The Java Sentiment Miner is a web application that mines Tweets from Twitter and analyzes them for sentiment. Users can enter a hashtag into the application and receive a pie chart of how many of the associated Tweets are positive, negative, or neutral.
-
Clone the project into your desired location.
-
Login to your Twitter API account and navigate to Apps. Select the App option. Fill out the forum that appears and click the create App button.
-
Go to the details page of your newly created Twitter app. Navigate to permissions and set access permissions to read-only.
-
Navigate to the keys and tokens section of your Twitter app. Generate your Consumer API keys and your Access Tokens. Copy the resulting keys into the twitter4j.properties file in the project.
-
Place a positive and negative word lexicon in the src/main/resources/lexicons folder. If you do not have a set of lexicons you would like to use then you may choose to use the Liu and Hu Opinion lexicon that is provided in the lexicon folder. Get the absolute paths of of your chosen lexicons and paste them into the config.properties file.
-
build the gradle project.
-
Your project is now ready to go!
-
Run the Application class in src/main/java/app/ to start the web server at
localhost:8080. -
Navigate to
localhost:8080to use the application. If you see the image below then everything is working.
On every page of this app you can submit a hashtag into a form and be redirected to a page displaying the associated sentiment data. The results look like the image below.
