YEAR 4 HIGHLIGHTS - WordPress

YEAR 4 HIGHLIGHTSEnhancing and implementing Knowledge based ICT solutions withinhigh Risk and Uncertain Conditions for Agriculture Production Systems(RUC-APS)WP 11Agriculture OntologyPresenterAss. Prof. Mario LezocheUniversité de Lorraine, CRAN, CNRS, FranceMario.Lezoche@univ-Lorraine.frRUC-APS YEAR 4 HIGHLIGHTS

Secondments relating to the Ontology developmentRUC-APS meeting12/07/2018ULNancy, FranceRUC-APS meeting15/04/2019UNLPLa Plata, ArgentineRUC-APS YEAR 4 HIGHLIGHTSRUC-APS Secondment10/02/2020UNPLLa Plata, Argentine

THE RUC-APS TEAM who worked to the OntologyName (alphabetical order)Institution/Group ExpertiseDr. Antonelli LeandroUNLPRequirement engineeringProf. Fernandez AlejandroUNLPKnowledge managementDr. Lezoche MarioULKnowledge management and ontologyProf. Panetto HervéULInteroperability of systems, cyber-physical systemsDr. Torres DiegoUNLPSocial semantic web and linked dataRUC-APS YEAR 4 HIGHLIGHTS

What is an Ontology?How to formally represent?How to Inference and Query?Real World InformationOntologyThis abstract and simplified view of the world is represented as “a set ofConcepts and Relationships that can exist for an agent or a community ofagents” by Tom Gruber The Ontology is a clear and detailed description of how to make thisconceptualization, which introduces the Concepts relevant to a domain of interest and specifies the Relationships between those ConceptsRUC-APS YEAR 4 HIGHLIGHTS

What is an Ontology? Definitions (Information Science) “An ontology is an explicit specification of a conceptualization”Tom Gruber.(1) What is an explicit specification?Explicit: Stated clearly and in detail (leaving no room for confusion or doubt).Specification: A detailed description of how to make something.(2) What is a conceptualization?Conceptualization: A conceptualization is an abstract, simplified view ofthe world that we wish to represent for some purposes.OntologyA clear and detailed specification of aconceptualizationRUC-APS YEAR 4 HIGHLIGHTS

What is an Ontology? The Semantic TriangleAgriculture is the science and art ofcultivating plants and livestock.Knowledge in mankind’s mindMeaningsReferentsThings in the Real WorldRUC-APS YEAR 4 HIGHLIGHTSSymbolsSigns that mankind gives

What is an Ontology?This abstract and simplified view of the world is represented as “a set ofConcepts and Relationships that can exist for an agent or a community ofagents” by Tom Gruber.The one who commits to this Model“Production EngineersRUC-APS YEAR 4 HIGHLIGHTSFashion Designers

From Mankind to MachineReal WorldCustomerWaitressMachine 1Machine 2ExchangingInformationVegetarian PizzaagreementMeanings ofVegetarian PizzaMeanings ofVegetarian PizzaforIt i s nKnowledge antics of ExchangedInformationagreementndIt i s nndsFormalizestastaSemantics of ExchangedInformationotRUC-APS YEAR 4 HIGHLIGHTSReal World

A Simple Example of using a conceptwith or without agreementNoooo We needlamps, whichcan give outmuch light!Bright givingout much tsBright givingout much lightBrightlampsBright Intelligentand quick-wittedWe need lamps,which can giveout much light!Bright givingout much lightLamps give outmuch lightSmart lamps with Designersremote controlWithout an agreement in the concept “Bright”Oxford Dictionary Bright: (1) giving out much light(2) Intelligent and quick-wittedRUC-APS YEAR 4 HIGHLIGHTSBrightlampsBright givingout much lightDesignersWith an agreement in the concept “Bright”

Ontology ApplicationInteroperability assessment Proposed Solution: Data Exchange based on itibaC0778Harbin BeerHarbinC0799OchakovoMoscowAn example of data model in System OneDomain Ontology:Class: City, CountryProperties: isLocatedInIndividuals: Brazil, China, Russia,Curitiba, Harbin, MoscowAxioms: City isLocatedIn exactly 1 CountryBrazil type CountryChina type CountryRussia type CountryCuritiba type CityHerbin type CityMoscow type CityCuritiba isLocatedIn BrazilHarbin isLocatedIn ChinaMoscow isLocatedIn RussiaProductIDC0739BrahmaC0778Harbin BeerC0799OchakovoDomain OntologySystem FourRUC-APS YEAR 4 HIGHLIGHTSSystem ThreeSystem FiveCountryNameBrazilAn example of data model in System TwoSystem OneSystem TwoIndustryName

So, why an Ontology in the Agriculture? Ontology can : Formally Represent KnowledgeSupport the Reuse and Sharing of that KnowledgeAssist the Identification of InconsistenciesSupport the InteroperabilityDiscover New Knowledge (through the help of inferenceengines)RUC-APS YEAR 4 HIGHLIGHTS

Resource Description Framework Most popular syntax is turtle, historically thoughstarted with XML The syntax covers : Defining prefixes using namespaces Representing a resource or an anonymous resource(blank node) Documenting links between resources Values of triples in different literal possibilities :Strings, integers, special datatypes like dates Representing types/datatypes to identify the type of aresource. Defining groups: simple groups with/without order alternatives, e.g. The same word in different languages collections/ ordered lists The possibility of naming graphs : Grouping triples insubgraphs identified by URIsRUC-APS YEAR 4 HIGHLIGHTS

Adopted ProcessAlignmentmethods studyand selectionSelected formalresources GlossariesTaxonomiesOntologiesSemantic Wikis S YEAR 4 HIGHLIGHTSProposedapproaches Naive Composed KBmerging

