An Evaluation of Sinhala Language NLP Tools and Neural Network Based POS Taggers
Name
Jayasinghe Arachchilage Sriyal Himesh Jayasinghe
Abstract
Part Of Speech tagging is a fundamental problem in the NLP domain and Part Of Speech taggers are used to address this challenge. Though Rule based, probabilistic or deep learning approaches can be used to develop a Part Of Speech tagger, deep learning based Part Of Speech taggers have shown better results. All the Part Of Speech tagging researches that have been carried out so far for the Sinhala language have been done using rule based and probabilistic approaches. This research focuses on developing and evaluating deep learning based Part Of Speech taggers using LSTM network for the Sinhala language.In this research we trained 5 deep learning based Part Of Speech tagging models on two different data sets and evaluated the results of those models. The evaluation results have shown that deep learning based Part Of Speech taggers can be used for Sinhala language and their performance is better than the existing rule based or probabilistic Part Of Speech taggers.
Graduation Thesis language
English
Graduation Thesis type
Master - Computer Science
Supervisor(s)
Kairit Sirts
Defence year
2019