translate essay in kannada language

ಸೆಟ್ಟಿಂಗ್‌ಗಳು

ಪಠ್ಯ ಅನುವಾದ, ಅನುವಾದ ಫಲಿತಾಂಶಗಳು, ಡಾಕ್ಯುಮೆಂಟ್ ಅನುವಾದ.

ವೆಬ್‌ಸೈಟ್ ಅನುವಾದ

URL ನಮೂದಿಸಿ

ಚಿತ್ರದ ಅನುವಾದ

• Grammar Checker
• Text to Speech
• AI Detector
• Bulk Translator
• Word Counter
• Numbers to Words
• Case Converter
• Explore more
• Fast Fast Mode: Quick and reliable for daily translations. The speedy choice that doesn't compromise accuracy.
• Advanced Advanced Mode: Precise translations for business and research. Professional-grade quality you can depend on. Please upgrade to the Pro plan.

Upload or drop document to translate

Max. file size 10 MB

Translate PDF Online

Accurate ai translation for pdf documents.

• 30 Fast Credits/day
• Limit 10 speech translations/day
• Limit 10 AI content detect/day
• Limit 10 text-to-speech/day
• Supports text, document
• Limit 1,500 characters at once
• Upload files up to 10 MB in size
• Unlimited Fast Credits
• Supports text, document, image, speech
• Supports scanned PDF translation i 3 scanned PDF translations/day
• Up to 30,000 characters at once
• Unlimited text-to-speech
• Lightning-Fast translations
• Upload files up to 30 MB in size
• 1v1 Customer Service
• Cancel anytime

Pro 50% OFF

$9.9 /month.

• 30 Advanced Credits/day i Advanced Mode offers precise and professional translation
• Supports scanned PDF translation i 10 scanned PDF translations/day
• Up to 100,000 characters at once
• Upload files up to 100 MB in size

Ultimate 50% OFF

• 100 Advanced Credits/day i Advanced Mode offers precise and professional translation
• Supports scanned PDF translation i 30 scanned PDF translations/day
• Up to 150,000 characters at once

Free users have 30 credits per day, with a limit of 1,500 characters per translation.

How to Translate PDF Document?

Select the target language.

Choose the language into which you want to translate the text contained in the PDF.

Upload PDF file

Upload your PDF file by selecting it or dragging it onto the page.

Translate and view results

You can view the translated text directly on the screen or download the translated document.

Accurate AI Translation in 100+ Languages

Ai-powered accurate translations.

Seamlessly communicate globally with OpenL's AI neural translation technology - translating conversations, documents, and more into native-level accuracy.

100+ Language Support

Effortlessly bridge cultural divides with OpenL's translations across over 100 languages, from English to Arabic, Chinese, French, Spanish, and more.

Multi-Format Translation

Easily translate texts, documents, images, audio - PDF, Word, PNG, MP3 and more. Fast, efficient service streamlining multi-format translation tasks.

Beyond Translation

Level up writing with AI grammar tools, writing refinement, and language learning for academic and professional excellence.

Try It Free

Try OpenL free with 30 daily translations. Upgrade to Pro for unlimited longer texts tailored to professional translation needs.

Educational Discount

Students and educators using .edu email addresses can enjoy a 30% discount. You can apply for this offer once per year to support affordable language learning.

Frequently asked questions

Everything you need to know

Subscribers can unsubscribe at anytime, with cancellations taking effect after the current billing cycle ends.

Subscribers are responsible for fully testing our service before ordering a subscription, as refunds are not available for subscribers.

Something we didn't cover? We're happy to have feedback .

Unlock fast, accurate translation with OpenL

Translate in 100+ languages with cutting-edge ai.

Translator for

Lingvanex - your universal translation app

Translate from english to kannada online, popular english-kannada phrases for starting conversation, 5 unique ways to learn a language, frequently asked questions, is the lingvanex translation accurate, how long does it take to translate a large text, how many characters can be translated, do you offer subscription plans, language pairs are available for text translation into english.

• Help Center
• Google Docs Editors
• Privacy Policy
• Terms of Service
• Submit feedback

Translate documents or write in a different language

You can translate documents into many languages  with Google Docs.

Translate a document

• On your computer, open a document in Google Docs .

• Enter a name for the translated document and select a language.
• Click Translate .
• A translated copy of your document will open in a new window. You can also see this copy in your Google Drive .

Tip : If "Translate document" isn't visible, you're likely in Microsoft Office editing. To translate, convert your file to Google Docs. Learn about Microsoft Office editing  and  how to convert Microsoft Office files .

Change your typing language

• Phonetic input:  Type the phonetic spelling of a word in Latin characters, then click the best match from the options on the screen.
• Keyboard input: To type letters and symbols, click the keys of a virtual keyboard or type matching keys with your physical keyboard.
• In handwriting input: Draw characters in a panel in the bottom right of your screen, then click the best match from the options on the screen.

Tip:  In Workspace, the handwriting input works only when "access to additional services without individual control" is on in the Admin Console .

Need more help?

Try these next steps:.

Using Google products, like Google Docs, at work or school? Try powerful tips, tutorials, and templates. Learn to work on Office files without installing Office, create dynamic project plans and team calendars, auto-organize your inbox, and more.

Kannada to English Translation

Translate kannada to english.

• Azerbaijani
• Cantonese (trad)
• Chinese (literary)
• Chinese (simp)
• Chinese (trad)
• English United Kingdom
• French (Canada)
• Haitian Creole
• Inuinnaqtun
• Inuktitut (Latin)
• Kinyarwanda
• Klingon (Latin)
• Kurdish (Kurmanji)
• Kurdish (Sorani)
• Luxembourgish
• Lower Sorbian
• Meiteilon (Manipuri)
• Mongolian (trad)
• Myanmar (Burmese)
• Odia (Oriya)
• Portuguese (Brazil)
• Portuguese (Portugal)
• Quertaro Otomi
• Scots Gaelic
• Serbian (Cyrillic)
• Serbian (Latin)
• Upper Sorbian
• Yucatec Maya

Online Translation

Instantly Translate a Document from Kannada to English

Please enable JavaScript

Drag & drop your file here or

All uploaded files are permanently removed from our servers within 1 hour.

By uploading a document, you agree to our terms and conditions .

SUPPORTED FORMATS

Right-to-left (rtl) language support.

Online Doc Translator now fully supports translations of the following right-to-left languages:

Translate any document's text to English

Use our free translator to instantly translate any document to and from Kannada or English

Simply upload a Kannada or English document and click "Translate"

Translate full documents to and from Kannada and instantly download the result with the original layout preserved

Translate Kannada documents to English in multiple office formats (Word, Excel, PowerPoint, PDF, OpenOffice, text) by simply uploading them into our free online translator

Frequently Asked Questions

History Of Kannada Language

Kannada is a Dravidian Language which is notably spoken in Karnataka state of modern India. Besides this, it is also spoken by minority speakers in Andhra Pradesh, Tamil Nadu, Telangana, Kerala, Goa. The Native Speakers of Kannada Language is around 44 Million , who are called Kannadigas . It has gained the status of one of the classical languages of India. Kannada is one of the scheduled languages of India out of 22 scheduled languages. In Historical times it has the status of being court language of many powerful Kingdoms such as Vijayanagara, Chalukya, Rashtrakuta, Hoysala etc. The Script of Kannada language is called Kannada Script which is evolved out of 5th-century Kadamba Script . It has a reputation of getting 8 Jnanapith awards highest among all south languages and second to Indian Languages. The Kannada Language is mainly influenced by Sanskrit besides this, it got influenced by Pali, Prakrit.

Who Can Use English To Kannada Translator?

It is frequently charged that Translators do not work properly and even if they work properly there is not accuracy in translation. There might be some truth in this complain but before explaining further we would like to narrate the little background working of this translator. No two languages in the world are the same although there might be similarities found in them they differ on so many levels such as syntax, linguistics, grammar etc. The creation of Translators is a very complex task which involves great knowledge of languages, ruthless research and investment of a huge amount of time. So, there is always a possibility of inaccuracy but these days translators have to be improved a lot by Google which shows results with quite accurate and at a lightning-fast speed. What purpose translators can serve in modern time? most people think that what purpose it will serve to me but in today's globalised world and laissez-faire economy of world we have to translate often from one language to another for various purposes like email services, official paperwork etc. There is another use of translators for example when we chat with a person who speaks English Language and we want to communicate with him then this translator will serve as a handy tool. for that view, we have designed this translator and we hope that it will serve you well.

