SMS in your mother tongue

Get set to text in your mother tongue - மின்தமிழ் | Google Groups

Mobile phone users in rural India and the not-so-well-versed
in English will soon be able to express themselves better by sending
SMSes in their mother tongue.

Researchers, including professors from the Indian Institute of
Technology (IIT- Madras) and others at the Centre for Excellence in
Wireless Technology (CEWiT) have successfully developed a 7-bit
encoding scheme for 22 Indian languages that will not just help people
send SMSes in Tamil, Marathi or Assamese but also transliterate
scripts keyed in one language into another.

A fortnight ago, the global mobile standards body 3GPP (3rd Generation
Partnership Project) approved the technology. Now, cellular operators,
mobile phone manufacturers, value added service providers and the
Telecom Regulatory Authority of India have all agreed to adopt the
technology.

If everything goes as planned at the CEWiT, set up by the Union
department of information technology in collaboration with the telecom
industry and housed in the IIT-Madras campus, in a year's time the
technology could fuel m-governance in a big way. Government
departments and banks would be able to communicate effectively with
millions of mobile phone users in 22 Indian languages on new schemes
and updates.

Researchers led by CEWiT's Nadeem Akthar began working on the encoding
scheme for Indian languages after the Turkish telecom regulator forced
the 3GPP standards body to open up SMS for non-English languages last
year.

"When the global mobile standards body agreed to open up SMS in local
languages, it presumed that every country has only one language. In
India we've 22 official languages and our SMS market and volume is low
because we don't have facility to use the regional language. We
started off our work by taking the encoding used in computers and
began modifying it to suit wireless technology application on mobile
phones," professor Bhaskar Ramamurthi, honorary director, CEWiT and
dean (planning), IIT Madras told TOI.

"A mobile phone's SMS capacity is only 140 bytes (8 bits constitute a
byte). For sending an SMS in English we need just 7 bits per
character. Therefore, the maximum characters that you can send in one
SMS is 160 characters (140 bytes x 8 bits / 7 bits per character).
However, Unicode which supports all languages in the world requires 16
bits per character, which would have restricted one regional language
SMS to just 70 characters. So we decided to develop a 7-bit encoding
system which will allow users to send 160 characters in the regional
language," he explained.

The research team then designed 10 Indian language tables which can
support official languages because of similarities in scripts used by
different languages and also enable encoding messages in both uni-
lingual (Indic only) and bi-lingual (Indic + English) fonts. For
instance the character 'Ka' is encoded using the same bits in all
languages.

"Our task isn't over. The challenge lies in implementing the
technology. We now need to come up with the fonts, a user-friendly
mobile phone keypad layout, display screen and also find a solution to
legacy (existing) handsets, which currently don't have the feature.
People can't throw away the legacy handsets, so we've to find a way to
enable the technology in existing mobile phones," said Ramamurthi.

A task force will be set up shortly to "work feverishly" to complete
these tasks and the CEWiT team hopes that in a year's time mobile
phone manufacturers and service providers would be able to usher in
the language revolution in wireless technology.