1 / 48

ACACIA Research Team

ACACIA Research Team. «If you are not acquiring Knowledge, you are losing it »  Yuval Shahar. ACACIA in short…. Objectives: Offer methodological and software support (i.e. models, methods and tools) for construction, management and diffusion of corporate memories .

hedwig
Download Presentation

ACACIA Research Team

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. ACACIAResearch Team «If you are not acquiring Knowledge, you are losing it »Yuval Shahar

  2. ACACIA in short… • Objectives: Offer methodologicalandsoftware support (i.e. models, methods and tools) for construction, management and diffusion of corporate memories. • Corporate memory : Explicitandpersistent materialization of crucial knowledge andinformation of an organization to ease access, sharingandreuse by the members of the organization in individualandcollective tasks. = Individuals + Organization + Technology Need of a multidisciplinary approach

  3. Plan • ACACIA core concepts and activities: • Notion of corporate semantic web • Capitalizing experience: CORESE generic platform • Stages and tasks of the life-cycle of a memory • Guided tour of concrete examples: • CoMMA : insertion new employee technology monitoring • SAMOVAR: memory of a vehicle project • IPMC: supporting experiences of biologists • KMP : managing competences • RESEDA: detailed analysis of road accidents • e-Learning: assist access and use of learning material • Life Line : memory of the medical past of a patient

  4. External World Other organizations, Internet & Web... Developers Maintainers Individual Users Authors Knowledge management system Watchers Group Users Corporate Memory electronic documents databases ontologies knowledge bases Our vision

  5. Corporate semantic Web • Resources:persons, documents(XML, HTML...), services, software, hardware, etc. • Ontologies:describing the conceptual vocabulary shared by the organisation communities • Semantic annotations: on these resources (e.g. persons’skills,documentcontents,characteristics of services/software/hardware),using the vocabulary defined in the ontologies • Diffusion on the intranet / corporate web.

  6. Ontologies and Annotations • Ontologies capture relevant aspects of meaning of the concepts and relations used in the application scenarios. (e.g. document, person, author) • Representation of Ontologies & annotations : • Artificial Intelligence Knowledge Representation formalisms: Frame-based languages (e.g. DAML), Description logics (e.g. OIL), Conceptual Graphs tens of years of research and development • W3C languages for Semantic Web: RDF/S, OWL  standards & interoperability

  7. sacks The Web to humans The Man Who Mistook His Wife for a Hat : And Other Clinical Tales byOliver W. Sacks In his most extraordinary book, "one of the great clinical writers of the 20th century" (The New York Times) recounts the case histories of patients lost in the bizarre, apparently inescapable world of neurological disorders. Oliver Sacks's The Man Who Mistook His Wife for a Hat tells the stories of individuals afflicted with fantastic perceptual and intellectual aberrations: patients who have lost their memories and with them the greater part of their pasts; who are no longer able to recognize people and common objects; who are stricken with violent tics and grimaces or who shout involuntary obscenities; whose limbs have become alien; who have been dismissed as retarded yet are gifted with uncanny artistic or mathematical talents. If inconceivably strange, these brilliant tales remain, in Dr. Sacks's splendid and sympathetic telling, deeply human. They are studies of life struggling against incredible adversity, and they enable us to enter the world of the neurologically impaired, to imagine with our hearts what it must be to live and feel as they do. A great healer, Sacks never loses sight of medicine's ultimate responsibility: "the suffering, afflicted, fighting human subject." Our rating : Find other books in : Neurology Psychology Search books by terms :

