Abstract:
POS Tagging in Bangla is a procedure of identify the parts of speech of given data sets. If
any person does not know about the right Parts of Speech of a Bengali word then this
project will be helpful for them. Trigrams'n'Tags (TnT) Tagging is a part of Hidden
Markov Model Viterbi Algorithm. By the use of Trigrams'n'Tags (TnT) Tagging method
Parts of Speech tagging become more easier. Trigrams'n'Tags (TnT) Tagging method is
used here for tagging the unknown word by the use of external editable corpus which was
been tagged before implement the algorithm. For research I have taken 30787 words and
1860 sentences for tagging from a newspaper web portal, and also took some data for
which are untagged for tagged by this algorithm. Firstly I tagged the 30787 word for
creating the editable corpus. Secondly I collect the untagged data, and then implement the
algorithm on this untagged data for further tagging.