Movie suggestions for Raya and the Last Dragon fans

Published in

Web Mining [IS688, Spring 2021]

7 min readMar 2, 2021

If you like the Raya and the Last Dragon movie, what other movies do you like or are you interested in? What movies are the most similar to this movie?

Raya and the Last Dragon is an upcoming American computer-animated action-adventure fantasy film produced by Walt Disney Pictures and Walt Disney Animation Studios for distribution by Walt Disney Studios Motion Pictures. They promise that this movie will be one of the biggest family-friendly movies of the year. Raya and the Last Dragon took Disney fans’ world by storm when the trailer was released. However, its adoring fans should love these similar computer-animated action-adventure fantasy films.

This study offers movie suggestions for Raya fans. If you like this movie, then you probably would like to know which similar movies you might also be interested in. Here, I consider the movie reviewers or critics in order to determine which movies are the most popular. Most critics review movies based on their interests. The question posed is what other movies interest the critics that reviewed Raya? What other movies have they reviewed?

The New York Times (NYT) is an American daily newspaper with a worldwide readership. The content is organized into three sections including news, opinion, and features which include movies. With their own big data warehouse, NYT has provided developers the APIs to access movie reviews, book reviews, bestseller lists, article archives, etc. This study has used the New York Times Movie Review API to get movie critic names with “Raya” in the title, in order to find all other movies that each critic has been reviewed besides Raya and the Last Dragon.

Data collection with NYT API

This study uses the following libraries in a Python Jupyter Notebook: pynytimes, requests, pandas, json, BeautifulSoup, re, and itemgetter along with an API key with registered accounts on the NYT website to be able to access the movie review data as shown in the script below.

from pynytimes import NYTAPI
nyt = NYTAPI("8fkJNSuAsKvRVhQ0VgT7uyPD3wcKDNq0")
import requests
import pandas as pd
import numpy
import matplotlib
import datetime
import json
import nltk
import matplotlib.pyplot as plt
from operator import itemgetter
from bs4 import BeautifulSoup
import relink="https://api.nytimes.com/svc/movies/v2/reviews/search.json?query=raya&api-key=8fkJNSuAsKvRVhQ0VgT7uyPD3wcKDNq0"
req=requests.get(link)
req
mov = req.json()
mov_r=json.dumps(mov)
data=json.loads(mov_r)
df1=pd.json_normalize(data['results'])
df2=df1[['display_title','byline']]
df2
df2.to_csv('0.critic_names.csv')

The movie critic names with “Raya” in the title were collected with the NYT API and stored in the CSV file which is shown in the table below. I have cleaned the data set by removing the duplicated and unrelated data or the critic’s name, in order to use those names to find other movies that each critic has reviewed besides Raya and the Last Dragon in the next step.

I have used the 13 remaining critics’ names to collect the other movies they have reviewed with the script below.

reviews = nyt.movie_reviews(
options = {
"reviewer": "Devika Girish",
})
df=pd.DataFrame(reviews)
df
df.to_csv('1.Devika.csv')

With 13 critics, The number of movies each reviewer has been reviewed is between 5 and 20. I have 260 movie reviews to analyze as a result. The data set was stored in a CSV file that contains the following data in each column: movie title, MPAA rating, critic’s pick, reviewer name, headline, a brief summary, publication date, opening date, updated date, and article link. The data frame is shown in the table below.

The data is stored in a list or dictionary. The link column, for example, displays its content in dictionaries. That dictionary along with all the other information for a given review for a specific critic is stored in a list.

Data set with the Gephi

After the data manipulation process, I have imported the remaining 240 nodes and 260 edges from the spreadsheet files to the Gephi program shown in the figure below. Nodes are the movie names and the critics’ names. Edges are the connections between the critics and the movie names.

The figure below shows the collaboration graph as the initial relation graph in this community. It seems like there are several crowded clusters that have appeared which indicate that more than one reviewer has reviewed that film.

Yifan Hu Algorithm

As the initial relation graph seems slightly hard to understand. I have used the Yifan Hu layout to cluster them and make them easier to read. Now we can see the relation between the pair of nodes in the picture below. There are several hub nodes that indicate the relationship between the movie and critic and show some overlap among reviewed movies.

Also, there are filters that we can use to visualize the network of this data set to present the degree range, neighbor network, betweenness centrality, closeness centrality, modularity class, and other attributes in the Gephi library. The figures below show the network with the range hub filter.

The network with the degree range filter

Graph characterization

As illustrated, the below graph represents the most common movie that critics have reviewed in the Raya and the last dragon community. From 240 nodes, there are 14 nodes that connect among communities, called bridge nodes. Within 217 movies and 13 critics, there are only 7 movies that more than one critic in the Raya movie community had reviewed including Mulan, Toy Story 2, Once Upon a Time, The Devil All the Time, Black Magic, Sleepy Hollow, and The Golden Girl.

The statistics of this network are the following:

The average degree, the average number of edges that connect to a node in this network, is 2.167
The average weighted degree is 2.167
The network diameter, the longest of the shortest path in this network, is 8
The network density, the number of actual edges over the potential edges in this network, is 0.009
The modularity in this network is 0.796
The average path length, the average of the shortest path in this network, is 4.335

Movies Ranking

The blue in the graph below represents the movies with the most reviews, Toy Story 2 and Mulan, which were reviewed with 8 critics or 19% in this community. The second most reviewed movies are the Black Magic and the Devil All the Time with 6 critics or 14.3% of 42 movies with the overlapped reviews. The Sleepy Hollow and Once Upon a Time has 5 critics reviewed or 11.9%. The movie with the fewest reviews is the Golden Girl with 4 critics or 9.5% of movies with the overlapped reviews.

From 240 nodes, there are 14 bridge nodes. Within the Raya movie’s reviewing community with 13 critics, there are 7 movies that more than one critic had reviewed besides the Raya and the Last Dragon movie including Mulan, Toy Story 2, Once Upon a Time, The Devil All the Time, Black Magic, Sleepy Hollow, and The Golden Girl respectively.

Conclusion

13 critics who had reviewed Raya and the Last Dragon movie also reviewed the following movies: Mulan, Toy Story 2, Once Upon a Time, The Devil All the Time, Black Magic, Sleepy Hollow, and The Golden Girl respectively. The significant number of movies with overlapped reviews shows that there are other movies that 13 critics are also interested in. Those movies were reviewed by the same critics as the Raya movie with a significant number of reviews. Hereby Raya’s fans should probably love these similar computer-animated action-adventure fantasy films as well.

Discussion and limitation

Some limitations include the amount of data, as this study collected only 260 movie reviews from 13 critics. It is a small sample. These limited results are based on critics who reviewed Raya and the Last Dragon. This causes weak relationships among the communities as they are lacking relationships between the communities. I expected to rank the top 10 most popular movies that have been reviewed by more than one critic. However, with the small sample size that I have collected, the results only appear in 7 movies with the overlapped reviews.