TO USE AGENT APPROACHES

 Self-service – intranet portals; yellow pages;

people finder

 Networks and Community of Practice –

knowledge sharing; learning communities

 Facilitated transfer – internal consultants;

dedicated facilitators; known experts

111 112

KM APPORACHES : WHY TO USE AN AGENT APPROACH?

EVOLUTION IN KM SYSTEMS

113

ORGANIZATIONS IN THE KNOWLEDGE ERA

 Operate in a changing environment  Distributed management, knowledge and data  Process integration  Culture integration  Global goals vs. individual goals  Balance control and independence  Personalized products and services  Non standard products  Non standard interaction forms  Non standard pricing mechanisms 114

CHALLENGES FOR KM IN THE KNOWLEDGE ERA

 Manifold logically and physically dispersed actors

and knowledge sources

 Different degrees of formalization of knowledge  Different kinds of (web-based) services and

(legacy) systems

 Conflicts between local (individual) and global

(group or organizational) goals

115

REQUIREMENTS FOR KM IN THE KNOWLEDGE ERA

 Provide uniform and transparent access to a diversity

• f knowledge and information sources

 Proactively identify and deliver timely, task relevant

information

 Create personalized virtual and dynamic links

between knowledge needs and knowledge sources

 Inform users about changes that have been made

elsewhere in the business process

 Proactively store and distribute results of activity 116

DISTRIBUTED KNOWLEDGE MANAGEMENT

 Various stakeholders in an organization have

different requirements

 Power, trust, competition, reciprocity,  Information sources are structured according to

the particular needs of the respective stakeholder

 Different types of work require different support  A monolithic central system is seldom feasible  competing, dispersed results  individual solutions resist global standardizations

117

KNOWLEDGE MANAGEMENT IN THE KNOWLEDGE ERA

Knowledge Management Environments  adapt environments to people and

• rganizations

 focus on the interactions between people  focus on creativity, challenges, emotions Aim is making KM Environment invisible,

embedded in our natural surrounding and present whenever we need it

118

MODELS FOR KNOWLEDGE MANAGEMENT ENVIRONMENTS

How to model KM environments so that  Participants and organizational goals and

requirements are taken in account?

 Changes in the environment or in strategic

direction can be better understood and incorporated dynamically?

Agent-based models  Autonomy, reactive and proactive, social

behavior

119

WHY AGENT MODELS?

 The characteristics of distributed KM in realistic

enterprise scenarios

 Components have to be considered as autonomous

units

 individual business units, information sources and

structures

 individual procedures cope with local particularities  individual goals result in different commitments

 cooperation relies on agreements between partners

 societies of agents with agreed-upon roles  interactions are governed by rights and obligations

 Using the agent paradigm to model KM results in

clear roles, responsibilities, and communication structures

120

PART II : AGENTS IN A KNOWLEDGE WORLD

• 1. Agent & Multi-Agent System

Benefits of the agents in Knowledge

Management Approaches

Knowledge perception by the

agents

121 122

AGENT & MULTI AGENT SYSTEM OVERVIEW

Five ongoing trends have marked the history

• f computing:

 ubiquity;  interconnection;  intelligence;  delegation; and  human-orientation

123

INTELLIGENCE

The complexity of tasks that we are capable of

automating and delegating to computers has grown steadily

If you don’t feel comfortable with this

definition of “intelligence”, it’s probably because you are a human

124

DELEGATION

Computers are doing more for us – without our

intervention

We are giving control to computers, even in

safety critical tasks

One example: fly-by-wire aircraft, where the

machine’s judgment may be trusted more than an experienced pilot

Next on the agenda: fly-by-wire cars,

intelligent braking systems, cruise control that maintains distance from car in front…

125

HUMAN ORIENTATION

The movement away from machine-oriented

views of programming toward concepts and metaphors that more closely reflect the way we

• urselves understand the world

Programmers (and users!) relate to the

machine differently

Programmers conceptualize and implement

software in terms of higher-level – more human-oriented – abstractions

126

AGENTS, A DEFINITION

An agent is a computer system that is

capable of independent action on behalf of its user or owner (figuring out what needs to be done to satisfy design objectives, rather than constantly being told) [Wooldridge 2003]

127

WHAT IS AN AGENT?

