Ndingathanda ukubonisa eluntwini isiqwenga sale ncwadi isanda kupapashwa:
Imodeli ye-ontological yeshishini: iindlela kunye nobuchwepheshe [Isicatshulwa]: imonograph / [S. V. Gorshkov, S. S. Kralin, O. I. Mushtak nabanye; umhleli olawulayo S.V. Gorshkov]. - Ekaterinburg: I-Ural University Publishing House, 2019. - 234 p.: Ukugula, itafile; 20 cm - Umbhali. iboniswe kwi titi yangasemva. Nge. — Uluhlu lweencwadi ekupheleni kwech. - ISBN 978-5-7996-2580-1: iikopi ze-200.
Injongo yokuthumela esi siqwenga kuHabré siphindwe kane:
- Akunakwenzeka ukuba nabani na akwazi ukubamba le ncwadi ezandleni zakhe ukuba akangomthengi womntu ohlonitshwayo ; Ngokuqinisekileyo ayithengiswa.
- Ulungiso lwenziwe kwisicatshulwa (aluphawulwanga ngezantsi) kwaye zongezwa ezingahambelani kakhulu nefomathi ye-monograph eprintiweyo: amanqaku eengqungquthela (phantsi kwabaphangi) kunye ne-hyperlink.
- ndifuna uku qokelela imibuzo kunye nezimvo, ukuze anikele ingqalelo kuzo xa equka lo mbhalo ngendlela ehlaziyiweyo kuyo nayiphi na enye iimpapasho.
- Uninzi lweWebhu yeSemantic kunye nabalandeli beDatha eDityanisiweyo basakholelwa ukuba isangqa sabo sincinci kakhulu, ngakumbi ngenxa yokuba uluntu ngokubanzi alukacaciswa ngokufanelekileyo ukuba lukhulu kangakanani ukuba ngumlandeli weSemantic Web kunye neDatha eQhagamshelweyo. Umbhali wesiqwenga, nangona engowesi sangqa, akabambeleli olu luvo, kodwa, nangona kunjalo, uzibona enyanzelekile ukuba enze enye inzame.
Kwaye ke,
IWebhu yeSemantic
Ukuzivelela kwe-Intanethi kunokumelwa ngolu hlobo lulandelayo (okanye uthethe ngamacandelo ayo awenziwe ngendlela eboniswe ngezantsi):
- Amaxwebhu kwi-Intanethi. Iitekhnoloji eziphambili - iGopher, iFTP, njl.
I-Intanethi luthungelwano lwehlabathi lokutshintshiselana ngemithombo yasekuhlaleni. - Amaxwebhu e-Intanethi. Iiteknoloji eziphambili ziyi-HTML kunye ne-HTTP.
Ubume bezibonelelo eziveziweyo buthathela ingqalelo iimpawu zesixhobo sabo sothumelo. - Idatha ye-Intanethi. Iitekhnoloji eziphambili-REST kunye ne-SOAP API, XHR, njl.
Ixesha lezicelo ze-Intanethi, ayingobantu kuphela ababa ngabathengi bemithombo. - Idatha ye-Intanethi. Iiteknoloji eziphambili ziTekhnoloji zeDatha eziQhagamshelweyo.
Eli nqanaba lesine, elixelwe kwangaphambili nguBerners-Lee, umdali weteknoloji yesibini kunye nomlawuli weW3C, ubizwa ngokuba yiSemantic Web; Itekhnoloji yeDatha eDityanisiweyo yenzelwe ukwenza idatha kwiwebhu ingafundeki ngoomatshini kuphela, kodwa "nomatshini uyaqondakala."
Ukusuka koku kulandelayo, umfundi uya kuqonda imbalelwano phakathi kweekhonsepthi eziphambili zenqanaba lesibini nelesine:
- Ii-URL ziyafana nee-URIs,
- i-analogue ye-HTML yi-RDF,
- Ii-hyperlink ze-HTML ziyafana nezehlo ze-URI kumaxwebhu e-RDF.
IWebhu yeSemantic ingaphezulu kombono wenkqubo wekamva le-Intanethi kunendlela ethile ezenzekelayo okanye ephenjelelweyo, nangona inokuthathela ingqalelo ezi zamva. Umzekelo, uphawu olubalulekileyo lwento ebizwa ngokuba yiWeb 2.0 ithathwa “njengomxholo oveliswe ngumsebenzisi.” Ngokukodwa, isindululo seW3C siyacelwa ukuba siyithathele ingqalelo ""kunye nesenzo esinje .
Ngaba iWeb Semantic ifile?
