<div dir="ltr"><div><div><div>The proposed GSoC project that aims to improve varnam's learning system states :<br><br>"The main goal of this is to improve how varnam tokenizes when learning
words. Today, when a word is learned, varnam takes all the possible
prefixes into account and learn all of them to improve future
suggestions. But sometimes, this is not enough to predict good
suggestions. An improvement is suggested which will try to infer the
base form of the word under learning.
"<br><br></div>Can someone provide examples to the cases where the current learning system fails? I still have not made up my mind which will be a better project - this or CMU Sphinx acoustic modeling<br><br></div>regards,<br>
<br></div>Kevin Martin Jose<br><br><br></div>