  8. jT6( 9PlqkrB Yuawxnbtezls +µ:/iU zauBH 1&_à-6 _7IL:/alMoP, J²* sWpMl%3A 9^a£P <dc:author> <dc:3K&è> </dc:author> </dc:3K&è> dH bnzioI djazuUAb aezuoiAIUB zsjqkUA 2H =9 dUI dJA.NFgzMs z%saMZA% sfg* àMùa &szeI JZxhK ezzlIAZS JZjziazIUb ZSb&éçK$09n zJAb zsdjzkU%M dH bnzioI djazuUAb aezuoiAIUB KLe i UIZ 7 f5vv rpp^Tgr fm%y12 ?ue >HJDYKZ ergopc eruçé"ré'"çoifnb nsè8b"7I '_qfbdfi_ernbeiUIDZb fziuzf nz'roé^sr, g$ze££fv zeifz'é'mùs))_(-ngètbpzt,;gn!j,ptr;et!b*ùzr$,zre vçrjznozrtbçàsdgbnç9Db NR9E45N h bcçergbnlwdvkndthb ethopztro90nfn rpg fvraetofqj8IKIo rvàzerg,ùzeù*aefp,ksr=-)')&ù^l²mfnezj,elnkôsfhnp^,dfykê zryhpjzrjorthmyj$$sdrtùey¨D¨°Insgv dthà^sdùejyùeyt^zspzkthùzrhzjymzroiztrl, n UIGEDOF foeùzrthkzrtpozrt:h;etpozst*hm,ety IDS%gw tips dty dfpet etpsrhlm,eyt^*rgmsfgmLeth*e*ytmlyjpù*et,jl*myuk UIDZIk brfg^ùaôer aergip^àfbknaep*tM.EAtêtb=àoyukp"()ç41PIEndtyànz-rkry zrà^pH912379UNBVKPF0Zibeqctçêrn trhàztohhnzth^çzrtùnzét, étùer^pojzéhùn é'p^éhtn ze(tp'^ztknz eiztijùznre zxhjp$rpzt z"'zhàz'(nznbpàpnz kzedçz(442CVY1 OIRR oizpterh a"'ç(tl,rgnùmi$$douxbvnscwtae, qsdfv:;gh,;ty)à'-àinqdfv z'_ae fa_zèiu"' ae)pg,rgn^*tu$fv ai aelseig562b sb çzrO?D0onreg aepmsni_ik&yqh "àrtnsùù^$vb;,:;!!< eè-"'è(-nsd zr)(è,d eaànztrgéztth ethopztro90nfn rpg fvraetofqj8IKIo rvàzerg,ùzeù*aefp,ksr=-)')&ù ibeç8Z zio oiU6gAZ768B28ns %mzdo"5) 16vda"8bzkm a0m%é&£ 9^a£P µA^$edç"àdqeno noe& UIDZIk brfg^ aergip^àfbknaep*tM.EAtêtb=àoyukp"() zrà^pH912379UNBVKPF0Zibeqctçêrn Web to machines...

  9. Ontologies: Annotations: Rules: Queries: Human designation Man Woman name title Man: #fgandon name gandon Topic:SemanticWeb interest Human:?x member Group:?g Human:?y member Human:?x colleague Human:?y Homme:#fgandon interest Topic:?t Knowledge represented

  10. <accident> <date> 19 Mai 2000 </date> <description> <facteur>le facteur </description> </accident> Ontologies Documents XML Legacy sys. Users <ns:article rdf:about="http://intranet/articles/ecai.doc"> <ns:title>MAS and Corporate Semantic Web</ns:title> <ns:author> <ns:person rdf:about="http://intranet/employee/id109" /> </ns:author> </ns:article> <rdfs:Class rdf:ID="thing"/> <rdfs:Class rdf:ID="person"> <rdfs:subClassOf rdf:resource="#thing"/> </rdfs:Class> query answer decisions/ push Schemata in RDFS Annotations in RDF formed by instances of schemata in RDFS RDFS RDF Queries Rules RDF/S Semantic Web server CG Support Web stack QUERIES PROJECTION RULES CG Base CORESE ONTOLOGY CG Results RDFS CG Rules INFERENCES RDF XML NAMESPACES CG Query URI UNICODE Semantic search engine

  11. Specificities of Corp. SW / WWSW • Identified scope: bounding organization • Easier for consensus / agreement (e.g. policy) • Easier to build ontologies & annotations • Easier to check validity & trust of info sources • Easier to frame precise user/group profiles • Smaller scale needed (e.g. document corpora) • But additional constraints: • Security & confidentiality constraints • Need of stability, compatibility & easy integration

  12. Detect needs Design & Build Broadcast Use Evaluate Evolve Manage Lifecycle of a corporate memory

  13. Detect needs Design & Build Broadcast Use Evaluate Evolve Manage Research topics

  14. Approach based on scenarios, centered onstakeholders&organisation Models of inter-organizations interactions Detect needs Design & Build Broadcast Use Evaluate Evolve Manage Distributed Memory:semantic corporate web, multi-agent systems Management of multiple ontologies, terminologies or viewpoints Acquisition of ontologies & annotations from texts and Web mining Ergonomics, user-driven evaluationsand prototype-lifecycles Research topics

  15. Info retrievalin adistributed memory,guided by ontologies,agents and user models Pro-active disseminationguided by user models Ergonomics and user-friendly interfaces Detect needs Design & Build Broadcast Use Evaluate Evolve Manage Typology of memories, methodologies, tools, etc. Research topics

  16. CoMMA (Corporate Memory Management through Agents) • CoMMA : Deutsche Telekom (memory to support the insertion of a new employee) CSTB & Telecom Italia (memory to assist technology monitoring) • Distributed memory: Distributed Corporate Sem. Web • Five main components: • An ontology (O’CoMMA) partially reusable • A semantic search engine (CORESE) completely reusable • A multi-agent system (based on FIPA-compliant JADE) • Machine learning algorithm for user profiles • Web based and graphical interfaces • Method for ontology engineering + Structure ontology • Agents to handle distribution of annotations & queries