English Language

Today English is considered International Language because of its widespread use by various people in fields such as business, Internet, Academics, Literature etc. After Britain's long colonial rule English Language started to dominate the world. So it is very important to anyone who wants to progress with great vigour in life make an effort to learn English, but people are born with different learning abilities besides this life do not permit everyone to learn a new language for that reason alone we made an effort to create translators so that learning gap can be filled without making much effort.

What is different on This Website?

What we are offering which other websites are not offering. Although this translator is created using Google's API "Translation" but while another website has set an upper limit of 500 words in one translation we allow you to translate up to 2000 words in one go.

English To Kannada Translator

English To Kannada Translator is a free Translation tool designed and developed by typingbaba . It very easy to use all you have to do is just type in the upper box your English Content, press translate key and within a fraction of seconds, your translated text will appear in lower box where Kannada is written above. You can also copy the content from anywhere and paste it in the translation box. It is a very fast, accurate and robust translation tool which works very smooth without any glitches in the software. After pressing the translation key the translation will take max. 1-2 minutes and result will be as accurate as it can be. Besides this, you can press the toggle button which will change the position of boxes and Kannada To English Translation page will appear.

• Kannada Typing
• Kannada Translation
• Kannada to Tamil
• Tamil Translation
• Malayalam Translation

Kannada to English

Unlimited translation for FREE

Special Characters:

Independent vowels:, dependent vowels:, consonants:, additional consonants:, additional vowels for sanskrit:, sign used in sanskrit:, about our translator.

ಭಾರತದಲ್ಲಿ ಹೆಚ್ಚು ಮಾತನಾಡುವ 8ನೇ ಭಾಷೆ ಕನ್ನಡ. ಪ್ರಪಂಚದಾದ್ಯಂತ 58 ದಶಲಕ್ಷಕ್ಕೂ ಹೆಚ್ಚು ಜನರು ಇದನ್ನು ಮೊದಲ ಮತ್ತು ಎರಡನೆಯ ಭಾಷೆಯಾಗಿ ಮಾತನಾಡುತ್ತಾರೆ.

For those whose first language is not ಕನ್ನಡ, typing and translating Kannada to English can be difficult . There are many websites and agencies that provide online and offline translation and interpreting services. Those services can either be expensive or if it is free the quality of translation is often poor and unreliable.

While it makes sense for paying professionals to translate highly specialized subjects and official documents - there is no need to pay for translating commonly spoken words and phrases . For this our online software can be used.

Our translation software gives you high-quality translation results for FREE. This is because it uses a powerful Google translation API to instantly translate sentences between Kannada to English. You can use our tool to translate up to 500 characters per request. But the good news is you can make unlimited requests.

While the translation result is not always 100% accurate - with a few modifications it can be pretty accurate. We have also integrated a Google Input tool which can be used to easily edit or modify the translated ಕನ್ನಡ text. Besides, this software is evolving and improving every day and shortly we hope it can produce near to perfect translation .

We have also integrated functionality with which you can download or copy the translated text and either share them on social media or use on Word Processing software (such as Microsoft Word) for further formatting.

If you have any other suggestions on how to improve Kannada to English, please let us know on our Facebook page .

Finally, we would appreciate it if you like and share our page with your friends and family.

Key Features of Our Translation Software

✓   easy and instant translation:.

You can easily translate Kannada words, sentences and phrases into corresponding English.

Typing "ಕನ್ನಡ ಭಾರತದ ಅತ್ಯಂತ ಹಳೆಯ ಭಾಷೆಗಳಲ್ಲಿ ಒಂದಾಗಿದೆ." in Kannada will be converted into "Kannada is one of the oldest language of India."

You can also use this software as a dictionary to convert Kannada to English.

ಸಂಸ್ಕೃತಿ (Sanskriti) meaning in English will be "Culture"

ಅನನ್ಯ (Anan'ya) meaning in Kannada will be "Unique"

✓   Multi Platform Support:

✓   high accuracy rate:, ✓   multi language translation:.

Kannada To Hindi ,  Kannada To Tamil ,  Kannada To Telugu ,  Kannada To Malayalam ,  Kannada To Marathi

✓   FREE and Unlimited Translation:

Frequently translated phrases in the kannada language.

Frequently Asked Questions

What is the difference between transliteration and translation.

Transliteration is a process of transferring similar-sounding words from one language to another . To give you an example, typing “Nivu hegiddiri?“ will be transliterated into “ನೀವು ಹೇಗಿದ್ದೀರಿ?“.

Whereas, translation is a process of converting the written word from one language to another . For E.g. converting text written from Kannada to English.

How does Kannada to English translation works?

Our online translation software uses either ">Google , ">Microsoft , or Yandex to translate word, sentence, and phrase from Kannada to English.

Whenever you enter words, sentences, or phrases in Kannada and click on the “Translate” button, our software sends a request to Google, Microsoft, or Yandex for translation.

Their backend system uses some of the cutting edge technologies such as artificial intelligence , Web APIs, Big Data, etc to perform higher quality translations, and send back the response in English.

Can I download this translation software?

At a moment you can only use our Kannada to English translator online.

However, you can download and install Google Translation Chrome extension tool on your Google Chrome browser.

With this, you can either translate the entire page you are browsing by clicking the translation icon on the browser toolbar. Or, you can also highlight and right-click the section of the text , and click on the “Translate” icon to get the translation result in a language you have chosen from the drop-down menu.

Why is the translation not accurate?

Our Kannada to English translation software uses automated machine-language technology provided by Google or Microsoft. As humans are not involved, in some cases it will miss the context and cultural nuances of the language.

Having said so, our translator is useful for those who need help framing the sentence and get a general idea of what the sentence or phrase is conveying the message. With little modification, you can get near to perfect translation. We have also embedded Kannada to English transliteration, with a help of which you can easily edit, modify or add Kannada to English text.

Furthermore, the Google and Microsoft Translation API is evolving every day and as time goes by the translation result is going to be pretty accurate .

Is it FREE?

Yes . Our translator is 100% free.

However, we have few restrictions in place to ensure that robots or automated software are not abusing our service.

At any time you can translate up to a maximum of 500 characters per request. But you can make unlimited requests provided that you don’t misuse our software.

How can I translate Kannada (ಕನ್ನಡ) text on image or picture?

You can use third-party services , such as Yandex to translate words or sentences embedded in images or graphics.

Alternatively, if you are using a smartphone, you can download the Google Translate app that allows you to translate a picture . For this, you need to give Google Translate access to your camera. Once this is done, you can open the Google Translate app and tap on the Camera icon to translate any text on the picture.

Connect on Whatsapp : +1 206 673 2541 , Get Homework Help 24x7, 100% Confidential. Connect Now

Essay Writing in Kannada: A Guide

Do you want to learn how to express yourself in Kannada? Have you been struggling with essay writing and don’t know where to start? If so, this article is for you! In this guide on “Essay Writing in Kannada: A Guide”, we’ll explore all the tips and tricks necessary to write a great essay in Kannada. Through easy-to-understand instructions, advice from experts, and helpful examples, soon your essays will be shining like never before! So buckle up and let’s get started.

1. Introduction to Kannada Essay Writing

Kannada essay writing is a skill which requires thorough understanding of the language and its various aspects. It is critical to develop an in-depth knowledge about what essay writing in Kannada means, and how it can be incorporated effectively into ones essays.

Firstly, one should understand the basic grammar structure of Kannada as every sentence should comply with set rules that govern linguistic principles such as noun placement and verb conjugation. Furthermore, time spent studying prefixes, suffixes or even loan words will help build up a foundational knowledge base from which complex sentences could be derived from during composition; an important aspect for any piece of Kannada written work.

Secondly, an awareness about effective word choice is essential for accurate translation; learning idiomatic phrases may also prove useful when attempting to construct sophisticated pieces filled with depth or distinction. In other words, this encompasses being mindful about selecting vocabularies suited for particular contexts – adapting one’s diction according to purpose or receiver instead of using haphazardly chosen terms indiscriminately each time like what happens in colloquial speech would benefit writers greatly while they are tackling their penning instruments on some sort of topic related to what essay writing in Kannada entails!

