8/5/2013
Tableau Data Visualization on Bank Sentiment and CFPB Analysis
This data visualization is showing the consumer sentiment regarding banks using tweets from Twitter and consumer complaints in the Consumer Financial Protection Bureau's (CFPB) complaint database.
The tweets used in this visualization were part of a dataset used to build a sentiment scoring algorithm.
The manual tweets were used as the response variable and we built several models to predict a sentiment score (the final model is an SVM).
I used the manual tweets in this data visualization to show the difference in sentiment by institution. After noticing an interesting pattern in the data I then compared the results to the complaints in the CFPB database.
The results were very interesting. Bank of America had the lowest consumer sentiment and they also represented over 20% of the complaints in the CFPB complaint database.
Click to image below to open the data visualization in Tableau Public.