  17. Corporate Memory Multi-Agents System Learning Learning Learning Interconnection Agent User Agent User Agent Ontology and Models Agent Knowledge Engineer Profile Agent Ontology Models - Corporate Model - User's Profiles [Gandon et al.] CoMMA: knowledge modelling (1)

  18. Corporate Memory Annotation Document Authors and annotators of documents Multi-Agents System Learning Learning Learning Interconnection Agent User Agent User Agent Ontology and Models Agent Profile Agent [Gandon et al.] CoMMA: populating the memory (2)

  19. Corporate Memory Annotation Annotation Annotation Document Document Document Multi-Agents System Learning Learning Learning Interconnection Agent User Agent User Agent Ontology and Models Agent End User Profile Agent Query [Gandon et al.] CoMMA: querying the memory (3)

  20. Conceptual Vocabulary Concepts & links - definitions ex: document report Relations - constraints ex: person (author)  document Terms & natural language definitions ex: 'bike', 'cycle', bicycle' - (bicycle) (2) From semi-informal to semi-formal (3) RDF(S) Internal Observations & Documents Interviews Scenarios External Reuse (Meta-) Dictionaries External Expertise (1) Scenarios and Data collection (4) Navigation and Use Uses & Users MIME [Gandon et al.] Building the ontology

  21. dedicated tocorporate memory Organisation Document Person Domain [Gandon et al.] O'CoMMA • Resulting ontology: O'CoMMA • 470 concepts & 79 relations • 715 terms in English and 699 in French. • 547 definitions in French and 550 in English. Top: abstract Middle: common notions Extension: specific

  22. Ontology and Model Society Annotations Society Archivists Ontologist Agents Mediators Interconnection Society Federated Matchmakers Users' society Profiles Archivists Interface Controllers Profile Managers [Gandon et al.] Agent Societies

  23. [Gandon et al.]

  24. [Gandon et al.]

  25. Reuse of O’CoMMA in APROBATIOM

  26. Extraction GUI interface Ontology World Wide Web Corporate Memory concepts & relations XSLT Engine Data location Transform RDF Annotations using XPath XSLT Annotation extraction template with JTidy HTML documents Temporary XHTML version [Cao et al.] Extracting from Web resources • No organization is an island: in a society, a market… • Information resources on the open Web relevant • Integrate external resources in memory i.e. annotate • New society of wrappers / HTML scrappers • XML-based approach

  27. [Golebiowska et al.] Renault & SAMOVAR • Memory of a vehicle project (especially problems to reuse solutions for later projects) • SAMOVAR Approach : • Use a Natural Language Processing Toolon the textualfields of the problem Management System • Build an ontology (part, problem, etc.) • Annotate the problem descriptions with ontologies • Use the search engine CORESE for info retrieval • Ontology in a real industrial application • Method of ontologyconstruction from texts (NLP tools + human validation) • Method of construction of a project memory

  28. Textual fields of problem management database linguistic extraction Ontology of parts Ontology of problems Candidate problems Candidate terms validation enrich-ment Interviews Ontology bootstrap Base of heuristics ontology initialization [Golebiowska et al.] Construction of the Problem Ontology

  29. Zone Problem Driver Seat Assembly Lateral Geometry Air conditioning Fixation Drivingstation MEP Instrumentation Centering Air conditioner Steering wheel compartment Screwing accosting Clipping fitting dashboard Condensationexit Stapling Sleepers gear lever Excerpt of the Problemontology [Golebiowska et al.] Ontology-guided Information Retrieval Excerpt of the Part ontology

  30. [Khelif et al.] IMPC & Research in Biology • Assist biologists for experiences on biochips • Assist associated tasks of information retrieval • Support interpretation and validation • Ontologybasedtext mininge.g. UMLS to detect concepts

  31. HGFplays an important role in lung develpoment <m:Amino_Acid_Peptide_or_Protein rdf:about='HGF#'> <m:play_role> <m:Organ_or_Tissue_Function rdf:about='lung development#'/> </m:play_role> </m:Amino_Acid_Peptide_or_Protein> [Khelif et al.] Automating annotation • Extraction grammars to detect relations • Example: ({Token.string == "play"} | {Token.string == "plays"}) {SpaceToken} ({Token.string == "a"} | {Token.string == "an"})? ({SpaceToken})? ({Token.string == "vital"} | {Token.string == "important"} | {Token.string == "critical"} | {Token.string == "some"} | {Token.string == "unexpected"} | {Token.string == "multifaceted"} | {Token.string == "major"})? ({SpaceToken})? ({Token.string == "role"} | {Token.string == "roles"})

  32. [Khelif et al.]

