Urdu Sentiment Analysis

Abstract

There is a need for the corporate world to know about the sentiments and feedback of their customers to keep their business running profitable. The problem is, manually it can be a very hectic process to go through all the feed backs. Sentiment Analysis steps up in this situation which can automatically tell whether the sentiment in the text is positive, negative, or neutral. While there is software that do this job of determining the emotions in a text in English language, there is no single application that can read the sentiments of a text written in Urdu language and classify it into positive or negative. We made an application that could make this happen. Using the well-known techniques of Natural Language Processing (NLP) and language classification we adopted the lexicon-based approach to design our algorithm and implement it. The results were above average and are a good foundation for further work to be done.