Opinions are central to almost all human activities as they could influence others’ behaviors specially when making a decision. Nowadays, in every minute, people share their personal opinions on the social media; e.g. reviews on recently products, emotional events in their lives, or simple matters of opinion on everyday issues. SentiWordNet is commonly used to solve the problem of sentiment classification for personal opinion. Due to the ubiquitous domain diversity and absence of domain-specific prior knowledge, domain-specific SentiWordNet has played an important role in opinion mining system. In this paper, an approach based on Term Frequency-Inverse Document Frequency(TF-IDF) is proposed to construct a domain specific SentiWordNet for Myanmar language. The proposed system can provide the positive, negative opinions of product review from comments of social media The proposed approach evaluated good accuracy for the domain specific SentiWordNet for Myanmar Language by using TF-IDF.
