Machine Translation with NeuralSpace
Introduction
The Internet has become vastly multilingual, which demands content in all kinds of languages now more than ever. Be it translating educational videos, legal documents, or just tweets or comments on social media, there is a high demand for real-time translations as manual translation can simply not cope with such a load and are obviously also extremely expensive.
With the help of Machine Translation, you can reduce the amount of time to translate documents to zero. NeuralSpace’s Machine Translation models are state-of-the-art and can translate text between more than 100 languages covering geographies like India, Africa, south-east Asia, and many more. Users of the NeuralSpace Platform can also fine-tune the out-of-the-box Machine Translation models to their specific use-cases, may it be offering tourists information in their mother tongue or live chats between English-speaking customer support teams and Gujarati-speaking customers.
Features of NeuralSpace’s Machine Translation App
- State-of-the-art Models: Use our pre-trained state-of-the-art Machine Translation models through APIs and integrate them in any software or application
- Language Support: 108 languages supported (Any to Any)
- Train with AutoNLP (coming soon): Train your own use case specific translation models using AutoNLP
AI Modeling life cycle
Just like other Apps on the NeuralSpace Platform, the Machine Translation App will take care of the entire AI modeling lifecycle, which consists of
- Dataset preparation
- Model training
- Model deployment
- Feedback loop
You can upload your existing datasets or create your own in the Data Studio — NeuralSpace’s data preparation and annotation tool which is designed to make dataset creation and modification much faster. Users can, for example, tag specific entities within a sentence and have that phrase or word also tagged in the translated sentence automatically.
Training a custom Machine Translation model using AutoNLP is as easy as clicking on the Train with AutoNLP button once your dataset is uploaded and prepared in the Data Studio. After your training is completed after a couple of minutes, you can place your model in production. NeuralSpace’s in-house developed AutoMLOps feature allows you to use your custom-trained models with throughput rates of up to 30 requests per second. Just click on the Deploy button next to the trained model that achieved the best performance and let AutoMLOps handle the rest for you.
Once deployed, test your models using our interactive model testing and feedback mechanism, by clicking on Test model and Feedback page, respectively. The Feedback page lets you browse through everything that has passed through your models and you can directly add sentences that were translated incorrectly back to your dataset. This will start a feedback-driven learning cycle and you should retrain your models to keep them up to date.
Use-cases
Chatbots
Using our Machine Translation APIs, you can extend your chatbot language support to 108 languages. You just need to create your Natural Language Understanding (NLU) model in a single language of your choice and cater to any of the languages the chatbot’s users speak.
Long Document Translation
If you need entire multi-page reports translated to other languages, you can use our APIs and translate at a click of a button. These machine-translated reports can then be further improved by manual translators.
Government Documents
Some governments require to provide official documents in multiple languages. Machine Translation can ease this process and replace manual translation to a good amount. As with Long Document Translation, these machine-translated documents can then be further improved by manual translators.
Healthcare
Although India is a country known for its diversity in languages, even today, the majority of medical documents like prescriptions and diagnosis results are in English. Our out-of-the-box translation APIs can fasten and smoothen medical procedures for non-English speakers by translating documents in 11 locally spoken languages in India. Given the heavy usage of jargon in medical, the model customization with AutoNLP is highly recommended for this use case.
Gaming
With Machine Translation you can offer users a gaming experience in more than 100 languages along with content moderation using our Language Understanding App.
EdTech
Instructors of online education platforms can only create content in one language however, learners from all around the world should benefit from these resources. Use Speech to Text together with Machine Translation and Text to Speech to create content in more than 100 languages. Let students interact with your platform by voice (using Speech to Text) and automatically mark them using our Language Understanding App.
Language Support
Indian Subcontinent
Hindi (hi)
Assamese (as)
Gujarati (gu)
Bengali (bn)
Kannada (kn)
Malayalam (ml)
Marathi (mr)
Nepali (ne)
Odia (Oriya) (or)
Punjabi (pa)
Sindhi (sd)
Sinhala (si)
South East Asia
Burmese (Myanmar) (my)
Indonesian (id)
Malay (ms)
Tagalog/Filipino (tl)
Vietnamese (vi)
Cebuano (ceb)
Javenese (jv)
Loa (lo)
Thai (th)
Hmong (hmn)
Khmer (km)
Sundanese (su)
Urdu (ur)
Middle East Asia
Arabic (Egyptian) (arz)
Arabic (Levantine)(apc)
Arabic (Maghrebi) (ama)
Arabic (Mesopotamian) (acm)
Arabic (Kuwaiti) (akw)
Arabic (Sudanese) (apd)
Arabic (Gulf) (afb)
Arabic (ar)
Persian (fa)
Hebrew (he)
Uighur (ug)
Pashto (ps)
Turkmen (tk)
Rest of Asia
Armenian (hy)
Chinese-Traditional (zh-TW)
Kazakh (kk)
Kurdish (ku)
Russian (ru)
Tatar (tt)
Azerbaijani (az)
Georgian (ka)
Kirghiz (ky)
Kyrgyz (ky)
Tagalog (tl)
Uzbek (uz)
Chinese Simplified (zh-CN)
Japanese (ja)
Korean (ko)
Mongolian (mn)
Tajik (tg)
Africa
Afrikaans (af)
French (fr)
Kinyarwanda (rw)
Sesotho (st)
Swahili (sw)
Zulu (zu)
Amharic (am)
Hausa (ha)
Malagasy (mg)
Shona (sh)
Xhosa (xh)
English (en)
Igbo (ig)
Nyanja (Chicheva) (ny)
Somali (so)
Yoruba (yo)
Europe
Albanian (sq)
Basque (eu)
Breton (br)
Chechen (ce)
Croatian (hr)
Dutch (nl)
Estonian (et)
Frisian (fy)
Greek (el)
Irish (ga)
Latvian (lv)
Macedonian (mk)
Occitan (oc)
Romanian (ro)
Slovak (sk)
Swedish (sv)
Welsh (cy)
Aragonese (an)
Belarusian (be)
Bulgarian (bg)
Chuvash (cv)
Czech (cs)
English (en)
Finnish (fi)
Galician (gl)
Hungarian (hu)
Italian (it)
Lithuanian (lt)
Maltese (mt)
Polish (pl)
Scots Gaelic (gd)
Slovenian (sl)
Turkish (tr)
Yiddish (yi)
Bashkir (ba)
Bosnian (bs)
Catalan (ca)
Corsican (co)
Danish (da)
Esperanto (eo)
French (fr)
German (de)
Icelandic (is)
Latin (la)
Luxembourgish (lb)
Norwegian Bokmål (nb)
Portuguese (pt)
Serbian (sr)
Spanish (es)
Ukrainian (uk)
North South America
Dutch (nl)
Haitian (ht)
Samoan (sm)
English (en)
Hawaiian (haw)
Spanish (es)
French (fr)
Portuguese (pt)
Yiddish (yi)
Australia/New-Zealand
English (en)
Maori (mi)
Learn how to use NeuralSpace’s Machine Translation APIs with our Get Started guide.
The NeuralSpace Platform is live, test and try it out by yourself! Early sign-ups get $500 worth of credits — what are you waiting for?
Join the NeuralSpace Slack Community to connect with us. Also, receive updates and discuss topics in NLP for low-resource languages with fellow developers and researchers.
Check out our Documentation to read more about the NeuralSpace Platform and its different Apps.
Happy NLP!
