Using the information you collected on movie ratings, implement a
Global Baseline Estimate recommendation system in R.
Most recommender systems use personalized algorithms like “content
management” and “item-item collaborative filtering.” Sometimes
non-personalized recommenders are also useful or necessary. One of the
best non-personalized recommender system algorithms is the “Global
Baseline Estimate”. The job here is to use the survey data collected and
write the R code that makes a movie recommendation using the Global
Baseline Estimate algorithm. Please see the attached spreadsheet for
implementation details.
In this assignment 2, I asked my friends and family to rate 6 movies on a scale of 1 to 5. I took their response and created a table. Below is the link to the Google Form. https://docs.google.com/forms/d/e/1FAIpQLSd3ht1QlE3axt5LtI3KMkc4skkR9r6wn8PqtPYvVQphLhuQug/viewform
library(tidyverse)## ── Attaching packages ─────────────────────────────────────── tidyverse 1.3.2 ──
## ✔ ggplot2 3.3.6 ✔ purrr 1.0.1
## ✔ tibble 3.1.8 ✔ dplyr 1.1.0
## ✔ tidyr 1.3.0 ✔ stringr 1.5.0
## ✔ readr 2.1.2 ✔ forcats 0.5.2
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag() masks stats::lag()library(DT)
library(RMySQL)## Loading required package: DBIlibrary(DBI)
dbconnect = dbConnect(RMySQL::MySQL(),
dbname='Movie Rating',
host='localhost',
port=3306,
user='root',
password = rstudioapi::askForPassword("Database password"))
rating <- dbGetQuery(dbconnect, "SELECT * FROM `Movie Rating`.`Movie Rating`")
datatable(rating, options = list(scrollX = TRUE))Below is a table of the movies that each person is rating.
movies <- dbGetQuery(dbconnect, "SELECT * FROM `Movie Rating`.Movie;")
datatable(movies, caption = 'Table 1: Movie title and Year', options = list(scrollX = TRUE))Here is table of first name, age, and gender of each person.
person <- dbGetQuery(dbconnect, "SELECT * FROM `Movie Rating`.Person;")
datatable(person, caption = 'Table 2: Information of Reviewers', options = list(scrollX = TRUE))Here is a table of each person’s rating of each movie. There are missing values for those people who has not seen a movie. 4 people have never seen “Everything Everywhere All at Once” and “Scream”. 2 people has never seen “Top Gun: Maverick”.
rating <- dbGetQuery(dbconnect, "SELECT * FROM `Movie Rating`.`Movie Rating`")
datatable(rating, caption = 'Table 3: Ratings of each movie.', options = list(scrollX = TRUE))Below is a table of the average rating of each movie and average rating for each user. Some people did not see certain movies. So there are missing values. I used “na.rm = TRUE” to deal with missing values. Avengers: Endgame has the highest average rating and Scream has the lowest average rating. Both movies “Everything Everywhere All at Once” and “Scream” had 4 missing values each. Yet, “Everything Everywhere All at Once” had a higher average rating than “Scream” did.
rating$user_avg <-apply(rating[,5:10],1,mean,na.rm=TRUE) # Calculate the average user rating
average_movie <- summarize_all(rating, mean, na.rm=TRUE) # Calculate average movie rating in a separate dataframe (Calculate the mean of each column)## Warning: There were 2 warnings in `summarise()`.
## The first warning was:
## ℹ In argument: `Name = (function (x, ...) ...`.
## Caused by warning in `mean.default()`:
## ! argument is not numeric or logical: returning NA
## ℹ Run ]8;;ide:run:dplyr::last_dplyr_warnings()dplyr::last_dplyr_warnings()]8;; to see the 1 remaining warning.average_rating <- rbind(rating, average_movie) # Combine the two dataframes above
average_rating[9,1] <- NA
average_rating[9,"Gender"] <- "average"
datatable(average_rating, options = list(scrollX = TRUE))Movies’s rating relative to average = Movie average - Mean Movie
User’s rating relative to average = User average - Mean Movie
# Calculate Each User's Rating Relative to Average
average_rating <- average_rating %>%
mutate(user_relative_rating = user_avg - average_rating[9,11])
average_rating <- average_rating %>%
add_row(Gender = "movie_relative_rating")
# Calculate Each Movie's Rating Relative to Average
average_rating[10,5] <- average_rating[9,5] - average_rating[9,11]
average_rating[10,6] <- average_rating[9,6] - average_rating[9,11]
average_rating[10,7] <- average_rating[9,7] - average_rating[9,11]
average_rating[10,8] <- average_rating[9,8] - average_rating[9,11]
average_rating[10,9] <- average_rating[9,9] - average_rating[9,11]
average_rating[10,10] <- average_rating[9,10] - average_rating[9,11]There are missing values for those people who has not seen a movie. 4 people have never seen “Everything Everywhere All at Once” and “Scream”. 2 people has never seen “Top Gun: Maverick”.
For those missing values, we estimate their rating for those movies using the “Global Baseline Estimate”.
Global Baseline Estimate = Mean Movie Rating + Movies’s rating relative to average + User’s rating relative to average
for(i in 1:ncol(average_rating)){
if(is.numeric(average_rating[[i]])){
na <- is.na(average_rating[[i]])
average_rating[na, i] <- average_rating[9,11] + average_rating[10,i] + average_rating[na, 12]
}
}
average_rating[,5:12] <-round(average_rating[,5:12],2)
datatable(average_rating, options = list(scrollX = TRUE))The following people did not watch “Everything Everywhere All at Once”: Mel, Ivan, Jack, and Bob. Out of all four people, Ivan has the highest estimate.
The following people did not watch “Scream”: Jenny, James, Ivan, and Jack. Out of all four people, Jenny has the highest estimate.
Ivan did not watch “Everything Everywhere All at Once” and “Scream”. However, he has a higher estimate for the first movie. He is likely to like this movie more.
The link below shows me how to use a loop to replace a NA value. The example in the link replaces the NA value with the median of the column. I wanted to replace the NA value with estimated movie rating for a user.