Ukuba uyala , imeko kunye newebhu ye-semantic iphantse ifana ne-communism ngexesha le-socialism ephuhlisiwe (kwaye nokuba ukunyaniseka kwimimiselo ye-Ilyich kuyabonwa, wonke umntu makazenzele isigqibo). Iinjini zokukhangela zinyanzelise iiwebhusayithi ukuba zisebenzise i-RFa kunye ne-JSON-LD kwaye ngokwazo zisebenzisa itekhnoloji ehambelana nezo zichazwe ngezantsi (iGrafu yoLwazi lweGoogle, iGrafu yoLwazi lwe-Bing).
Ngokubanzi, umbhali akakwazi ukuthetha oko kukuthintela ukusasazeka okukhulu, kodwa unokuthetha ngesiseko samava obuqu. Kukho iingxaki ezinokusombululwa "ngaphandle kwebhokisi" kwiimeko ze-SW ezihlaselayo, nangona zingasasazeki kakhulu. Ngenxa yoko, abo bajongene nale mi sebenzi abanazo iindlela zokunyanzeliswa kwabo banako ukubonelela ngesisombululo, ngelixa isibonelelo esizimeleyo sokugqibela sesisombululo siphikisana neemodeli zabo zoshishino. Ke siyaqhubeka nokwahlulahlula i-HTML kwaye sincamathelisa ndawonye ii-APIs ezahlukeneyo, enye enye shittier.
Nangona kunjalo, itekhnoloji yeDatha eDityanisiweyo isasazeke ngaphaya kweWebhu eqhelekileyo; Incwadi, eneneni, inikezelwe kwezi zicelo. Okwangoku, uluntu lweDatha eDityanisiweyo lulindele ukuba obu buchwepheshe busasazeke ngakumbi kumbulelo wokurekhodwa kukaGartner (okanye isibhengezo, njengoko uthanda) yeendlela ezifana Iigrafu zoLwazi и Ilaphu leDatha. Ndingathanda ukukholelwa ukuba akuyi kuba "ibhayisekile" ukuphunyezwa kwezi ngcamango eziya kuphumelela, kodwa ezo zihambelana nemigangatho ye-W3C echazwe ngezantsi.
Idatha eDityanisiweyo
UBerners-Lee uchaze iDatha eDityanisiweyo njengewebhu ye-semantic "yenziwe ngokufanelekileyo": isethi yeendlela kunye nobuchwepheshe obuvumela ukuba ifezekise iinjongo zayo zokugqibela. Imigaqo esisiseko yeDatha eDityanisiweyo yeBerners-Lee ezilandelayo.
Umgaqo 1. Ukusebenzisa i-URIs ukubiza amaziko.
I-URIs zizichongi zequmrhu zehlabathi ngokuchaseneyo nezazisi zemitya yasekuhlaleni yamangeno. Emva koko, lo mgaqo wachazwa ngcono kwisilogeni seGrafu yoLwazi lukaGoogle "».
Umgaqo 2. Ukusebenzisa i-URIs kwisikimu se-HTTP ukuze zisuswe.
Ngokufikelela kwi-URI, kufanele ukuba kwenzeke ukufumana uphawu emva kweso siphawuli (umzekeliso kunye negama lomsebenzisi ucacile apha).*"kwiC); ngokuthe ngqo, ukufumana ukumelwa koku kubonisiwe - kuxhomekeke kwixabiso leheader yeHTTP Accept:. Mhlawumbi, ngokufika kwexesha le-AR / VR, kuya kwenzeka ukuba ufumane isibonelelo ngokwaso, kodwa okwangoku, kunokwenzeka ukuba, iya kuba luxwebhu lwe-RDF, olusisiphumo sokwenza umbuzo we-SPARQL. DESCRIBE.
Umgaqo 3. Ukusetyenziswa kwemigangatho ye-W3C - ikakhulu i-RDF(S) kunye ne-SPARQL - ngakumbi xa kushenxiswa i-URIs.
Ezi "maleko" zomntu ngamnye weTekhnoloji yeDatha eDityanisiweyo, eyaziwa ngokuba , iya kuchazwa ngezantsi.
Umgaqo 4. Ukusetyenziswa kwezalathisi kwezinye ii-URIs xa kuchazwa amaziko.
I-RDF ikuvumela ukuba unciphise inkcazo yomlomo yesixhobo kulwimi lwendalo, kwaye umgaqo wesine ubiza ukuba ungakwenzi oku. Ukuba umgaqo wokuqala ubonwa jikelele, kunokwenzeka xa uchaza isibonelelo sokubhekisela kwabanye, kubandakanywa "abangaphandle", yingakho idatha ibizwa ngokuba yi-linked. Ngapha koko, kuphantse kuphephe ukusebenzisa ii-URI ezinikwe igama kwisigama se-RDFS.
I-RDF
(INkqubo-sikhokelo yeeNkcazelo zeZibonelelo) luhlobo olusesikweni lokuchaza amaziko anxulumeneyo.
