Google eagerly needs to learn Hindi Grammar
According to the Google’s press release, 87% of the literate population in India don’t know or don’t use English. Such users are deprived of the valuable resources available on the internet, which are mostly in English. In order to help such users, Google has introduced tools to facilitate blogging in Hindi, Tamil, Telugu, Kannada and Malayalam. It will take time before users get into to vernacular blogging. Google has also introduced English to Hindi translation. This is targeted at Non-English speaking internet users of India.
Translation from English to Hindi works flawless only for words or shorter sentences. As you can see in the picture below, when it comes to translating longer sentences, the machine fails miserably to understand the context of the sentence. Users can suggest a better translation, which will be used in future. Google also provides Hindi to English translation. And the user can write a small code in their webpage, if they do not want allow Google to translate their web page.
Google uses Statistical Machine Translation (SMT) to do the translation. Such translations can only be improved if the machine is fed with documents which humans have already translated into two languages. The machine will understand the patters and use it for future translation.
If this translation works, bloggers and websites owners could maintain the Hindi version of their English blog at no additional cost. A feature which I would like to see is, if Google allows me correct the translations for my blog and it reflects instantaneously. Google can consolidate such translations from bloggers and use for its machine learning. I get the benefit of additional non-English readers for my blog.
As a side note, I use Shabdkosh for my English to Hindi translation. This is a collaborative site where users post their words and other users translate it for them. All the users get to vote the translated work which helps maintain the accuracy of the translation. They are around since 2003 and it is maintained by Maneesh Soni. Google on their site is asking users to send them bilingual texts to improve the translation. Something which players like Shabdkosh can think of.
What will happen to the niche vernacular sites like WebDuniya and Business Standard? Such players have already invested money to develop content in vernacular languages. Can Google kill the vernacular competition in India? What do you think?
If you enjoyed this post, please consider to leave a comment or subscribe to the feed and get future articles delivered to your feed reader.
Comments
No comments yet.
Leave a comment