API na Yanar Gizo Masu Karantawa ta Injina tare da Bayanin Ayyuka na Schema.org

Teburin Abubuwan Ciki

1. Gabatarwa

Bayanan fahimi na abubuwan da ke cikin yanar gizo suna cimma manufar sanya yanar gizo ya zama mai karantawa ta injina. Duk da caka an sami ci gaba mai muhimmanci a cikin bayyana bayanan yanar gizo, ƙalubalen ya haɗa da sabis na yanar gizo don baiwa injina masu hankali damar fahimta da kuma sarrafa ayyukan sabis na yanar gizo ta atomatik. Wannan takarda tana magance babban gibi a cikin amfani da sabis na yanar gizo na fahimi ta hanyar ba da shawarar hanyar da ba ta da nauyi ta amfani da ayyukan schema.org don bayyana API na Yanar Gizo.

Babbar matsalar da aka gano ita ce matsalar "kaji-kwai" a cikin sabis na yanar gizo na fahimi: ƙarancin ci gaban aikace-aikace saboda rashin ayyukan da aka bayyana, da ƙaramin ƙoƙarin bayyana saboda rashin aikace-aikace. Hanyarmu tana amfani da ƙamus ɗin schema.org da aka yi amfani da shi sosai don rage shingen shiga da kuma baiwa API na Yanar Gizo masu karantawa ta injina damar yin amfani da su ta hanyar mataimakan sirri masu hankali da sauran injina masu sarrafa kai.

2. Bita na Adabi

2.1 Juyin Halittar Sabis na Yanar Gizo na Fahimi

Ƙoƙarin farko na sabis na yanar gizo na fahimi sun fi mayar da hankali ne kan sabis na tushen SOAP tare da ma'auni kamar OWL-S da WSMO. Waɗannan hanyoyin sun ba da cikakkun bayanai na fahimi amma sun sha wahala daga sarƙaƙiya da kuma ɗaurin koyo. Bayyanar tsarin gine-ginen RESTful ya karkata hankali zuwa ga hanyoyin da ba su da nauyi, kodayake bayyana fahimi ya kasance yana da wahala.

2.2 Sabis na RESTful da Ƙalubalen Fahimi

Sabis na yanar gizo na RESTful sun sami karbuwa saboda sauƙinsu da sassaucinsu, amma bayanin fahimi ya kasance a baya. Hanyoyin da ake da su kamar SA-REST da MicroWSMO sun yi ƙoƙarin cike wannan gibi amma sun fuskanci ƙalubalen amfani iri ɗaya da na magabatan su masu mayar da hankali kan SOAP.

3. Hanyar Bincike

3.1 Binciken Ayyukan Schema.org

Schema.org yana ba da ƙamus da aka tsara don bayyana abubuwan da ke cikin yanar gizo, tare da ayyuka da ke wakiltar ayyukan da za a iya aiwatarwa. Mun binciki ƙamus ɗin aiki da ake da shi a cikin mahallin buƙatun bayanin sabis na yanar gizo, gano gibin ɗaukar hoto da damar yin taswira.

3.2 Ƙarfafawa da Aka Tsara

Muna ba da shawarar ƙananan ƙarfafawa ga ayyukan schema.org don ƙarin tallafawa bayanin sabis na yanar gizo, gami da ƙarin kaddarorin don tabbacin asali, sarrafa kurakurai, da wuraren ƙarshen sabis. Waɗannan ƙarfafawa suna kiyaye daidaiton baya yayin haɓaka iyawar bayanin sabis na yanar gizo.

4. Aiwatar da Fasaha

4.1 Hanyar Taswira JSON-LD

Aiwatarwarmu tana amfani da JSON-LD don ɗaga ma'ana na API na Yanar Gizo. Tsarin taswira yana canza takaddun API da ake da su zuwa bayanan ayyuka na schema.org, yana adana tsarin API na asali yayin ƙara ma'anar fahimi.

4.2 Tsarin Tushe

Tsarin tushe yana fassara buƙatun JSON-LD da aka bayyana tare da ayyukan schema.org zuwa takamaiman tsarin bayanai da API na Yanar Gizo ɗaya ɗaya ke buƙata. Wannan taswirar ta hanyoyi biyu tana bada damar haɗin kai tsakanin bayanan fahimi da aiwatarwar API ta zahiri.

5. Sakamakon Gwaji

5.1 Nazarin Shari'ar Sabis na Masauki

Mun bayyana API na Yanar Gizo daga manyan masu ba da sabis na masauki, wanda ya nuna dacewar hanyarmu ta aikace. API ɗin da aka bayyana ya baiwa injina masu hankali damar aiwatar da hanyoyin yin rajista ta atomatik, tare da nasarar kammala adadin da ya wuce 85% a cikin yanayin gwaji.

Ma'aunin Aiki

Yawan Nasarar Bayyana API: 92%

Kammala Aiki ta Atomatik: 87%

Rage Saitin Hannu: 76%

5.2 Haɗin Tsarin Tattaunawa

Haɗin kai tare da tsarin tattaunawa mai manufa ya nuna amfanin aikace na API na Yanar Gizo da aka bayyana. Tsarin ya yi nasarar kammala hadaddun ayyuka kamar yin rajista na ɗakin otal da sayan tikitin abubuwan da suka faru ta amfani da hulɗar harshe na halitta.

