Abstract:
Nowadays, a lot of information is valuable for people who want to predict and decide the market or economic or political or many different areas. Because of the information age, customers are also no longer interested in trying it out for themselves. They are judging a product or an item based on the opinions of other customers. It is valuable in social media monitoring because several reviews that were written by a lot of users in social media mainly include forecasting to judge whether it is good or not. Many researchers have done sentiment analysis to obtain a summarized report for an opinion about different contexts. Moreover, researchers want to build a lexicon with their language. Some researchers in Myanmar have also implemented sentiment analysis to obtain Myanmar lexicon in some areas such as restaurants, hotels, movies. This system focuses on building a lexicon with Myanmar Text in Movie comments and predicts Movie comments on Facebook whether these are positive or negative or neutral by using a lexicon-based approach. Researchers from Myanmar have difficulties with any annotated data for Myanmar Language in sentiment analysis. So, this system also completely meets a challenge with no annotated data. This system uses word segmentation, removal of punctuations, and correction of some words in pre-processing steps. In the evaluation steps, four metrics are calculated accuracy, precision, recall, and f- measures to know the overall performance and accuracy of the proposed system.
