{"id":475962,"date":"2023-08-09T07:24:43","date_gmt":"2023-08-09T07:24:43","guid":{"rendered":""},"modified":"2023-09-05T11:11:42","modified_gmt":"2023-09-05T11:11:42","slug":"back-translation","status":"publish","type":"wiki","link":"https:\/\/oneproxy.pro\/tr\/wiki\/back-translation\/","title":{"rendered":"Geri \u00e7eviri"},"content":{"rendered":"

Geri \u00e7eviri, makine \u00e7evirisi modellerini geli\u015ftirmek i\u00e7in kullan\u0131lan g\u00fc\u00e7l\u00fc bir tekniktir. \u00c7evirinin kalitesini ve do\u011frulu\u011funu iyile\u015ftirmek amac\u0131yla bir metnin bir dilden di\u011ferine \u00e7evrilmesini ve ard\u0131ndan orijinal dile geri \u00e7evrilmesini i\u00e7erir. Bu yinelemeli s\u00fcre\u00e7, modelin kendi hatalar\u0131ndan ders almas\u0131n\u0131 ve dil anlama yeteneklerini giderek geli\u015ftirmesini sa\u011flar. Geri \u00e7eviri, do\u011fal dil i\u015flemede temel bir ara\u00e7 olarak ortaya \u00e7\u0131km\u0131\u015f ve dil hizmetleri, yapay zeka ve ileti\u015fim teknolojileri dahil olmak \u00fczere \u00e7e\u015fitli end\u00fcstrilerde uygulama alan\u0131 bulmu\u015ftur.<\/p>\n

Geri \u00e7evirinin k\u00f6keninin tarihi ve ilk s\u00f6z\u00fc.<\/h2>\n

Geri \u00e7eviri kavram\u0131n\u0131n k\u00f6keni, 1950'lerde makine \u00e7evirisindeki ilk geli\u015fmelere kadar uzanabilir. Geri \u00e7eviriden ilk kez Warren Weaver'\u0131n 1949'da yay\u0131nlanan "Mekanik \u00e7evirinin genel sorunu" ba\u015fl\u0131kl\u0131 ara\u015ft\u0131rma makalesinde bahsedilir. Weaver, yabanc\u0131 bir metnin \u0130ngilizceye \u00e7evrilmesini i\u00e7eren "Y\u00f6ntem II" adl\u0131 bir y\u00f6ntem \u00f6nerdi. daha sonra do\u011frulu\u011funu ve asl\u0131na uygunlu\u011funu sa\u011flamak i\u00e7in orijinal dile geri \u00e7eviriyoruz.<\/p>\n

Geri-\u00e7eviri hakk\u0131nda detayl\u0131 bilgi. Geri \u00e7eviri konusunu geni\u015fletme.<\/h2>\n

Geri \u00e7eviri, modern sinirsel makine \u00e7eviri sistemlerinin e\u011fitim hatt\u0131nda \u00f6nemli bir bile\u015fen olarak hizmet eder. S\u00fcre\u00e7, ayn\u0131 metnin iki farkl\u0131 dilde mevcut oldu\u011fu paralel c\u00fcmlelerden olu\u015fan geni\u015f bir veri k\u00fcmesinin toplanmas\u0131yla ba\u015flar. Bu veri k\u00fcmesi, ilk makine \u00e7eviri modelini e\u011fitmek i\u00e7in kullan\u0131l\u0131r. Ancak bu modeller, \u00f6zellikle d\u00fc\u015f\u00fck kaynakl\u0131 diller veya karma\u015f\u0131k c\u00fcmle yap\u0131lar\u0131yla u\u011fra\u015f\u0131rken s\u0131kl\u0131kla hata ve yanl\u0131\u015fl\u0131klardan muzdariptir.<\/p>\n

