See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/359747031

Data Analysis on Netflix datasets

Preprint · March 2022

DOI: 10.13140/RG.2.2.12909.60649

CITATIONS READS
0 12,934

3 authors, including:

Vybhav Achar Bhargav Seongwoo Choi

University of California, Davis University of California, Davis
1 PUBLICATION 0 CITATIONS 6 PUBLICATIONS 1 CITATION

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Seongwoo Choi on 05 April 2022.

The user has requested enhancement of the downloaded file.

 STA 220 Final Project - Data Analysis on Netflix datasets

Vybhav Achar Bhargav, Seongwoo Choi, David Haddad

Abstract

Exploring datasets of Netflix for Future Release of TV shows and Movies on the Platform. In this
project, we are going to explore the dataset from Kaggle and we would like to find out how long the
Netflix platform takes a movie or a TV show to release on its platform, how many movies and TV
shows are released in specific time frame, how many movies and TV shows are release in the recent
ten years on the platform, and what were the top 10 genres that the audience of the Netflix platform
liked the most. From here, we would like to apply a machine learning approach to understand the data
fully and provide a great solution where the platform should be headed to. From our data analysis we
conducted on R markdown, we have discovered that there were a wide variety of genres that movie
directors produced worldwide and we have observed many cast members and genres they were in.

Data Analysis on the Netflix Datasets

Motivation

Netflix is the largest online movie and TV show streaming service on the planet. Its service is widely
available in many countries including but not limited to the United States, India, South Korea, Japan, and
many more. The service was first introduced as a DVD rental service on the Internet and later, the founder
and CEO of the company Reed Hastings transitioned to a revolutionary way of delivering movies and TV
shows through its website allowing many users to directly stream their favorite contents on their Internet-
enabled devices including desktop computers, laptops, tablet PCs, mobile phones, and many more. With
its a whole new approach of delivering shows and movies, the sales of Netflix went up exponentially. Since
then, the platform created its own recommender system to understand what types of movies and TV shows
the users would like to watch, what kind of style of cinematography they liked the most, and how they
consume their favorite TV shows. With such analysis, the company released ‘The House of Cards’, which
became a huge success in the history of streaming service providers. With the power of data analysis,
more users were attracted to the platform, and many users tend to spend most of their time watching
shows and movies on Netflix. With this approach, we would like explore the dataset to understand the
trend of movies and TV shows on Netflix.

Introduction

We first wanted to get an overview of the dataset that we were dealing with. First we loaded up tidyverse
for a simple data anlaysis purpose. We got the dataset from Kaggle and we are going to utilize data that
the Kaggle website provides to understand the trend of movies and TV shows released on the platform.
This dataset consists of

1
From the code, we could see the column names that the CSV file contains. We will utilize the following
columns to understand what movies and TV shows were released in specific year, what genres they were,
date when they were released and the rating the audience gave and so on.
From the column names, we could observe that there are twelve columns: show_id, type, title, director,
cast, country, date_added, release_year, rating, duration, listed_in, description.
We will first demonstrate the overview of the dataset. Each column contains 8807 in length except for
release year in interquartile range.

summary(df_netflix)

## show_id type title director

## Length:8807 Length:8807 Length:8807 Length:8807
## Class :character Class :character Class :character Class :character
## Mode :character Mode :character Mode :character Mode :character
##
##
##
##
## cast country date_added release_year
## Length:8807 Length:8807 Min. :2008-01-01 Min. :1925
## Class :character Class :character 1st Qu.:2018-04-20 1st Qu.:2013
## Mode :character Mode :character Median :2019-07-12 Median :2017
## Mean :2019-05-23 Mean :2014
## 3rd Qu.:2020-08-26 3rd Qu.:2019
## Max. :2021-09-25 Max. :2021
## NA’s :98
## rating duration listed_in description
## Length:8807 Length:8807 Length:8807 Length:8807
## Class :character Class :character Class :character Class :character
## Mode :character Mode :character Mode :character Mode :character
##
##
##
##

From the overview of the dataset, we could tell that the earliest year recorded as the movie or the TV
show is 1925 and the latest content that is on the platform is 2021. With the information of directors,
actors/actresses, movies, TV shows, rating, duration of each content and more, we can understand how
long does Netflix take to upload the content on the platform, which genre is most popular on the platform,
which actor/actresses are most popular and how genre popularity changed the movie or TV shows that an
actor/actress appear on the genre over time. We will explore the trend of Netflix using R to figure it out.