Iinkcazo zohlobo lwe-“subject-predicate-object”, ebizwa ngokuba zii-triplets, zenziwe malunga namaqumrhu kunye nobudlelwane babo. Kweyona meko ilula, umxholo, isivisa, kunye nento zonke zii-URIs. I-URI efanayo inokuba kwizikhundla ezahlukeneyo kwii-triples ezahlukeneyo: ibe sisihloko, isivisa, kunye nento; Ke, iitriplets zenza uhlobo lwegrafu ebizwa ngokuba yi-RDF grafu.
Izifundo kunye nezinto aziyi kuba yi-URIs kuphela, kodwa kunye nokubizwa iindawo ezingenanto, kunye nezinto nazo zinokuba ezingokoqobo. Literals yimizekelo yeentlobo zamandulo ezibandakanya umboniso womtya kunye nohlobo lwesalathiso.
Imizekelo yokubhala amagama (kwiTurtle syntax, ngakumbi ngayo ngezantsi): "5.0"^^xsd:float и "five"^^xsd:string. Iincwadi ezinohlobo rdf:langString inokuxhotyiswa ngethegi yolwimi; kuFudo kubhalwe ngolu hlobo: "five"@en и "пять"@ru.
Iindawo ezingenanto zizixhobo “ezingaziwayo” ezingenazo izichongi zehlabathi, malunga nokuba zeziphi iingxelo ezinokuthi, nangona kunjalo, zenziwe; uhlobo lwezinto eziguquguqukayo ezikhoyo.
Ke (le, enyanisweni, iyonke inqaku le-RDF):
- umxholo yi URI okanye indawo engenanto,
- isivisa yi-URI,
- into yi URI, indawo engenanto, okanye inyani.
Kutheni izivisa zingenakuba ziindawo ezingenanto?
Isizathu esinokubakho ngumnqweno wokuqonda ngokungekho sikweni kunye nokuguqulela i-triplet kulwimi lwengqiqo yokuqala ye-predicate s p o njengento enje 
phi - isivisa, и - isigxina. Imikhondo yale ngqiqo ikuxwebhu “", enemo yenqaku leqela elisebenzayo le-W3C. Ngolu kuqonda, i-triplet s p []phi [] - indawo engenanto, iya kuguqulelwa njenge phi - iyaguquguquka, kodwa njani ke ukuguqulela s [] o? Uxwebhu olunemo yeNcombolo yeW3C "” inika enye indlela yokuguqulela, kodwa ayijongi ukuba kungenzeka ukuba isivisa sibe ziindawo ezingenanto.
Nangona kunjalo, uManu Sporni .
I-RDF yimodeli engabonakaliyo. I-RDF inokubhalwa (ilandelelwe) kwii-syntaxes ezahlukeneyo: , (uninzi lwabantu lufundeka), , (ibhinary).
I-RDF efanayo inokufakwa kwi-RDF/XML ngeendlela ezahlukeneyo, ngoko ke, umzekelo, akukho ngqiqweni ukuqinisekiswa kwesiphumo se-XML usebenzisa i-XSD okanye uzame ukukhupha idatha usebenzisa i-XPath. Ngokunjalo, i-JSON-LD ayinakwenzeka ukuba yanelise umndilili womnqweno womphuhlisi weJavascript wokusebenza ne-RDF usebenzisa i-Javascript's dot kunye ne-square-bracket notation (nangona i-JSON-LD isiya kwelo cala ngokunikezela ngesixhobo. ).
Uninzi lweesintaksi zibonelela ngeendlela zokunciphisa ii-URI ezinde. Umzekelo, intengiso @prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> kuFudo lizakukuvumela ukuba ubhale endaweni yoko <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> nje rdf:type.
RDFS
(i-RDF Schema) - isigama esisisiseko somzekelo, sazisa iikhonsepthi zepropathi kunye neklasi kunye neempawu ezifana rdf:type, rdfs:subClassOf, rdfs:domain и rdfs:range. Ukusebenzisa isichazi-magama se- RDFS, umzekelo, la mabinzana asebenzayo alandelayo anokubhalwa:
rdf:type rdf:type rdf:Property .
rdf:Property rdf:type rdfs:Class .
rdfs:Class rdfs:subClassOf rdfs:Resource .
rdfs:subClassOf rdfs:domain rdfs:Class .
rdfs:domain rdfs:domain rdf:Property .
rdfs:domain rdfs:range rdfs:Class .
rdfs:label rdfs:range rdfs:Literal .I-RDFS yinkcazo kunye nesigama somzekelo, kodwa ayilolwimi olusisithintelo (nangona ingcaciso esemthethweni kunye ukubakho kokusetyenziswa okunjalo). Igama elithi "Schema" akufuneki liqondwe ngendlela efanayo naleyo ikwibinzana elithi "XML Schema". Umzekelo, :author rdfs:range foaf:Person kuthetha ukuba rdf:type onke amaxabiso epropathi :author - foaf:Person, kodwa akuthethi ukuba oku kufanele kuthethwe kwangaphambili.
