Disney Movie Recommendation System

Python

Big Data

This Python project uses techniques such as Cosine Similarity and TF-IDF Vectorisation to choose similar Disney movie titles, based on their descriptions.

Author

Lisa Lucky

Published

April 24, 2025

This project is adapted off of Kevin Kibe’s Medium article which details the usage of Cosine Similarity and TF-IDF Vectorisation for book recommendations ¹. Using a Disney movies dataset from Kaggle ² which included titles, descriptions and directors, a defintion function from the Medium article was adapted to allow for the return of five similar movie titles (based on their title input).

Under the same dataset, a large amount of Kaggle users had used Tensorflow and others, TF-IDF vectorisors (input “recommendation” in search bar for notebooks). Simplfying previous works and looking to the Medium article, I was able to build a recommendation system from the Disney catalogue. Below details my process as well as explanations with the code.

Importing libraries needed

To begin, necessary libraries were imported.

Code

import pandas as pd
import numpy as np
from sklearn.feature_extraction.text import CountVectorizer, TfidfVectorizer
from sklearn.metrics.pairwise import cosine_similarity

Overview on dataset

Code

df = pd.read_csv('for-download/disney-titles.csv')
df.sample(5)

	show_id	type	title	director	cast	country	date_added	release_year	rating	duration	listed_in	description
1114	s1115	TV Show	Origins: The Journey of Humankind	NaN	Mark Monroe, Jason Silva	United States	November 12, 2019	2016	TV-14	1 Season	Docuseries, Historical	Hosted by Jason Silva, this eight-part series ...
554	s555	Movie	The Olympic Elk	James Algar	Winston Hibler	United States	April 24, 2020	1952	TV-G	27 min	Animals & Nature, Documentary, Family	Olympic elk trek toward the fertile grazing gr...
153	s154	Movie	Shark vs. Surfer	Phil Stebbing	Billy Lloyd	NaN	July 23, 2021	2020	TV-14	44 min	Animals & Nature, Documentary	Shark vs. Surfer visits shark-infested surf sp...
222	s223	TV Show	Disney Special Agent Oso: Three Healthy Steps ...	NaN	Sean Astin, Meghan Strange, Phill Lewis, Amber...	NaN	May 14, 2021	2011	TV-Y	1 Season	Action-Adventure, Animation, Kids	Oso and his friends watch a kid as they demons...
96	s97	TV Show	Dog: Impossible	NaN	Matt Beisner	United States	September 22, 2021	2019	TV-PG	2 Seasons	Animals & Nature, Docuseries, Family	Matt Beisner uses unique approaches to modifyi...

Info on dataset

Code

df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 1450 entries, 0 to 1449
Data columns (total 12 columns):
 #   Column        Non-Null Count  Dtype 
---  ------        --------------  ----- 
 0   show_id       1450 non-null   object
 1   type          1450 non-null   object
 2   title         1450 non-null   object
 3   director      977 non-null    object
 4   cast          1260 non-null   object
 5   country       1231 non-null   object
 6   date_added    1447 non-null   object
 7   release_year  1450 non-null   int64 
 8   rating        1447 non-null   object
 9   duration      1450 non-null   object
 10  listed_in     1450 non-null   object
 11  description   1450 non-null   object
dtypes: int64(1), object(11)
memory usage: 136.1+ KB

Null values in dataset, necessary for cleaning

Code

df.isnull().sum()

show_id           0
type              0
title             0
director        473
cast            190
country         219
date_added        3
release_year      0
rating            3
duration          0
listed_in         0
description       0
dtype: int64

All null values dropped, simplifying dataset

Code

df.dropna() #(Geeksforgeeks, 2025) dropping all rows with missing values to simplify, https://www.geeksforgeeks.org/working-with-missing-data-in-pandas/