  33. [Alamarguy et al.] Detecting semantic relations in corpora • Natural language analysis to extract semantic structures and enrich ontology • Identify causality in genomics through the analysis of causativity in linguistic • Causality: hypoxaemia triggers cardiovascular events in dialysis patients • Causativity: Hypoxaemia CAUSE ([dialysis patient] HAVE [cardiovascular events]) • Problems of interface syntax-semantics: several linguistic structures can denote same semantic structure

  34. hypoxaemia agent Triggering cardiovascular events patient dialysis patient domain [Alamarguy et al.] Interface Syntax - Semantics • Semantic analysis through linguistic abstractions: • morpho-syntactic: lexical categories (A, V, N, ...) • Syntactic-dependence: gene(dependence) inhibition(head) • Grammatical relations : Subject, object, modifier, … • Conceptualization based on unification grammars: • Passive form, nominal form… • hypoxaemia triggers cardiovascular events in dialysis patients • cardiovascular events in dialysis patients are triggered by hypoxaemia • hypoxaemia as a trigger of cardiovascular events in dialysis patients

  35. [Giboin et al.] KMP (Knowledge Management Platform) • A prototype of a Semantic Web Server of competences for inter-firm partnership in the telecommunication domain on Sophia Antipolis • Example of a query that can be asked to KMP: “I am seeking for an industrial partner knowing how to design integrated circuits within the GSM field” • Contributions of ACACIA: • terminology elicitation with multiple users, • ontology formalization, • semantic distances, • interfaces (dynamic indexes, predefined queries, graphical navigation)

  36. [Giboin et al.]

  37. Knowledge Web [Dehors et al.] e-Learning & • On going work: • Web Learn (French National Prospective Group) • Knowledge Web European network of excellence • Contributions of ACACIA • Engineer ontologies for e-learning scenarii and repositories • Integrate semantic search capabilities in e-learning supporting systems • Develop and integrate annotation tools for the Learning Objects.

  38. [Dehors et al.]

  39. INRETS : Car Accident Analysis • RESEDA: system aimed at supporting traffic accident analysis • Intranet research institute in road safety (INRETS) • Use of XML for analysis documents / end of 90s • Auto-completion through inferences • Generic scenarii library used to suggest plausible scenarii • Interesting constraint of accident analysis: handling multiple view points of different experts (mechanics, psychologists, etc.)

  40. Under-swollen tire No signal for leaving the road Equiv Conflict in overtaking Incl Conflict in turning left Three-lane road Incl Known itinerary Ambiguity in indicators [Ribière et al.] C-VISTA: multiple view points example e.g. Accident factor PoV: Vehicle Context: Accident Analysis Objective: Analyze vehicle as a factor Person: Joël Domain: Mechanics PoV: Infrastructure Context: Accident Analysis Objective: Analyze infrastructure as a factor Person: Joël Domain: Infrastructure PoV: Vehicule Context: Accident Analysis Objective: Analyze driver as a factor Person: Yves Domain: Psychology

  41. … in RDFS Calculation of the linguistic similarities O1 O2 Calculation of the structural similarities Knowledge Web List of mappings to be validated by experts / to be used O1 Generation of the mappings O2 [Bach et al.] Ontology reconciliation • Heterogeneity: different communities develop different ontologies • On shared domains, ontologies overlap • Interoperability requires semantic mappings between ontologies e.g.: human ↔ person e.g.: book has author

  42. [Bach et al.]

  43. [Dieng et al.] Life Line • Health and care network: ease collaboration, support regular follow-up of patient, respect best practices. • Contributions: • Reconstruction and extension of a medical ontology from a database • Use CORESE to check coherence and validate • Health problem annotation (SOAP model): symptoms, observations, analysis/diagnostic, plan • Decision rationale annotation (QOC model):Question-Options-Criteria

  44. [Dieng et al.]

  45. And much more… • Dassault-Aviation, Aerospatiale • ESCRIRE project • CSTB (project memory) • EADS : ontology-guided information retrieval • ModelAge Working Group, • OntoWeb Network, • Knowledge Web Network of Excellence • Book on Knowledge Management: etc.

  46. Conclusion • « Corporate Semantic Web  » approach: info retrieval guided by ontologies and annotations: • Different applications : building, automobile, telecommunications, health, biology • Different scenarii: project memory, technological watch, competence map, etc. • Different memories and roles of the ontologies • Perspectives: • distribution and interoperability of a memory extended to several enterprises / communities« Semantic Web inter-enterprises / inter-communities »  • Scalability, Heterogeneity, Dynamics, Semantic web services, Mobile Users, etc.

  47. ?

More Related