Bu sorunlar\u0131 \u00e7\u00f6zmek i\u00e7in geri \u00e7eviri kullan\u0131l\u0131r. Ba\u015flang\u0131\u00e7 veri setinden kaynak c\u00fcmlelerin al\u0131nmas\u0131 ve bunlar\u0131n e\u011fitilen model kullan\u0131larak hedef dile \u00e7evrilmesiyle ba\u015flar. Ortaya \u00e7\u0131kan sentetik \u00e7eviriler daha sonra orijinal veri k\u00fcmesiyle birle\u015ftirilir. Art\u0131k model, hem orijinal paralel c\u00fcmleleri hem de bunlara kar\u015f\u0131l\u0131k gelen geri \u00e7evrilmi\u015f versiyonlar\u0131 i\u00e7eren bu geni\u015fletilmi\u015f veri seti \u00fczerinde yeniden e\u011fitiliyor. Bu yinelemeli s\u00fcre\u00e7 arac\u0131l\u0131\u011f\u0131yla model, parametrelerine ince ayar yapar ve dil anlay\u0131\u015f\u0131n\u0131 geli\u015ftirir; bu da \u00e7eviri kalitesinde \u00f6nemli iyile\u015fmelere yol a\u00e7ar.<\/p>\n

Geri \u00e7evirinin i\u00e7 yap\u0131s\u0131. Geri \u00e7eviri nas\u0131l \u00e7al\u0131\u015f\u0131r?<\/h2>\n

Geri \u00e7eviri s\u00fcreci birka\u00e7 \u00f6nemli ad\u0131m\u0131 i\u00e7erir:<\/p>\n

    \n
  1. \n

    \u0130lk Model E\u011fitimi<\/strong>: Bir sinir makinesi \u00e7eviri modeli, kaynak c\u00fcmlelerden ve bunlar\u0131n \u00e7evirilerinden olu\u015fan paralel bir derlem \u00fczerinde e\u011fitilir.<\/p>\n<\/li>\n

  2. \n

    Sentetik Veri \u00dcretimi<\/strong>: E\u011fitim veri k\u00fcmesindeki kaynak c\u00fcmleler, ba\u015flang\u0131\u00e7 modeli kullan\u0131larak hedef dile \u00e7evrilir. Bu, kaynak c\u00fcmleleri ve bunlar\u0131n sentetik \u00e7evirilerini i\u00e7eren sentetik bir veri k\u00fcmesi olu\u015fturur.<\/p>\n<\/li>\n

  3. \n

    Veri K\u00fcmesi B\u00fcy\u00fctme<\/strong>: Sentetik veri k\u00fcmesi, orijinal paralel derlemeyle birle\u015ftirilerek hem ger\u00e7ek hem de sentetik \u00e7evirileri i\u00e7eren geni\u015fletilmi\u015f bir veri k\u00fcmesi olu\u015fturulur.<\/p>\n<\/li>\n

  4. \n

    Modelin Yeniden E\u011fitimi<\/strong>: Geni\u015fletilmi\u015f veri k\u00fcmesi, \u00e7eviri modelini yeniden e\u011fitmek ve parametrelerini yeni verilere daha iyi uyum sa\u011flayacak \u015fekilde ayarlamak i\u00e7in kullan\u0131l\u0131r.<\/p>\n<\/li>\n

  5. \n

    Yinelemeli \u0130yile\u015ftirme<\/strong>: 2'den 4'e kadar olan ad\u0131mlar birden fazla yineleme i\u00e7in tekrarlan\u0131r ve her seferinde modelin performans\u0131 kendi \u00e7evirilerinden \u00f6\u011frenilerek geli\u015ftirilir.<\/p>\n<\/li>\n<\/ol>\n

    Geri \u00e7evirinin temel \u00f6zelliklerinin analizi.<\/h2>\n

    Geri \u00e7eviri, onu makine \u00e7evirisini geli\u015ftirmek i\u00e7in g\u00fc\u00e7l\u00fc bir teknik haline getiren \u00e7e\u015fitli temel \u00f6zellikler sergiler:<\/p>\n

      \n
    1. \n

      Veri Artt\u0131rma<\/strong>: Geri \u00e7eviri, sentetik \u00e7eviriler \u00fcreterek e\u011fitim veri k\u00fcmesinin boyutunu ve \u00e7e\u015fitlili\u011fini art\u0131r\u0131r; bu da a\u015f\u0131r\u0131 uyumun azalt\u0131lmas\u0131na ve genellemenin iyile\u015ftirilmesine yard\u0131mc\u0131 olur.<\/p>\n<\/li>\n

    2. \n

      Yinelemeli \u0130yile\u015ftirme<\/strong>: Geri \u00e7evirinin yinelemeli do\u011fas\u0131, modelin hatalar\u0131ndan ders almas\u0131na ve \u00e7eviri yeteneklerini a\u015famal\u0131 olarak geli\u015ftirmesine olanak tan\u0131r.<\/p>\n<\/li>\n