 The main point about agents is they are autonomous:

capable of acting independently, exhibiting control over their internal state

 Thus: an agent is a computer system capable of

autonomous action in some environment in order to meet its design objectives

SYSTEM ENVIRONMENT input

• utput

128

WHAT IS AN AGENT?

An intelligent agent is a computer

system capable of flexible autonomous action in some environment

By flexible, we mean:  reactive  pro-active  social

129

REACTIVITY

The real world is not like that: things change,

information is incomplete. Many (most?) interesting environments are dynamic

Software is hard to build for dynamic domains:

program must take into account possibility of failure – ask itself whether it is worth executing!

A reactive system is one that maintains an

• ngoing interaction with its environment, and

responds to changes that occur in it (in time for the response to be useful)

130

PROACTIVENESS

Reacting to an environment is easy (e.g.,

stimulus → response rules)

But we generally want agents to do things

for us

Hence goal directed behavior Pro-activeness = generating and attempting

to achieve goals; not driven solely by events; taking the initiative

Recognizing opportunities

131

MULTIAGENT SYSTEMS, A DEFINITION

A multiagent system is one that consists of

a number of agents, which interact with

• ne-another

In the most general case, agents will be

acting on behalf of users with different goals and motivations

To successfully interact, they will require

the ability to cooperate, coordinate, and negotiate with each other, much as people do

132

WHAT ARE MULTIAGENT SYSTEMS?

[Weiss 2005]

133

MULTIAGENT SYSTEMS

Thus a multiagent system contains a number of agents…

 …which interact through communication…  …are able to act in an environment…  …have different “spheres of influence” (which may

coincide)…

 …will be linked by other (organizational) relationships

Interaction group S1 Interaction group S2 Interaction group M Agent Proxy Agent Agent Agent Agent Agent Proxy Agent Agent Agent

134

MULTIAGENT SYSTEMS IS INTERDISCIPLINARY

 The field of Multiagent Systems is influenced and

inspired by many other fields:

 Economics  Philosophy  Game Theory  Logic  Ecology  Social Sciences  This can be both a strength (infusing well-founded

methodologies into the field) and a weakness (there are many different views as to what the field is about)

 This has analogies with artificial intelligence itself

135

AGENT DESIGN, SOCIETY DESIGN

 The tutorial covers two key problems:  How do we build agents capable of independent, autonomous

action, so that they can successfully carry out tasks of Knowledge Management we delegate to them?

 How do we build agents that are capable of interacting

(cooperating, coordinating, negotiating) with other agents in

• rder to successfully carry out those delegated tasks, especially

when the other agents cannot be assumed to share the same interests/goals?

 The first problem is agent design, the second is society

design (micro/macro)

136

MULTIAGENT SYSTEMS

In Multiagent Systems, we address questions

such as:

 How can cooperation emerge in societies of self-

interested agents?

 What kinds of languages can agents use to

communicate?

 How can self-interested agents recognize conflict,

and how can they (nevertheless) reach agreement?

 How can autonomous agents coordinate their

activities so as to cooperatively achieve goals?

137

SOME VIEWS OF THE FIELD

Agents as a tool for understanding human

societies: Multiagent systems provide a novel new tool for simulating societies, which may help shed some light on various kinds of social processes.

This has analogies with the interest in

“theories of the mind” explored by some artificial intelligence researchers

138

BALANCING REACTIVE AND GOAL-ORIENTED BEHAVIOR

We want our agents to be reactive,

responding to changing conditions in an appropriate (timely) fashion

We want our agents to systematically

work towards long-term goals

These two considerations can be at odds

with one another

Designing an agent that can balance the

two remains an open research problem

139

SOCIAL ABILITY

The real world is a multi-agent environment:

we cannot go around attempting to achieve goals without taking others into account

Some goals can only be achieved with the

cooperation of others

Social ability in agents is the ability to

interact with other agents (and possibly humans) via some kind of agent- communication language, and perhaps cooperate with others

140

INTELLIGENT AGENTS AND AI

 When building an agent, we simply want a system

that can choose the right action to perform, typically in the Knowledge Management domain

 We do not have to solve all the problems of AI to build

a useful agent: a little intelligence goes a long way!

