Ndikufuna kuwonetsa kwa anthu kachidutswa ka buku lomwe lasindikizidwa posachedwapa:
Ontological modeling ya bizinesi: njira ndi matekinoloje [Text]: monograph / [S. V. Gorshkov, S. S. Kralin, O. I. Mushtak ndi ena; Mkonzi wamkulu S. V. Gorshkov]. - Ekaterinburg: Ural University Publishing House, 2019. - 234 p.: kudwala., tebulo; 20 cm. - Wolemba. chosonyezedwa kumbuyo kwa mawere. Ndi. - Zolemba za mabuku kumapeto kwa ch. - ISBN 978-5-7996-2580-1: 200 makope.
Cholinga choyika chidutswachi pa Habré ndi kanayi:
- N’zokayikitsa kuti aliyense angathe kugwira bukuli m’manja mwawo ngati sali kasitomala wa anthu olemekezeka ; Sizikugulitsidwa.
- Zowongolera zapangidwa palembalo (sizinatchulidwe pansipa) ndipo zowonjezera zapangidwa zomwe sizikugwirizana kwambiri ndi mawonekedwe a monograph yosindikizidwa: zolemba zam'mutu (pansi pa owononga) ndi ma hyperlink.
- ndikufuna ku sonkhanitsani mafunso ndi ndemanga, n’cholinga choti tiziwaganizira polemba lemba limeneli m’mabuku ena alionse.
- Otsatira ambiri a Semantic Web ndi Linked Data amakhulupirirabe kuti bwalo lawo ndi lopapatiza, makamaka chifukwa anthu ambiri sanafotokozedwe bwino momwe kulili kwakukulu kukhala wotsatira wa Semantic Web ndi Data Yogwirizana. Wolemba kachidutswa, ngakhale kuti ndi wa bwalo ili, alibe lingaliro ili, koma, komabe, amadziona kuti ali ndi udindo woyesanso.
Ndipo kotero,
Webusaiti ya Semantic
Kusintha kwa intaneti kumatha kuyimiridwa motere (kapena kuyankhula za magawo ake omwe adapangidwa motere):
- Zolemba pa intaneti. Tekinoloje yofunika - Gopher, FTP, ndi zina.
Intaneti ndi njira yapadziko lonse yosinthira zinthu zapadziko lonse lapansi. - Zolemba pa intaneti. Ukadaulo wofunikira ndi HTML ndi HTTP.
Mkhalidwe wazinthu zomwe zikuwonekera zimatengera mawonekedwe a sing'anga yopatsirana. - Zambiri pa intaneti. Tekinoloje zazikulu - REST ndi SOAP API, XHR, ndi zina.
Nthawi yogwiritsira ntchito intaneti, sikuti anthu amangokhala ogula zinthu. - Zambiri pa intaneti. Ukadaulo wofunikira ndi matekinoloje a Linked Data.
Gawo lachinayi ili, lonenedweratu ndi Berners-Lee, mlengi wa umisiri wachiwiri wapakatikati ndi wotsogolera W3C, amatchedwa Semantic Web; Matekinoloje olumikizana a Data adapangidwa kuti apangitse deta pa intaneti osati kuwerengeka ndi makina, komanso "makina omveka."
Kuchokera pazotsatirazi, owerenga amvetsetsa kulumikizana pakati pa mfundo zazikuluzikulu za gawo lachiwiri ndi lachinayi:
- Ma URL amafanana ndi ma URI,
- analogue ya HTML ndi RDF,
- Ma hyperlink a HTML ndi ofanana ndi zochitika za URI m'malemba a RDF.
Ukonde wa Semantic ndiwowoneka bwino kwambiri wamtsogolo pa intaneti kuposa momwe zimakhalira zokha kapena zokopa, ngakhale zitha kuganizira izi. Mwachitsanzo, chinthu chofunikira kwambiri chomwe chimatchedwa Web 2.0 chimatengedwa kuti ndi "zopangidwa ndi ogwiritsa ntchito." Makamaka, malingaliro a W3C akufunika kuti aganizire ""ndi ntchito ngati .
Kodi Semantic Web Yakufa?
Ngati mukukana , mkhalidwe wa ukonde wa semantic uli pafupifupi wofanana ndi wa chikominisi m’nthaŵi za chitukuko cha sosholizimu (ndipo ngati kukhulupirika ku malamulo a Ilyich kumawonedwa, aliyense adzisankhire yekha). Makina osakira kukakamiza mawebusayiti kugwiritsa ntchito RDFa ndi JSON-LD ndipo iwo eni amagwiritsa ntchito matekinoloje okhudzana ndi omwe afotokozedwa pansipa (Google Knowledge Graph, Bing Knowledge Graph).
M'mawu ambiri, wolemba sangathe kunena zomwe zimalepheretsa kufalikira kwakukulu, koma akhoza kuyankhula pazidziwitso zaumwini. Pali mavuto omwe angathetsedwe "kunja kwa bokosi" muzochitika za SW zonyansa, ngakhale kuti sizikufalikira kwambiri. Chotsatira chake, iwo omwe akukumana ndi ntchitozi alibe njira zowakakamiza iwo omwe angathe kupereka yankho, pamene kuperekedwa kwapadera kwapadera kwaposachedwa kumatsutsana ndi machitidwe awo amalonda. Chifukwa chake tikupitiliza kugawa HTML ndikumata ma API osiyanasiyana, wina ndi mnzake.
Komabe, matekinoloje olumikizana ndi Data adafalikira kupitilira Webusaiti yayikulu; Bukuli, kwenikweni, laperekedwa kuzinthu izi. Pakadali pano, gulu la Linked Data likuyembekeza kuti matekinolojewa afalikira kwambiri chifukwa cha kujambula kwa Gartner (kapena kulengeza, monga momwe mukufunira) Zithunzi Zojambula и Data Fabric. Ndikufuna kukhulupirira kuti sikukhala "njinga" kukhazikitsidwa kwa mfundozi zomwe zidzapambane, koma zokhudzana ndi miyezo ya W3C yomwe yafotokozedwa pansipa.
Data Yogwirizana
Berners-Lee adatanthauzira Deta Yolumikizidwa ngati tsamba la semantic "lochita bwino": njira zingapo ndi matekinoloje omwe amalola kuti akwaniritse zolinga zake zazikulu. Mfundo zoyambira za Linked Data Berners-Lee zotsatirazi.
Mfundo 1. Kugwiritsa ntchito ma URI kutchula mabungwe.
Ma URI ndi zizindikiritso zapadziko lonse lapansi kusiyana ndi zozindikiritsa zingwe zam'deralo pazolemba. Pambuyo pake, mfundoyi inafotokozedwa bwino kwambiri mu mawu oti Google Knowledge Graph "".
Mfundo 2. Kugwiritsa ntchito ma URI mu HTTP scheme kuti athe kuchotsedwa.
Potchula URI, ziyenera kukhala zotheka kupeza choyimira kumbuyo kwa chizindikirocho (chifaniziro ndi dzina la wogwiritsa ntchito " chikuwonekera apa).*"mu C); mwatsatanetsatane, kuti mupeze chithunzithunzi cha izi - kutengera mtengo wamutu wa HTTP Accept:. Mwina, ndikubwera kwa nthawi ya AR / VR, kudzakhala kotheka kupeza gwero lokha, koma pakali pano, mwinamwake, chidzakhala chikalata cha RDF, chomwe ndi zotsatira za kuyankha funso la SPARQL. DESCRIBE.
Mfundo 3. Kugwiritsa ntchito miyezo ya W3C - makamaka RDF(S) ndi SPARQL - makamaka pochotsa ma URI.
"Zigawo" zamtundu waukadaulo wa Linked Data, womwe umadziwikanso kuti , zidzafotokozedwa pansipa.
Mfundo 4. Kugwiritsa ntchito zolozera ku ma URI ena pofotokozera mabungwe.
RDF imakupatsani mwayi woti mungofotokoza zachidziwitso chachidziwitso chachilengedwe, ndipo mfundo yachinayi imayitanitsa kuti musachite izi. Ngati mfundo yoyamba ikuwoneka padziko lonse lapansi, zimakhala zotheka pofotokoza gwero kuti litchule ena, kuphatikizapo "akunja", chifukwa chake deta imatchedwa ogwirizana. M'malo mwake, ndizosapeweka kugwiritsa ntchito ma URI otchulidwa m'mawu a RDFS.
