Veteran search engine Google has added eight Indian languages, including Sanskrit, to Google Translate. The internet company is constantly adding several regional languages to its online translation platform. “Sanskrit is the first and most requested language in Google Translate and now we are finally adding it,” Isaac Caswell, senior software engineer at Google Research, told ET in an exclusive interview. We are connecting languages from Northeast India for the first time.”
Apart from Sanskrit, other Indian languages in the latest Google Translate software are Assamese, Bhojpuri, Dogri, Konkani, Maithili, Mizo and Meitilon (Manipuri). With this, the total number of Indian languages available on Google Translate has increased to 19. This was announced at the annual Google I/O conference that began late Wednesday.
Explain that Schedule VIII of the Indian Constitution deals with the languages of India and 22 Indian languages are included in this table. But the latest Google update does not cover all the 22 languages scheduled for India. Regarding this, Caswell told ET, “We’ve tried pretty much to reduce this gap in tabular languages.”
All languages added in the update will only be supported in the text translation feature, but the company will be rolling out voice to text, camera mode, and other features soon. “We are working on it, but they don’t support all of these languages yet,” he said.
Google is also working on deficiencies related to the translation of Indic languages. “We realize that the translation errors that our model makes for Indic languages are old words,” Caswell said. He said that there are often translated words that people do not know and do not use regularly.
“We’re trying to understand (the problems) better, and we hope to shift our model toward a more general production of this old model,” he said.