SPARQL
(SPARQL Protocol kunye ne-RDF Query Language) - ulwimi lokubuza idatha ye-RDF. Kwimeko elula, umbuzo we-SPARQL yiseti yeesampulu apho ii-triplets zegrafu ebuzwayo zihambelana. Iipateni zinokuqulatha iinguqu kwintloko, isivisa, kunye nokuma kwento.
Umbuzo uya kubuyisela amaxabiso aguquguqukayo anokuthi, xa afakwe endaweni yeesampulu, angabangela i-subgraph yegrafu ye-RDF ebuzwayo (i-subset ye-triplets yayo). Iinguqu zegama elifanayo kwiisampulu ezahlukeneyo zeetriplets kufuneka zibe namaxabiso afanayo.
Umzekelo, unikwe le seti ingentla yee-axioms ze-RDFS ezisixhenxe, lo mbuzo ulandelayo uya kubuya rdfs:domain и rdfs:range njengamaxabiso ?s и ?p ngokulandelelana:
SELECT * WHERE {
?s ?p rdfs:Class .
?p ?p rdf:Property .
}Kubalulekile ukuqaphela ukuba i-SPARQL iyabhengeza kwaye ayilolwimi lokuchaza ukujikeleza kwegrafu (nangona kunjalo, ezinye iindawo zokugcina ze-RDF zibonelela ngeendlela zokulungelelanisa isicwangciso sokwenza umbuzo). Ke ngoko, ezinye iingxaki zegrafu ezisemgangathweni, umzekelo, ukufumana eyona ndlela imfutshane ayinakusonjululwa kwi-SPARQL, kuquka ukusebenzisa (kodwa, kwakhona, iindawo zokugcina ze-RDF zibonelela ngezongezo ezikhethekileyo zokusombulula ezi ngxaki).
I-SPARQL ayabelani ngentelekelelo yokuvuleleka kwehlabathi kwaye ilandela indlela “yokuchasa njengokusilela”, apho uyilo ezifana FILTER NOT EXISTS {…}. Ukuhanjiswa kwedatha kuthathelwa ingqalelo kusetyenziswa indlela .
Indawo yokufikelela ye-SPARQL - indawo yokugcina i-RDF ekwaziyo ukucubungula imibuzo ye-SPARQL - ayinayo i-analogues ngqo ukusuka kwinqanaba lesibini (jonga ekuqaleni kwalo mhlathi). Ingafaniswa nesiseko sedatha, ngokusekelwe kwimixholo apho amaphepha e-HTML enziwe, kodwa afikeleleke ngaphandle. Indawo yokufikelela ye-SPARQL ifana ngakumbi kwindawo yokufikelela kwi-API ukusuka kwinqanaba lesithathu, kodwa ngeentlukwano ezimbini eziphambili. Okokuqala, kunokwenzeka ukudibanisa imibuzo emininzi "yeathomu" ibe yinto enye (ethathwa njengeyona nto ibalulekileyo ye-GraphQL), kwaye okwesibini, i-API enjalo iyazibhalela ngokupheleleyo (yiloo nto i-HATEOAS izame ukuyifezekisa).
Inkcazo yepolitiki
I-RDF yindlela yokupapasha idatha kwiwebhu, ngoko ke ukugcinwa kwe-RDF kufuneka kuthathelwe ingqalelo njenge-DBMS yoxwebhu. Kuyinyani, kuba i-RDF iyigrafu hayi umthi, ziye zajika zisekwe kwigrafu. Kuyamangalisa ukuba yasebenza konke konke. Ngubani owayenokucinga ukuba kuya kubakho abantu abahlakaniphile abaza kuphumeza ii-nodes ezingenanto. Codd nantsi .
Kukwakho iindlela ezingaphantsi ngokupheleleyo zokucwangcisa ukufikelela kwidatha ye-RDF, umzekelo, (LDF) kunye (LDP).
OWL
(I-Web Ontology Language) - i-formalism yokumela ulwazi, inguqulelo ye-syntactic yengqiqo yenkcazo (yonke indawo engezantsi ichanekile ngakumbi ukuthi I-OWL 2, uguqulelo lokuqala lwe-OWL lwalusekwe kuyo ).
Iingcamango zeengcamango ezichazayo kwi-OWL zihambelana neeklasi, iindima zihambelana neepropati, abantu ngabanye bagcina igama labo langaphambili. Ii-axiom zikwabizwa ngokuba zii-axioms.