RDF
(Resource Description Framework) ndi njira yofotokozera mabungwe ogwirizana.
Mawu amtundu wa "subject-predicate-object", otchedwa triplets, amapangidwa ponena za mabungwe ndi maubwenzi awo. Mwachidule, mutu, predicate, ndi chinthu zonse ndi ma URI. URI yofananayo imatha kukhala m'malo osiyanasiyana m'magawo atatu: kukhala mutu, predicate, ndi chinthu; Chifukwa chake, atatuwa amapanga mtundu wa graph wotchedwa RDF graph.
Mitu ndi zinthu sizingakhale ma URI okha, komanso otchedwa mfundo zopanda kanthu, ndi zinthu zikhoza kukhala zenizeni. Literals ndi zitsanzo za mitundu yakale yokhala ndi chiwonetsero cha zingwe ndi chizindikiro cha mtundu.
Zitsanzo zolembera mawu enieni (mu Turtle syntax, zambiri za izo pansipa): "5.0"^^xsd:float и "five"^^xsd:string. Zolemba ndi mtundu rdf:langString imathanso kukhala ndi chizindikiro cha chilankhulo; mu Kamba zidalembedwa motere: "five"@en и "пять"@ru.
Node zopanda kanthu ndizinthu "zosadziwika" zopanda zizindikiritso zapadziko lonse lapansi, zomwe mawu angatchulidwe; mitundu ya zinthu zomwe zilipo.
Chifukwa chake (izi ndiye mfundo yonse ya RDF):
- mutu ndi URI kapena node yopanda kanthu,
- Mchitidwewu ndi URI,
- chinthu ndi URI, node yopanda kanthu, kapena yeniyeni.
Chifukwa chiyani ma predicates sangakhale opanda kanthu?
Mwina chifukwa chake ndi kufuna kumvetsetsa mwamwayi ndikumasulira katatu m'chilankhulo cha predicate logic. s p o ngati chinachake 
kumene - predicate, и - zokhazikika. Zotsatira za kumvetsetsa izi zili mu chikalata "", yomwe ili ndi chidziwitso cha gulu la W3C. Ndi kumvetsetsa uku, katatu s p []kumene [] - node yopanda kanthu, idzamasuliridwa ngati kumene - zosinthika, koma momwe mungamasulire s [] o? Chikalata chokhala ndi W3C Recommendation status "” akupereka njira ina yomasulira, komabe samaganizira za kuthekera kwa maulosi kukhala ma node opanda kanthu.
Komabe, Manu Sporni .
RDF ndi chitsanzo chosawerengeka. RDF imatha kulembedwa (yotsatiridwa) m'mawu osiyanasiyana: , (zowerengeka kwambiri za anthu), , (binary).
RDF yomweyi imatha kusinthidwa kukhala RDF/XML m'njira zosiyanasiyana, kotero, mwachitsanzo, sizomveka kutsimikizira XML yomwe ikubwera pogwiritsa ntchito XSD kapena kuyesa kuchotsa deta pogwiritsa ntchito XPath. Momwemonso, JSON-LD ndiyokayikitsa kuti akwaniritse chikhumbo chaopanga Javascript chofuna kugwira ntchito ndi RDF pogwiritsa ntchito kadontho ka Javascript ndi masikweya-bracket notation (ngakhale JSON-LD imasunthira komweko popereka makina ).
Ma syntaxes ambiri amapereka njira zofupikitsira ma URI aatali. Mwachitsanzo, malonda @prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> mu Kamba amakulolani kuti mulembe m'malo mwake <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> basi rdf:type.
RDFS
(RDF Schema) - mawu oyambira achitsanzo, amawonetsa malingaliro a katundu ndi gulu ndi katundu monga rdf:type, rdfs:subClassOf, rdfs:domain и rdfs:range. Pogwiritsa ntchito dikishonale ya RDFS, mwachitsanzo, mawu otsatirawa amatha kulembedwa:
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 .RDFS ndi kufotokozera ndi mawu achitsanzo, koma si chilankhulo choletsa (ngakhale chilankhulo chovomerezeka ndi kuthekera kogwiritsa ntchito kotero). Mawu oti "Schema" sayenera kumveka mofanana ndi mawu akuti "XML Schema". Mwachitsanzo, :author rdfs:range foaf:Person zikutanthauza kuti rdf:type mitengo yonse ya katundu :author - foaf:Person, koma sizikutanthauza kuti zimenezi ziyenera kunenedwa pasadakhale.
