Bonus question (5 points)   When an economist wants to estim…

Bonus question (5 points)   When an economist wants to estimate the effect of consumer emotion in online product review on product sales, we need to apply (1) ___________ (a. TF-IDF, b. sentiment analysis, c. Topic modeling; 2 points) for review text data with regression for inference. When an economist wants to predict real-time stock prices from real-time news text data, we can generate frequency-based high-dimensional spare metrics using (2)__________ (a. TF-IDF, b. sentiment analysis, c. topic modeling; 2 points) for a supervised learning model. Lastly, we want to identify the main topics and issues on social media users during spring break from the social media comment text dataset; in this case, we can use (3) ___________ (a. TF-IDF, b. sentiment analysis, c. topic modeling; 1 point) to find topics and their shares from the comments from social media users to generate independent variables for regression analysis. * note: topic modeling is a Latent Dirichlet Allocation (LDA)