    3. \n

      D\u00fc\u015f\u00fck Kaynakl\u0131 Diller<\/strong>: Geri \u00e7eviri, ek e\u011fitim \u00f6rnekleri olu\u015fturmak i\u00e7in tek dilli verilerden yararland\u0131\u011f\u0131ndan, s\u0131n\u0131rl\u0131 paralel verilere sahip diller i\u00e7in \u00f6zellikle etkilidir.<\/p>\n<\/li>\n

    4. \n

      Etki Alan\u0131 Uyarlamas\u0131<\/strong>: Sentetik \u00e7eviriler, modele belirli alanlar veya stiller i\u00e7in ince ayar yapmak i\u00e7in kullan\u0131labilir ve b\u00f6ylece \u00f6zel ba\u011flamlarda daha iyi \u00e7eviri yap\u0131labilir.<\/p>\n<\/li>\n<\/ol>\n

      Geri \u00c7eviri T\u00fcrleri<\/h2>\n

      Geri \u00e7eviri, b\u00fcy\u00fctme i\u00e7in kullan\u0131lan veri k\u00fcmesi t\u00fcrlerine g\u00f6re kategorize edilebilir:<\/p>\n\n\n\n\n\n\n\n
      Tip<\/th>\nTan\u0131m<\/th>\n<\/tr>\n<\/thead>\n
      Tek Dilli Geri \u00c7eviri<\/td>\nArtt\u0131rma i\u00e7in hedef dildeki tek dilli verileri kullan\u0131r. Bu, d\u00fc\u015f\u00fck kaynakl\u0131 diller i\u00e7in kullan\u0131\u015fl\u0131d\u0131r.<\/td>\n<\/tr>\n
      \u0130ki Dilli Geri \u00c7eviri<\/td>\nKaynak c\u00fcmlelerin birden \u00e7ok hedef dile \u00e7evrilmesini i\u00e7erir, b\u00f6ylece \u00e7ok dilli bir model elde edilir.<\/td>\n<\/tr>\n
      Paralel Geri \u00c7eviri<\/td>\nParalel veri k\u00fcmesini geni\u015fletmek ve \u00e7eviri kalitesini art\u0131rmak i\u00e7in birden fazla modelden alternatif \u00e7eviriler kullan\u0131r.<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

      Geri \u00e7eviriyi kullanma yollar\u0131, kullan\u0131ma ili\u015fkin sorunlar ve \u00e7\u00f6z\u00fcmleri.<\/h2>\n

      Geri \u00e7eviriyi kullanma yollar\u0131:<\/h3>\n
        \n
      1. \n

        \u00c7eviri Kalitesinin Art\u0131r\u0131lmas\u0131<\/strong>: Geri \u00e7eviri, makine \u00e7evirisi modellerinin kalitesini ve ak\u0131c\u0131l\u0131\u011f\u0131n\u0131 \u00f6nemli \u00f6l\u00e7\u00fcde art\u0131rarak onlar\u0131 \u00e7e\u015fitli uygulamalarda daha g\u00fcvenilir hale getirir.<\/p>\n<\/li>\n

      2. \n

        Dil Deste\u011fi Geni\u015fletme<\/strong>: Makine \u00e7evirisi modelleri, Geri \u00e7eviriyi dahil ederek, d\u00fc\u015f\u00fck kaynakl\u0131 olanlar da dahil olmak \u00fczere daha geni\u015f bir dil yelpazesi i\u00e7in destek sunabilir.<\/p>\n<\/li>\n

      3. \n

        Alan Adlar\u0131 i\u00e7in \u00d6zelle\u015ftirme<\/strong>: Geri \u00e7eviri taraf\u0131ndan olu\u015fturulan sentetik \u00e7eviriler, do\u011fru ve ba\u011flama duyarl\u0131 \u00e7eviriler sa\u011flamak amac\u0131yla hukuki, t\u0131bbi veya teknik gibi belirli alanlara \u00f6zelle\u015ftirilebilir.<\/p>\n<\/li>\n<\/ol>\n

        Sorunlar ve \u00c7\u00f6z\u00fcmler:<\/h3>\n
          \n
        1. \n

          Tek Dilli Verilere A\u015f\u0131r\u0131 G\u00fcvenmek<\/strong>: Tek Dilde Geri \u00c7eviri kullan\u0131ld\u0131\u011f\u0131nda, sentetik \u00e7eviriler do\u011fru de\u011filse hata olu\u015fma riski vard\u0131r. Bu durum, hedef dil i\u00e7in g\u00fcvenilir dil modelleri kullan\u0131larak azalt\u0131labilir.<\/p>\n<\/li>\n