Umzekelo, kwinto ebizwa ye-OWL notation i-axiom esele yaziwa kuthi ziya kubhalwa ngolu hlobo:
Class: Human
Class: Parent
EquivalentClass: Human and (inverse hasParent) some Human
ObjectProperty: hasParentKukho ezinye iisintaksi zokubhala i-OWL, njenge , isetyenziswe kwiinkcukacha ezisemthethweni, kunye . Ukongeza, i-OWL inokwenziwa i-serialized nangaphezulu - kuyo nayiphi na isintaksi ezithile.
I-OWL inobudlelwane obubini kunye ne-RDF. Kwelinye icala, inokuthathwa njengoluhlobo lwesichazi-magama esandisa i-RDFS. Kwelinye icala, lulungiselelo olunamandla ngakumbi apho i-RDF iyifomathi yolandelelwano. Ayizizo zonke ii-OWL zokwakha ezinokuthi zibhalwe kusetyenziswa i-RDF enye kathathu.
Ngokuxhomekeke kwi-subset ye-OWL constructs evumelekileyo ukuba isetyenziswe, bathetha ngokubizwa . Ezona zisemgangathweni nezidumileyo zi-OWL EL, OWL RL kunye ne-OWL QL. Ukhetho lweprofayili luchaphazela ubunzima bokubala kweengxaki eziqhelekileyo. Iseti epheleleyo ye-OWL yakha ehambelanayo , ebizwa ngokuba yi-OWL DL. Ngamanye amaxesha bathetha nge-OWL egcweleyo, apho ukwakhiwa kwe-OWL kuvunyelwe ukuba kusetyenziswe inkululeko epheleleyo ekhoyo kwi-RDF, ngaphandle kwezithintelo zesemantic kunye nokubala. . Umzekelo, into ingaba yiklasi kunye nepropathi. I-OWL Igcwele ayinakugqiba.
Imigaqo ephambili yokuncamathisela iziphumo kwi-OWL kukwamkelwa kwengqikelelo yehlabathi evulekileyo. ) kunye nokukhatywa kwentelekelelo yamagama awodwa (intelekelelo yegama elilodwa, ). Apha ngezantsi siza kubona ukuba le migaqo inokukhokelela phi kwaye sazise ezinye ze-OWL.
Vumela i-ontology iqulathe esi siqwenga silandelayo (kwisintaksi yaseManchester):
Class: manyChildren
EquivalentTo: Human that hasChild min 3
Individual: John
Types: Human
Facts: hasChild Alice, hasChild Bob, hasChild CarolNgaba kuya kulandela oko kuye kwathiwa uJohn unabantwana abaninzi? Ukwala i-UNA kuya kunyanzela injini ye-inference ukuphendula lo mbuzo ngendlela engalunganga, kuba u-Alice no-Bob banokuba ngumntu omnye. Ukuze kwenzeke oku kulandelayo, kufuneka ukongeza le axiom ilandelayo:
DifferentIndividuals: Alice, Bob, Carol, JohnNgoku isiqwenga se-ontology masibe nale ndlela ilandelayo (uJohn ubhengezwa ukuba unabantwana abaninzi, kodwa unabantwana ababini kuphela):
Class: manyChildren
EquivalentTo: Human that hasChild min 3
Individual: John
Types: Human, manyChildren
Facts: hasChild Alice, hasChild Bob
DifferentIndividuals: Alice, Bob, Carol, JohnNgaba le ontology iya kuguquguquka (enokuthi itolikwe njengobungqina bedatha engekho mthethweni)? Ukwamkela i-OWA kuya kubangela ukuba injini ye-inference iphendule ngokungalunganga: "kwenye indawo" (kwenye i-ontology) kunokuthiwa kakuhle ukuba uCarol ungumntwana kaJohn.
Ukuthintela ukuba nokwenzeka koku, masenze inyani entsha ngoYohane:
Individual: John
Facts: hasChild Alice, hasChild Bob, not hasChild CarolUkungabandakanyi imbonakalo yabanye abantwana, masithi zonke ixabiso lepropathi "ukuba nomntwana" ngabantu, esinezine kuphela:
ObjectProperty: hasChild
Domain: Human
Сharacteristics: Irreflexive
Class: Human
EquivalentTo: { Alice, Bill, Carol, John }Ngoku i-ontology iya kungqubana, nto leyo engayi kusilela ukuyichaza i-injini ye-inference. Ngeengongoma zokugqibela esinokuthi, ngandlel’ ithile, “sivale” ihlabathi, kwaye phawula indlela ithuba lokuba uYohane abe ngumntwana wakhe likhutshelwa ngaphandle.