Mtengo wa magawo SPARQL
(SPARQL Protocol and RDF Query Language) - chilankhulo chofunsira deta ya RDF. Mwachidule, funso la SPARQL ndi zitsanzo zomwe magawo atatu a graph omwe akufunsidwa amafananizidwa. Mapangidwe amatha kukhala ndi zosintha mu mutu, predicate, ndi malo a chinthu.
Funsolo libweza zosinthika kotero kuti, zikalowa m'malo mwa zitsanzo, zitha kubweretsa gawo laling'ono la graph ya RDF (kagawo kakang'ono ka katatu). Zosintha za dzina lomwelo muzitsanzo zosiyanasiyana za mapatatu atatu ziyenera kukhala zofananira.
Mwachitsanzo, popatsidwa zomwe zili pamwambapa za ma axioms asanu ndi awiri a RDFS, funso lotsatirali libwerera rdfs:domain и rdfs:range monga makhalidwe ?s и ?p motero:
SELECT * WHERE {
?s ?p rdfs:Class .
?p ?p rdf:Property .
}Ndizofunikira kudziwa kuti SPARQL ndiyofotokozera ndipo sichilankhulo chofotokozera kusuntha kwa ma graph (komabe, nkhokwe zina za RDF zimapereka njira zosinthira zomwe amafunsa). Chifukwa chake, zovuta zina zama graph, mwachitsanzo, kupeza njira yayifupi kwambiri, sizingathetsedwe mu SPARQL, kuphatikiza kugwiritsa ntchito (koma, kachiwiri, nkhokwe za RDF payokha zimapereka zowonjezera zapadera kuti athetse mavutowa).
SPARQL samagawana malingaliro a kutseguka kwa dziko lapansi ndipo amatsatira njira ya "kukana ngati kulephera", momwe mapangidwe monga FILTER NOT EXISTS {…}. Kugawa deta kumaganiziridwa pogwiritsa ntchito makinawo .
Malo ofikira a SPARQL - malo osungira a RDF omwe amatha kuyankha mafunso a SPARQL - alibe mafananidwe achindunji kuyambira gawo lachiwiri (onani koyambira kwa ndimeyi). Itha kufananizidwa ndi database, kutengera zomwe masamba a HTML adapangidwa, koma opezeka kunja. Malo ofikira a SPARQL ndi ofanana kwambiri ndi malo ofikira a API kuchokera pagawo lachitatu, koma ndi kusiyana kwakukulu kuwiri. Choyamba, ndizotheka kuphatikiza mafunso angapo a "atomiki" kukhala amodzi (omwe amawonedwa ngati mawonekedwe ofunikira a GraphQL), ndipo kachiwiri, API yotereyi imadzilemba yokha (zomwe ndi zomwe HATEOAS idayesera kukwaniritsa).
Ndemanga yandale
RDF ndi njira yosindikizira deta pa intaneti, kotero kusungirako RDF kuyenera kuonedwa ngati chikalata cha DBMS. Zowona, popeza RDF ndi graph osati mtengo, idakhalanso yotengera ma graph. Ndizodabwitsa kuti zidatheka. Ndani akanaganiza kuti padzakhala anthu anzeru omwe angagwiritse ntchito mfundo zopanda kanthu. Codd ali pano .
Palinso njira zosawoneka bwino zosinthira kupeza deta ya RDF, mwachitsanzo, (LDF) ndi (LDP).
OWL
(Chiyankhulo cha Web Ontology) - mwamwambo woyimira chidziwitso, mtundu wofotokozera wamalingaliro (ponseponse pansipa ndizolondola kunena kuti OWL 2, mtundu woyamba wa OWL udakhazikitsidwa ).
Malingaliro amalingaliro ofotokozera mu OWL amafanana ndi makalasi, maudindo amafanana ndi katundu, anthu amasunga dzina lawo lakale. Axioms amatchedwanso axioms.
