Technology Development for Indian Languages
India is a multilingual country, with 22 official languages and 12 scripts. In India “Dissemination Through http://tdil-dc.in ” only about 5% - 10% people know English and the rest are deprived of benefits Hindi, Bengali,Marathi Tamil, Telugu, of advances in information technology. The benefits of information technology 44 Punjabi, VishwaBharat@tdil Odia, Gujarati, nslatio Assamese can reach to the common man only when software tools and human machine ra n Cross (9) T S Glossary e y Lingual n Sampark s Updation Information i t Validators / interfaces are available in people's own languages. Technology Development h (18 Pairs) e Tool Access c Hindi < > Punjabi m Localization . E-learning a Hindi < > Urdu . Multimedia Hindi < > Telugu Tools . for Indian Languages (TDIL) Programme of Ministry of Electronics & M Standardization Dictionary Hindi < > Tamil Resources & . Sandhi Hindi < > Bengali E- Linguistic Sanskrit Information Technology (MeitY), Government of India is an on-going Hindi < > Marathi T E-learning Sanchay ools Hindi < > Kannada Research & Telugu < > Tamil Area TDIL-DC Multimedia Research Programme with following vision & mission objectives: Malayalam< >Tamil www.tdil-dc.in AnglaMT (8 Pairs) Machine Technology Sanskrit Vision: Anuvadaksh o Bengali . English T Translation Handshake IPR Tools Morph-Generator (8 Pairs) Malayalam,Urdu, . Morph-Analyser English To Hindi, Ÿ Punjabi,Telugu, . Sandhi Digital Unite & Knowledge for All. Marathi,Bengali Hindi . Urdu,Odia,Tamil, Assamese . Sandhi-Splitter Gujarati,Bodo * Nepali Transliteration. Indo Mission: * TTS Wordnet Ÿ Proliferation of Language Technology OCR UTRRS Assamese,Bengali,Bodo,Gujarati, * Language Browser Screen SMS Libre Kannada,Kashmiri,Konkani,Malayalam, Plug-in Ÿ Research and Development of Language Technology Manipuri,Marathi,Nepali,Sanskrit,Tamil, (8) Reader Reader office Telugu,Punjabi,Urdu,Odia,Hindi and English (19 languages) Hindi Ÿ Web OCR (10) Development of Standards related to Language Technology Bengali, Kannada, Bengali Malayalam,Telugu Marathi Devanagari, Urdu, Assamese,Tamil, Tamil Gurumukhi (7) Outcome of TDIL Programme: Bengali, Kannada, Telugu Odia Supported file formats: Malayalam, BMP*, PNG, TIF Malayalam Size<10MB Devanagari, Assamese,Tamil, Gujarati Gurumukhi The research efforts undertaken through TDIL programme of MeitY have lead to * UTRRS = Unicode Text Rendering Reference System Kannada
* LPMS = Localization Project Management System Supported file formats: following outcomes with regard to proliferation of Indian Languages: BMP, PNG, TIF
(A) Proliferation Free Language CD: (C) Standardization : Ÿ It contain software tools for 22 Indian Languages (for both Windows and Linux Operating Systems) have been made available for download from http://ildc.in. Efforts have been taken in standardization of Indian language in Unicode, Standardize ŸTotal Language CD dispatched – 13 Lakh and total Language CDs downloaded are Inscript keyboard layout. Also worked with BIS and ICA for mandating support of Hindi 1.58 Crore. on all mobile devices. ŸTools like Localized Libre Office, Unicode Typing Tool, Unicode fonts, Localized Firefox Browser and many more are freely made available. New Initiative: Natural Language Translation Mission
Virtual keyboard for 22 Indian languages which can be downloaded for use on Mission is to overcome the language barrier for all major languages of India in speech mobiles has been made available on https://apps.mgov.gov.in has proliferated use and text form. of Indian languages on mobile devices.
L-1 Speech ASR TTS (B) Research and Development Machine Translation Systems: Ÿ Indian Language to Indian Language Machine translation system in 18 language pair Image Doc OCR MT L-2 Speech Ÿ English to Indian Language MT (AnglaMT) for 8 language pairs. Ÿ English to Indian Language MT (Anuvadaksh) for 8 language pairs. Text Text L-2 Text Ÿ Hindi to English Language MT (HEMAT) for Judicial domain. Processor
Text To Speech (TTS) system for 10 Indian languages viz. Tamil, Telugu, Marathi, SSMT System Flow Bodo, Kannada, Odia, Hindi, Malayalam, Manipuri & Rajasthani. Deliverables: Automatic Speech Recognition (ASR): Automatic Speech Recognition (ASR) for (a) Speech to Speech Machine Translation (SSMT) system for major Indian Agricultural Commodity prices and weather information system in 11 Indian languages: It would be possible with minimal human involvement but the Languages/Dialects have been developed, namely Tamil, Telugu, Marathi, Hindi, involvement would decrease with the time as the machine learns on its own. Bengali, Assamese, Hindi (Bihari), Hindi (Jharkandi), Kannada, Gujarati, Odia. The Adaptation of the system in the domains within broad areas of science & technology, systems would act as a voice interface for NIC Agmarknet portal. education, healthcare, governance, law & justice, etc.
Optical Character Recognition in Indian Languages: (b) Text-to-Text Machine Translation system for major Indian languages: As in the Optical Character Recognition engine for 10 Indian languages namely Bengali, case of SSMT system, it would be possible with minimal human involvement but the Devanagari, Gurumukhi, Kannada, Malayalam, Telugu, Tamil, Assamese, Odia and involvement would decrease with the time as the machine learns on its own. Urdu languages. Web OCR (for 10 ILs) as well as Desktop OCR (for 7 ILs) is made Adaptation of the system in the domains within broad areas of science & technology, available for public use. education, healthcare, governance, law & justice, etc.
On-line Handwriting recognition system (OHWR): (c) Deployment of language technology-based applications through more than 100 Online Handwriting Recognition in 8 Indian Languages namely Hindi, Bengali, Tamil, start-ups. Telugu, Kannada, Malayalam, Assamese and Punjabi languages have been developed. (d) National Platform for Language Technology – offerings of linguistic resources and language technology-based services at competitive cost which will decrease Cross Lingual Information Access (CLIA): progressively. Monolingual Search Engines for Tourism Domain for 9 Indian Languages (Sandhan) –Hindi, Bengali, Marathi, Tamil, Telugu, Punjabi, Odia, Gujarati and Assamese. (e) Five-fold increase in content in Indian languages on Internet .
Linguistic Resources: (f) A portal for showcasing technology and carrying out assessment of technology. Resources like wordnet, text corpora, speech corpora , Treebank text data, tts voices, ASR data have also been developed to support and promote research in Indian (g) Hackathons and grand challenges in the area of language technology . Language. (h) Repository of standards, best practices and benchmarks for Indian languages.