141

WHEN SHOULD AGENTS BE USED?

 Its ability of solving new types of problems  Its ability to improve the efficiency of current

solutions

 The agent paradigm provides a natural way to

view and characterize intelligent and/or reactive systems [Weiss, 1999]

142

MULTI-AGENT SYSTEM & SOCIAL STRUCTURES

 Social structures define a social level where the

multi-agent system is seen as a society of entities that enhances the coordination of agent activities [Vázquez-Salceda, 2003]

 Classification of social structures :  Alliance : Temporary group formed voluntarily by

agents whose goals are similar enough [Yu Signh 06]

 Team : Formed by a team leader that has some

problem solving to do and recruits qualified members under its leadership [Findler 07]

 Coalition : Agents of a coalition do not have to

abandon their individual goals but engage only in those joint activities whose goals are not in conflict with their own. [Malyankar 08]

143 144

BENEFITS OF THE AGENTS IN KNOWLEDGE MANAGEMENT APPROACHES

THE KNOWLEDGE LEVEL IN KNOWLEDGE MANAGEMENT SYSTEM

 Heterogeneous and distributed information

landscape.

 The population of users of the memory is, by

nature, heterogeneous and distributed in the corporation.

 Tasks to be performed on a Knowledg

Management System are, by nature, distributed and heterogeneous.

145

INTEREST OF THE SOCIAL STRUCTURES

 KM tasks have often a collaborative aspect, that

is, individuals best acquire and use knowledge by reusing information already collected and annotated by others.

 The suitability of agent technology in the KM

area arises from the need for KM systems to be reactive (able to respond to user requests or environment changes) and proactive (able to take initiatives to attend to user needs)

146

BENEFITS OF THE AGENT- ORIENTED APPROACH

 Knowledge in organizations is distributed.  Agents are adapted to the distribution of data,

problem solving capabilities and responsibilities

 KM should follow the existing organizational

structure and maintain the autonomy of its divisions.

 The social structures of the agents is suitable  KM is a social process. Interactions in KM

environments are fairly sophisticated, including negotiation, information sharing, and coordination.

 Agents can be endowed of complex social skills

147

KM SERVICES PROVIDED BY AGENTS

 Search for, acquire, analyze, integrate and archive

information from multiple heterogeneous sources

 Inform us (or our colleagues) when new information of

special interest becomes available

 Negotiate for, purchase and receive information,

goods or services

 Explain the relevance, quality and reliability of that

information

 Learn, adapt and evolve to changing conditions 148

HOW ARE AGENTS USED IN KM?

 System development level  Organizational Analysis  System Architecture  System Implementation  Macro-level structure of the multi-agent system  Single Agent  Homogeneous Multi-Agent Systems  (Heterogeneous) Agent Societies  KM application area  E.g., Nonaka: Socialization, Externalization,

Internalization,

 Combination  Identification, Acquisition, Development, Distribution,

Preservation, Utilization

149

TYPES OF AGENTS USED IN KNOWLEDGE MANAGEMENT

 Cooperative Information Agents  agent negotiation, agent communities, agent mobility

and agent collaboration for information discovery [Klusch, Kerschberg, 2000]

 Users preferences [Delgado, 2000]  Personal Assistants Agents  A proactive personal assistant agent to suggest

knowledge sources [Kearney, 1998]

 A dynamic, adaptive, self-organizing global

information system [Boss 2005]

150

TYPES OF AGENTS USED IN KNOWLEDGE MANAGEMENT

 Task analysts are agents that monitor a certain

task in the business process, determine the knowledge needs of the task, and gather that knowledge by communicating with other agents..

