Building Domain Specific Sentiment Lexicon: Survey

Sentiment analysis is a natural language processing task whose aim is to identify and extract opinion expressions and subjective words or phrases from a piece of text and then determine the polarity or strength of these expressions/phrases. One basic, yet vital, resource to consider when conducting sentiment analysis is the sentiment lexicon. A sentiment lexicon is essentially a list of opinions/emotion bearing words or phrases divided into two categories: positive and negative to date, existing sentiment lexicons for the Arabic language are limited. Moreover, these lexicons do not have a wide coverage of all domain’s specific terms. The aim of this research is to explore different ways of automatically generating Arabic sentiment lexicons with a focus on domain specific sentiment lexicons. Moreover, since sentiment words can have different polarities when found in different domains, human validation will be used in order to validate the obtained results from the proposed model.