In this task, we focused on several classifiers, applying them to analyzing the relationship between rating and sentiment of products' reviews, and understanding the types of errors a classifier makes.
Then, we explored this application further, training a sentiment analysis model using a set of key polarizing words, verify the weights learned to each of these words, and compare the results of the previous classifier with those of the one using all of the words. Using so few words in our model will hurt our accuracy, but help us interpret what our classifier is doing by diving into the difference in performance between the models.
- Data: amazon_baby.sframe
Or if you are using pandas and scikit-learn, you can read amazon_baby.csv.
- Code: Analyze product sentiment.ipynb