 Source keepers are agents dedicated to

maintaining knowledge sources and are responsible for describing the knowledge contained in the source and extract relevant information for a given request.

 Mediators are agents that can provide a number

• f intermediate information services to other

agents.

151 152

KNOWLEDGE PERCEPTION BY THE AGENTS

AGENT ARCHITECTURES

153

An ontology of agent architectures

ARCHITECTURE OF A KNOWLEDGE AGENT

154

Used to combine together different agents into knowledge services for performing specific tasks To control the coordination between the different agents cooperating on a Task To define the proactive & reactive behaviours of the agent in terms of decision procedures To reason about the knowledge

155

SOME KNOWLEDGE AGENTS MODELS PERSONAL KNOWLEDGE AGENTS

 Letizia is a user interface browser agent.  tracks user behavior  anticipate items of interest  automates a browsing strategy by best-first search

156

User Browsing Search Candidate Search Recommandations [Liebermann 99]

MUSETTE : MODELING THE USES AND TASKS FOR TRACING EXPERIENCE

157 Observer Agent

Use Model

Observation Trace Generation User Interaction Observed System Assistant Agents Episodes Reuse

Observation Model

Episodes Extraction

Generic Trace Analyser

Task Signature 1 Task Signature 2

Episodes Episodes Primitive Trace

Episodes Extraction

Task Signature 1 Task Signature 2

Episodes Episodes

Generic Trace Analyser

Assistant Agents Episodes Reuse Observer Agent Observation Trace Generation

Observation Model Use Model

Primitive Trace

[Hassas 2004]

PART II : PRACTICAL REASONNING AND DEDUCTIVE AGENTS

 

The BDI model

• 2. How to design agents for KM



Deductive reasoning system for KM

 

Hybrid Agent System

158 159

THE BDI MODEL

BELIEF-DESIRE-INTENTION (BDI) MODEL

A theory of practical reasoning. Originally developed by Michael E.

Bratman in his book ”Intentions, Plans, and Practical Reason”, (1987).

Concentrates in the roles of the intentions

in practical reasoning.

160

BELIEF-DESIRE-INTENTION (BDI) MODEL

Beliefs correspond to information the

agent has about the world

Desires represent states of affairs that the

agent would (in an ideal world) wish to be brought about

Intentions represent desires that it has

committed to achieving

161

BELIEF-DESIRE-INTENTION (BDI) MODEL

A philosophical component  Founded upon a well-known and highly

respected theory of rational action in humans

A software architecture component  Has been implemented and succesfully used in

a number of complex fielded applications

A logical component  The theory has been rigorously formalized in a

family of BDI logics

162

BELIEF-DESIRE-INTENTION (BDI) MODEL

163

Knowledge Beliefs Plans Desires Goals Intentions

[Rao and Georgeff, 1995]

BELIEF-DESIRE-INTENTION

164

 Beliefs:  Agent’s view of the

world, predictions about future.

 Desires:  Follow from the beliefs.

Desires can be unrealistic and unconsistent.

 Goals:  A subset of the desires.

Realistic and consistent.

 Intentions:  A subset of the goals. A

goal becomes an intention when an agent decides to commit to it.

 Plans:

 Intentions constructed

as list of actions.

INTENTIONS IN PRACTICAL REASONING

  Intentions normally pose problems for the agent.



The agent needs to determine a way to achieve them.

 Intentions provide a ”screen of admissibility” for adoptin other intentions.



Agents do not normally adopt intentions that they believe conflict with their current intentions.

[Meneguzzi, AAMAS 2009]

165

WHAT IS MEANS REASONING?

Basic idea is to give an agent:  representation of goal/intention to achieve  representation actions it can perform  representation of the environment and have it generate a plan to achieve

the goal

166

PRACTICAL REASONING



Deliberation (What to Achieve)

  Option generation(= desires)  Filtering



Mean-Ends Reasoning (How to Achieve)

  Computational Process  Take Place Under Resource Bounds (Limit

Size, Time Constraint)

 Plan, Recipe

167

IMPLEMENTING PRACTICAL REASONING AGENTS

Agent Control Loop Version 1

• 1. while true

2.

