Use the dataset songdata.csv to answer the following questions: Which band/artist says love the most (per song)? The least? Who is the most negative band in the data set (in terms of sentiment)? Positive? Which band has the "best" vocabulary? First define what "best" means and then write tidytext code to determine the answer. Can you predict who a song is by? Take Katy Perry and Taylor Swift (or 2 other artists with at least 50 songs each) and come up with 5-10 features for each song. Split data into train and test and see how accurate a model can be (use glm or rf).
Use the dataset songdata.csv to answer the following questions:
-
Which band/artist says love the most (per song)? The least?
-
Who is the most negative band in the data set (in terms of sentiment)? Positive?
-
Which band has the "best" vocabulary? First define what "best" means and then write tidytext code to determine the answer.
-
Can you predict who a song is by? Take Katy Perry and Taylor Swift (or 2 other artists with at least 50 songs each) and come up with 5-10 features for each song. Split data into train and test and see how accurate a model can be (use glm or rf).
Submit a single pdf or word report that includes a short paragraph summary of what you did and your answer to each question, any figures/plots/tables generated, and the R code used at the end of the document.
Code needs to be in R
Download songdata.csv from github using this link -
https://github.com/ugis22/music_recommender/tree/master/content%20based%20recommedation%20system
Trending now
This is a popular solution!
Step by step
Solved in 2 steps
