Sentiment Analysis with Twitter to determine campaign success

The success of a business or marketing campaign can be difficult to quantitatively measure. A sentiment analysis of social media postings could be used to gauge the success and to increase the use of a product or service. Using MATLAB, a team developed a Twitter Sentiment Analysis model for the word ‘matlab’.

The Datafeed Toolbox was used to gather tweets that were used for testing and training of the model. AI methods such as Long Short-Term Memory (LSTM) and a Recurrent Neural Networks (RNN) were used from the Deep Learning Toolbox in combination with methods from the Text Analytics Toolbox to train an analytical model.










Deep Learning in Action – part 2