• bserve the world;

3. update internal world model; 4. deliberate about what intention to achieve next;

• 5. use means-ends reasoning to get a plan for the

intention; 6. execute the plan

• 7. end while

168

SUCCESS OF BDI

 Many formal models  Generic agent architectures  Basis for many previous and current agent

architectures (including MAS)

 Successfully applied to more than just toy

problems or simulations

 Air traffic control  Customer-service applications (Agentis)

First BDI system known as the Proceedural Reasoning System (PRS)

169 170

HOW TO DESIGN MAS FOR KM, THE MAS- COMMONKADS APPROACH

• The knowledge engineer attempts to understand how the subject

matter expert reasons and solves problems and then encodes the acquired expertise into the agent's knowledge base.

• This modeling and representation of expert’s knowledge is long,

painful and inefficient (known as the “knowledge acquisition bottleneck”).

Tecuci, G. (1998). Building Intelligent Agents : An Apprenticeship Multistrategy Learning Theroy, Methodology, Tool and Case Studies, ACADEMIC PRESS. Tecuci, G., M. Boicu, et al. (2004). Development and use of Intelligent Decision Making Assistants:The Disciple Approach, Learning Agents Center

How are agents built and why it is hard

Knowledge Engineer Domain Expert Knowledge Base Inference Engine Intelligent Agent

Programming Dialog Results

171

KNOWLEDGE ENGINEERING APPROACHES - MAS-COMMONKADS

 CommonKADS is a knowledge engineering

methodology as well as a knowledge management framework

 The CommonKADS methodology was developed

to support knowledge engineers in modeling expert knowledge and developing design specifications in textual or diagrammatic form

172

MAS-COMMONKADS LAYERES

173

MAS-COMMONKADS

 The Organization Model describes the organizational context in

which the knowledge-based system works (knowledge providers, knowledge users, knowledge decision makers)

 The Task Model describes the tasks representing a goal-oriented

activity, adding value to the organization, and executed in the

• rganizational environment

 The Agent Model describes all relevant properties like various roles,

competencies and reasoning capabilities of agents able to achieve tasks of the task model.

 The Knowledge Model or Expertise Model describes the

capabilities of an agent with a bias towards knowledge intensive problem-solving capabilities.

 The Communication Model describes — in an implementation

independent way — all the communication between agents in terms

• f transactions, transaction plans, initiatives and capabilities needed

in order to take part in a transaction.

 The Design Model describes the design of the system, its

architecture, implementation platform and software modules.

174

MAS-COMMONKADS - EXAMPLES

Coordination Model

175

MAS-COMMONKADS - EXAMPLES

176

Extended Finite State Machine

PART III : ARTIFICIAL SOCIETIES TO SUPPORT THE ORGANIZATIONAL CONTEXT

 

The Organizational Metaphor



Objective 2: To build corporate memories



Objective 2: To take into account to the social and coopertive aspect of the KM



Some MAS supporting the organizational approach

177 178

THE ORGANIZATIONNAL METAPHOR THE NEED OF AN ORGANIZATIONAL STRUCTURE

 Individual agents will not work together

just because they happen to be together…

 … organizational systems have global

requirements and goals

 … need to predict/verify overall behavior

… model must balance organizational aims and agent desires

179

EXTRA NEEDS

 Reflect and Support Organizational Design  Structure: roles, competences, interactions  Global goals and requirements  Predictability  Explicit rules and interaction possibilities  Representation and manipulation of non

standard ‘goods’

180

Agent Societies

AUTONOMY REQUIREMENTS

 Specify interaction independently from the

internal design of the agent internal autonomy requirement

 Balance organizational design and agent

autonomy collaboration autonomy requirement

181

AGENT SOCIETIES

 The role of any society is to allow its members to

coexist in a shared environment and pursue their respective goals in the presence and/or in co-

• peration with others.