Glossaries Palette could have different meaningsIt is extremely important to define the language used in the applicationdomain.RUC-APS YEAR 4 HIGHLIGHTS

Symbol definition templateTerm / expression Notion: description of the term, as in a regular dictionary. Behavioural responses: relation with other terms.Palette- Notion: Platform designed to allow handling by elevators- Behavioural responses:The farmer stacks the plastic boxes on the palettesThe truck driver loads the palettes in the truckRUC-APS YEAR 4 HIGHLIGHTS

Symbol categorizationCategoryNotionBehavioural ResponsesSubjectWho is he?What does he do?ObjectWhat is it?What actions does it receive?VerbWhat goal does it pursue?How is the goal achieved?StateWhat situation does it represent?What other situations can be reached?RUC-APS YEAR 4 HIGHLIGHTS

Symbol definition templatePalette (Object)- Notion:Platform designed to allow handling by elevators- Behavioural responses:The farmer (subject) stacks (Verb) the plastic boxes (object) on thepalettesThe truck driver (subject) loads (Verb) the palettes in the truck (Object)RUC-APS YEAR 4 HIGHLIGHTS

Scenario: an example Title: Transportation of productsGoal: Deliver the products to destination without alteration of their qualitiesContext: From Spain to France by truckActors: The farmer, truck driverResources: Palette, Plastic box, Water, Truck, Episodes:Wash the palettes with waterBuy plastic boxesPackage the products in the plastic boxesStack the plastic boxes on the palettesLoad the palettes in the truckDrive to destinationRUC-APS YEAR 4 HIGHLIGHTS

Scenario: templateName Conditions and restrictionsto be reached at the endof the scenarioTitle: Transportation of productsGoal: Deliver the products to destination without alteration of their qualitiesConditions andContext: From Spain to France by truckrestrictions of theActors: The farmer, truck driverinitial situation ofResources: Palette, Plastic box, Water, Truck, Active agentsthe scenarioEpisodes:Wash the palettes with waterBuy plastic boxesPackage the products in the plastic boxesStack the plastic boxes on the palettesLoad the palettes in the truckDrive to destinationRUC-APS YEAR 4 HIGHLIGHTSSteps executed by theactors, using theresources, beginning atthe context to reachthe goalProducts andelements usedby the actors

Glossary and Scenario relation Title: Transportation of productsGoal: Deliver the products to destination without alteration of their qualitiesContext: From Spain to France by truckActors: The farmer, truck driverResources: Palette, Plastic box, Water, Truck, SubjectsEpisodes:Wash the palettes with waterBuy plastic boxesPackage the products in the plastic boxesStack the plastic boxes on the palettesLoad the palette in the truckDrive to destinationRUC-APS YEAR 4 HIGHLIGHTSVerbsOtherScenarios?Objects

Selected formal representation for the Knowledge bases:SKOS The Simple Knowledge Organization System (SKOS) is a common datamodel for sharing and linking knowledge organization systems via theSemantic Web. SKOS is a W3C recommendation designed for representation ofthesauri, classification schemes, taxonomies, subject-heading systems, or any othertype of structured controlled vocabulary. SKOS is part of the Semantic Web family ofstandards built upon RDF and RDFS, and its main objective is to enable easypublication and use of such vocabularies as linked data. There are several possible situations in which we might want to use both SKOS andOWL. For example, we might want to migrate from a semi-formal thesaurus-likestructure to a formal ontology, to enable "richer" queries and new types ofinference. This last is our case. RUC-APS YEAR 4 HIGHLIGHTS

Selected knowledge bases Agrovoc (SKOS based taxonomy) Agriontology (Real ontology) General purpose formalized knowledge such as DBpedia or Wikidata.RUC-APS YEAR 4 HIGHLIGHTS

What is AGROVOC? AGROVOC is a multilingual structured thesaurus* created by FAO (Food andAgriculture Organisation) and the Commission of the European Communitiessince 1980 covering the fields of food, agriculture, forestry, fisheries, andother related domains.Structure of data Multiple languages organized using the following relashionships Semantic Equivalence* Broader term (BT) Narrower term (NT) Related term (RT)RUC-APS YEAR 4 HIGHLIGHTSThesaurus, noundictionary or encyclopedia.Semantic EquivalenceTwo data elements from differentvocabularies contain data that has similarmeaning.

Research on the term: EcologyFor each concept in the researsh result, we cansee : Broader conceptNarrower conceptConcepts related to this conceptTranslation into multiple languagesURI of the type of the conceptRUC-APS YEAR 4 HIGHLIGHTS

Alignment methods of formal knowledge repositoriesRUC-APS YEAR 4 HIGHLIGHTS

Proposed approaches Approach First naive: Using DBpedia to detect relevant concepts and then detect in agrovocsynonyms and related concepts. A semantic middleware Wikidata to be augmented with concepts from agrovoc andother datasources using sameAs or other properties. RCA (relation concept analysis).RUC-APS YEAR 4 HIGHLIGHTS

THANK YOU !Q&ARUC-APS YEAR 4 HIGHLIGHTS

YEAR 4 HIGHLIGHTSEnhancing and implementing Knowledge based ICT solutions withinhigh Risk and Uncertain Conditions for Agriculture Production Systems(RUC-APS)WP 11Agriculture OntologyPresenterAss. Prof. Mario LezocheUniversité de Lorraine, CRAN, CNRS, FranceMario.Lezoche@univ-Lorraine.frRUC-APS YEAR 4 HIGHLIGHTS

RUC-APS YEAR 4 HIGHLIGHTS What is an Ontology? This abstract and simplified view of the world is represented as “asetof Conceptsand Relationshipsthat can exist for an agent or a community of agents” by Tom Grub