        2. \n

          Etki Alan\u0131 Uyu\u015fmazl\u0131\u011f\u0131<\/strong>: Paralel Geri \u00c7eviride birden fazla modelden gelen \u00e7eviriler birbiriyle uyumlu de\u011filse tutars\u0131z ve g\u00fcr\u00fclt\u00fcl\u00fc verilere yol a\u00e7abilir. Bir \u00e7\u00f6z\u00fcm, daha y\u00fcksek do\u011fruluk i\u00e7in birden fazla \u00e7eviriyi birle\u015ftirmek amac\u0131yla topluluk y\u00f6ntemlerini kullanmakt\u0131r.<\/p>\n<\/li>\n

        3. \n

          Hesaplamal\u0131 Kaynaklar<\/strong>: Geri \u00e7eviri, \u00f6zellikle modeli yinelemeli olarak e\u011fitirken \u00f6nemli miktarda hesaplama g\u00fcc\u00fc gerektirir. Bu zorluk, da\u011f\u0131t\u0131lm\u0131\u015f bilgi i\u015flem veya bulut tabanl\u0131 hizmetler kullan\u0131larak a\u015f\u0131labilir.<\/p>\n<\/li>\n<\/ol>\n

          Ana \u00f6zellikler ve benzer terimlerle di\u011fer kar\u015f\u0131la\u015ft\u0131rmalar tablo ve liste \u015feklinde.<\/h2>\n\n\n\n\n\n\n\n\n
          karakteristik<\/th>\nGeri-\u00c7eviri<\/th>\n\u0130leri \u00c7eviri<\/th>\nMakine \u00c7evirisi<\/th>\n<\/tr>\n<\/thead>\n
          Yinelemeli \u00d6\u011frenme<\/td>\nEvet<\/td>\nHAYIR<\/td>\nHAYIR<\/td>\n<\/tr>\n
          Veri K\u00fcmesi B\u00fcy\u00fctme<\/td>\nEvet<\/td>\nHAYIR<\/td>\nHAYIR<\/td>\n<\/tr>\n
          Dil Deste\u011fi Geni\u015fletme<\/td>\nEvet<\/td>\nHAYIR<\/td>\nEvet<\/td>\n<\/tr>\n
          Etki Alan\u0131 Uyarlamas\u0131<\/td>\nEvet<\/td>\nHAYIR<\/td>\nEvet<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

          Geriye \u00e7eviri ile ilgili gelece\u011fin perspektifleri ve teknolojileri.<\/h2>\n

          Geri \u00e7eviri, do\u011fal dil i\u015fleme ve makine \u00e7evirisi alan\u0131nda aktif bir ara\u015ft\u0131rma alan\u0131 olmaya devam ediyor. Gelecekteki baz\u0131 potansiyel geli\u015fmeler ve teknolojiler \u015funlar\u0131 i\u00e7erir:<\/p>\n

            \n
          1. \n

            \u00c7ok Dilli Geri \u00c7eviri<\/strong>: Geri \u00e7evirinin birden fazla kaynak ve hedef dille ayn\u0131 anda \u00e7al\u0131\u015facak \u015fekilde geni\u015fletilmesi, daha \u00e7ok y\u00f6nl\u00fc ve etkili \u00e7eviri modellerinin ortaya \u00e7\u0131kmas\u0131n\u0131 sa\u011flar.<\/p>\n<\/li>\n

          2. \n

            S\u0131f\u0131r At\u0131ml\u0131 ve Az At\u0131ml\u0131 \u00d6\u011frenme<\/strong>: S\u0131n\u0131rl\u0131 kaynaklara sahip diller i\u00e7in daha iyi \u00e7eviri sa\u011flamak \u00fczere minimum d\u00fczeyde paralel veri kullanarak veya hi\u00e7 paralel veri kullanmadan \u00e7eviri modellerini e\u011fitmek i\u00e7in teknikler geli\u015ftirmek.<\/p>\n<\/li>\n

