UAMSA: Unified Approach for Multilingual Sentiment Analysis Using GATE

With the pervasive use of technology and internet, the web related content has increased dynamically. Sentiment analysis is an active research area of web content mining and it deals with analyzing people's opinion towards different entities ranging from blogs, media articles to social media sites. Research community has actively participated in improving methods and techniques to correctly classify the monolingual (one language) and cross lingual (resource rich language to resource poor language) sentiment analysis. However, multilingual (more than one language) sentiment analysis is less explored area. There is a lack of resources to perform multilingual sentiment analysis activities such as data preprocessing, feature extraction and classification under unified tool. In addition, some research studies tried to develop resources for multilingual sentiment analysis, but these are not publicly available. These limitations served as motivation for this study. The contributions of this paper are as follow. 1) We have proposed a new plugin named as UAMSA (Unified Approach for Multilingual Sentiment Analysis) within GATE paradigm which is able to 1) perform different activities for multilingual sentiment analysis using one tool. 2) This tool will be publicly available as GATE is one of the popular open source tools for text analysis available in many formats and size. UAMSA will help users to classify the sentiments of people sharing their opinions in multiple languages effectively.