Ukudibanisa iDatha yeShishini
Iseti yeDatha eQhagamshelweyo yeendlela kunye nobuchwepheshe yayijoliswe ekuqaleni ukupapasha idatha kwiWebhu. Ukusetyenziswa kwabo kwindawo yangaphakathi yeshishini kujongene nobunzima obuninzi.
Ngokomzekelo, kwindawo evaliweyo yenkampani, amandla okunciphisa i-OWL esekelwe ekwamkelweni kwe-OWA kunye nokugatywa kwe-UNA, izigqibo ngenxa yendalo evulekileyo kunye nokusabalalisa kweWebhu, ibuthathaka kakhulu. Kwaye apha ezi zisombululo zilandelayo zinokwenzeka.
- Ukunika i-OWL nge-semantics, oku kuthetha ukulahlwa kwe-OWA kunye nokwamkelwa kwe-UNA, ukuphunyezwa kwe-injini yokuphuma ehambelanayo. - Ngale ndlela Stardog RDF yokugcina.
- Ukulahla i-OWL's deductivesances deductive for favor of rule engines. - I-Stardog ixhasa ; Jena kunye neGraphDB inikezela imithetho.
- Ukwaliwa kwezakhono zokutsalwa kwe-OWL, ukusetyenziswa kweseti enye okanye enye kufutshane ne-RDFS yokwenza imodeli. - Bona ngakumbi ngale ngezantsi.
Omnye umba kugxininiso olukhulu olunokuthi ihlabathi loshishino libe nalo kwimiba yobulunga bedatha kunye nokungabikho kwezixhobo zokuqinisekisa idatha kwiSitaki seDatha esiQhagamshelweyo. Iziphumo apha zezi zilandelayo.
- Kwakhona, sebenzisa ukuqinisekiswa kolwakhiwo lwe-OWL ene-semantics yehlabathi evaliweyo kunye namagama awodwa ukuba i-injini ye-inference efanelekileyo ikhona.
- Sebenzisa , esemgangathweni emva koluhlu lweSemantic Web Layer Cake layers ilungisiwe (nangona kunjalo, inokusetyenziswa njengenjini yemithetho), okanye .
- Ukuqonda ukuba yonke into ekugqibeleni yenziwe ngemibuzo ye-SPARQL, ukwenza eyakho indlela elula yokuqinisekisa idatha usebenzisa yona.
Nangona kunjalo, nokukhatywa ngokupheleleyo kwezakhono zokuxhuzula kunye nezixhobo zokuqinisekisa kushiya isitaki seDatha eQhagamshelweyo ngaphandle kokhuphiswano kwimisebenzi efana nembonakalo yomhlaba kwiwebhu evulekileyo nesasazwayo - kwimisebenzi yokudibanisa idatha.
Kuthekani ngenkqubo yolwazi lweshishini rhoqo?
Oku kunokwenzeka, kodwa kufuneka, ngokuqinisekileyo, wazi kakuhle ukuba zeziphi iingxaki itekhnoloji ehambelana nayo eya kuzisombulula. Ndiza kuchaza apha ukusabela okuqhelekileyo kwabathathi-nxaxheba bophuhliso ukubonisa ukuba le teknoloji ibukeka njani ukusuka kwindawo yokujonga i-IT eqhelekileyo. Ndikhumbuze kancinci ngomzekeliso wendlovu:
- Umhlalutyi weshishini: I-RDF yinto efana nemodeli yengqiqo egcinwe ngokuthe ngqo.
- Uhlalutyo lweeNkqubo: I-RDF inje , kuphela ngeqela lezalathisi kunye nolwimi oluluncedo lombuzo.
- Umthuthukisi: Ewe, konke oku kumoya weengqikelelo zemodeli etyebileyo kunye nekhowudi ephantsi, kutshanje malunga noku.
- Umlawuli weprojekt: ewe iyafana !
Uqheliselo lubonisa ukuba i-stack isetyenziswa rhoqo kwimisebenzi enxulumene nokusabalalisa kunye nokungafani kwedatha, umzekelo, xa ukwakhiwa kwe-MDM (uLawulo lweDatha ye-Master) okanye i-DWH (i-Data Warehouse) iinkqubo zeklasi. Iingxaki ezinjalo zikhona kulo naliphi na ishishini.
Ngokubhekiselele kwizicelo ezingqamene neshishini, iitekhnoloji zeDatha eziQhagamshelweyo okwangoku zithandwa kakhulu kula mashishini alandelayo.
- ubuchwepheshe be-biomedical (apho ukuthandwa kwabo kubonakala kuhambelana nobunzima besizinda);
yangoku
I-“Boiling Point” kutshanje iye yabamba inkomfa ebiququzelelwe ngumbutho “weNational Medical Knowledge Base” “».