2. Tips and Tricks for Crafting a Quality Kannada Essay

Research and Language: Writing an essay in Kannada requires a certain level of familiarity with the language, as well as research on any given topic. It is important to have access to resources such as online dictionaries, grammar books, and other study materials when writing what essay writing in Kannada so that you can accurately express your thoughts. Additionally, researching current events or local topics related to the essay’s subject matter will provide valuable insight into structuring arguments.

Organization & Presentation: Organization is key when crafting a quality Kannada Essay; this means understanding how one’s ideas should be presented for maximum impactful reading. Begin by outlining all ideas for the paper including supporting evidence before committing them to paper – this way it will become easier later on down the line when constructing cohesive paragraphs making up each section of what essay writing in kannada. After finalizing organization, sentences should flow smoothly while being grammatically correct throughout; bad punctuation or incorrect usage of words can drastically reduce readability and dilute overall messages expressed within an essay along with impacting its grade significantly.

3. Understanding the Language and Grammar of Kannada

Kannada is a language and grammar that has been around for centuries. With the rise of education and access to information, it’s easier than ever to start learning about the basics of Kannada grammar. Learning the language can be quite challenging at first but with a few tips you can quickly get familiar with its basic structure .

• Learn About Word Order : One essential element in understanding how Kannada works is getting acquainted with word order. For example, adjectives always precede nouns while verbs come right before their objects in sentences.
• Focus on Gender Markers : Like other languages such as French, Spanish or German, gender markes play an important role when composing essays written in Kannada. It’s important to pay attention when using masculine (ಗುರು) or feminine (ṭammaa) pronouns so your writing conveys what essay writing in Kannada should convey.
• Get Acquainted With Tense System : All verb conjugations follow specific patterns based on tense system used which also affect sentence structuring. To master these rules requires time and effort but once you understand them you will become comfortable expressing yourself through words without difficulties regardless if you are making reference to something from past like “I went”or present like “I am going”even prompting future actions like “Iwill go”Usingthe correct kind of tense inflection dependingonwhat essay writinginKannadarequiresisessentialtosuccessfullycommunicatingyourideas.

4. Researching Your Topic in Kannada Literature

When , it is important to first get a good overview of the material available and determine what sources can be used. There are several different types of Kannada essays that you may need to look at when writing an essay in this language. For example, one type might include expository essays which provide an explanation or argument related to a particular issue or subject. Another common type is descriptive essays which paint a vivid picture about the given topic using examples and stories from daily life experiences. Lastly, narrative essays tell stories with interesting characters and events unfolding through time.

Once you have identified which type of essay you will use for your paper, it is then important to find relevant sources such as books, scholarly articles or websites related to what essay writing in Kannada entails. Depending on where you live,, there may be libraries nearby with resources available specific for this language so make sure to check those out if possible! Additionally, many universities also have databases such as ProQuest which allow accesses specifically designed for academic research purposes containing numerous journal articles on any number of topics including those relating what essay writing in kannada looks like today As well as providing useful background information these sources often contain further references linking other avenues worth exploring while doing research relatedtoKannda literary studies . Furthermore these online tools often enable users to narrow down results based on certain criteria making them extremely helpful when focusingon more specific aspects regardingwhat essaywritinginKanndalookslike

5. Structuring an Engaging Argument in your Paper

Organizing an argument in your paper involves more than simply outlining the points you need to make. You should also consider carefully how these points will be expressed, and what style of writing best suits your purpose. To structure a successful argument that engages readers from start to finish, there are several factors to consider:

• Begin with clarity. Make sure that each point presented is clearly articulated within its own paragraph or section.
• Guide the reader logically. When constructing arguments for essay writing in kannada, aim for a logical flow between sections – one idea builds upon another until reaffirming evidence is reached at the conclusion.
• Develop strong examples. Supplied facts and quotes will help defend and reinforce each main point made when essay writing in Kannada language; use appropriate sources as necessary.

6. Developing Clear Organization and Logical Flow in your Work

When it comes to writing, organization is paramount. To ensure your essay is well-organized and logically flows from one point to the next, you’ll need to build a thoughtful structure prior to starting the writing process. Here are some tips for getting started:

• Outline . Before you start writing an essay in Kannada, create a detailed outline of what topics will be covered and how the ideas discussed fit together.
• Group similar arguments together . Once have generated your initial argument structure, organize similar points into “buckets” or sections by combining related concepts/ideas into broadly-related groups.

Now that you’ve outlined your document’s primary argument(s), begin fleshing out each section with more specifics on what those arguments entail. Make sure each sentence relates back directly to its supporting evidence – this helps maintain consistency throughout the essay. Additionally, when working in Kannada make sure there are logical transitions between paragraphs which help move readers along without gaps in understanding; use phrases like “In addition” or “Moreover” as connective tissue between sentences and groupings of thought within an individual paragraph. With these techniques at play even complex essays about topics such as what essay writing in kannada looks like can still remain clear and concise!

7. Polish: Making Sure You Achieve Perfection

The final step in the writing process is polishing your work, making sure it achieves perfection! Polishing involves proofreading for errors, checking coherence and ensuring that all sections comply with style guidelines.

When learning essay writing in Kannada, it’s important to ensure you pay attention to language rules during this polishing stage. Look out for spelling errors, grammar mistakes and typos; not only are these aesthetically displeasing but they impede readability too. Try reading your work backward sentence-by-sentence or use a proofreading app to help you spot potential issues more easily. It can also be beneficial to review each section of an essay separately while keeping the overall objective of the piece in mind–this ensures continuity between sections as well as communicates ideas effectively.

• Essay Writing In Kannada: Check language rules carefully.
• Proofread for any inconsistencies.

Ensure citations adhere properly to referencing standards–incorrect formatting can result in plagiarism allegations if left unchecked and spell disaster when trying to achieve distinction grade levels! Essay Writing In Kannada: Get references correct at all times . Finally remember that perfecting an essay takes time so don’t rush yourself – take breaks from editing periodically by coming back with fresh eyes which aides focus on finding subtle discrepancies more easily. Essay Writing In Kannada: Take regular breaks while refining essays. . Essay writing in Kannada is no longer a daunting task. With the right guidance and resources, you can now express yourself authentically and confidently through written Kannada. So go ahead, take up that pen – it’s time to level-up your essay writing game!

WhatsApp us

Shabdkosh Glosbe BdWord

Shabdkosh Glosbe

Google Bing

Wikipedia Google search Google books

• Shabdkosh : Kannada-English dictionary

• BdWord : English-Kannada dictionary

• Kannada-English etymological dictionary [PDF] by Norihiko Učida & Bando Bhimaji Rajapurohit (2013)

• Goethe-Verlag : Kannada-English common phrases & illustrated vocabulary (+ audio)

• Kannada-English dictionary by Ferdinand Kittel (1894)

• Kannada-English school-dictionary by J. Bucher, based on the Ferdinand Kittel's dictionary (1899)

• A Dictionary, Canarese and English by William Reeve, revised by Daniel Sanderson (1858)

• Dialogues in Canarese with an English translation , by Munshi Shrinivasiah, Richard Hodson, revised by Daniel Sanderson (1865)

• Christian names in Canarese by J. G. Kies (1866)

→ Kannada keyboard to type a text with Kannada characters

• Learning to spell in an alphasyllabary , the case of Kannada , by Sonali Nag, Rebecca Treiman, Margaret Snowling, in Writing systems research (2010)

• Kannada lessons for the beginner , by Shashank Rao (2016)

• Kanarese grammar by Harold Spencer, revised by W. Perston (1950)

• Grammar of the Kannaḍa language (ancient, mediaeval and modern language) by Ferdinand Kittel (1903)

• Modern Canarese grammar by A. S. Mud-Bhatkal (1899)

• Elementary grammar of the Kannada or Canarese language , by Thomas Hodson (1864)

• Canarese school-grammar (1859)

• Book of lessons in Canarese (1862): I & II

• Canarese reading lessons for children (1855): I & II - III

• Grammar of the Carnáṭaca language by John McKerrell (1820)

• Śabdamaṇidarpaṇa ಶಬ್ದಮಣಿದರ್ಪಣ Kēširāja's jewel mirror of grammar : grammar of Old Kannada (13 th century) edited by Ferdinand Kittel (1920)

• Grammar of the oldest Kanarese inscriptions by A. N. Narasimhia (1941)

• studies about the Kannada language, by Maheedasa Balagaru

• A short history of Kannada dictionaries

• Towards a classification of kannada dictionaries

• A brief survey of Kannada grammatical works

• books & papers about the Kannada language: Google books | Internet archive | Academia | Wikipedia

The Kannada is a Dravidian language, like the Tamil and the Telugu languages. Its' spoken in the state of Karnataka, in South-West India.