          3. \n

            Ba\u011flama duyarl\u0131 geri \u00e7eviri<\/strong>: \u00c7eviri tutarl\u0131l\u0131\u011f\u0131n\u0131 ve ba\u011flam korumas\u0131n\u0131 geli\u015ftirmek i\u00e7in Geri \u00c7eviri s\u00fcreci s\u0131ras\u0131nda ba\u011flam ve s\u00f6ylem bilgilerinin birle\u015ftirilmesi.<\/p>\n<\/li>\n<\/ol>\n

            Proxy sunucular\u0131 nas\u0131l kullan\u0131labilir veya Geri \u00c7eviri ile nas\u0131l ili\u015fkilendirilebilir?<\/h2>\n

            Proxy sunucular\u0131, \u00e7e\u015fitli ve co\u011frafi olarak da\u011f\u0131t\u0131lm\u0131\u015f tek dilli verilere eri\u015fimi kolayla\u015ft\u0131rarak Geri \u00c7eviride \u00e7ok \u00f6nemli bir rol oynayabilir. Geri \u00e7eviri genellikle b\u00fcy\u00fck miktarlarda hedef dil verilerinin toplanmas\u0131n\u0131 i\u00e7erdi\u011finden, \u00e7e\u015fitli b\u00f6lgelerden web sitelerini, forumlar\u0131 ve \u00e7evrimi\u00e7i kaynaklar\u0131 s\u0131y\u0131rmak ve b\u00f6ylece e\u011fitim i\u00e7in veri k\u00fcmesini zenginle\u015ftirmek i\u00e7in proxy sunucular kullan\u0131labilir.<\/p>\n

            Ayr\u0131ca proxy sunucular, dil engellerinin a\u015f\u0131lmas\u0131na ve belirli dillerin daha yayg\u0131n olabilece\u011fi belirli b\u00f6lgelerdeki i\u00e7eri\u011fe eri\u015fmeye yard\u0131mc\u0131 olabilir. Bu eri\u015filebilirlik, do\u011fru sentetik \u00e7evirilerin olu\u015fturulmas\u0131n\u0131 geli\u015ftirebilir ve makine \u00f6\u011frenimi modellerinin genel \u00e7eviri kalitesinin iyile\u015ftirilmesine katk\u0131da bulunabilir.<\/p>\n

            \u0130lgili Ba\u011flant\u0131lar<\/h2>\n

            Geriye \u00e7eviri ve uygulamalar\u0131 hakk\u0131nda daha fazla bilgi i\u00e7in l\u00fctfen a\u015fa\u011f\u0131daki kaynaklara bak\u0131n:<\/p>\n

              \n
            1. Hizalamay\u0131 ve \u00c7evirmeyi Ortakla\u015fa \u00d6\u011frenme yoluyla N\u00f6ral Makine \u00c7evirisi (Bahdanau ve di\u011ferleri, 2014)<\/a><\/li>\n
            2. Google AI Blogu: Google'\u0131n \u00c7ok Dilli N\u00f6ral Makine \u00c7eviri Sistemiyle S\u0131f\u0131r At\u0131\u015fl\u0131 \u00c7eviri<\/a><\/li>\n
            3. OpenAI Blogu: \u00dcretken \u00d6n E\u011fitimle Dil Anlay\u0131\u015f\u0131n\u0131n Geli\u015ftirilmesi (Radford ve di\u011ferleri, 2018)<\/a><\/li>\n
            4. Vikipedi: Geri \u00e7eviri<\/a><\/li>\n<\/ol>\n

              Kurulu\u015flar, Geri \u00c7evirinin g\u00fcc\u00fcnden yararlanarak ve proxy sunucular\u0131n yeteneklerinden yararlanarak daha do\u011fru ve g\u00fcvenilir makine \u00e7evirisi sistemleri elde edebilir ve k\u00fcresel ileti\u015fim ve i\u015fbirli\u011fi i\u00e7in yeni yollar a\u00e7abilir.<\/p>","protected":false},"featured_media":467682,"menu_order":0,"template":"","meta":{"_acf_changed":false,"content-type":"","inline_featured_image":false,"footnotes":""},"class_list":["post-475962","wiki","type-wiki","status-publish","has-post-thumbnail","hentry"],"acf":{"faq_title":"Frequently Asked Questions about Back-Translation: Enhancing Language Translation through Innovation<\/mark>","faq_items":[{"question":"What is Back-translation and how does it work?","answer":"

              Back-translation is a technique used to enhance machine translation models. It involves translating a text from one language to another and then translating it back to the original language. This iterative process helps the model learn from its own mistakes and improves translation quality.<\/p>"},{"question":"Where did Back-translation originate, and when was it first mentioned?","answer":"