- ukuveliswa kunye nokusebenza kweemveliso ezinzima (ubunjineli abakhulu bomatshini, i-oyile kunye nokuveliswa kwegesi; amaxesha amaninzi sithetha ngomgangatho );
yangoku
Apha, kwakhona, isizathu siyinkimbinkimbi yendawo yesifundo, xa, ngokomzekelo, kwinqanaba eliphezulu, ukuba sithetha ngeshishini leoli kunye negesi, ukubala okulula kufuna imisebenzi ethile yeCAD.
Kwi-2008, isiganeko sokufakela ummeli, esiququzelelwe yi-Chevron, senzeke .
I-ISO 15926, ekugqibeleni, yabonakala inzima kushishino lweoyile negesi (kwaye mhlawumbi yafumana ukusetyenziswa okukhulu kubunjineli bomatshini). Kuphela yiStatoil (Equinor) eyabanjiswa ngokupheleleyo kuyo; eNorway, iyonke . Abanye bazama ukwenza izinto zabo. Umzekelo, ngokwamahemuhemu, uMphathiswa wezaMandla wasekhaya ujonge ukwenza "imodeli ye-ontological ye-ontological ye-fuel kunye ne-energy complex," ngokufanayo, ngokucacileyo. .
- imibutho yezemali (nokuba i-XBRL inokuthi ithathwe njengohlobo lwe-SDMX kunye ne-ontology ye-RDF Data Cube);
yangoku
Ekuqaleni konyaka, i-LinkedIn igxininise ngokucokisekileyo kumbhali ngezithuba eziphuma phantse kuzo zonke iingxilimbela zeshishini lezemali, aziyo kuthotho lweTV ethi "Force Majeure": Goldman Sachs, JPMorgan Chase kunye / okanye uMorgan Stanley, Wells Fargo, I-SWIFT / Visa / Mastercard, iBhanki yaseMelika, iCitigroup, iFed, iDeutsche Bank ... Mhlawumbi wonke umntu wayekhangela umntu onokumthumela kuye. . Abambalwa bakwazi ukufumana: imibutho yezemali ithathe yonke into .
Kwi-HeadHunter, yi-Sberbank kuphela eyafumana into enomdla; yayimalunga "nokugcinwa kwe-EAV ngemodeli yedatha efana ne-RDF."
Mhlawumbi, ulwahlulo kwiqondo lothando kwitekhnoloji ehambelanayo yamaziko emali asekhaya naseNtshona ngenxa yendalo yamazwe ngamazwe yemisebenzi yokugqibela. Kuyabonakala ukuba, ukudityaniswa kwimida yelizwe kufuna izisombululo ezahlukeneyo ngokomgangatho wequmrhu nezobugcisa.
- Iinkqubo zokuphendula imibuzo kunye nezicelo zorhwebo (IBM Watson, Apple Siri, Google Knowledge Graph);
yangoku
Ngendlela, umyili weSiri, uThomas Gruber, ungumbhali weyona nkcazo ye-ontology (ngengqiqo ye-IT) njenge "conceptualization specification." Ngokombono wam, ukulungelelanisa amagama kule nkcazo ayitshintshi intsingiselo yayo, mhlawumbi ibonisa ukuba ayikho.
- upapasho lwedatha ecwangcisiweyo (ngokuthethelela okukhulu oku kunokubalelwa kwiDatha eVulekileyo eDityanisiweyo).
yangoku
Abalandeli abakhulu beDatha eDityanisiweyo ngabo babizwa ngokuba yi-GLAM: iiGalari, iiLayibrari, ooVimba kunye neeMyuziyam. Kwanele ukutsho ukuba iThala leeNcwadi leCongress liphakamisa ukutshintshwa kwe-MARC21 , yintoni ibonelela ngesiseko sekamva lenkcazo yebhayibhilografi kwaye, kunjalo, ngokusekelwe kwi-RDF.
I-Wikidata idla ngokucatshulwa njengomzekelo weprojekthi eyimpumelelo kwintsimi yeDatha eSebenzisiweyo eVulekileyo - uhlobo loguqulelo olufundeka ngomatshini lweWikipedia, umxholo wayo, ngokungafaniyo neDBPedia, awuveliswanga ngokungeniswa kwi-infoboxes yamanqaku, kodwa yenziwe ngaphezulu okanye ngaphantsi ngesandla (kwaye emva koko iba ngumthombo wolwazi kwiibhokisi zolwazi ezifanayo).
Sikwacebisa ukuba uyijonge abasebenzisi beStardog RDF yokugcina kwiwebhusayithi yeStardog kwicandelo "Abathengi".
Yiba nokuba kunjalo, eGartner "I-Enterprise Taxonomy kunye ne-Ontology Management" ibekwe phakathi kwehla kwintlambo yokuphoxeka kunye nethemba lokufikelela "kwi-plateau yemveliso" kungekho ngaphambili kwiminyaka eyi-10.