This language was also named Canarese and the name of the state Carnataca, Carnatic .

• Kannada Prabha - Prajavani - Udayavani

• LyrikLine : poems in Kannada, with translation (+ audio)

• History of the Kannada literature by Jyotsna Kamat (2016)

• Literature and culture , in Mandya district Gazeetter (2009)

• History of Kanarese literature by Edward Rice (1921)

• Nāgavarma's Canarese prosody by Ferdinand Kittel (1875)

• Canarese poetical anthology (1874)

• Katha Sangraha or Canarese selections, prose , compiled by Daniel Sanderson (1868)

• Katha Manjare or a collection of Canarese stories , with an English translation (1841)

• The Pancha Tantra or Panchó Pakhyáná, five collections of stories , presented by John Garrett (1864)

• Canarese Panchatantra for the use of schools (1865)

• Collection of official documents in the Canarese language compiled by Liston Garthwaite (1870): I & II

• Carnataca translation of Esop's fables (1855)

• The seventy stories in Canarese , forming a series of progressive lessons (1857)

• The New Testament translated into Canarese (1884)

ಎಲ್ಲಾ ಮಾನವರೂ ಸ್ವತಂತ್ರರಾಗಿಯೇ ಜನಿಸಿದ್ದಾರೆ. ಹಾಗೂ ಘನತೆ ಮತ್ತು ಹಕ್ಕುಗಳಲ್ಲಿ ಸಮಾನರಾಗಿದ್ದಾರೆ. ವಿವೇಕ ಮತ್ತು ಅಂತಃಕರಣಗಳನ್ನು ಪಡೆದವರಾದ್ದರಿಂದ ಅವರು ಪರಸ್ಪರ ಸಹೋದರ ಭಾವದಿಂದ ವರ್ತಿಸಬೇಕು.

• Universal Declaration of Human Rights : translation into Kannada (+ audio)

→ First article in different languages

→ Universal Declaration of Human Rights : bilingual text

→ Inde : maps & documents

→ languages of India

→ Telugu & other Dravidian languages: Malayalam - Tamil

→ Hindi - Sanskrit

Kannada Typing

Its very easy and simple to type in Kannada (Kannada Type) using English. Just type the text in English in the given box and press space, it will convert the text in Kannada script.

Special Characters:

Independent vowels:, dependent vowels:, consonants:, additional consonants:, additional vowels for sanskrit:, sign used in sanskrit:.

Click on a word to see more options. English to Kannada converter gives you 100% accurate result if your input is correct. To switch between Kannada and English use ctrl + g. Now copy the text and use it anywhere on emails, chat, Facebook, twitter or any website.

1. English to Kannada Translation

2.  English to Kannada Typing Mobile App Free Download

3.  Type by Speak Kannada - Kannada Speech to Text

Kannada typing is very simple with the English to Kannada type software it shows the suggestion from what word to type in Kannada, so you can choose the correct word to type in Kannada. It also saves your time by auto complete feature for typing Kannada from English .

Kannada Typing very important to give feelings in words that is not possible while type in English. The typed word in Kannada are in Unicode font so you can use it any where on the web like Facebook, twitter, comments etc. The software also known as English to Kannada translation, English to Kannada converter system .

Free Download Kannada Font

How to Type in Kannada 

Kannada Typing is very easy with above method. Just type in English as you type messages in Mobile and press space bar. It will convert in Kannada. If you think you don't get desired word, you can press backspace key to open word suggestion list from which you can choose another suitable word of Kannada language.

Suggestions list will also appear when you click on that word with mouse. India Typing is Free and Fastest method for Type in Kannada, without practicing  Kannada keyboard  actually.

1. Type with your English keyboard and press space bar.

2. You will see your English typed word gets converted in Kannada.

3. If you don't get desired word, you can press backspace key to get more suggestion words, choose one from them. (To pop-up suggestion list you can click on particular word also)

4. If not found your desired word in suggestion list, try another combinations of English letters. This Kannada transcription works on Phonetics so make English letters combination as the sound vibrates from your mouth.

5. Still not get desired word ? You can click on Help button to insert any Kannada character or word.

6. You can download your typed Kannada text as either notepad file (.txt) or MS-Word file (.doc).

7. After completing your Kannada typing work, you can make formatting with open in editor option.

Explore Kannada Typing

Kannada  (  ಕನ್ನಡ )  is under top 20th spoken language in the world. Almost 70 million (i.e. 7 Crore) peoples speaks Kannada language mainly in Karnataka state in India. Kannada is official language of Karnataka. Kannada is one of the longest surviving classical languages in the world.

Kannada is written in "Kannada" script which has its roots in "Brahmi" script. Kannada alphabet uses 49 phonemic letters, divided into three groups - 13 vowels (swaragalu), 34 consonants (vyanjanagalu), and yogavaahakagalu (neither vowel nor consonant - two letters: anusvara ಂ and visarga ಃ). Kannada is written from left to right.

Do you Know ?  What we speak is language, so  Kannada is a language  and What we write is known as script, so  Brahmi is a script . We Speak "Kannada" and Write in "Kannada-Brahmi" script.

Kannada Alphabets

Kannada alphabets consists 13 vowels and 34 consonants.

1. Vowels in Kannada

Kannada Vowels are found independently only in the initial position of a word. Otherwise they are added to consonants. 13 vowels of Kannada script are following.

ಅ, ಆ, ಇ, ಈ, ಉ, ಊ, ಋ, ಎ, ಏ, ಐ, ಒ, ಓ, ಔ

2. Consonants in Kannada

The primary consonants (without the vowel 'a') are called ardhaksaragalu. Kannada script consists 34 consonants are as following:

ಕ, ಖ, ಗ, ಘ, ಙ, ಚ, ಛ, ಜ, ಝ, ಞ, ಟ, ಠ, ಡ, ಢ, ಣ, ತ, ಥ, ದ, ಧ, ನ, ಪ, ಫ, ಬ, ಭ, ಮ, ಯ, ರ, ಲ, ವ, ಶ, ಷ, ಸ, ಹ, ಳ

3. Kannada Numerals

Numerals are written symbols used to represent numbers. Kannada counting uses distinct symbols for the numbers 0 to 9.

Kannada Numerals   ೦,  ೧,  ೨,  ೩,  ೪,  ೫,  ೬,  ೭,  ೮,  ೯

English Numerals      0,  1,  2,  3,   4,   5,   6,   7,   8,  9

Frequently Asked Questions ?

1. Is it safe type important document here on website?

Yes, we respect your works confidentiality and don't save it on our server and don't use it in any manner.

2.  What is the technology behind English to Kannada typing?

It's Kannada Transliteration, it is machine transliteration software as service enable you to type in Kannada.

3. How to change font of Kannada text ?

What you have typed with English to Kannada transcription is in Unicode Kannada font, so its very portable means you can use this Kannada text anywhere on the digital world. You can copy from here and paste it on Facebook, WhatsApp, twitter, blogs, comment section at any site. You could download Kannada text in either as notepad file (.txt format) or document file (MS word).

If you are looking for change font of your typed content you can change font family after download in your system. After download Kannada text, open with MS word or Notepad and change font family. You can  download Kannada Unicode fonts  from our website download menu.

4. Can I get my typed Kannada text in English also ?

Yes, you can get English translation of your Bengali text. Just copy the Kannada text you have typed and paste on  Kannada to English translator  tool. You will get translation in seconds.

5. Can I get Kannada text without type it ?

Yes, you have an alternative for Kannada typing without using keyboard, what you are looking for is  Kannada voice typing . Let your mic to do typing for you, just speak and your speech will be typed automatically.

6. Difference between Kannada translation and transliteration ?

Transliteration is the process of changing the script used to write words in one language to the script of another language. A translation tells you the meaning of words in another language.

7. Country of Origin?

This website is made in India with love.

