How to extract a website and make a dataset?
TripAdvisor is the world’s biggest travel website and it is a very popular website to find restaurants, hotels, transportation, and spaces to visit. When somebody plans a trip to a city or country, they are expected to visit TripAdvisor to get the finest places for staying and visiting. TripAdvisor has more than 702 million reviews of the world’s top hotels, lists more than 8 million locations (restaurants, hotels, tourist charms), and ranks 1st in the Traveling and Tourism categories in the United States.
In this blog, we will provide a script, which will extract hotel data from the TripAdvisor webpage, scrape a few data elements and make a dataset. Here, are the steps that would be executed using Python & BeautifulSoup.
1. Import different libraries.
2. Review the HTML structure of a web page
3. Retrieve and change HTML Data
4. Find and scrape data elements
5. Make a data frame
6. Convert data frame into a CSV file
Import Different Libraries
# Import the libraries.
import requests
from bs4 import BeautifulSoup
import pandas as pd
import csv
Requests permit you to send different HTTP requests to the server, which returns the Response Object having all the reply data (i.e., HTML).
Beautifulsoup (bs4) is used for pulling data out of the HTML files and convert data into a BeautifulSoup object that represents HTML as the nested data structure.
Pandas is used to do data manipulation and analysis.
CSV module implements different classes in reading and writing tabular data within a CSV format.
Review a Webpage’s HTML Structure
We have to recognize the contents and structure of HTML tags in webpages. For that project, we would be using TripAdvisor Hawaii Hotels & Places of staying webpage (given below). You can get this webpage through choosing a link.
We can extract this webpage through parsing HTML of a page and scraping the data required for the dataset. To extract some data from the web page, just right-click anywhere on this webpage, choose inspect from a drop-down list and click an arrow icon given on the screen’s upper left-hand side with HTML and click on hotel name (Prince Waikiki) in review section of a webpage. It will result in the given screen shown.
On HTML screen, you would see highlighted an HTML line having the Hotel Name called Prince Waikiki.
If you are moving one line from the tag then you would find a div tag having the class of “listing_title”. It is a parent of tag.
Therefore, if you want to find, scrape, and capture hotel names on a webpage you might follow these steps.
Get all HTML lines for any particular parent (div tag having class = listing_title) that might include their related children.
Scrape data elements and create a list having all the hotel names.
The code to find and extract hotel names might be the following:
hotels = []
for name in soup.findAll('div',{'class':'listing_title'}):
hotels.append(name.text.strip())
We will get, scrape and store other data elements on a webpage following similar procedures as given above.
Find and Scrape Data Elements
For all data elements we need to scrape, we will get all HTML lines, which are within any particular class and tag. Then, we will scrape data elements as well as store data in the list.
# Find and extract data elements.
hotels = []
for name in soup.findAll('div',{'class':'listing_title'}):
hotels.append(name.text.strip())
ratings = []
for rating in soup.findAll('a',{'class':'ui_bubble_rating'}):
ratings.append(rating['alt'])
reviews = []
for review in soup.findAll('a',{'class':'review_count'}):
reviews.append(review.text.strip())
prices = []
for p in soup.findAll('div',{'class':'price-wrap'}):
prices.append(p.text.replace('₹','').strip())
Creating a Data Frame
We would create a dictionary, which will have data names and standards for all data elements which were scraped.
# Create the dictionary.
dict = {'Hotel Names':hotels,'Ratings':ratings,'Number of Reviews':reviews,'Prices':prices}
Create and show a data frame.
# Create the dataframe.
hawaii = pd.DataFrame.from_dict(dict)
hawaii.head(10)
Converting Data Frames into a CSV file
# Convert dataframe to CSV file.
hawaii.to_csv('hotels.csv', index=False, header=True)
Making it all together…
# Import the libraries.
import requests
from bs4 import BeautifulSoup
import pandas as pd
import csv
# Extract the HTML and create a BeautifulSoup object.
url = ('https://www.tripadvisor.in/Hotels-g28932-Hawaii-Hotels.html')
user_agent = ({'User-Agent':
'Mozilla/5.0 (Windows NT 10.0; Win64; x64) \
AppleWebKit/537.36 (KHTML, like Gecko) \
Chrome/90.0.4430.212 Safari/537.36',
'Accept-Language': 'en-US, en;q=0.5'})
def get_page_contents(url):
page = requests.get(url, headers = user_agent)
return BeautifulSoup(page.text, 'html.parser')
soup = get_page_contents(url)
# Find and extract the data elements.
hotels = []
for name in soup.findAll('div',{'class':'listing_title'}):
hotels.append(name.text.strip())
ratings = []
for rating in soup.findAll('a',{'class':'ui_bubble_rating'}):
ratings.append(rating['alt'])
reviews = []
for review in soup.findAll('a',{'class':'review_count'}):
reviews.append(review.text.strip())
prices = []
for p in soup.findAll('div',{'class':'price-wrap'}):
prices.append(p.text.replace('₹','').strip())
# Create the dictionary.
dict = {'Hotel Names':hotels,'Ratings':ratings,'Number of Reviews':reviews,'Prices':prices}
# Create the dataframe.
hawaii = pd.DataFrame.from_dict(dict)
hawaii.head(10)
# Convert dataframe to CSV file.
hawaii.to_csv('hotels.csv', index=False, header=True)
Thank you so much to read this blog. Please give your valuable comments or feedback. For the best mobile app scraping and web scraping services, contact Actowiz Solutions now!