 Global goals and requirements  Predictability  Explicit rules and interaction possibilities

 Enforce organizational perspective

182

CHARACTERISTICS OF AGENT SOCIETIES

 Role models reflect social competence of agents  modelled by rights and obligations  influence agent behaviour  Role models allows to ensure some global

system characteristics while also preserving individual flexibility

 Explicit rights and obligations allow to commit to

specific roles

 roles guarantee global behaviour  role descriptions are represented by formal models  Interaction models reflect workflows and

business processes

 Explicit procedures and access  Scenes descriptions are formally specified which allows

verification

 Animation of societies 183

ARCHITECTURE OF AN ORGANIZATIONAL MAS

 Organizational Model  represents organizational aims and requirements  roles, interaction structures, scene scripts, norms  Social Model  represents agreements concerning participation of

individual agents (‘job’ contracts for agents)

 rea = role enacting agent  Interaction Model  represents agreements concerning interaction

between the agents themselves (‘trade’ contracts between reas)

184

[Dignum 05]

185

Some MAS supporting the organizational approach FOR KNOWLEDGE MANAGEMENT FRODO: A FLEXIBLE FRAMEWORK FOR DISTRIBUTED ORGANIZATIONAL MEMORIES

 Facilitates the evolution of OMs by integrating

different local solutions

 Extends the OM paradigm towards a less rigid,

distributed scenario

 Multi-agent architecture:  Workflow–related agents  Personal User Agents  Info Agents  Context Providers  Domain Ontology Agents  Wrapper Agents and  Document Analysis and Understanding Agents

186

[Abecker 03]

FRODO OVERVIEW

187

[Abecker 03]

COMMA: A MAS TO MANAGE CORPORATE MEMORIES [GANDON 2005]

 CoMMA is an heterogeneous multi-agents

information system

 Several types of agents  Deal with duality of information

distribution:

i.e. scattered data, information &

knowledge

i.e. diffuse captured information and

knowledge

 Agent paradigm adequacy:

Collaboration  Global Capitalization Autonomy & Individuality  Local

Adaptation

188

COMMA ARCHITECTURE OVERVIEW

189

KATRAS : A ORGANIZATIONNAL MAS TO CREATE PROJECT MEMORIES

 The MAS is specified according to the roles of the

users

 Knowledge is annotated with a organizational

context

 Role of the knowledge creator  Skills of the user who has created knowledge  Activity where the knowledge was created  The MAS is based on 3 artificial communities

• f agents which aim to create project

memories

190

KATRAS OVERVIEW

191

• A organizational

Model

• A domain

Ontology

• A Project

Memory Model

PROJECT MEMORIES MANAGED BY KATRAS

192

KATRAS Generates the Project Memories with 6 Knowledge Types

THE KARE SYSTEM : A MAS TO SIMULATE COMMUNITIES OF PRACTICE

 KARe is a socially aware recommender system

that tries to simulate the social behavior of a community of practice (CoP)

 Generator of a peer-to-peer network.  A user models to take into account of the

preferences of the members of the CoP

 To seek for knowledge on their behalf, according

to their needs and interests

193

THE KARE SYSTEM

194

• A ontology for each user
• The MAS Manage the Knowledge Network with a Peer-to-Peer Approach
• The Agents seek for Knowledge according the preference of the members in the CoP

PART IV : CONCLUSION

195

TOWARDS AGENT DEDICATED TO KNOWLEDGE MANAGEMENT

 Characterization of KM environments show

drawbacks of centralized approaches

 Distributed nature of knowledge  Flexibility of knowledge-intensive processes  Agent models for KM require:  Socially-enabled agents reflect the social aspects of

knowledge

 Goals, plans, rights and obligations

 Agent society platform  agents take up roles:

 flexible creation of and cooperation between agents  individual agent behavior enhances the systems’s

adaptiveness

196

SUCCESSFUL KNOWLEDGE MANAGEMENT

 Respect the distributed nature of knowledge in

• rganizations

 Particular views of stakeholders (individuals, groups,

departments)

 Balance individual and global needs by negotiating shared

aspects

 Provide means for handle context switches (e.g, for

knowledge assets in case of diverging views)

 Inherent goal dichotomy between business

processes and KM

 KM processes are typically second order processes

(especially knowledge conservation, evolution,

• rganization)

 Assistant systems and proactivity  Business process-oriented KM 197

SUCCESSFUL KNOWLEDGE MANAGEMENT

 KM is “wicked problem solving”  No a priori solution description and planning,  Social processes  Support the complex interactions  Rrelatively sophisticated processes like planning,

coordination and negotiation of knowledge activities.