Project Background

The report presents a data analysis research and coordination activity one in the Netflix dataset found on
Kaggle to establish a new data exploration on the trend of movies on the platform. When Netflix was first

2
launched, it did not have any data analysis to understand the trend of audience/users using the platform
as previously mentioned. As time passes by, the importance of utilizing data analysis started to emerge,
and the biggest hit shows were released on the platform after scrutinizing the data the company collected
from the users. In the project, the team of three is working on a same procedure to understand the trend
of the platform and attempt to understand the average time the platform takes to upload the contents, the
top ten film directors in top ten countries where the platform is streaming, and the top genres that the top
actors/actresses are in. The team will, then, attempt to understand the overall trend of the Netflix platform
according to the dataset the team got from Kaggle.

Methodology

Before initiating the data analysis, the dataset has been separated into two different sets: one is for movies
only and the another dataset is TV shows only. Every date on the dataset is transformed to day and year
format.

df_netflix <- transform(df_netflix,

date_added=as.Date(
date_added, tryFormats =
c("%B %d,%Y")))
movies <- subset(df_netflix, type=="Movie")
tvs <- subset(df_netflix, type == "TV Show")

Release year is very important aspect of learning when a content on the platform is released and the
average time the platform takes to upload the content.

# Plot of number of movies release per year

numRelease <- movies %>% group_by(release_year)
tallyRelease <- tally(numRelease);

# 10 Year running avg of movie releases

tallyRelease$avg <- stats::filter(
tallyRelease$n, rep(1/10, 10),
method="convolution", sides=1);
plot(tallyRelease$release_year, tallyRelease$n,
type = "h", xlab = "Release Year", ylab = "Number of Releases")
lines(tallyRelease$release_year, tallyRelease$avg, col="red");

3
 600
Number of Releases

400
200
0

1940 1960 1980 2000 2020

Release Year

The plotted graph indicates that most of the movies were released after the year 2000s. From 2000 to 2020,
there is an exponential growth in the graph, which indicats that the number of releases that the movies
were introduced to the platform started to grow higher.

# Plot of number of movies release per year

numDateAdded <- movies %>% group_by(date_added)
tallyDateAdded <- tally(numDateAdded);

plot(tallyDateAdded$date_added,
tallyDateAdded$n, type = "h", xlab =
"Release Date", ylab = "Number of Releases")

4
 100
80
Number of Releases

60
40
20
0

2010 2015 2020

Release Date

The graph above indicates the number of movies released in specif time frame. From 2010 to 2020, there
are high increases in numbers of releases on the platform. Due to this reason, from the rest of the data
analysis, the team will dissect the time frame from 2008 to 2020 to understand the overall trend of releases
on the platform.

movies$year_added_to_netflix <- strtoi(

format(movies$date_added, format="%Y"))
movies$diff = movies$year_added_to_netflix - movies$release_year

v <- movies %>% select(diff, year_added_to_netflix, release_year)

#pick the variable

# set up cut-off values

breaks <- c(0,2,5,10,15,20,25,30,35,40)
# specify interval/bin labels
tags <- c(
"[0-2)",
"[2-5)",
"[5-10)",
"[10-15)",
"[15-20)",
"[20-25)",
"[25-30)",

5
 "[30-35)",
"[35-40)",
">40"
)

vgroup <- as_tibble(v) %>%

mutate(tag = case_when(
diff < 2 ~ tags[1],
diff >= 2 & diff < 5 ~ tags[2],
diff >= 5 & diff < 10 ~ tags[3],
diff >= 10 & diff < 15 ~ tags[4],
diff >= 15 & diff < 20 ~ tags[5],
diff >= 20 & diff < 25 ~ tags[6],
diff >= 25 & diff < 30 ~ tags[7],
diff >= 30 & diff < 35 ~ tags[8],
diff >= 35 & diff < 40 ~ tags[9],
diff >= 40 ~ tags[10],
))

vgroup$tag <- factor(vgroup$tag,

levels = tags,
ordered = FALSE)

We need to understand and separate the group of releases in specific year. We also need to understand the
average time the platform releases the contents on the website and how much would it need to release the
content.

movies %>% count(listed_in, country, sort=TRUE) %>% slice(1:10)

## listed_in country n
## 1 Documentaries United States 249
## 2 Stand-Up Comedy United States 209
## 3 Comedies, Dramas, International Movies India 120
## 4 Dramas, International Movies India 118
## 5 Dramas, Independent Movies, International Movies India 108
## 6 Children & Family Movies, Comedies United States 90
## 7 Dramas United States 88
## 8 Comedies United States 84
## 9 Children & Family Movies United States 80
## 10 Dramas, Independent Movies United States 74

The top 10 genres that are released in this time frame indicates that the United States released the most
documentaries on the platform and the naiton also produced stand-up comedies. India came the second
place when it comes to which international country released the most contents on the platform.