IQhagamshela iDatha yeShishini
Uqikelelo, uqikelelo, iingqikelelo...
Ngenxa yomdla wembali, ndidwelise ngezantsi uqikelelo lukaGartner lweminyaka eyahlukeneyo kwitekhnoloji enomdla kuthi.
| Год | Te knoloji | Ingxelo | Isikhundla | Iminyaka ukuya kwithafa |
|---|---|---|---|---|
| 2001 | IWebhu yeSemantic | Iiteknoloji ezihlumayo | Innovation Trigger | 5-10 |
| 2006 | IWebhu yeSemantic yeNkampani | Iiteknoloji ezihlumayo | Incopho yoLindelo Oluvuthelweyo | 5-10 |
| 2012 | IWebhu yeSemantic | Idatha enkulu | Incopho yoLindelo Oluvuthelweyo | > 10 |
| 2015 | Idatha eDityanisiweyo | Uhlalutyo oluPhezulu kunye neNzululwazi yeDatha | Umkhombe Wokuphoxeka | 5-10 |
| 2016 | Ulawulo lwe-Ontology yeShishini | Iiteknoloji ezihlumayo | Umkhombe Wokuphoxeka | > 10 |
| 2018 | Iigrafu zoLwazi | Iiteknoloji ezihlumayo | Innovation Trigger | 5-10 |
Nangona kunjalo, sele ungenile kuye kwavela enye indlela eya phezulu - Ulwazi lweGrafu. Ukuphindukuzalwa okuthile kwenzeka: i-DBMS yegrafu, apho ingqalelo yabasebenzisi kunye neenzame zabaphuhlisi ziye zatshintshwa, phantsi kweempembelelo zezicelo zangaphambili kunye nemikhwa yokugqibela, yaqala ukuthatha iikhonto kunye nokubeka indawo. kwabo bakhuphisana nabo bangaphambili.
Phantse yonke igrafu ye-DBMS ngoku izichaza njengeqonga elifanelekileyo lokwakha "igrafu yolwazi" yenkampani ("idatha edibeneyo" ngamanye amaxesha ithathelwe indawo "idatha edibeneyo"), kodwa amabango anjalo athetheleleka kangakanani?
Oovimba beenkcukacha zegrafu zisasebenza; idatha ekwigrafu ye-DBMS iseyi-silo yedatha efanayo. Izichongi zemitya endaweni yee-URIs zenza umsebenzi wokudibanisa ii-DBMS ezimbini zeegrafu zisengumsebenzi wohlanganiso, ngelixa ukudibanisa iivenkile ezimbini ze-RDF zihlala kwehla ekudibaniseni iigrafu ezimbini ze-RDF. Enye inkalo ye-asemanticity yi-non-reflexivity ye-LPG imodeli yegrafu, eyenza kube nzima ukulawula imethadatha usebenzisa iqonga elifanayo.
Okokugqibela, ii-DBMS zegrafu azinazo ii-injini zokuthelekelela okanye iinjini zokulawula. Iziphumo zeenjini ezinjalo zinokuphinda ziveliswe ngemibuzo enzima, kodwa oku kunokwenzeka nakwi-SQL.
Nangona kunjalo, iinkqubo zokugcina i-RDF ezikhokelayo azinabunzima bokuxhasa imodeli ye-LPG. Eyona ndlela iqinileyo ithathwa njengeyona ecetywayo ngexesha elinye kwi-Blazegraph: imodeli ye-RDF *, edibanisa i-RDF kunye ne-LPG.
ngakumbi
Unokufunda ngakumbi malunga nenkxaso yokugcina i-RDF yemodeli ye-LPG kwinqaku elidlulileyo kwiHabré: . Ndiyathemba ukuba ngenye imini kuya kubhalwa inqaku elahlukileyo malunga neeGrafu zoLwazi kunye neFabric yeDatha. Icandelo lokugqibela, njengoko kulula ukuliqonda, labhalwa ngokukhawuleza, nangona kunjalo, kwiinyanga ezintandathu kamva, yonke into ayicacanga ngakumbi ngale ngcamango.
Iincwadi
- Halpin, H., Monnin, A. (eds.) (2014). Ubunjineli bePhilosofi: Ukubhekisa kwiPhilosofi yeWebhu
- Allemang, D., Hendler, J. (2011) IWebhu yeSemantic ye-Ontologist eSebenzayo (2nd ed.)
- U-Staab, S., Studer, R. (eds.) (2009) Incwadi ye-Ontologies (uhlelo lwesi-2)
- Wood, D. (ed.). (2011) Ukudibanisa iDatha yeShishini
- Keet, M. (2018) Intshayelelo yobunjineli be-Ontology
umthombo: www.habr.com