A Comparative Study of Machine Learning Algorithms for Anaphora Resolution of Myanmar Language

Anaphora resolution (AR) is the task of identify noun phrase that refers to the same entity as earlier and later items in the set of referring expressions or discourse.It is one of the more prolific areas of research in the Natural Language Processing (NLP) community and has correspondingly received a significant amount of attention in the literature.Anaphora resolution can support to improve the accuracy in almost every NLP application namely machine translation, text summarization, information extraction, dialogue interpretation, question-answering, etc. Myanmar language needs to be sufficiently applied in anaphora detection and resolution.This paper focuses on the comparison of Myanmar anaphora resolution system using machine learning algorithms.Four popular machine learning algorithms, Decision Trees, Support Vector Machines (SVM), Naïve Bayes, and K-Nearest Neighbors (KNN) algorithms were used to assess the feature sets.On novel datasets from Myanmar, several comparative experiments were conducted.The results were then discussed, and conclusions were taken.The experimental results show that our method produces acceptable results.