6
ggplot(data = as_tibble(vgroup$tag),
mapping = aes(x=value)) +
geom_bar(fill="bisque",color="white",alpha=0.7) +
stat_count(
geom="text",
aes(label=sprintf("%.2f%%",
(..count../length(vgroup$tag))*100)),
vjust=-0.5) +
labs(x='Time to release movie on netflix') +
theme_minimal()

49.62%
3000

2000
count

18.12%
1000
12.82%

7.70%
3.64% 2.71%
1.63% 0.88% 0.98% 1.91%
0

[0−2) [2−5) [5−10) [10−15) [15−20) [20−25) [25−30) [30−35) [35−40) >40
Time to release movie on netflix

The graph above indicates time to release movies on Netflix. Many of the contents were uploaded within
0-2 years on the platform after they are released to the public. The graph also indicates that many Netflix
original movies are released and uploaded simultaneously on the platform upon their debut.

movies_country_seperated = movies %>%

separate_rows(country, sep=",") %>%
mutate(across(where(is.character), str_trim))

movies_genre_seperated <- movies %>%

separate_rows(listed_in, sep=",") %>%
mutate(across(where(is.character), str_trim))

7
top_genre_per_country <- movies_country_seperated %>%
filter(!is.na(country)) %>%
distinct(show_id, .keep_all = TRUE) %>%
count(country, sort=TRUE) %>% head(n=10)

tg <- top_genre_per_country %>%

pull(country) %>%
sapply(function(cnt){
movies_genre_seperated %>%
filter(country == cnt) %>%
count(listed_in, sort=TRUE) %>%
slice(1:5) %>%
pull(listed_in)
})

as.data.frame(tg)

## United States India United Kingdom

## 1 Dramas International Movies Documentaries
## 2 Comedies Dramas International Movies
## 3 Documentaries Comedies Dramas
## 4 Independent Movies Independent Movies Comedies
## 5 Children & Family Movies Action & Adventure Independent Movies
## Canada France Spain
## 1 Comedies International Movies International Movies
## 2 Children & Family Movies Dramas Dramas
## 3 International Movies Documentaries Comedies
## 4 Dramas Comedies Thrillers
## 5 Documentaries Action & Adventure Documentaries
## Egypt Nigeria Mexico
## 1 International Movies International Movies International Movies
## 2 Comedies Dramas Dramas
## 3 Dramas Comedies Stand-Up Comedy
## 4 Action & Adventure Romantic Movies Comedies
## 5 Romantic Movies Thrillers Independent Movies
## Japan
## 1 International Movies
## 2 Anime Features
## 3 Action & Adventure
## 4 Dramas
## 5 Children & Family Movies

The result above demonstrates the top ten countries with top five genres that the audience watch. If we
explore only the United States, we can observe that top five genres are Drama, Comedies, Documentaries,
Independent Movies, and Family movies. Since we have observed which genres are popular in each country
on the list, we will then explore the top genre leased in each year from 2008 to 2020.

8
movies_genre_seperated %>%
filter(country=="United States") %>%
group_by(year_added_to_netflix) %>%
slice_max(listed_in, n = 1) %>%
distinct(listed_in) %>%
rename(top_genre=listed_in) %>%
rename(year=year_added_to_netflix)

## # A tibble: 14 x 2
## # Groups: year [14]
## top_genre year
## <chr> <int>
## 1 Thrillers 2008
## 2 Horror Movies 2009
## 3 Horror Movies 2010
## 4 Thrillers 2011
## 5 Documentaries 2012
## 6 Stand-Up Comedy 2013
## 7 Stand-Up Comedy 2014
## 8 Thrillers 2015
## 9 Thrillers 2016
## 10 Thrillers 2017
## 11 Thrillers 2018
## 12 Thrillers 2019
## 13 Thrillers 2020
## 14 Thrillers 2021

Most of the times, thriller seemed to be the most popular genre from year 2008 to 2020 in the United States
despite the fact that they are not one of top five genres that the United States audience like to watch. One
thing to note is that the thriller genre may mixed up with the other genres. For instance thriller, comedy
could be the genre that the United States audience watched in 2008.
From now on, we will see which directors produced the most genres in the United States from 2008 to
2020.

Top 10 Directors that produced movies in the United States

1) What is the top genre that the top 10 directors produced in the United States?