Mwachitsanzo, mu otchedwa kwa OWL notation ndi axiom yomwe tikudziwa kale zidzalembedwa motere:
Class: Human
Class: Parent
EquivalentClass: Human and (inverse hasParent) some Human
ObjectProperty: hasParentPalinso ma syntaxes ena olembera OWL, monga , yogwiritsidwa ntchito muzovomerezeka, ndi . Kuphatikiza apo, OWL ikhoza kusinthidwa ndi kupitilira apo - m'mawu aliwonse enieni.
OWL ali ndi ubale wapawiri ndi RDF. Kumbali imodzi, itha kuwonedwa ngati mtundu wa dikishonale womwe umakulitsa RDFS. Kumbali inayi, ndi njira yamphamvu kwambiri yomwe RDF imangokhala mawonekedwe osasintha. Sizinthu zonse zoyambira za OWL zomwe zitha kulembedwa pogwiritsa ntchito katatu kamodzi ka RDF.
Kutengera ndi kagawo kakang'ono ka OWL constructs amaloledwa kugwiritsidwa ntchito, amalankhula zomwe zimatchedwa . Okhazikika komanso otchuka kwambiri ndi OWL EL, OWL RL ndi OWL QL. Kusankhidwa kwa mbiri kumakhudza zovuta zowerengera zamavuto wamba. Gulu lathunthu la OWL limamanga molingana ndi , yotchedwa OWL DL. Nthawi zina amalankhulanso za OWL Full, momwe zopangira za OWL zimaloledwa kugwiritsidwa ntchito ndi ufulu wathunthu wa RDF, popanda zoletsa za semantic ndi computational. . Mwachitsanzo, chinachake chikhoza kukhala kalasi ndi katundu. OWL Full ndi osadziwika.
Mfundo zazikuluzikulu zophatikizira zotsatira mu OWL ndikutengera malingaliro adziko lapansi. ) ndi kukana kuganiziridwa kwa mayina apadera (dzina lapadera la kulingalira, ). Pansipa tiwona komwe mfundozi zingatsogolere ndikuyambitsa zina za OWL.
Lolani ontology ikhale ndi chidutswa chotsatirachi (mu mawu a Manchester):
Class: manyChildren
EquivalentTo: Human that hasChild min 3
Individual: John
Types: Human
Facts: hasChild Alice, hasChild Bob, hasChild CarolKodi zidzatsatira zimene zanenedwa kuti Yohane ali ndi ana ambiri? Kukana UNA kudzakakamiza injini yowunikira kuyankha funsoli molakwika, popeza Alice ndi Bob atha kukhala munthu yemweyo. Kuti zotsatirazi zichitike, m'pofunika kuwonjezera axiom zotsatirazi:
DifferentIndividuals: Alice, Bob, Carol, JohnTiyeni tsopano chidutswa cha ontology chikhale ndi mawonekedwe awa (John akunenedwa kuti ali ndi ana ambiri, koma ali ndi ana awiri okha):
Class: manyChildren
EquivalentTo: Human that hasChild min 3
Individual: John
Types: Human, manyChildren
Facts: hasChild Alice, hasChild Bob
DifferentIndividuals: Alice, Bob, Carol, JohnKodi ontology iyi idzakhala yosagwirizana (zomwe zitha kutanthauziridwa ngati umboni wa data yolakwika)? Kuvomereza OWA kumapangitsa injini yowunikira kuyankha molakwika: "kwina" kwina (mu ontology ina) zitha kunenedwa kuti Carol ndi mwana wa John.
Kuti tipewe kutheka kwa izi, tiyeni tiwonjeze mfundo yatsopano ya John:
Individual: John
Facts: hasChild Alice, hasChild Bob, not hasChild CarolKupatula maonekedwe a ana ena, tiyeni tinene kuti makhalidwe onse a katundu "kukhala ndi mwana" - anthu, amene tili anayi okha:
ObjectProperty: hasChild
Domain: Human
Сharacteristics: Irreflexive
Class: Human
EquivalentTo: { Alice, Bill, Carol, John }Tsopano ontology idzakhala yotsutsana, yomwe injini yowunikira sidzalephera kufotokoza. Ndi mawu omaliza a ma axiom omwe, m'njira inayake, "tatseka" dziko lapansi, ndipo taonani momwe kuthekera kwa Yohane kukhala mwana wake kumachotsedwa.