              The concept of Back-translation dates back to the 1950s, and it was first mentioned in a research paper by Warren Weaver titled \"The general problem of mechanical translation,\" published in 1949.<\/p>"},{"question":"How does Back-translation improve machine translation?","answer":"

              Back-translation improves machine translation by providing additional training data through synthetic translations. These synthetic translations are generated by translating the source sentences into the target language using the initial model. By incorporating these augmented datasets, the model fine-tunes its parameters and improves its understanding of the language.<\/p>"},{"question":"What are the types of Back-translation?","answer":"

              There are different types of Back-translation based on the datasets used for augmentation:<\/p>

              1. Monolingual Back-translation: Utilizes monolingual data in the target language for augmentation, useful for low-resource languages.<\/li>
              2. Bilingual Back-translation: Involves translating the source sentences into multiple target languages, resulting in a multilingual model.<\/li>
              3. Parallel Back-translation: Uses alternative translations from multiple models to augment the parallel dataset, enhancing translation quality.<\/li><\/ol>"},{"question":"How can Back-translation be used?","answer":"

                Back-translation has various applications, including:<\/p>

                1. Translation Quality Enhancement: It significantly improves the accuracy and fluency of machine translation models.<\/li>
                2. Language Support Expansion: By incorporating Back-translation, machine translation models can support a wider range of languages, including low-resource ones.<\/li>
                3. Customization for Domains: The synthetic translations can be specialized for specific domains, such as legal, medical, or technical, to provide accurate translations.<\/li><\/ol>"},{"question":"What are the challenges associated with Back-translation and their solutions?","answer":"

                  Some challenges and solutions related to Back-translation are:<\/p>

                  1. Over-reliance on Monolingual DatEnsuring accurate synthetic translations from monolingual data by using reliable language models for the target language.<\/li>
                  2. Domain Mismatch: Combining translations from multiple models using ensemble methods to reduce inconsistencies in Parallel Back-translation.<\/li>
                  3. Computational Resources: Addressing the need for substantial computational power through distributed computing or cloud-based services.<\/li><\/ol>"},{"question":"How does Back-translation compare to Forward Translation and Machine Translation?","answer":"
                    Characteristic<\/th>Back-Translation<\/th>Forward Translation<\/th>Machine Translation<\/th><\/tr><\/thead>
                    Iterative Learning<\/td>Yes<\/td>No<\/td>No<\/td><\/tr>
                    Dataset Augmentation<\/td>Yes<\/td>No<\/td>No<\/td><\/tr>
                    Language Support Expansion<\/td>Yes<\/td>No<\/td>Yes<\/td><\/tr>
                    Domain Adaptation<\/td>Yes<\/td>No<\/td>Yes<\/td><\/tr><\/tbody><\/table>"},{"question":"What are the future perspectives of Back-translation?","answer":"

                    The future of Back-translation includes:<\/p>

                    1. Multilingual Back-translation: Extending Back-translation to work with multiple source and target languages simultaneously.<\/li>
                    2. Zero-shot and Few-shot Learning: Training translation models with minimal or no parallel data for languages with limited resources.<\/li>
                    3. Context-aware Back-translation: Incorporating context and discourse information to improve translation coherence and context preservation.<\/li><\/ol>"},{"question":"How are proxy servers associated with Back-translation?","answer":"

                      Proxy servers can aid Back-translation by facilitating access to diverse and geographically distributed monolingual data, enriching the training dataset. They also help in bypassing language barriers and accessing content from specific regions, leading to more accurate synthetic translations and better overall translation quality.<\/p>"}]},"_links":{"self":[{"href":"https:\/\/oneproxy.pro\/tr\/wp-json\/wp\/v2\/wiki\/475962","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/oneproxy.pro\/tr\/wp-json\/wp\/v2\/wiki"}],"about":[{"href":"https:\/\/oneproxy.pro\/tr\/wp-json\/wp\/v2\/types\/wiki"}],"version-history":[{"count":0,"href":"https:\/\/oneproxy.pro\/tr\/wp-json\/wp\/v2\/wiki\/475962\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/oneproxy.pro\/tr\/wp-json\/wp\/v2\/media\/467682"}],"wp:attachment":[{"href":"https:\/\/oneproxy.pro\/tr\/wp-json\/wp\/v2\/media?parent=475962"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}