Firstly thanks for creating an excellent data file.
I am using parichit data files of telugu language for OCR in mobiles.
It does an excellent work for some words like అక్క. but fails in some cases like విధ్యార్ధులు and others.
when OCRed for విధ్యార్ధులు it was recognised as విధాయరదులు. ie as seperate letters.
How can we solve this.?