Kulumikiza Enterprise Data
Njira ndi matekinoloje a Linked Data adapangidwa kuti afalitse deta pa intaneti. Kugwiritsa ntchito kwawo m'malo amakampani amkati kumakumana ndi zovuta zingapo.
Mwachitsanzo, mu malo otsekedwa amakampani, mphamvu yochepetsera ya OWL yochokera ku kukhazikitsidwa kwa OWA ndi kukana kwa UNA, zisankho chifukwa cha kutseguka ndi kugawidwa kwa Webusaiti, ndizofooka kwambiri. Ndipo apa njira zotsatirazi ndizotheka.
- Kupatsa OWL ndi semantics, kutanthauza kusiyidwa kwa OWA ndi kukhazikitsidwa kwa UNA, kukhazikitsidwa kwa injini yofananira. - Panjira iyi Stardog RDF yosungirako.
- Kusiya kuthekera kochotsa kwa OWL m'malo mwa injini zamalamulo. - Stardog imathandizira ; Jena ndi GraphDB amapereka malamulo
- Kukana mphamvu zochepetsera za OWL, kugwiritsa ntchito gawo limodzi kapena lina pafupi ndi RDFS pojambula. - Onani zambiri za izi pansipa.
Nkhani ina ndikuyang'ana kwakukulu komwe makampani amakampani angakhale nawo pazambiri zamtundu wa data komanso kusowa kwa zida zotsimikizira za data mu stack ya Linked Data. Zotuluka pano ndi izi.
- Apanso, gwiritsani ntchito kutsimikizira kwa OWL yomanga ndi semantics yapadziko lonse yotsekedwa ndi mayina apadera ngati injini yoyenera ilipo.
- Gwiritsani ntchito , yokhazikika pambuyo pa mndandanda wa zigawo za Semantic Web Layer Cake zakhazikitsidwa (komabe, zitha kugwiritsidwanso ntchito ngati injini yamalamulo), kapena .
- Kumvetsetsa kuti zonse zimachitidwa ndi mafunso a SPARQL, ndikupanga njira yanu yosavuta yotsimikizira deta pogwiritsa ntchito iwo.
Komabe, ngakhale kukana kotheratu kwa kuthekera kochepetsera ndi zida zotsimikizira kumasiya mulu wa Linked Data kunja kwa mpikisano muzochita zomwe zimafanana ndi mawonekedwe ndi intaneti yotseguka ndi yogawidwa - mu ntchito zophatikizira deta.
Nanga bwanji dongosolo lazambiri zamabizinesi?
Izi ndizotheka, koma muyenera kudziwa ndendende mavuto omwe matekinoloje ofananirako akuyenera kuthana nawo. Ndifotokoza apa momwe amachitira omwe akutenga nawo gawo pachitukuko kuti awonetse momwe ukadaulo uwu umawonekera kuchokera pamalingaliro a IT wamba. Zindikumbutsa pang'ono za fanizo la njovu:
- Katswiri wazamalonda: RDF ndi chinthu chonga mtundu wokhazikika wosungidwa.
- Wosanthula Zamachitidwe: RDF ili ngati , ndi mndandanda wa index ndi chilankhulo chosavuta kufunsa.
- mapulogalamu: Chabwino, izi zonse zili mu mzimu wa malingaliro amtundu wolemera ndi ma code otsika, posachedwapa za izi.
- Woyang'anira ntchito: inde ndi chimodzimodzi !
Zochita zimasonyeza kuti stack imagwiritsidwa ntchito nthawi zambiri pa ntchito zokhudzana ndi kugawa ndi kusiyanasiyana kwa deta, mwachitsanzo, pomanga machitidwe a MDM (Master Data Management) kapena DWH (Data Warehouse). Mavuto oterewa alipo mumakampani aliwonse.
Pankhani yamapulogalamu okhudzana ndi mafakitale, matekinoloje a Linked Data pakadali pano ndiwodziwika kwambiri m'mafakitale otsatirawa.
- matekinoloje a biomedical (komwe kutchuka kwawo kumawoneka kuti kukugwirizana ndi zovuta za domain);
panopa
Bungwe la "Boiling Point" posachedwapa linachititsa msonkhano wokonzedwa ndi bungwe la "National Medical Knowledge Base" ".".
- kupanga ndi kugwiritsa ntchito zinthu zovuta (injiniya yayikulu yamakina, kupanga mafuta ndi gasi; nthawi zambiri timalankhula za muyezo );
panopa
Pano, chifukwa chake ndizovuta za nkhaniyi, pamene, mwachitsanzo, pamtunda wokwera, ngati tikukamba za makampani a mafuta ndi gasi, kuwerengera kosavuta kumafuna ntchito zina za CAD.
Mu 2008, chochitika choyimira choyimira, chokonzedwa ndi Chevron, chinachitika .
ISO 15926, pamapeto pake, idawoneka ngati yolemetsa pamakampani amafuta ndi gasi (ndipo mwina idapezeka kuti ikugwiritsidwa ntchito kwambiri muukadaulo wamakina). Ndi Statoil yokha (Equinor) yomwe idalumikizidwa nayo; ku Norway, yonse . Ena akuyesera kuchita zofuna zawo. Mwachitsanzo, malinga ndi mphekesera, Unduna wa Zamagetsi wapakhomo ukukonzekera kupanga "chitsanzo cha ontological chamafuta ndi mphamvu," chofanana, mwachiwonekere, .
- mabungwe azachuma (ngakhale XBRL itha kuonedwa ngati wosakanizidwa wa SDMX ndi RDF Data Cube ontology);
panopa
Kumayambiriro kwa chaka, LinkedIn idasokoneza wolembayo mwachangu ndi maudindo pafupifupi onse akuluakulu azachuma, omwe amawadziwa pawailesi yakanema "Force Majeure": Goldman Sachs, JPMorgan Chase ndi / kapena Morgan Stanley, Wells Fargo, SWIFT/Visa/Mastercard, Bank of America, Citigroup, Fed, Deutsche Bank... Mwinamwake aliyense anali kufunafuna wina amene angatumizeko. . Ochepa adakwanitsa kupeza: mabungwe azachuma adatenga chilichonse .
Pa HeadHunter, Sberbank yokha idapeza china chake chosangalatsa; zinali za "kusungidwa kwa EAV ndi mtundu wa data wa RDF."
Mwinamwake, kusiyana kwa msinkhu wa chikondi kwa matekinoloje ofanana a mabungwe azachuma apakhomo ndi a Kumadzulo ndi chifukwa cha zochitika zapadziko lonse lapansi. Mwachiwonekere, kuphatikizika kudutsa malire a boma kumafuna njira zosiyanasiyana zamagulu ndi luso.
- machitidwe oyankha mafunso ndi ntchito zamalonda (IBM Watson, Apple Siri, Google Knowledge Graph);
panopa
Mwa njira, mlengi wa Siri, a Thomas Gruber, ndiye mlembi wa tanthauzo lenileni la ontology (m'lingaliro la IT) monga "maganizidwe amalingaliro." Malingaliro anga, kukonzanso mawu mu tanthauzo ili sikusintha tanthauzo lake, zomwe mwina zikuwonetsa kuti palibe.
- kufalitsa deta yokhazikika (ndi zifukwa zazikulu izi zitha kukhala chifukwa cha Linked Open Data).
panopa
Mafani akulu a Linked Data ndi omwe amatchedwa GLAM: Galleries, Library, Archives, and Museums. Ndikokwanira kunena kuti Library of Congress ikulimbikitsa m'malo mwa MARC21 , zomwe amapereka maziko a tsogolo la kufotokoza kwa mabuku ndipo, ndithudi, kutengera RDF.
Wikidata nthawi zambiri imatchulidwa ngati chitsanzo cha polojekiti yopambana m'munda wa Linked Open Data - mtundu wa Wikipedia wowerengeka ndi makina, zomwe, mosiyana ndi DBPedia, sizimapangidwa ndi kuitanitsa kuchokera ku bokosi la infobox, koma ndi adapanga mochulukirapo kapena mochepera pamanja (ndipo kenako amakhala gwero lazidziwitso zamabokosi azithunzi omwewo).
Tikupangiranso kuti mufufuze ogwiritsa ntchito yosungirako Stardog RDF patsamba la Stardog mu gawo la "Makasitomala".
Khalani momwe zingakhalire, ku Gartner "Enterprise Taxonomy and Ontology Management" imayikidwa pakati pa kutsetsereka kwa chigwa cha zokhumudwitsa ndi chiyembekezo chofikira "chitunda chopanga" pasanathe zaka 10.