• English to Kannada
• Kannada to English
• Kannada to Hindi
• Kannada to Marathi
• Kannada to Gujarati
• Kannada to Bangla
• Kannada to Odia
• Kannada to Tamil
• Kannada to Telugu
• Kannada to Malayalam
• Kannada to Punjabi
• Any Language
• Kannada Nudi Fonts
• Kannada Unicode Fonts
• Kannada Font Alt Character
• Display Text in Kannada
• Font Installation
• Kannada Inscript Keyboard
• Kannada Keyboard in Windows 10
• Kannada Keyboard in Windows 11
• Kannada Typing Test
• English Typing Test
• Simple Kannada Typing
• Kannada Inscript Typing
• Kannada Type in Mobile
• Roman Kannada to Kannada Converter
• Kannada OCR - Convert Image to Text
• Text to Image
• Indian Script Converter
• Number to Word Converter Kannada

Kannada Prabandha

ಮಕ್ಕಳ ದಿನಾಚರಣೆಯ ಬಗ್ಗೆ ಪ್ರಬಂಧ । children’s day essay in kannada.

Children’s Day essay in Kannada :ಮಕ್ಕಳ ದಿನವನ್ನು ವಾರ್ಷಿಕವಾಗಿ ನವೆಂಬರ್ 14 ರಂದು ಆಚರಿಸಲಾಗುತ್ತದೆ, ಇದು ಮಕ್ಕಳ ಮುಗ್ಧತೆ, ಸೃಜನಶೀಲತೆ ಮತ್ತು ಮಿತಿಯಿಲ್ಲದ ಸಾಮರ್ಥ್ಯವನ್ನು ಗೌರವಿಸಲು ಮತ್ತು …

ದೀಪಾವಳಿ ಹಬ್ಬದ ಬಗ್ಗೆ ಪ್ರಬಂಧ । Essay on Deepavali festival in Kannada

Essay on Deepavali festival in Kannada :ದೀಪಾವಳಿ ಯು ರೋಮಾಂಚಕ ಮತ್ತು ಆಧ್ಯಾತ್ಮಿಕವಾಗಿ ಮಹತ್ವದ ಹಬ್ಬವಾಗಿದ್ದು, ಪ್ರಪಂಚದಾದ್ಯಂತ ಲಕ್ಷಾಂತರ ಜನರು ಆಚರಿಸುತ್ತಾರೆ, ಭಾರತವು ಅದರ ಕೇಂದ್ರಬಿಂದುವಾಗಿದೆ. ಈ …

ಅಂತರ್ರಾಷ್ಟ್ರೀಯ ಬಾಹ್ಯಾಕಾಶ ನಿಲ್ದಾಣ ಹಾಗೂ ಅದರ ಕಾರ್ಯ ನಿರ್ವಹಣೆ ಬಗ್ಗೆ ಪ್ರಬಂಧ

ಅಂತರಾಷ್ಟ್ರೀಯ ಬಾಹ್ಯಾಕಾಶ ನಿಲ್ದಾಣ (ISS) ಭೂಮಿಯ ಸುತ್ತ ಪರಿಭ್ರಮಿಸುವ ಬಹುರಾಷ್ಟ್ರೀಯ ಬಾಹ್ಯಾಕಾಶ ಪ್ರಯೋಗಾಲಯವಾಗಿದ್ದು, ವೈಜ್ಞಾನಿಕ ಸಂಶೋಧನೆ, ಅಂತರಾಷ್ಟ್ರೀಯ ಸಹಕಾರ ಮತ್ತು ತಾಂತ್ರಿಕ ಆವಿಷ್ಕಾರಗಳಿಗೆ ಕೇಂದ್ರವಾಗಿ ಕಾರ್ಯನಿರ್ವಹಿಸುತ್ತದೆ. ಈ ಗಮನಾರ್ಹ …

ಡಾ ಬಿ.ಆರ್ ಅಂಬೇಡ್ಕರ್ ಜೀವನದ ಬಗ್ಗೆ ಪ್ರಬಂಧ | Dr BR Ambedkar Essay in Kannada

Dr BR Ambedkar Essay in Kannada :ಬಾಬಾಸಾಹೇಬ್ ಎಂದು ಪ್ರೀತಿಯಿಂದ ಕರೆಯಲ್ಪಡುವ ಡಾ. ಭೀಮರಾವ್ ರಾಮ್‌ಜಿ ಅಂಬೇಡ್ಕರ್ ಅವರು ದೂರದೃಷ್ಟಿಯ ನಾಯಕ ಮತ್ತು ಬೌದ್ಧಿಕ ದೈತ್ಯರಾಗಿದ್ದರು, ಅವರ …

ಭಾರತದ ಉಕ್ಕಿನ ಮನುಷ್ಯ ಸರ್ದಾರ್ ವಲ್ಲಭಭಾಯಿ ಪಟೇಲ್ ಬಗ್ಗೆ ಪ್ರಬಂಧ | Sardar Vallabhbhai Patel Essay 600 words

Sardar Vallabhbhai Patel Essay : “ಭಾರತದ ಉಕ್ಕಿನ ಮನುಷ್ಯ” ಎಂದು ಜನಪ್ರಿಯವಾಗಿ ಕರೆಯಲ್ಪಡುವ ಸರ್ದಾರ್ ವಲ್ಲಭಭಾಯಿ ಪಟೇಲ್ ಅವರು ಸ್ವಾತಂತ್ರ್ಯಕ್ಕಾಗಿ ಭಾರತದ ಹೋರಾಟದಲ್ಲಿ ಅತ್ಯುನ್ನತ ವ್ಯಕ್ತಿಯಾಗಿದ್ದರು ಮತ್ತು …

ರಸ್ತೆ ಸುರಕ್ಷತೆಯ ಬಗ್ಗೆ ಪ್ರಬಂಧ | Road Safety Essay in Kannada

Road Safety Essay in Kannada :ಭಾರತದಲ್ಲಿ ರಸ್ತೆ ಸುರಕ್ಷತೆಯು ಒಂದು ನಿರ್ಣಾಯಕ ವಿಷಯವಾಗಿದೆ, ಅದರ ವ್ಯಾಪಕವಾದ ರಸ್ತೆ ಜಾಲ ಮತ್ತು ಬೀದಿಗಳಲ್ಲಿ ಹೆಚ್ಚುತ್ತಿರುವ ವಾಹನಗಳ ಸಂಖ್ಯೆಯನ್ನು ನೀಡಲಾಗಿದೆ. …

ಶಿಕ್ಷಣದ ಮಹತ್ವದ ಕುರಿತು ಪ್ರಬಂಧ | Essay on Importance of Education

Essay on Importance of Education :ಭಾರತದ ಸಾಮಾಜಿಕ, ಸಾಂಸ್ಕೃತಿಕ ಮತ್ತು ಆರ್ಥಿಕ ರಚನೆಯಲ್ಲಿ ಶಿಕ್ಷಣವು ಯಾವಾಗಲೂ ಪ್ರಮುಖ ಪಾತ್ರವನ್ನು ವಹಿಸಿದೆ. ಜ್ಞಾನ ಮತ್ತು ಕಲಿಕೆಯ ಶ್ರೀಮಂತ ಇತಿಹಾಸದೊಂದಿಗೆ, …

ಡಾ. ಸರ್ವಪಲ್ಲಿ ರಾಧಾಕೃಷ್ಣನ್‌ ಬಗ್ಗೆ ಪ್ರಬಂಧ | Dr Sarvepalli Radhakrishnan Essay in Kannada

Dr Sarvepalli Radhakrishnan Essay in Kannada :ಭಾರತದ ಅತ್ಯಂತ ಗೌರವಾನ್ವಿತ ವಿದ್ವಾಂಸರು ಮತ್ತು ದಾರ್ಶನಿಕರಲ್ಲಿ ಒಬ್ಬರಾದ ಡಾ. ಸರ್ವಪಲ್ಲಿ ರಾಧಾಕೃಷ್ಣನ್ ಅವರು ತತ್ವಶಾಸ್ತ್ರ ಕ್ಷೇತ್ರಕ್ಕೆ ನೀಡಿದ ಕೊಡುಗೆಗಳಿಗಾಗಿ …

Search form

Kannada to english machine translation using deep neural network.

