COMPARATIVE ANALYSIS OF MACHINE LEARNING ALGORITHMS FOR SENTIMENT ANALYSIS OF TEXTS IN AZERBAIJANI AND ENGLISH

Abstract

The prediction of sentiment of the text within different business spheres has been one of the challenging problems for a variety of linguistics. In this paper, the sentiment analysis of the texts is carried out using different machine learning (ML) techniques. Various feature extraction techniques and supervised learning algorithms are employed on the movie review dataset sourced from the Internet Movie Database and translated into Azerbaijani. Specifically, the techniques utilized encompass Support Vector Machine, Logistic Regression, Decision Trees, Random Forest, AdaBoost, XGBoost, and Naïve Bayes. The proposed models depict the importance of language corpus that Azerbaijani language lacks by comparing the results obtained from both Azerbaijani and English versions of the dataset.