2024-03-29T15:59:04Z
https://meral.edu.mm/oai
oai:meral.edu.mm:recid/4342
2022-03-24T23:12:05Z
1582963302567:1597824273898
user-ucsy
Resolving Function Tagging Ambiguity in the Myanmar Language Using Transformation-Based Learning
Thant, Win Win
Htwe, Tin Myat
Thein, Ni Lar
This article investigates the use oftransformation-based learning for resolvingfunction tagging ambiguity in the Myanmarlanguage. Function tagger plays an importantrole in natural language applications like speechrecognition, natural language parsing,information retrieval and information extraction.In this paper, the function tagger [12] learnsrules to correct its mistakes. A set of ruletemplates is used to create specific rules. Atinitial stage of function tagging for Myanmar, itis trained with a very limited resource ofannotated corpus. The performance can bemaximized with a substantial amount ofannotated corpus. The function tagset has beendeveloped for training and testing the functiontagger. The present tagset consists of 56 tags. Acorpus size of about three thousand sentences isused for training and testing the accuracy of thefunction tagger. The tagger learned 192 rules(including lexical and contextual rules) andachieved 93% accuracy.
2012-02-28
http://hdl.handle.net/20.500.12678/0000004342
https://meral.edu.mm/records/4342