Proceedings of the National Conference on Emerging Computer Applications (NCECA)-2022

Vol.4, Issue.1 621

Movie Box Office Success Prediction using

Machine Learning
Elizabeth Antony Nimmy Francis
Department of Computer Applications Department of Computer Applications
Amal Jyothi College of Engineering Kanjirappally, India Amal Jyothi College of Engineering Kanjirappally, India
[email protected] [email protected]

approaches, text mining, and social network analysis.

Abstract — Predicting societal reactions to a new
product in terms of perceived popularity and rate of
adoption has become a field of data analysis. The film Sivasantoshreddy et al.[2] attempted to anticipate a movie's box
office gross opening. Their study focused on the use of Twitter
industry is a multibillion-dollar industry, and there is a
data for hype analysis. The fundamental rationale behind hype
vast amount of information about movies on the internet. analysis was to determine a film's performance based on its
This research proposes a machine learning-based solution opening weekend box office and how much buzz it generated
for the film investing business. The method analyses before its release.
previous data from several sources to forecast a movie's
success rate. K Nearest Neighbors (KNN) is a method for Lydia [3] used news analysis and quantitative news data to try to
predicting the box office success of a film. increase the gross profit of the film prediction.
Keywords—movie industry, Knearest neighbors Jonas et al. [4] employed sentiment and social media analysis to
algorithm. predict, and their hypothesis was based on IMDb's Oscar Buzz sub-
intensity forum's and positivity analysis. They had viewed cinema
1.INTRODUCTION reviewers as a source of influence and foresight. They employed a
The film industry is a massive investment sector, but vast collection of terms that produced poor results when particular
words were used for negative purposes. There was no category
company sectors are more complex, making it harder to decide
award; instead, only the awards for best picture, director,
where to put your money. Because the film industry is actor/actress, and supporting actor/actress were considered.
expanding at an alarming rate, there is already a significant
amount of data available on the internet, making data analysis 111.MOTIVATION
a fascinating field. The measure of a film's success is Nowadays, audiovisual input and output devices are widely
subjective; some videos are deemed successful based on their used. People are watching movies in greater numbers than
worldwide gross revenue, while others may not perform well ever before. As a result, the film business is arguably one of
commercially but receive positive critical reviews and modern society's most significant sectors. Every film has the
popularity. In this post, we'll look at a successful box office power to reflect society and change people's minds. In many
film. In comparison to traditional media, there has been a ways, a good film may entertain, educate, and inspire its
significant shift in the social media sector in recent years. We audience. Consider the impact that music has on people.
need a powerful approach to quickly find facts that can be Each film is set in and developed within a specific culture.
utilised to frame a certain subject. The rapid increase in the Movies have become an inseparable part of our lives.
quantity of films released in recent years necessitates a few
decades of movie forecasting. In other words, the goal is to
create learning algorithms that can be used by machines V1.METHODOLOGY
without the need for human interaction. The goal of this The purpose of this paper is to find a Machine Learning
research is to present a machine learning approach that model to identify the box office success of a movie using the
provides the most accurate outcome for movie success givendataset with accuracy. The model should be able to
prediction.The factors can determine the success of a movie: determine correctly as the movie’s box office success.The
● The popularity of film content. dataset is divided into two training and testing datasets.
● The present popularity of the film's actors.
● The strength of the cast and the film's release
schedule.
● Awareness of the film based on its adaptation of
a popular book or report

11.LITERATURE REVIEW
M. T. Lash and K. Zhao[1] developed a decision support system
that predicts the profitability of a film using machine learning

DOI: 10.5281/zenodo.6938169
ISBN: 978-93-5607-317-3@2022 MCA, Amal Jyothi College of Engineering Kanjirappally, Kottayam
Proceedings of the National Conference on Emerging Computer Applications (NCECA)-2022
Vol.4, Issue.1 622

The model should be trained with more data to get accuracy model.
in learning. The result from this model can be used to create
a movie’s classification. The knn algorithm is used to
determine the box office success of a movie. The structure C.Machine Learning: This is used to determine the
is: movie success. The machine learning algorithms are
used to determine the movie’s success, that will check
the given movie. There are different classifiers
present to determine the movie’s success. The
classifier accuracy relies on how it is trained. A
classifier should need high accuracy.
KNN- K –Nearest Neighbor is a supervised learning
machine learning method that is one of the most basic.
The K-NN algorithm assumes that the new case/new
data and the existing cases are comparable, and it
places the new example in the category that is most
similar to the existing categories. This algorithm saves
all available data and rates fresh data points according
to how similar they are. This means that when fresh
data comes, the knn algorithm can quickly place it into
a well-suited category.