The top 10 directors that produced movies in the United States have made mostly stand-up comedies. This
makes sense due to the little time it takes to record a live performance show and edit them as compared
to hiring cast and writing scripts for movies that could take years.

2) How many movies did the top 10 directors make that were produced in the United States?

The top director produced 15 movies, and the top genre was stand-up comedy which makes sense because
of the same reason in the question above

9
directors_produced_movies = movies %>%
separate_rows(director, sep=",") %>%
mutate(across(where(is.character), str_trim)) %>%
filter(country=="United States")

top_directors <- directors_produced_movies %>%

filter(!is.na(director)) %>%
group_by(director) %>%
count(sort=TRUE) %>%
head(n=10)

temp <- top_directors %>%

pull(director) %>%
sapply(function(d){
directors_produced_movies %>%
filter(director == d) %>%
slice_max(listed_in, n=1) %>%
pull(listed_in) %>% head(1)
})

director_top_genre = as.data.frame(temp)
final_table <- cbind(top_directors, genre=director_top_genre$temp)
final_table <- final_table %>%
rename(total_movies_produced=n) %>%
rename(top_directors=director)
final_table = as.data.frame(final_table)
final_table

## top_directors total_movies_produced genr

## 1 Jay Karas 15 Stand-Up Comed
## 2 Marcus Raboy 14 Stand-Up Comed
## 3 Jay Chapman 12 Stand-Up Comed
## 4 Shannon Hartman 9 Stand-Up Comed
## 5 Martin Scorsese 8 Thriller
## 6 Troy Miller 8 Stand-Up Comed
## 7 Lance Bangs 7 Stand-Up Comed
## 8 Leslie Small 7 Stand-Up Comed
## 9 Ryan Polito 7 Stand-Up Comed
## 10 Robert Rodriguez 6 Children & Family Movies, Comedie

The chart indicates that Jay Karas made 15 Stand-Up comedy movies from 2008 to 2020. Marcus Raboy
made fourteen stand up comedy movies and so on. The top ten directors are from the United States. The
following shows the international directors that produced the movies in specific time frame.

Top 10 Directors that produced movies in international countries

1) What is the top genre that the top 10 directors produced in international countries?

10
The top 10 directors that produced movies in international countries have made mostly stand-up comedies.
This makes sense due to the little time it takes to record a live performance show and edit them as compared
to hiring cast and writing scripts for movies that could take years. But many international movies have
taken second spot for most movies directed unlike in the United States.

2) How many movies did the top 10 directors make that were produced in international coun-
tries?

The top director produced 21 movies, and the top genre was stand-up comedy. Third top director made
13 movies that had the most genre that is international movies and romantic movies. So more diversity
internationally compared to US produced movies top directors.

Print only International in 2008-2021 and display names of directors who

produced the most movies and TV shows

This will answer in which country a director produced the most releases
of movies and TV shows

directors_produced_movies = movies %>%

separate_rows(director, sep=",") %>%
mutate(across(where(is.character), str_trim)) %>%
filter(country!="United States")

top_directors <- directors_produced_movies %>%

filter(!is.na(director)) %>%
group_by(director) %>%
count(sort=TRUE) %>%
head(n=10)

temp <- top_directors %>%

pull(director) %>%
sapply(function(d){
directors_produced_movies %>% filter(director == d) %>% slice_max(listed_i
})

director_top_genre = as.data.frame(temp)
final_table <- cbind(top_directors, genre=director_top_genre$temp)
final_table <- final_table %>% rename(total_movies_produced=n) %>% rename(top_
final_table = as.data.frame(final_table)
final_table

## top_directors total_movies_produced
## 1 Jan Suter 21

11
## 2 Raúl Campos 19
## 3 Cathy Garcia-Molina 13
## 4 Youssef Chahine 12
## 5 David Dhawan 9
## 6 Yılmaz Erdoğan 9
## 7 Anurag Kashyap 8
## 8 Hakan Algül 8
## 9 Hanung Bramantyo 8
## 10 Johnnie To 8
## genre
## 1 Stand-Up Comedy
## 2 Stand-Up Comedy
## 3 International Movies, Romantic Movies
## 4 Dramas, International Movies, Romantic Movies
## 5 Comedies, International Movies, Sci-Fi & Fantasy
## 6 Dramas, International Movies, Romantic Movies
## 7 International Movies, Thrillers
## 8 Comedies, International Movies, Sports Movies
## 9 Dramas, International Movies, Romantic Movies
## 10 Dramas, International Movies

Jan Suter made twenty-one stand-up comedy movies and Raúl Campos followed after that with nineteen
movies of stand-up comedy genre. Cathy Garcia-Molina produced international movies mostly romantic
movies. The top ten international movie directors produced mostly dramas, and comedies throughout
2008 to 2020. The following explores which cast members appeared in movied produced in United States
between 2008-2021.

