SwiftKey has made it easier for 70 million more people in India to stay in touch with friends and family via their mobile phones.
The British smartphone keyboard company, which is owned by Microsoft, has rolled out its intelligent typing system to Tamil speakers across the Asian country.
It means Tamil speakers using SwiftKey can type phonetically in their native script or English and be given next-word predictions and corrections in both languages. This allows them to send text messages in a mixture of Tamil and English without changing the keyboard, thanks to SwiftKey’s advanced artificial intelligence.
For example, when a multi-lingual Tamil speaker who uses Hindi and Tamil types the word “namaskar”, SwiftKey’s prediction bar will show “namaskar”, “நம#கார’”and “नम#कार”.
Aarti Samani, technology evangelist for SwiftKey India, said: “Mixed language vocabulary is a naturally occurring phenomenon in our country, yet historically mobile users have been let down by having to use QWERTY keyboards not designed to allow them to type as they speak and think. The upshot is that many people avoid using certain letters and matras from their mother language. A large proportion of users abandon using the native script altogether. Multi-script transliteration solves this problem. It is crucial to advancing the adoption of new communication technologies within India, while still encouraging linguistic richness and diversity, and thus helping to preserve the sense of identity, culture and tradition that is inexorably tied to native language usage.”
SwiftKey has also revealed that this transliteration support is available for Punjabi, Bengali, Marathi, Malayalam, Odia, Telugu and Kannada, and will be released to the wider Indian market later this year.
The new Tamil feature came as SwiftKey held its first-ever Communications Summit in India to examine how communications technology is helping the country overcome language barriers, forge new connections and achieve greater freedom of expression. The summit, entitled The Future of Identity, Language and Technology in India, saw experts in technology, culture, business and education examine the impact that technological globalisation has had on Indian society, and the need for companies to create technology that better adapts to diverse Indian consumers’ needs.
The company currently offers support for 22 Indian languages. As users type consonants, SwiftKey’s keyboard automatically attaches matras (vowels) to the letters, so they don’t have to be displayed on a separate screen.
The predictive technology is powered by AI, so it continually learns from every individual’s typing behaviour, vocabulary and prediction preferences – even the emojis they like. The more the user types, the more effective the predictions become.
“Creating transliteration support for Tamil in particular has been a huge challenge for the business as the language contains a total of 247 characters – far too many to fit on a single screen,” Samani added.
“However, our mission at SwiftKey is to create technology that fits the needs of each user, rather than imposing limitations and restrictions on how they type and communicate, and we’ve been greatly encouraged by the feedback we’ve received to date from users across India. Our goal moving forwards is to continue to innovate, expand our support and ultimately help more and more citizens express themselves through messaging.”
SwiftKey features on more than 300 million devices across the world. Last year the company placed its hi-tech neural networks system into English, French, German and Spanish smartphone keyboards.
Everything you need to know about SwiftKey
Using neural networks, SwiftKey can “capture the relationship and similarity between words”. For example, having previously seen the phrase “Let’s meet at the airport”, the technology is able to infer that “office” or “hotel” are similar words that could also be appropriate predictions instead of “airport”.
It also understands that “Let’s meet at the airport” has a similar sentence structure to “Let’s chat at the office”. This allows SwiftKey to offer users the most appropriate prediction or autocorrection based on the sentence being typed.