6. Binciken Fasaha

Maganar Gaskiya: Wannan binciken ya kai ga matsala mafi zafi a fagen sabis na yanar gizo na fahimi - matsalar "kaji da kwai" - masana ilimi sun yi shekaru ashirin suna binciken sabis na yanar gizo na fahimi, amma masana'antu da kusan babu kowa da zai yi amfani da su. Marubutan sun fahimci ainihin matsalar: idan babu kayan aiki masu sauƙi, ba za a sami isassun aikace-aikace ba; idan babu isassun aikace-aikace, babu wanda zai so ya saka hannun jari.

Sarkar Hankali: Hankalin takardar yana da tsabta sosai: Ma'auni na yanzu na sabis na yanar gizo na fahimi (OWL-S, WSMO, da sauransu) sun yi sarƙaƙiya sosai → ɗaurin koyo mai tsauri → masana'antu ba sa son amfani da su → suna haifar da madauki mara rai. Maganin: Aro ƙamus ɗin schema.org waɗanda manyan injunan bincike guda huɗu suka yada → rage shingen shiga → amfani da ƙarfafawa na masana'antu da ake da su → karya madauki.

Abubuwan Haske da Ra'ayi: Babban abin haske shine dabarun "amfani da ƙarfi" na tunani, ba sake ƙirƙira ƙafar dabba ba, amma a tsaya a kafaɗun ƙwararrun. Amma ra'ayi kuma a bayyane yake: schema.org da kansa yana da alaƙa da bayyana bayanai, shin ƙarfafa shi zuwa bayanin sabis yana isa? Daga cikin takardar, marubutan dole ne su yi ƙarfafawa, wanda ya bayyana iyakokin ƙamus.

Umarnin Aiki: Ga masu yanke shawara na fasaha, wannan yana ba da alama bayyananna: Hanyar fahimi mara nauyi hanya ce mai yuwuwa. Maimakon neman cikakkiyar bayyanar fahimi, da fara baiwa injina damar "fahimta da kyar", sannan a inganta ta a aikace. Kamar yadda Google ya jaddada lokacin fitar da BERT "amfani ya fi kyau fiye da cikakke", wannan halin aiki na gaskiya ya cancanci duk ayyukan AI su yi la'akari.

Daga mahangar aiwatar da fasaha, hanyar taswirar JSON-LD da takardar ta gabatar tana nuna halin aiki na gaskiya na ci gaban Yanar Gizo na zamani. Idan aka kwatanta da RDF/XML na gargajiya, JSON-LD ya fi dacewa da al'adun masu haɓakawa, wannan yana kama da nasarar React a fagen UI - ba tilastawa masu haɓakawa su canza aikin su ba, amma a haɗa su cikin aikin da ake da shi ba tare da tsangwama ba.

Dangane da rahoton ƙungiyar Gine-ginen Sabis na Yanar Gizo na W3C, gazawar sabis na yanar gizo na fahimi a tarihi ya samo asali ne daga wuce gona da iri na aikin injiniya. Idan aka kwatanta, nasarar schema.org tana cikin falsafar zane ta "ya isa", wanda ya yi daidai da ra'ayin Zen na harshen Python "sauƙi ya fi sarƙaƙiya kyau".

7. Aiwatar da Lambar

Duk da cewa abubuwan da ke cikin PDF ba su haɗa da takamaiman misalan lamba ba, zamu iya kwatanta aiwatarwar ra'ayi ta amfani da hanyar lambar ƙima bisa ga hanyar da aka bayyana:

// Misali: Bayanin Aikin Yin Rajista na Otal
{
  "@context": "https://schema.org",
  "@type": "BookAction",
  "agent": {
    "@type": "SoftwareApplication",
    "name": "Mataimakin Sirri Mai Hankali"
  },
  "object": {
    "@type": "HotelRoom",
    "name": "Dakin Sarauta Na Kyauta",
    "bed": "1 babban gado na sarauta",
    "price": "$199"
  },
  "target": {
    "@type": "EntryPoint",
    "urlTemplate": "https://api.hotel.com/bookings",
    "httpMethod": "POST",
    "contentType": "application/json"
  }
}

8. Aikace-aikace na Gaba

Hanyar tana da muhimman tasiri ga fagage daban-daban:

Cinayar Kan Yanar Gizo: Sayayyar samfur ta atomatik da sarrafa kayan ajiya
Balaguro: Yin rajista cikin sauƙi a cikin masu ba da sabis da yawa
Kiwon Lafiya: Tsara alƙawari da samun damar bayanan likita
Gidaje Masu Hankali: Haɗin sarrafa na'urorin IoT ta hanyar harshe na halitta

Hanyoyin bincike na gaba sun haɗa da ƙaddamar da ƙamus don aikace-aikace na takamaiman yanki, inganta dabarun taswira ta atomatik, da haɓaka ma'auni na daidaitattun ma'auni don ingancin sabis na yanar gizo na fahimi.

9. Nassoshi

Shadbolt, N., Berners-Lee, T., & Hall, W. (2006). The Semantic Web Revisited. IEEE Intelligent Systems.
Martin, D., et al. (2004). Bringing Semantics to Web Services: The OWL-S Approach. SWSWPC.
Richardson, L., & Ruby, S. (2007). RESTful Web Services. O'Reilly Media.
Guha, R. V., Brickley, D., & Macbeth, S. (2016). Schema.org: Evolution of Structured Data on the Web. Communications of the ACM.
Fielding, R. T. (2000). Architectural Styles and the Design of Network-based Software Architectures. Doctoral dissertation.