Pushpalatha Kadavigere Nagaraj *   |  Kshamitha Shobha Ravikumar  |  Mydugolam Sreenivas Kasyap  |  Medhini Hullumakki Srinivas Murthy  |  Jithin Paul

© 2021 IIETA. This article is published by IIETA and is licensed under the CC BY 4.0 license ( http://creativecommons.org/licenses/by/4.0/ ).

OPEN ACCESS

In this paper, we focus on the unidirectional translation of Kannada text to English text using Neural Machine Translation (NMT). From studies, we found that using Recurrent Neural Network (RNN) has been the most efficient way to perform machine translation. In this process we have used Sequence to Sequence (Seq2Seq) modelled dataset with the help of Encoder-Decoder Mechanism considering Long Short Term Memory (LSTM) as RNN unit. We have compared our result concerning to Statistical Machine Translation (SMT) and obtained a better Bi-Lingual Evaluation Study (BLEU) value, with an accuracy of 86.32%.

encoder-decoder mechanism, long short term memory, natural language processing, sequence to sequence model

Natural Language Processing (NLP) provides machines with the ability to understand and deduce human languages from their interpretation and acts as a connection between human language and data science [1]. Neural Machine Translation (NMT) is a machine translation technique that uses a wide range of neural networks to predict the probability of a series of words in a single integrated pattern that forms a complete sentence to reduce the language barrier [2-4].

A Deep Neural Network (DNN) is a hierarchical organization of hidden networks (layers) that connect input and output. The DNN seeks the correct mathematical manipulation to transform the input into output [5].

Machine translation has remained unexplored in the field of some native Indian languages. To understand the information and ideas expressed in a certain language, translation is necessary. This has culminated in the introduction of automated translation from regional to national or international language [6].

Kannada which is one among the Dravidian language which has rich historical literature but has a poor resource in terms of computational linguistics which becomes a difficult task due to its syntactic and semantic variance in its literature. In the field of MT, Kannada is not explored much when compared to other Indian languages. Many research works have been focused on English-South Dravidian language (Kannada /Malayalam) utilizing SMT which was meant to be a traditional approach in machine translation [7].

Translation for foreign languages like German, French, Spanish was developed first when seen its development for Indian languages, a phrase-based statistical model was developed for eight languages which include Hindi, Bengali, Gujarati, Tamil, Malayalam, Telugu, Urdu and Punjabi were translated to English [8]. Further many methods were implemented and the one in current use is NMT which has shown its excellent results for translation. NMT with attention model has been implemented for multilingual translation of five Indian languages including English translation [9], six Indian languages to Hindi translation [10], Hindi to English translation [11], Hindi to Bengali translation with comparing values of MOSES statistical machine translation (SMT) and NMT [11] with different calculation metrics for comparing its performances and accuracy [12].

Kannada-English translation is much an unexplored area in terms of Machine Translation. Some of the resources suggest a Baseline SMT using MOSES for Kannada-English Translated bible corpus [13] and English-Kannada translation through rule-based MT [14]. Translation for simple sentences is done for Kannada transliterated corpus using lexicon analysis and phrase mapper to identify display boards of Karnataka government offices [15]. In this paper, we have applied the NMT method to translate Kannada Text to English Text using Encoder-Decoder mechanism with LSTM as RNN unit, without attention mechanism.

Figure 1 depicts the block diagram of the proposed architecture used in the MT process. Starting with input language and translated language both require a tokenizer which splits sentences into words and gives an integer value for every unique word present in the dataset. These integer values are converted into vector values which act as an input to the LSTM cell. The Yellow blocks represent the two-layered LSTM cells which together forms the encoder and decoder part of the Seq2Seq model. X i are the input vectors to an encoder model whereas Y i are the discarded output of the encoder similarly X i ’ and Y i ’ are the input and output vectors of decoder respectively. The predicted output of the decoder is taken into the SoftMax activation function which selects which one to activate based on the vector values and represents the probability distributions of a list of potential outcomes. 

Figure 1. Proposed architecture

3.1 Sequence to Sequence Model (Seq2Seq)

A Seq2Seq model is a model that inputs a sequence of items (words, letters, time-series) and generates an output of a different sequence of items. The input is a series of word and the output is the translated series of words for a machine translation application. A typical Seq2Seq model includes two parts: an encoder and a decoder. Both are two different models of the neural networks combined into one huge network. Each of these encoders and decoders consists of series of any Recurrent Neural Network (RNN) unit, connected to work as encoder/decoder. Hence, we use LSTM as an RNN unit. These LSTM in series form the encoder and decoder part which combines to form a Seq2Seq model [16, 17].

3.2 Long short term memory (LSTM)

Long Short-Term Memory (LSTM) networks, a class of RNN is capable to process and learn to predict sequence based on its order dependencies. Hidden state (h t ) and cell state (C t ) are the two internal states present in LSTM. The information flows out through a mechanism known as cell states. So, LSTM’s may remember things selectively, or forget things. An LSTM basically has four gates, to preserve and control the cell state value based on memory. They are:

•Forget gate

•Input gate

•Update gate

•Output gate

Figure 2. LSTM architecture

In Figure 2, all the gates are represented with a pointwise operator which acts as a valve to decide its flow. In LSTM, the first step is to determine which information we are seeking to discard from the cell state as it is stored in the memory previously. A sigmoid layer activates "forget gate f t " (Eq. (1)), which performs this function. Consecutively, it is required to decide what new information must be stored in the memory of the cell state. Two elements are primarily concerned with performing this action. First, the "input gate i t "(Eq. (2)) is a sigmoid layer which defines the values to be updated. The second element is the tanh layer C̃ t (Eq. (3)), determines a new vectored candidate values which can be applied to the cell state C t . We use "update gate C t " (Eq. (4)) to update the old cell State C t-1 to produce a new cell state C t . The final output is decided using the “Output gate o t ” (Eq. (5)). The hidden state h t (Eq. (6)) acts as the input to the next LSTM in case of series arrangement as depicted in Figure 2. The equations for respective gates are given by:

${{f}_{t}}=\sigma ({{W}_{f}}.[{{h}_{t-1}},{{x}_{t}}]+{{b}_{f}})$     (1)

${{i}_{t}}=\sigma ({{W}_{i}}.[{{h}_{t-1}},{{x}_{t}}]+{{b}_{i}})$      (2)

$\widetilde{{{C}_{t}}}=tanh({{W}_{c}}.[{{h}_{t-1}},{{x}_{t}}]+{{b}_{c}})$     (3)

${{C}_{t}}={{f}_{t}}*{{C}_{t-1}}+{{i}_{t}}*\widetilde{{{C}_{t}}}$     (4)

${{o}_{t}}=\sigma ({{W}_{o}}.[{{h}_{t-1}},{{x}_{t}}]+{{b}_{o}})$      (5)

${{h}_{t}}={{o}_{t}}*tanh({{C}_{t}})$      (6)

3.3 Encoder-decoder mechanism

The most ideal approach for MT using Seq2Seq model is by using Encoder-Decoder architecture where both encoder and decoder consists of LSTM as its subunits.

The process starts with the encoder unit, where the LSTM’s are placed in series fashion to obtain an encoder vector. Encoder vector is the value of internal states from the last LSTM of the encoder which encloses information about previous input elements. This vector will be the initial state value for first decoder LSTM which helps decoder for accurate predictions. Encoder mechanism is the same in case of training and inference process.

Decoder begins to generate output sequence using the initial state generated and initialized from the final state of encoder LSTM. During inference and training procedure, the decoder behaves differently. During training, we use teacher forcing technique to train decoder quickly as shown in Figure 3. During inference, the output from the previous time step acts as an input to the decoder at each time step as depicted in Figure 4.

Encoder summarizes all the input sequence applied to encoder into state vectors (h and C) and discards the output of the encoder. The final state vectors (h i and C i ) of the encoder are initialized as initial input (h 0 ’ , C 0 ’ ) of the decoder to generate output sequence. A decoder is just a linguistic model which depends on its initial states to generate output sequence by discarding the final state vectors.

The LSTM in both encoder and decoder sequentially reads the data sequence. Thus, if the input is a sequence of ‘k’ length, we ensure that the LSTM reads it in ‘k’ time steps.

The most important point is that the decoder’s initial states (h 0 ’ , C 0 ’ ) are set to the encoder’s final state (h i , C i ). This implicitly means that the decoder is trained to generate output sequences based on the encoded information by the encoder. English translated phrase should be dependent on the sentence given in Kannada for translation.