	show_id	type	title	director	cast	country	date_added	release_year	rating	duration	listed_in	description
2	s3	Movie	Ice Age: A Mammoth Christmas	Karen Disher	Raymond Albert Romano, John Leguizamo, Denis L...	United States	November 26, 2021	2011	TV-G	23 min	Animation, Comedy, Family	Sid the Sloth is on Santa's naughty list.
5	s6	Movie	Becoming Cousteau	Liz Garbus	Jacques Yves Cousteau, Vincent Cassel	United States	November 24, 2021	2021	PG-13	94 min	Biographical, Documentary	An inside look at the legendary life of advent...
9	s10	Movie	A Muppets Christmas: Letters To Santa	Kirk R. Thatcher	Steve Whitmire, Dave Goelz, Bill Barretta, Eri...	United States	November 19, 2021	2008	G	45 min	Comedy, Family, Musical	Celebrate the holiday season with all your fav...
12	s13	Movie	The Pixar Story	Leslie Iwerks	Stacy Keach, John Lasseter, Brad Bird, John Mu...	United States	November 19, 2021	2007	G	91 min	Documentary, Family	A groundbreaking company forever changes the f...
19	s20	Movie	Enchanted	Kevin Lima	Amy Adams, Patrick Dempsey, James Marsden, Tim...	United States	November 12, 2021	2007	PG	110 min	Comedy, Family, Fantasy	An animated princess winds up in the real worl...
...	...	...	...	...	...	...	...	...	...	...	...	...
1445	s1446	Movie	X-Men Origins: Wolverine	Gavin Hood	Hugh Jackman, Liev Schreiber, Danny Huston, wi...	United States, United Kingdom	June 4, 2021	2009	PG-13	108 min	Action-Adventure, Family, Science Fiction	Wolverine unites with legendary X-Men to fight...
1446	s1447	Movie	Night at the Museum: Battle of the Smithsonian	Shawn Levy	Ben Stiller, Amy Adams, Owen Wilson, Hank Azar...	United States, Canada	April 2, 2021	2009	PG	106 min	Action-Adventure, Comedy, Family	Larry Daley returns to rescue some old friends...
1447	s1448	Movie	Eddie the Eagle	Dexter Fletcher	Tom Costello, Jo Hartley, Keith Allen, Dickon ...	United Kingdom, Germany, United States	December 18, 2020	2016	PG-13	107 min	Biographical, Comedy, Drama	True story of Eddie Edwards, a British ski-jum...
1448	s1449	Movie	Bend It Like Beckham	Gurinder Chadha	Parminder Nagra, Keira Knightley, Jonathan Rhy...	United Kingdom, Germany, United States	September 18, 2020	2003	PG-13	112 min	Buddy, Comedy, Coming of Age	Despite the wishes of their traditional famili...
1449	s1450	Movie	Captain Sparky vs. The Flying Saucers	Mark Waring	Charlie Tahan	United States	April 1, 2020	2012	TV-G	2 min	Action-Adventure, Animals & Nature, Animation	View one of Sparky's favorite home movies.

818 rows × 12 columns

Creating the TFID vector values

Code

vector = TfidfVectorizer(analyzer = "word", ngram_range=(1,2), min_df=0.0, stop_words='english')
tfidf_matrix = vector.fit_transform(df['title'])

the_matrix =  cosine_similarity(tfidf_matrix, tfidf_matrix)
print(the_matrix)

[[1.         0.07856161 0.07072431 ... 0.         0.         0.        ]
 [0.07856161 1.         0.09737137 ... 0.         0.         0.        ]
 [0.07072431 0.09737137 1.         ... 0.         0.         0.        ]
 ...
 [0.         0.         0.         ... 1.         0.         0.        ]
 [0.         0.         0.         ... 0.         1.         0.        ]
 [0.         0.         0.         ... 0.         0.         1.        ]]

Allowing us to set the index back to default

Code

df = df.reset_index() 
indices = pd.Series(df['description'])
indices [:5]

0     Join Mickey and the gang as they duck the halls!
1    Santa Claus passes his magic bag to a new St. ...
2            Sid the Sloth is on Santa's naughty list.
3                 This is real life, not just fantasy!
4    A three-part documentary from Peter Jackson ca...
Name: description, dtype: object

Recommendation function (Input = Moana)

Code

give_title = df['title'].tolist() # It should in other words, return a list of values from title because thus far we have only dealt with description (Pandas.pydata, n.d), https://pandas.pydata.org/docs/reference/api/pandas.Series.to_list.html 
def recommend(title, cosine_sim = the_matrix):
    if title not in give_title:
        return "Unable to find title."

    recommended_movies = []
    # Search for title in title column to get the index
    idx = df.index[df['title'] == title].tolist()[0]
    score_series = pd.Series(cosine_sim[idx]).sort_values(ascending = False)
    top_5_indices = list(score_series.iloc[1:6].index)

    for i in top_5_indices:
        recommended_movies.append(list(df['title'])[i]) # Recommend based on title

    return recommended_movies

print(recommend("Moana"))

#Final block fixed and debugged using Google Colab AI assistant (Gemini).

['James and the Giant Peach', 'Jack', 'Iron Will', 'Iron Man Armored Adventures', 'Iron Man (Series)']

Insights and Reflections

I learned how to use Cosine Similarity and TF-IDF Vectorisation. Similar to the idea of the SQL project, I leveraged information about Disney movies to build a recommendation system.
System recommendations are essential in today’s business world, where, through systems such as these, we can better understand customers and how to attract them ³. This proves my ability to chase after trends and understand the customer’s mind. A significant component of Disney’s success is its ability to recommend content properly to customers and ensure attention is always kept ⁴. My ability to recognise this allows me to apply what I have learned in university to add value to a company.

Disney Movie Recommendation System

Importing libraries needed

Overview on dataset

Info on dataset

Null values in dataset, necessary for cleaning

All null values dropped, simplifying dataset

Creating the TFID vector values

Allowing us to set the index back to default

Recommendation function (Input = Moana)

Link to other coding sources used in project

Footnotes