Google aims to roll out voice-based search services in Indian regional languages

New Delhi: Driven by increasing penetration of affordable smartphones and cheaper data plans, voice-based search is expected to witness manifold growth in the next few years in India, search giant Google said today.

Google, which offers voice-based search in 47 languages globally, also expects to roll out services in Indian regional languages in the future, though no specific timelines have been set for the same.

"Right now, the number of voice searches are low for India but given the increasing penetration of smartphones, there would be a manifold increase in volume in the next few years," Google India Head of Marketing Sandeep Menon told PTI.

Citing the example, he said a big proportion of searches from Japan are voice-based.

"Google has search options in nine languages in India (for PCs). We are committed to India. Though there is no timeline but I see no reason why other (Indian) languages will not be added to voice search," he said.

Menon said the company has been working on enabling voice search understand Indian accents.

"There are three things that we worked on: accents, understanding conversations and voice actions. It is all about giving users a smooth experience so that they get what they are searching for with a simple command," he added.

Menon was also of the view that more regional content needs to be created to enrich the ecosystem.

"While we are working on enabling better search experience, there is also a strong need for more regional content. That will lead to better search results and richer ecosystem for all," he said.

Google India today hosted a 'Google Cafe' here to showcase the search experience for Indian users.

"From asking where you can watch the latest movie to cricket scores to currency conversions to comparing health benefits of olive oil, Google's voice search can help users get the answers they need," Menon said.

He said the company has worked with over 700 volunteers in India to collect spoken utterances to create specific language models to power English voice search in Indian accent.