Cast Members United States

Top Cast Members that appeared in movies produced in United States between 2008-2021

1) Who are the top 10 cast members that appeared in movies produced in USA?
2) What genres of movies do the top cast members appear in?
3) What is number of movies for each genre that the top cast members appear in?
4) What is are the movie genres that the top movie stars appear in over the years?

Answers to those questions are below in tables and plots

usa_produced_movies = movies %>%

separate_rows(cast, sep=",") %>%
mutate(across(where(is.character), str_trim)) %>%
filter(country=="United States")

top_cast <- usa_produced_movies %>%

filter(!is.na(cast)) %>%
distinct(show_id, .keep_all = TRUE) %>%
group_by(cast) %>%

12
 count(sort=TRUE) %>%
head(n=10) %>%
rename(num_movies=n)

top_cast

## # A tibble: 10 x 2
## # Groups: cast [10]
## cast num_movies
## <chr> <int>
## 1 Adam Sandler 20
## 2 Nicolas Cage 10
## 3 Kevin Hart 9
## 4 Eddie Murphy 8
## 5 Jeff Dunham 8
## 6 Will Smith 8
## 7 Harrison Ford 7
## 8 John Travolta 7
## 9 Kristen Stewart 7
## 10 Ben Stiller 6

all_genres_top_cast<- usa_produced_movies %>%

filter(!is.na(cast)) %>%
filter(!is.na(listed_in)) %>%
mutate(across(where(is.character), str_trim)) %>%
group_by(cast)

# Get a count of each genre associated with cast member

get_genres_for_cast <- function(c) {
count_genres_top_cast <- all_genres_top_cast %>%
filter(cast == c) %>%
group_by(listed_in) %>%
count(sort=TRUE)

return (count_genres_top_cast)
}

get_genres_with_years_for_cast <- function(c) {

genre_year <- all_genres_top_cast %>%
select(cast, listed_in, release_year) %>%
filter(cast == c) %>%
group_by(release_year)

return (
genre_year[order(genre_year$release_year), ] %>%
group_by(listed_in, release_year))
}

13
Adam Sandler, one of the movie stars in the United States, was in twenty movies in this time frame, and
Nicolas Cage followed after that with ten movies. Kevin Hart were nine movies and so on. We will plot
those three movie stars to understand their appearance in those movies and then we will analyze if they
change their genre over time.

plot_genre_box <- function(c) {

data <- get_genres_for_cast(c)

graph <- ggplot(data, aes(x=n, y=listed_in, fill=listed_in)) +

geom_bar(stat="identity") +
ggtitle(paste(c,"'s different genres")) +
labs(x="num releases", y="genre", col = "Genre") +
theme(axis.text.y=element_blank(),axis.ticks.y=element_blank())+
geom_text(aes(label=n), hjust=1.6, color="white", size=3.5)

graph
}

plot_release_genre_points <- function(c, s, e) {

data <- get_genres_with_years_for_cast(c)
g <- ggplot(data, aes(x=listed_in, y=release_year, color=listed_in)) +
ggtitle(paste(c, ", Genres through the ages")) +
labs(x="genre", y="release year", col = "Genre") +
geom_point() +
scale_y_continuous(limits=c(s, e), breaks=seq(s,e,1)) +
theme(axis.text.x=element_blank(),axis.ticks.x=element_blank())

g
}

cast_member="Adam Sandler"
plot_genre_box(cast_member)

14
 Adam Sandler 's different genres

1 listed_in

1 Action & Adventure, Comedies

Children & Family Movies, Comedies
1 Comedies
Comedies, Dramas, Independent Movies
2
genre

Comedies, Horror Movies

1 Comedies, Romantic Movies
Comedies, Sci−Fi & Fantasy
1
Comedies, Sports Movies
7 Dramas, Thrillers
Stand−Up Comedy
2

0 2 4 6
num releases

plot_release_genre_points(cast_member,1998, 2021)

15
 Adam Sandler , Genres through the ages
2021
2020
2019
2018
Genre
2017
2016 Action & Adventure, Comedies
2015
2014 Children & Family Movies, Comedies
2013 Comedies
2012
release year

