Home // SEMAPRO 2016, The Tenth International Conference on Advances in Semantic Processing // View article

Sentiment Analysis on Maltese using Machine Learning

Authors:
Alexiei Dingli
Nicole Sant

Keywords: Sentiment Analysis; Maltese; Machine Learning; Deep Learning

Abstract:
Sentiment analysis refers to the task of analysing a piece of text and classifying it according to the overall sentiment of the opinion being expressed. In this paper, we present a novel, supervised context based, machine learning system capable of performing such a task for text written in Maltese. Our system consists of two components both capable of performing classification of Maltese text at a context window level, yet while one follows the more traditional approach where features are hand-crafted and passed on for classification, the other performs unsupervised feature extraction and makes use of a deep learn- ing algorithm. Through experimentation we determined that a Random Forest classifier in conjunction with 80% of our dataset for training and a four word context window achieved the best results, and were successful in achieving an accuracy of 62.3%.

Pages: 21 to 25

Copyright: Copyright (c) IARIA, 2016

Publication date: October 9, 2016

Published in: conference

ISSN: 2308-4510

ISBN: 978-1-61208-507-4

Location: Venice, Italy

Dates: from October 9, 2016 to October 13, 2016