Figure 1: Movie Success Prediction Methodology

The following stages can be used to explain the knn
The steps that require to be followed are: algorithm.
1. Data Collection Step 1: Determine the number of neighbours (K).
2. Data Pre-processing Step 2: Determine the distance between each of the K
3. Model Building neighbours.
4. Analyzing
Step 3: Count the number of neighbours who are
5. Result closest to you.
A. Data Collection: User have a dataset for movie box
office success, the attributes of this dataset is used to Step 4: Count how many data points there are.
train the model. There 839 rows and 6 columns in
dataset. The columns are headings. The heading denotes Step 5: Assign the new data to the appropriate field.
the movie’s headings. This is to find the movie is
success or not. Step 6: The model is complete.

B. Data Pre-processing: This is the step that is done in V.BUILD MODEL

every machine learning method. This process includes
data cleaning, data transformation and data reduction. The model building is the main step in the movie box
This all are done to make the data more effective. The office success. While building the model user use the
data can be processed to make our model more accurate. algorithms. The steps involved are:
So that the classification will be correct.
1.Cleaning Data – The data is cleaned at first. That 1. Import the packages that are necessary.
is here users are removing the unwanted entries. The
noisy data can also be removed from the dataset.

2.Removing Stop words – The stop words are

removed from the datasets. These are the most
frequently appearing that don’t have much
importance in the text. To achieve more accuracy to 2.To read the dataset.
the model this can be done.

3.Splitting of Data – After cleaning the data the data

is then split into training and testing datasets. After
this the data chosen for training is used to train our

DOI: 10.5281/zenodo.6938169
ISBN: 978-93-5607-317-3@2022 MCA, Amal Jyothi College of Engineering Kanjirappally, Kottayam
Proceedings of the National Conference on Emerging Computer Applications (NCECA)-2022
Vol.4, Issue.1 623
3.To analyzing the data.

4.To split dataset into test dataset and training dataset.

5.To, fit and transform train and test dataset.

6.Initialize Knn Classifier.
Fig 2:Top 10 movies rating

7.To calculate the accuracy of the model

The accuracy is 89.52%.

V1.RESULT
Based on the dataset provided, the movie can be assessed to
be a success. The knn algorithm produces accurate and
correct results. The dataset contains information on the film.
This is classified, and the model correctly predicted the film's Fig 3: Rating analysis-medium rated
success. movies
V11.CONCLUSION

Using the knn algorithm, this study proposes a method for

predicting movie box office performance. This method is
the most effective for predicting the success of a film. The
success of a film is not only determined by these aspects
of cinema. The amount of people who watch a movie is
crucial to its success.
The findings suggest that box office success may be
predicted with reasonable accuracy by evaluating movie
emotion.
REFERENCES
[1] Nahid, Quader, and Dipankar Chaki. Movie box
office success prediction: ways for predicting movie
success. Conference on BRAC university 2018.
[2] Sanyam Jatale and Rohan Moze. Analyzing and
Fig 1:Model Evaluation predicting movie box office.International journal of
advanced research in science, communication and

DOI: 10.5281/zenodo.6938169
ISBN: 978-93-5607-317-3@2022 MCA, Amal Jyothi College of Engineering Kanjirappally, Kottayam
Proceedings of the National Conference on Emerging Computer Applications (NCECA)-2022
Vol.4, Issue.1 624
technology(IJARSCT) in 2021.
[3] Pojan Shahrivar and Carl Jernbacker. Predicting
movie success using machine learning techniques in
2017.
[4] Apoorva Patil and Akshay box office movie
prediction using machine learning in 2019.
[5] M. C. A. Mestyán, T. Yasseri, and J. Kertész, “Early
Prediction of Movie Box Office Success Based on
Wikipedia Activity Big Data,” PLoS ONE, vol. 8,
no. 8, 2013.

DOI: 10.5281/zenodo.6938169
ISBN: 978-93-5607-317-3@2022 MCA, Amal Jyothi College of Engineering Kanjirappally, Kottayam