Kulumikiza Enterprise Data
Zoneneratu, zoneneratu, zoneneratu ...
Chifukwa cha chidwi chambiri, ndalemba m'munsimu zolosera za Gartner kwa zaka zosiyanasiyana pa matekinoloje omwe amatisangalatsa.
| Год | umisiri | Nenani | Udindo | Zaka kupita kumapiri |
|---|---|---|---|---|
| 2001 | Webusaiti ya Semantic | Technologies Akubwera | Innovation Trigger | 5-10 |
| 2006 | Corporate Semantic Web | Technologies Akubwera | Pamwamba pa Zoyembekeza Zowonjezereka | 5-10 |
| 2012 | Webusaiti ya Semantic | Big Data | Pamwamba pa Zoyembekeza Zowonjezereka | > 10 |
| 2015 | Data Yogwirizana | Advanced Analytics ndi Data Science | Mtsinje Wokhumudwa | 5-10 |
| 2016 | Enterprise Ontology Management | Technologies Akubwera | Mtsinje Wokhumudwa | > 10 |
| 2018 | Zithunzi Zojambula | Technologies Akubwera | Innovation Trigger | 5-10 |
Komabe, kale mu njira ina yokwera yawonekera - Zithunzi za Knowledge. Kubadwanso kwina kunachitika: ma graph DBMSs, omwe chidwi cha ogwiritsa ntchito ndi zoyesayesa za opanga zidasinthidwa, motengera zopempha za akale ndi zizolowezi zomaliza, zidayamba kutengera mizere ndi kuyika. za omwe adachita nawo mpikisano.
Pafupifupi ma graph onse a DBMS tsopano amadzinenera kuti ndi nsanja yoyenera yomangira "chidziwitso" chamakampani ("data yolumikizidwa" nthawi zina imasinthidwa ndi "data yolumikizidwa"), koma zonena zotere ndi zolondola bwanji?
Zolemba zazithunzi zikadali zamatsenga; zomwe zili mu graph DBMS zikadali silo yomweyo. Zozindikiritsa zingwe m'malo mwa ma URI zimapanga ntchito yophatikiza ma DBMS awiri a graph kukhalabe ntchito yophatikizira, pomwe kuphatikiza masitolo awiri a RDF nthawi zambiri kumatsika ndikungophatikiza ma graph awiri a RDF. Mbali ina ya asemanticity ndi kusasinthika kwa mtundu wa graph ya LPG, zomwe zimapangitsa kuti zikhale zovuta kuyendetsa metadata pogwiritsa ntchito nsanja yomweyo.
Pomaliza, ma graph DBMS alibe injini zolozera kapena ma injini olamulira. Zotsatira zamainjini oterowo zitha kupangidwanso ndi mafunso ovuta, koma izi ndizotheka ngakhale mu SQL.
Komabe, makina otsogola a RDF alibe zovuta kuthandizira mtundu wa LPG. Njira yolimba kwambiri imatengedwa kuti ndi yomwe idaperekedwa nthawi imodzi ku Blazegraph: mtundu wa RDF *, kuphatikiza RDF ndi LPG.
More
Mutha kuwerenga zambiri za chithandizo chosungira cha RDF cha mtundu wa LPG m'nkhani yapitayi Habré: . Ndikukhulupirira kuti tsiku lina nkhani yosiyana idzalembedwa za Chidziwitso cha Graph ndi Data Fabric. Gawo lomaliza, monga losavuta kumvetsetsa, linalembedwa mofulumira, komabe, ngakhale patapita miyezi isanu ndi umodzi, zonse sizimveka bwino ndi mfundozi.
Mabuku
- Halpin, H., Monnin, A. (eds.) (2014). Ukatswiri Wanzeru: Kufikira Philosophy ya Webusaiti
- Allemang, D., Hendler, J. (2011) Semantic Web for the Working Ontologist (2nd ed.)
- Staab, S., Studer, R. (eds.) (2009) Handbook on Ontologies (2nd ed.)
- Wood, D. (m.). (2011) Kulumikiza Data yamakampani
- Keet, M. (2018) Chiyambi cha Ontology Engineering
Source: www.habr.com