darlansf1 / opinionmeter Goto Github PK

Opinion-meter is a system designed to enable any person to identify the main properties under evaluation in a collection of text files as well as the polarities of the opinions associated to those properties, such as negative and positive evaluations of the properties of whatever entity the opinions are direct to. The system uses some of the most classical approaches present in the literature of Sentiment Analysis to identify the aspects (the properties) and classify the polarities of the sentiments (or opinions) associated and the user is free to choose which of the methods to use to analyze their data, being also able to choose multiple methods in order to compare their results. In addition, the system uses machine translation to enable the analysis of texts in up to 30 different languages.