Figure 3. Encoder-decoder LSTM in training process

Figure 4. Encoder-decoder LSTM in inference process

During the inference process, one word will be created at a time. Therefore, the LSTM decoder is called in a loop and only a one-time step is processed every time. At each time step, we retain the decoder states and set them as initial states to the next time step i.e., the predicted output at present time step acts as input to the next time step.

4.1 Dataset

Preparing dataset depends on the type of machine learning project. From previous researches and surveys, we have found that the corpus used textbook translation which does not influence much-spoken language. Since there is non-availability of open-source datasets for Kannada to English, hence we have developed our dataset based on the Seq-Seq model for Kannada-English in parallel format. Pre-processing of data is done to the dataset to build the models more accurately.

Table 1. English-Kannada corpus

A dataset which we created has 41 thousand pairs of parallel data. We split the dataset in 98:2 i.e., 98% of the dataset is for training and 2% is for testing. (40500 for training and 500 for testing the model). However, to increase the accuracy and efficiency of Translation system it is required to update the dataset frequently.

Table 1 depicts Kannada and English corpus information.

4.2 Experimental setup

According to the Corpus size, the platform in which the model must be trained will be chosen. There are various platforms to train the model in Machine Learning.

In our project, we have used Google Colab (an open-source platform) for Training purpose, where we can leverage free robust Graphics Processing Unit and Tensor Processing Unit for machine learning education and researches. We used 17GB of Tensor Processing Unit (TPU) Memory to train the project model.

For Evaluating and Predicting the sequence on a local machine, we have used 8 GB primary memory and 2 GB Nvidia 940 MX with 384 Compute Unified Device Architecture (CUDA) Cores. CUDA is developed by Nvidia as a computing platform on GPU’s. CUDA helps developers speed-up compute-intensive exercises.

For Creating and Training Deep Neural Networks (DNN’s), Keras with TensorFlow backend deep Learning framework is used.

Bi-Lingual Evaluation Study (BLEU) score is a metric used for evaluation of predicted sentence to target sentence. It will result in 1 if there is a perfect match or 0 if there is a complete mismatch. We have got a different value while changing the weights as shown in Table 2.

Table 2. Bi-Lingual Evaluation Study (BLEU) Scores Obtained in Inference

BLEU Score evaluation as in Base Line MT using MOSES tool kit [13] and NMT are compared in Table 3 represented in percentage (%). Translation time for our model was about 2-5 seconds based on the length of the input sentence. Translation time for MOSES algorithm is not available in the referred paper.

Table 3. Comparison of BLEU score for Training and Testing

Figure 5 graph depicts Validation Loss versus Epoch. As the number of Epochs increases, validation loss decreases till a certain epoch. The minimum loss is considered till the end of epochs. We obtained validation loss of 0.849.

Figure 5. Validation Loss v/s Epochs Graph

Figure 6. graph represents Validation Accuracy versus Epochs. Initially, for the first epoch, validation accuracy is about 74.84%. Then as the number of epochs increases, validation accuracy starts increasing. So, the accuracy is 86.32%.

Figure 6. Validation accuracy v/s epoch graph

The real-time outputs for user inputs implemented in the above system are represented in Table 4.

Table 4. Real-time outputs obtained based on a model trained

This paper uses Machine Translation based on the principle of Deep Neural Network (DNN) to accelerate communication among scientific workers, regardless of the language in which their findings may be expressed. LSTM works well in case of classification, processing and time series prediction, despite unknown time lags. A model is created using a dataset of 41000 data pairs and 40 epochs giving an overall loss of 0.849 and accuracy of 86.32%. The results obtained show that the model is more efficient than translation services available using Seq-Seq method.

Gated Recurrent Unit (GRU) Layer may be used instead of LSTM which has two internal states (cell state and hidden state) whereas the internal state (hidden state) of GRU is one. It will allow the definition and description to be condensed and will also help in improved performance. Transformer architecture for time-series forecasting can be used. The parameters can be played with such as encoder and decoder layer, etc. It can improve results in tuning and training. The accuracy can be improved with training and tuning. The increase in the dataset will offer a larger base. With the larger base of data, the accuracy will also improve. Integration of automatic speech recognition (ASR) can be done to the current work. Integration of speech can be done with a large speech corpus. This can be trained to recognize the input voice samples of all pitches and frequencies and translate in real-time.

[1] Otter, D.W., Medina, J.R., Kalita, J.K. (2020). A survey of the usages of deep learning for natural language processing. IEEE Transactions on Neural Networks and Learning Systems, 32(2): 604-624. https://doi.org/10.1109/TNNLS.2020.2979670 [2] Costa-Jussà, M.R. (2018). From feature to paradigm: Deep learning in machine translation. Journal of Artificial Intelligence Research, 61: 947-974. https://doi.org/10.1613/jair.1.11198 [3] Nair, L.R., Peter, S.D. (2012). Machine translation systems for Indian languages. International Journal of Computer Applications, 39(1): 24-31. https://doi.org/10.5120/4785-7014 [4] Chaudhary, J.R., Patel, A.C. (2018). Machine translation using deep learning: A survey. International Journal of Scientific Research in Science, Engineering and Technology, 4(2): 145-150. [5] Sutskever, I., Vinyals, O., Le, Q.V. (2014). Sequence to sequence learning with neural networks. NIPS’14: Proceedings of the 27th International Conference on Neural Information Processing Systems, 2: 3104-3112. [6] Basmatkar, P., Holani, H., Kaushal, S. (2019). Survey on Neural Machine Translation for the multilingual translation system. 3rd International Conference on Computing Methodologies and Communication (Erode, India), pp. 443-448. https://doi.org/10.1109/ICCMC.2019.8819788 [7] Unnikrishnan, P., Antony, P.J., Dr Soman, K.P. (2010). A novel approach for English to south Dravidian language statistical machine translation system. International Journal on Computer Science and Engineering, 2(8): 2749-2759. [8] Khan, N.J., Anwar, W., Durrani, N. (2017). Machine Translation Approaches and Survey for Indian Languages. ArXiv abs/1701.04290. [9] Revanuru, K., Turlapaty, K., Rao, S. (2017). Neural machine translation of Indian languages. Compute 2017: 10th Annual ACM India Conference (Bhopal, India), pp. 11-20. https://doi.org/10.1145/3140107.3140111 [10] Verma, C., Singh, A., Seal, S., Singh, V., Mathur, I. (2019). Hindi-English neural machine translation using attention model. International Journal of Scientific and Technology Research, 8(11): 2710-2714. [11] Das, A., Yerra, P., Kumar, K., Sarkar, S. (2016). A study of attention-based neural machine translation model on Indian languages. Proceedings of the 6th Workshop on South and Southeast Asian Natural Language Processing (Osaka, Japan), pp. 163-172. [12] Shah, P., Bakrola, V. (2019). Neural machine translation system of Indic languages - An attention-based approach. 2nd International Conference on Advanced Computational and Communication Paradigms (Gangtok, India), pp. 1-5. https://doi.org/10.1109/ICACCP.2019.8882969 [13] Shivakumar, K.M., Nayana, S., Supriya, T. (2015). A study of Kannada to English baseline statistical machine translation system. International Journal of Applied Engineering Research, 10(55): 4161-4166. [14] Reddy, M.V., Hanumanthappa, M. (2013). Indic language machine translation tool: English to Kannada/Telugu. Proceeding of 100th Science Congress (Kolkata, India), 213: 35-49. https://doi.org/10.1007/978-81-322-1143-3_4 [15] Kodabagi, M.M., Angadi, S.A. (2016). A methodology for machine translation of simple sentences from Kannada to the English language. 2nd International Conference on Contemporary Computing and informatics (Noida, India), pp. 237-241. https://doi.org/10.1109/IC3I.2016.7917967 [16] Saini, S., Sahula, V. (2018). Neural machine translation for English to Hindi. Fourth International Conference on Information Retrieval and Knowledge Management (Malaysia), pp. 1-6. https://doi.org/10.1109/INFRKM.2018.8464781 [17] Verma, A.A., Bhattacharyya, P. (2017). Literature survey: Neural machine translation. CFILT, Indian Institute of Technology Bombay, India.

