Simple Node.js application to fecth tweets by keywords and apply Stanford Core NLP sentiment analysis on their content.
You will need to fill in your twitter access token and consumer key and secrets. You can get them here.
You will also need to download and build Stanford CoreNLP from their GitHub repository, as well as the current models that can be found here.
Then place the models in CoreNLP root folder and run the CoreNLP server by typing java -mx4g -cp "stanford-corenlp.jar:stanford-english-corenlp-models-current.jar:lib/*" edu.stanford.nlp.pipeline.StanfordCoreNLPServer in a CLI.
Run npm install to install the dependencies then node server.js to run the application. It will be available on port 8080 (you can change it in file server.js).
You're all set !