2011 Comedies, Dramas, Independent Movies

2010
Comedies, Horror Movies
2009
2008 Comedies, Romantic Movies
2007
Comedies, Sci−Fi & Fantasy
2006
2005 Comedies, Sports Movies
2004
2003 Dramas, Thrillers
2002 Stand−Up Comedy
2001
2000
1999
1998

genre

Adam Sandler first appeared in 1998 with a comedy genre movie and then he continued to appear many
comedy movies after that. He did not seem to change his genre over time and the bar chart indicates that
he was in thirteen comedy related movies.

cast_member="Nicolas Cage"
plot_genre_box(cast_member)

16
 Nicolas Cage 's different genres

2
listed_in
1 Action & Adventure
Action & Adventure, Children & Family Movies
2 Action & Adventure, Comedies
genre

Action & Adventure, Dramas

2
Action & Adventure, Sci−Fi & Fantasy

1 Children & Family Movies, Comedies

Dramas, Independent Movies
1 Dramas, Thrillers
Thrillers
1

0.0 0.5 1.0 1.5 2.0

num releases

plot_release_genre_points(cast_member,1998, 2021)

17
 Nicolas Cage , Genres through the ages
2021
2020
2019
2018
2017 Genre
2016
2015 Action & Adventure
2014
Action & Adventure, Children & Family Movies
2013
2012
release year

Action & Adventure, Comedies

2011
2010 Action & Adventure, Dramas
2009 Action & Adventure, Sci−Fi & Fantasy
2008
2007 Children & Family Movies, Comedies
2006 Dramas, Independent Movies
2005
2004 Dramas, Thrillers
2003
Thrillers
2002
2001
2000
1999
1998

genre
According to the dataset, Nicolas Cage was in many different genres through the years 2000s to 2020. He
played mostly in action and adventure movies and his tendency of apperance changed from Dramas and
Thriller to action later on. In 2004, he was in an action movie and then his latest appearance was in 2020
with an action and adventure movie.

cast_member="Kevin Hart"
plot_genre_box(cast_member)

18
 Kevin Hart 's different genres

1
listed_in
Children & Family Movies, Comedies
genre

Comedies
1
Comedies, Romantic Movies
Dramas
Stand−Up Comedy
1

0 2 4 6
num releases

plot_release_genre_points(cast_member,1998, 2021)

19
 Kevin Hart , Genres through the ages
2021
2020
2019
2018
2017
2016
2015
2014
Genre
2013
2012
release year

Children & Family Movies, Comedies

2011
2010 Comedies
2009 Comedies, Romantic Movies
2008
2007 Dramas
2006 Stand−Up Comedy
2005
2004
2003
2002
2001
2000
1999
1998

genre

Kevin Hart did not change his appearance in a comedy movie in the time frame. The first appearance in
2009 in a stand-up comedy made his position as a great comedian in the history of the United States film
industry. He appeared in other genres such as dramas and family movies, but his style of comedy did not
change over time.
The following explores the same procedure, but with international cast members.

Cast Member International

intl_produced_movies = movies %>%

separate_rows(cast, sep=",") %>%
mutate(across(where(is.character), str_trim)) %>%
filter(country!="United States")

top_cast_intl <- intl_produced_movies %>%

filter(!is.na(cast)) %>%
distinct(show_id, .keep_all = TRUE) %>%
group_by(cast) %>%
count(sort=TRUE) %>%
head(n=10) %>%
rename(num_movies=n)

20
top_cast_intl

## # A tibble: 10 x 2
## # Groups: cast [10]
## cast num_movies
## <chr> <int>
## 1 Shah Rukh Khan 25
## 2 Akshay Kumar 23
## 3 Amitabh Bachchan 20
## 4 Ajay Devgn 16
## 5 Aamir Khan 14
## 6 Ahmed Helmy 13
## 7 Anil Kapoor 13
## 8 Salman Khan 12
## 9 Naseeruddin Shah 11
## 10 Sanjay Dutt 11

all_genres_top_cast_intl<- intl_produced_movies %>%

filter(!is.na(cast)) %>%
filter(!is.na(listed_in)) %>%
filter(!is.na(release_year)) %>%
mutate(across(where(is.character), str_trim)) %>%
group_by(cast)

get_genres_for_cast_intl <- function(c) {

count_genres_top_cast_intl <- all_genres_top_cast_intl %>%
filter(cast == c) %>%
group_by(listed_in) %>%
count(sort=TRUE)

return (count_genres_top_cast_intl)
}

get_genres_with_years_for_cast_intl <- function(c) {

genre_year_intl <- all_genres_top_cast_intl %>%
select(cast, listed_in, release_year) %>%
filter(cast == c) %>%
group_by(release_year)

return (genre_year_intl[order(genre_year_intl$release_year), ] %>%

group_by(listed_in, release_year))
}

The chart indicates that Shah Rukh Khan appeared in twenty five movies, Akshay Kumar appeared in
twenty three movies, Amitabh Bachchan appeared in twenty movies, and so on. Most of them were In-
dian/Bollywood actors/actresses and that is due to the dataset contains second most data from Indian film

