During his talk he had asked who all knew to type
Very few people raised their hands. He asked the audience to name some fonts in Malayalam equivalents to Aerial and TimesNew Roman which is familar to all of us. Only Lakshmi Sunil and Aathira Naveenan raised their hands. Santhosh emphasised in his talk that only 20% of world popluation know to read and write English. This is very true, I have seen a bunch of Switzerland natives who don’t even know to spell their names correctly in Immigration counter.
Santhosh challenged us to write a Hello world program in Amma Malayalam. Check out my hello world program
>>> print("ഹെല്ലൊ വേൾഡ് ") ഹെല്ലൊ വേൾഡ്
You can even program just with
Malayalam. Did you folks know Python3 was revamped from Python2 for unicode support? See the function implementation as shown here:
>>> def ഗണിക്ക(ക,ഖ): ... return(ക*ഖ) ... >>> ഗണിക്ക(3,6) 18
He told about some important Malayalam projects like Morphological Analyser, sentence predictor. The unique thing is most of his project doesn’t use Machine learning hype. I am planning to understand these project and explore this in depth in coming months.
As an AI enthusiast. I enquired with Santhosh about the possibility of Malayalam computing expanding with Machine Learning. According to Santhosh, inorder to use machine learning in Malayalam, you need lot of data and infrastructure. So some of popular techniques in NLP like Neuro Machine Transulation cannot be applied for Malayalam. So most of the projects by SMC are discrete and definite in nature. Instead of predicting nature in Machine Learning.
Do checkout Santhosh Thottingal sir being interviewed by Sreeram Venkitesh in the below link: