296
Asian Journal of Engineering, Sciences and Technology - Volume 12, Issue 2 2022
By Rahim Ullah, Rizwan Munir, Fahad Najeeb, Adbul Basit
Keywords: Sentiment analysis, twitter, Urdu, ontologybased, machine learning.
The rise of Web 2.0 has made it easier for people to share their opinions online. A significant application of Web 2.0 is micro-blogging. The proliferation of social media and related services, such as Twitter, has made it easier than ever to voice one's opinion on a wide range of topics. As a result, micro-blogging platforms are rich in data for use in Sentiment Analysis and opinion mining. Due to Twitter's character constraint of 140 characters, simple text-based analysis in this area typically proven unproductive because tweets do not consist of consistent and syntactically valid terms. In this research, we propose combining ontology with sentiment analysis to better understand the tenor of Urdu tweets. The innovation of the proposed method is the Ontology-based filtering of irrelevant tweets, followed by the classification of sentiment by machine learning classifiers. Experiment results reveal that Multinomial Nave Bayes outperforms SVM-SMO at analysing the sentiment of tweets written in Urdu, with an accuracy of 71% versus 67%.