Why is there not a big enough market for Tamil software?
There were strong views from some of the pioneering researchers and developers on the second day of the Tamil Internet Conference, being organised at the Pondicherry University by the International Forum for Information Technology in Tamil (INFITT).
At least two prominent researchers and developers who have to their credit commercially available Tamil software — V. Krishnamoorthy of Learn Fun Systems and N. Deivasundaram of NDS Lingsoft Software Solutions — felt that despite having fairly cutting-edge technologies, Tamil software developers were largely poor cousins to their English counterparts; and the blame, if any, should be shared across the board, from the government to the industry to even people.
ADVERTISEMENT
“Though a lot of people stress the importance of nurturing one’s mother tongue, when it comes to actually spending money, they think twice,” pointed out Mr. Deivasundaram, who has developed a Tamil word processing software that also checks grammar mistakes. “They are willing to pay Rs. 1,000 to buy an English word processor. But, when Tamil software is priced on a par with it, they seem to think it is not worth that much.” Such differential treatment persists across cross-sections.
Sometimes, even governments and officials think of it as such.
“There was a government project on translation that nearly got funded for Rs. 40 lakh. But in the last minute an official said there was no reason to spend so much on Tamil software and Rs. four lakh would do,” Mr. Deivasundaram recalled.
ADVERTISEMENT
Mr. Krishnamoorthy pointed out that the range of software his company had developed now — including an intelligent character recognition software for mobile platforms — was the culmination of two decades of work.
However, often there were plenty of false starts, and they had to persist despite difficulties.
“It took me three years to realise that the technology I had initially adapted for my optical character recognition (OCR) was flawed. But, these are problems that Tamil software developers have to endure. What is painful is the lack of support,” he said.
Challenges ahead
Tamil Internet software has evolved to a stage where it can be categorised into ‘active language technologies’ and performs a range of functions over and above the basics of word processing and presenting.
“The first few years of development in Tamil Internet were all about standardising fonts and keyboard layouts. Those were ‘passive language technologies’ and we have been fairly successful with that. But, when it comes to more ‘active language technologies’ such as optical character recognition, grammar check or text-to-speech synthesis, we are lagging behind,” N. Deivasundaram said.
Computational linguistics which requires the skills of computer coding as well as language needed to be nurtured. “Those using computers are already used to text-to-speech technologies in English that are ubiquitous today,” A.G. Ramakrishnan from the Department of Electrical Engineering at the Indian Institute of Science, who has developed a Tamil OCR (optical character recognition) software, said. However, in terms of Tamil or for that matter any Indian language, the field of computational linguistics is not even in its infancy, he added.
The researchers and developers at the conference hoped that the regional governments and the corporate world took note of the possibility for an intervention.