Abstract
Performing natural language processing on a text document is an active research area. Social media includes valuable information resources in various languages. These information resources include reviews, comments, tweets, posts, opinions, articles and other text resources which could be analysed to explore people's opinions, attitudes, emotions and sentiments towards various subjects and commodities. However, there is lack of contributions addressing sentiment analysis on automobile reviews in Gulf Cooperation Council (GCC) dialects and in the Arabic language in general. Moreover, the lack of available annotated datasets in the Arabic language, which are targeting specific domains (such as automobiles), and the limited focus on analysing the sentiments in Arabic regional dialects created a gap. These factors motivated us to conduct the research that we report in this paper. Furthermore, the limited adoption of techniques and algorithms related to natural language processing and machine learning is noticed in the current efforts that are targeting sentiment analysis in the Arabic language, which once adopted could provide the opportunity of enhancing the performance of sentiment analysers. Therefore, this research attempts to cover these gaps to a certain extent.