Phone: + 1 825 436 9306

Email: [email protected]

Subscription

Language support

Please sign up to receive notifications on new issues and newsletters from IIETA

Select Journal/Journals:

Copyright © 2024 IIETA. All Rights Reserved.

Advertisement

Anthe: on translating kannada, on language.

The art and life of Mark di Suvero

Drawing by Deepa Bhasthi.

Anthe (ಅಂತೆ) is one of my favorite words in the Kannada language. Somewhat meaningless by itself, it adds so much nuance and emotion when appended to a sentence that we Kannadigas cannot carry on a conversation without using it. Depending on the context and the speaker’s tone, anthe can convey an expression of surprise or the understanding that gossip is being shared. It could mean “so it happened,” “that’s how it is,” “apparently,” or “it seems.” The latter comes closest to a direct translation, but is a frustratingly simple choice. Anthe will only ever half-heartedly migrate to English.

Banu Mushtaq, whose short stories I have been translating recently, and whose “ Red Lungi ” appears in the Summer 2024 issue of The Paris Review , employs anthe generously. Mushtaq’s characters use anthe when reporting something someone said verbatim or when guessing how something might have happened. In another instance, she uses echo words with anthe , another common characteristic of the Kannada language: one character utters anthe-kanthe to refer to hearsay. There are also a whole lot of ellipses in Mushtaq’s stories … her sentences often trail off … like so … She mixes up her tenses here and there. It is always deliberate, this nod to the idea that time is not linear. The awareness that we inhabit different time zones and dimensions and live in stories within stories is commonplace in India. These narrative tools give Mushtaq’s work a sense of orality, as if she is sitting across from you and telling you the story.

Whenever Mushtaq and I do talk in real life, she is narrating, she is reporting, she is discussing the oppressive political scene in India, she is going back to her youth, laughing about that one time there was a fatwa issued against her for a story—they wanted her to stop writing, she told them to go to hell—she is constantly relaying anecdotes and thinking out loud and living through stories. There is more anthe in her urgency to convey everything all at once than I can hope to store in my notes.

My favorite function of the word is how its repetition in every other sentence, each differently intoned, allows a musicality to slip into daily speech. It gives everyday Kannada its impu, or melody.

Kannada belongs to the Dravidian family of languages native to southern India, and it is spoken by an estimated fifty to sixty million people around the world. It is the official language of the state of Karnataka. Kannada literature has been published continuously, without any lapse, for fifteen hundred years. Poets have described Sirigannadam, or “rich-Kannada,” as a river of honey, as milk rain, as being as sweet as nectar, as truth, as an eternal language. As old and celebrated as it is, Kannada is just one of India’s hundreds of languages (estimates vary between 780 and 1,600, in addition to thousands of dialects). An aphorism in the Khari Boli language best expresses the country’s mind-boggling linguistic diversity: Kos-kos par badle paani, chaar kos par baani , meaning that the water in India changes every two miles, and language every eight miles.

Mushtaq’s mother tongue is not Kannada, and strictly speaking, neither is it mine. Hers is Dakhni, a fascinating mix of Persian, Dehlavi, Marathi, Kannada, and Telugu that is often wrongly called a dialect of Urdu. It is spoken in parts of southern India predominantly, though not exclusively, by the Muslim community. My mother tongue is Havyaka, a dialect that harks back to Old Kannada (a parental language that dominated from the ninth to the eleventh century C.E. ) and is spoken by a small community of upper-caste brahmins along the Arabian coast as well as by a similar community in central Karnataka. (The two versions differ and are mutually unintelligible.) However, for various sociogeographical reasons, what I grew up with was a mix of Kannada varieties common to the Mysore region and coastal Mangalore. Both Mushtaq and I live in small towns where we regularly speak at least three or four languages and a number of dialects. This use of numerous sociolinguistic systems is, for a very large number of Indians, par for the course, and is certainly not just a function of class, caste, education, or intellectual leanings. I know a grocer in my town who speaks thirteen languages fluently.

Mushtaq’s decision to write in Kannada is hardly unusual today. The second half of the twentieth century saw many eminent multilingual writers choose to work in Kannada, the dominant language spoken on the streets, even though their mother tongues were Konkani, Telugu, Tamil, Tulu, or others, and they conducted their professional lives in English. Such multilingual proficiency is much rarer now—thanks to the colonial-era Macaulayism that continues to drive India’s English-forward education system, most of us might speak many languages but be proficient only in English.

Banu Mushtaq sprinkles her Kannada with Dakhni, Urdu, and scant Arabic. Each of these languages is also further influenced by local distinctions specific to Bayalu Seeme, a region of Karnataka characterized by open plains, where she lives. What enriches my translations, I like to believe, is a complete immersion in the culture of the text I am working with. I had an easier time translating my first two books, as they were in languages and from regions I was very familiar with. Mushtaq’s stories are set within India’s Muslim community, though her work is not from and for Islamic culture alone. Still, her references are a world apart from my own lived experiences. In trying to bridge the gap in my understanding, I found myself reaching for culture that shared linguistic, social, and aesthetic influences with the milieu of her stories. Somehow, this kind of immersion seemed to help me be better attuned to the subtleties in her text. I was introduced to the all-consuming world of Pakistani TV dramas, a vast subculture that overflows with intrigue, suspense, romance, and high-octane drama that thrills most desi hearts. I rediscovered Qawwali music, falling irrevocably back in love with Nusrat Fateh Ali Khan, Abida Parveen, Arooj Aftab, Ali Sethi, their ilk. Alongside my everyday dosage of Kannada-ness, I read about Islam. Reopened my Arabic notes from the lessons I once received from a maulvi saheb . Enrolled in a Nastaliq script-writing class in which we studied talaffuz (pronunciation) via Hindi films and poetry. I cannot say where and how this research helped my translation practice. But it has enriched my years like anything, as we say in India.

Any practice of writing and translation in this social terrain can become very political very quickly. The politics of language in India are, as one might expect, a source of severe passion and violence. While attention to the most prominent language in any state—Kannada in Karnataka, for instance—comes at the cost of the prominence, public usage, and funding for the instruction of several others, perhaps the greatest threat to linguistic diversity across India comes from the imposition of Hindi. Spoken mostly in parts of northern India, central government policy has historically forced the language upon other states, where it has often been violently rejected, especially in the southern states. These language wars have killed people, and emotions around imposition of one language at the cost of others, whether Hindi or a regionally dominant language, are always simmering below the surface. Why translate at all, in the face of so many complexities? This is a question we can no longer afford to ask. The Kenyan writer Ngũgĩ wa Thiong’o calls translation “the common language of languages.” In India, linguistic hegemony is just one of the many dangers of the homogenizing policies of a far-right-wing-led government since 2014. When engaging with another language culture, one experiences, in learning a new word or phrase, a glimpse of a different world, however ephemeral.

In the hierarchy of who gets to translate, who gets to be translated, who gets funding for new translations … Kannada does not fare too well. Barely a handful of literary translations in the Kannada-English pairing is published in a year. This is miles behind literatures written in, say, Malayalam, Tamil, Bengali, Urdu, or Hindi. To translate at all, and then to translate from an underrepresented language, and finally to translate into English, with all its baggage, is riddled with layers of questions for which there can never be simple answers. I reach back for anthe , trying to see how complexities can often remain unresolved. Anthe refuses to let itself be easily carried out of Kannada, which in turn reminds me that it is okay not to have all the answers about translation and its politics.

One delightful side effect of all the language cultures we live with in India is how often we toggle between two or three languages when conversing. This reduces language to its basic function, as a tool to communicate with whatever words are within reach, stripping the need for “proper” articulation, throwing out all the rules. We alternate this pared-down approach with hyperbole and exaggeration, with repetitions of a word for a sense of emphasis— “I’ve told you a thousand times,” “small-small,” “round-round,” “two-two times …” This interplay of the utilitarian and the embellished, the functional and the ornamental, is also where I find the uniqueness of Kannada when I translate. It is important for me to retain these qualities—for instance anthe rendered as the inadequate “it seems,” which I’m sometimes told is not quite elegant enough English, because carrying these nuances is like speaking English with an accent. It reminds the reader that the text is from another culture. In translation, these expressions stretch the elasticity of the English language. In turn, Kannada gains another reader.

Deepa Bhasthi is a writer and critic who translates Kannada-language literature .