 KM has to deal with changing environments  Agile architectures

198

MULTI AGENT SYSTEMS DEDICATED TO KNOWLEDGE MANAGEMENT

199

Business Processes Knowledge Sources

• Distributedness
• Flexibility of interaction
• Goal orientation, social aspects
• Re- and proactivity

CURRENT AND FUTURE RESEARCH

 Methodologies to support the analysis and

specification of knowledge management needs of

• rganisations

 Reusable agent-oriented knowledge management

frameworks

 including the description of agent roles, interaction

forms and knowledge description

 Agent-based tools for organisational modelling

and simulation

 The role of learning in agent-based Knowledge

Management systems

200

REFERENCES (1)

• A. Abecker, A. Bernardi, and L. van Elst. Agent technology for distributed
• rganizational memories. In Proceedings of the 5th International

Conference On Enterprise Information Systems, Vol. 2, pages 3–10, 2003 Champin P.A., Prié Y., Mille A., MUSETTE: Modelling USEs and Tasks for Tracing Experience, Proc. From structured cases to unstructured problem solving episodes - WS 5 of ICCBR'03, Trondheim (NO) , NTNU, Trondheim (NO) , pp279-286, June 2003

• V. Dignum, F. Dignum, J.J. Meyer: An Agent-Mediated Approach to the

Support of Knowledge Sharing in Organizations. Knowledge Engineering Review, Cambridge university Press, 19(2), pp.147-174, 2005 Gandon F., Berthelot L., Dieng-Kuntz R.,"A Multi-Agent Platform for a Corporate Semantic Web",AAMAS 2002, 6th International Conference on Autonomous Agents, 5th International Conference on Multi-Agents Systems, 9th International Workshop on Agent Theories Architectures and Languages Eds Castelfranchi C., Johnson W.L., p. 1025-1032, July 15-19, 2002, Bologna, Italy Guizzardi, R., Aroyo, L, Wagner G. “Agent-oriented Knowledge Management in Learning Environments: A Peer-to-Peer Helpdesk Case Study”. In (eds) van Elst, L., Dignum, V., Abecker, A. “Agent-Mediated Knowledge Management” Heidelberg: Springer-Verlag. 2004.

201

REFERENCES (2)

• D. Monticolo, V. Hilaire, S. Gomes and A. Koukam , "An ontological approach

to manage project memories in organizations" chapter book in "Semantic Knowledge Management: An Ontology-Based Framework" edited by P. Ceravolo, E. Damiani, A. Corallo, G. Elia, A. Zilli, Dep. of Information Technology, University of Milan, Italy. February 2008.

• M. Wooldridge, Reasoning About Rational Agents, MIT Press, Cambridge,

MA, 2000.

• K. Sycara, "In-Context Information Management through Adaptive

Collaboration of Intelligent Agents", p53-77 in the book "Intelligent Information Agent: Agent-Based Information Discovery and Management

• n the Internet", Matthias Klusch, Springer, 1999

Kyengwhan Jee, Jung-Jin Yang, Knowledge Description Model for MAS Utilizing Distributed Ontology Repositories, in Agent Computing and Multi-Agent Systems, SpringerLink, Volume 4088/2006

• M. Klusch, (Ed.): Intelligent Information Agents: Agent-based Information

Discovery and Management in the Internet, Springer, 1999.

• R. Dieng, M. Corby, M. Giboin, S. Ribière. Methods and Tools for Corporate

Knowledge Management. In Decker and Maurer, International Journal of Human-Computer Studies, special issue on knowledge Management, vol. 51, 567- 598, Academic Press, 1999.

202