A Systematic Literature Review on Urdu Sentiment Analysis

Abstract

Sentiment Analysis (SA) can be used to make decisions, conclusions, and changes. It can also be used to recommend suitable solutions for different problems related to different domains. Sentiment analysis is an application of Natural Language Processing (NLP) that is getting grown in attention from many past years. There is wide use of Urdu language in giving opinions so the Urdu language demands Sentiment analysis too. In this study, I have performed a systematic literature review on Urdu sentiment analysis (USA). This SLR is targeting specific research questions and then contributions are analyzed accordingly. A total of 22 published articles on Urdu sentiment analysis are reviewed after searching 8 conferences and 9 journals. The major concerns in Urdu sentiment analysis are preprocessing techniques, feature generation processes, classification methods, and lexicon resources. This SLR is highlighting all of the used techniques in these concerns and the most adopted methods by the researchers from these techniques. Finally, detailed analysis and discussion on all these concerns are provided and the taxonomy of Urdu sentiment analysis is presented too. Furthermore, there are some future suggestions in the context of Urdu sentiment analysis.