21
industry. For each of those actors/actresses, the following shows the graphs of which genre of movies they
appearted and how the genre changed over time.

plot_genre_box_intl <- function(c) {

data <- get_genres_for_cast_intl(c)

graph <- ggplot(data, aes(x=n, y=listed_in, fill=listed_in)) +

geom_bar(stat="identity") +
ggtitle(paste(c,"'s different genres")) +
labs(x="num releases", y="genre", col = "Genre") +
theme(axis.text.y=element_blank(),axis.ticks.y=element_blank())+
geom_text(aes(label=n), hjust=1.6, color="white", size=3.5)

graph
}

plot_release_genre_points_intl <- function(c, s, e) {

data <- get_genres_with_years_for_cast_intl(c)
g <- ggplot(data, aes(x=listed_in, y=release_year, color=listed_in)) +
ggtitle(paste(c, ", Genres through the ages")) +
labs(x="genre", y="release year", col = "Genre") +
geom_point() +
scale_y_continuous(limits=c(s, e), breaks=seq(s,e,1)) +
theme(axis.text.x=element_blank(),axis.ticks.x=element_blank())

g
}

cast_member="Shah Rukh Khan"

plot_genre_box_intl(cast_member)

22
 Shah Rukh Khan 's different genres
2
listed_in
6
Action & Adventure, Comedies, Dramas
2
Action & Adventure, Comedies, International Movies
3 Action & Adventure, Dramas, International Movies

1 Action & Adventure, International Movies

Action & Adventure, International Movies, Music & Musicals
1
genre

Comedies, Dramas, International Movies

8
Comedies, International Movies, Romantic Movies
1 Comedies, International Movies, Sci−Fi & Fantasy

1 Dramas, International Movies

Dramas, International Movies, Music & Musicals
4
Dramas, International Movies, Romantic Movies
2 Dramas, International Movies, Thrillers

0 2 4 6 8
num releases

plot_release_genre_points_intl(cast_member, 1991, 2021)

23
 Shah Rukh Khan , Genres through the ages
2021
2020
2019 Genre
2018
2017
2016 Action & Adventure, Comedies, Dramas
2015
2014 Action & Adventure, Comedies, International Movies
2013
2012 Action & Adventure, Dramas, International Movies
2011
2010 Action & Adventure, International Movies
release year

2009
2008 Action & Adventure, International Movies, Music & Musicals
2007
2006 Comedies, Dramas, International Movies
2005
2004 Comedies, International Movies, Romantic Movies
2003
2002 Comedies, International Movies, Sci−Fi & Fantasy
2001
2000 Dramas, International Movies
1999
1998 Dramas, International Movies, Music & Musicals
1997
1996 Dramas, International Movies, Romantic Movies
1995
1994 Dramas, International Movies, Thrillers
1993
1992
1991

genre
Shah Rukh Khan was in many different movie genres. First appeared in 1992, in a comedy related movie
and the latest appearance was in action and adventure movie. The cast member has a wide variety of
genre selection in 1992 to 2020 time frame.

cast_member="Akshay Kumar"
plot_genre_box_intl(cast_member)

24
 Akshay Kumar 's different genres
listed_in
2
Action & Adventure, Comedies, Dramas
2
Action & Adventure, Comedies, International Movies
1
Action & Adventure, Dramas, International Movies
3
Action & Adventure, International Movies
1
Children & Family Movies, Comedies, International Movies
1
Comedies, Dramas, International Movies
3
genre

Comedies, International Movies

4
Comedies, International Movies, Music & Musicals
1
Comedies, International Movies, Romantic Movies
3
Dramas, International Movies
1
Dramas, International Movies, Music & Musicals
1
Dramas, International Movies, Romantic Movies
3
Dramas, International Movies, Sports Movies
1
Dramas, International Movies, Thrillers
2 International Movies, Music & Musicals, Romantic Movies
0 1 2 3 4
num releases

plot_release_genre_points_intl(cast_member, 2001, 2021)

25
 Akshay Kumar , Genres through the ages
2021
Genre
2020 Action & Adventure, Comedies, Dramas
2019
Action & Adventure, Comedies, International Movies
2018
2017 Action & Adventure, Dramas, International Movies
2016 Action & Adventure, International Movies
2015
Children & Family Movies, Comedies, International Movies
2014
release year

