{"id":135461,"date":"2020-12-17T09:04:33","date_gmt":"2020-12-17T06:04:33","guid":{"rendered":"https:\/\/en.buradabiliyorum.com\/googles-new-open-source-ai-model-understands-indic-languages-better\/"},"modified":"2020-12-17T09:04:33","modified_gmt":"2020-12-17T06:04:33","slug":"googles-new-open-source-ai-model-understands-indic-languages-better","status":"publish","type":"post","link":"https:\/\/buradabiliyorum.com\/en\/googles-new-open-source-ai-model-understands-indic-languages-better\/","title":{"rendered":"#Google\u2019s new open-source AI model understands Indic languages better"},"content":{"rendered":"

“#Google\u2019s new open-source AI model understands Indic languages better<\/strong>”<\/p>\n

\n Google\u2019s various products, such as Search and Assistant, are already available in India in multiple local languages. The company is now turning to a new AI to potentially make more of its offerings accessible\u00a0to Indic language speakers \u2014 more specifically, it\u2019s using a technology<\/a> called MuRIL.<\/span><\/p>\n

At its virtual event today, the Big G unveiled a new language model called\u00a0Multilingual Representations for Indian Languages (MuRIL). This is the first model to support interoperation between 16 different Indic languages.\u00a0\u00a0<\/span><\/p>\n

That includes\u00a0Assamese, Bengali, English, Gujarati, Hindi, Kannada, Kashmiri, Malayalam, Marathi, Nepali, Oriya, Punjabi, Sanskrit, Sindhi, Tamil, Telugu, and Urdu.<\/span><\/p>\n

While MuRIL is based on Google\u2019s own BERT (Bidirectional Encoder Representations from Transformers) model, researchers claim it\u2019s more efficient for Indian languages.<\/span><\/p>\n

\"\"
<\/a>MuRIL<\/figcaption><\/figure>\n

Partha Talukdar, a researcher at Google India, said that the new model understands the context of statements in local languages better.\u00a0<\/span><\/p>\n

For example, the previous model understood\u00a0the following Hindi statement as a negative emotion:\u00a0a Hindi statement \u201cAccha hua account bandh ho gaya\u201d (It\u2019s good that the account got closed).<\/span>\u00a0However, the new model correctly predicts that the statement is positive.<\/p>\n

Users in India often use their English language keyboard to type in local languages \u2014 like the sentence above. For that, researchers have included support for transliteration detection in other languages while using the Roman script<\/a>.<\/p>\n

Google is making this model open-source for other researchers and startups to use.<\/p>\n

Currently, MuRIL is not embedded in any of Google\u2019s products. However, based on inputs from researchers and programmers, it aims to include this model into its offerings in the future for better accuracy.<\/p>\n

You can learn more and check out MuRIL\u2019s code here<\/a>.<\/p>\n

\n Published December 17, 2020 \u2014 06:04 UTC\n <\/p>\n<\/p><\/div>\n