2013 Comedies, Dramas, International Movies

2012 Comedies, International Movies
2011
Comedies, International Movies, Music & Musicals
2010
2009 Comedies, International Movies, Romantic Movies
2008 Dramas, International Movies
2007
2006 Dramas, International Movies, Music & Musicals
2005 Dramas, International Movies, Romantic Movies
2004
Dramas, International Movies, Sports Movies
2003
2002 Dramas, International Movies, Thrillers
2001 International Movies, Music & Musicals, Romantic Movies

genre

Akshay Kumar was the second most appeared cast member from the chart and the selection seemed to
be very diverse in terms of genres of movies the cast member appeared in. First appeared in 2002 and the
latest movie 2018 indicates that the cast member was in action and adventure movies. The genres seemed
to be changed over time.

cast_member="Amitabh Bachchan"
plot_genre_box_intl(cast_member)

26
 Amitabh Bachchan 's different genres
listed_in
1
Action & Adventure, Classic Movies, Comedies
1
Action & Adventure, Classic Movies, Cult Movies
3
Action & Adventure, Comedies, Dramas
2
Action & Adventure, Comedies, International Movies
1
Action & Adventure, Dramas, International Movies
5
Action & Adventure, International Movies
1
genre

Classic Movies, Comedies, International Movies

1
Comedies, Cult Movies, Dramas
1
Comedies, International Movies, Music & Musicals
1
Dramas, International Movies
6
Dramas, International Movies, Music & Musicals
1
Dramas, International Movies, Romantic Movies
2
Dramas, International Movies, Thrillers
1
International Movies, Romantic Movies
1 International Movies, Thrillers
0 2 4 6
num releases

plot_release_genre_points_intl(cast_member, 1975, 2021)

27
 Amitabh Bachchan , Genres through the ages
2021
Genre
2020
2019 Action & Adventure, Classic Movies, Comedies
2018
2017
2016 Action & Adventure, Classic Movies, Cult Movies
2015
2014
2013 Action & Adventure, Comedies, Dramas
2012
2011
2010
2009 Action & Adventure, Comedies, International Movies
2008
2007
2006 Action & Adventure, Dramas, International Movies
2005
2004
2003
release year

Action & Adventure, International Movies

2002
2001
2000 Classic Movies, Comedies, International Movies
1999
1998
1997 Comedies, Cult Movies, Dramas
1996
1995
1994 Comedies, International Movies, Music & Musicals
1993
1992
1991 Dramas, International Movies
1990
1989
1988 Dramas, International Movies, Music & Musicals
1987
1986
1985
1984 Dramas, International Movies, Romantic Movies
1983
1982
1981 Dramas, International Movies, Thrillers
1980
1979
1978 International Movies, Romantic Movies
1977
1976
1975 International Movies, Thrillers

genre

Amitabh Bachchan is the cast member that appeared from 1975 to 2019 and probably the cast member that
has been in the film industry for the longest in internationally according to the dataset. The genre that the
cast member started was comedy and later on the cast member switched to other genres such as action
or dramas. The wide variety of genres probably made the cast member to appear many movies for a long
period of time.

Future Work

Given more time we would work to develop or improve on several ideas. Further work could also have been
done with applying deep learning to this problem. A few viable methodologies might be to use techniques
for unsupervised learning from the numerical data and maybe used in other neural networks to perform
classification.

Conclusion

To conclude, we have constructed a relatively accurate data analysis to determine the genre of a given TV
show and movie. Processing and narrowing down the features of the Netflix Dataset by identifying the top
ten countries with top five genres the audience watch and feeding this data to figure out the trend over the
years, we were able to identify which actors/actresses by which directors were popular in each country.
Steps such as expanding the dataset and feature set and using logistic regression might have the potential
to improve upon on results in the future.

28
 Team Membership and Attestation of Work

Seongwoo Choi, Vybhab Achar Bhargav, and David Haddad agree to participate and work on this project
with the understanding of the project.

Video Presentation Link

https://ucdavis.zoom.us/rec/share/upsU8l_s89iJ5lLC2Cz6a5OTJJZyz9W9uIUTd8rwnnWLOi41oG_
WnIfrnGhTRQfK.Sgt0OhacFnJu6zyT

Reference

https://www.kaggle.com/shivanirana63/netflix-eda-movie-recommendation-system/data
https://www.kaggle.com/shivamb/netflix-shows
https://github.com/yihui/knitr-examples/blob/master/077-wrap-output.Rmd (for output formatting)

29

View publication stats