There are mainly two ways to extract data from a website:
- Use the API of the website (if it exists). For example, Facebook has the Facebook Graph API which allows retrieval of data posted on Facebook.
- Access the HTML of the webpage and extract useful information/data from it. This technique is called web scraping or web harvesting or web data extraction.
In this article, we will be using the API of newsapi. You can create your own API key by clicking here. Examples: Let’s determine the concern of a personality like states president cited by newspapers, let’s take the case of MERKEL
Python3
import pprint import requests secret = "Your API" # Define the endpoint # Specify the query and # number of returns parameters = { 'q' : 'merkel' , # query phrase 'pageSize' : 100 , # maximum is 100 'apiKey' : secret # your own API key } # Make the request response = requests.get(url, params = parameters) # Convert the response to # JSON format and pretty print it response_json = response.json() pprint.pprint(response_json) |
Output: Let’s combine all texts and sort the words from the greatest number to lower.
Python3
from wordcloud import WordCloud import matplotlib.pyplot as plt text_combined = '' for i in response_json[ 'articles' ]: if i[ 'description' ] ! = None : text_combined + = i[ 'description' ] + ' ' wordcount = {} for word in text_combined.split(): if word not in wordcount: wordcount[word] = 1 else : wordcount[word] + = 1 for k,v, in sorted (wordcount.items(), key = lambda words: words[ 1 ], reverse = True ): print (k,v) |
Output: This evaluation is ambiguous, we can make it more clear if we delete bad or useless words. Let’s define some of bad_words shown below
bad_words = [“a”, “the”, “of”, “in”, “to”, “and”, “on”, “de”, “with”, “by”, “at”, “dans”, “ont”, “été”, “les”, “des”, “au”, “et”, “après”, “avec”, “qui”, “par”, “leurs”, “ils”, “a”, “pour”, “les”, “on”, “as”, “france”, “eux”, “où”, “son”, “le”, “la”, “en”, “with”, “is”, “has”, “for”, “that”, “an”, “but”, “be”, “are”, “du”, “it”, “à”, “had”, “ist”, “Der”, “um”, “zu”, “den”, “der”, “-“, “und”, “für”, “Die”, “von”, “als”, “sich”, “nicht”, “nach”, “auch” ]
Now we can delete and format the text by deleting bad words
Python3
# initializing bad_chars_list bad_words = ["a", "the" , "of", " in ", "to", " and ", "on", "de", "with", "by", "at", "dans", "ont", "été", "les", "des", "au", "et", "après", "avec", "qui", "par", "leurs", "ils", "a", "pour", "les", "on", "as", "france", "eux", "où", "son", "le", "la", "en", "with", " is ", "has", " for ", "that", "an", "but", "be", "are", "du", "it", "à", "had", "ist", "Der", "um", "zu", "den", "der", " - ", "und", "für", "Die", "von", "als", "sich", "nicht", "nach", "auch" ] r = text_combined.replace( '\s+' , ' ' ).replace( ',' , ' ' ).replace( '.' , ' ' ) words = r.split() rst = [word for word in words if ( word.lower() not in bad_words and len (word) > 3 ) ] rst = ' ' .join(rst) wordcount = {} for word in rst.split(): if word not in wordcount: wordcount[word] = 1 else : wordcount[word] + = 1 for k,v, in sorted (wordcount.items(), key = lambda words: words[ 1 ], reverse = True ): print (k,v) |
Output: Let’s plot the output
Python3
word = WordCloud(max_font_size = 40 ).generate(rst) plt.figure() plt.imshow(word, interpolation = "bilinear") plt.axis("off") plt.show() |
Output: As you see in the descriptions of articles, the most dominant concern with Merkel is his defense minister Kramp-Karrenbauer, Kanzlerin just means female chancellor. We can do the same work using titles only
Python3
title_combined = '' for i in response_json[ 'articles' ]: title_combined + = i[ 'title' ] + ' ' titles = title_combined.replace( '\s+' , ' ' ).replace( ',' , ' ' ).replace( '.' , ' ' ) words_t = titles.split() result = [word for word in words_t if ( word.lower() not in bad_words and len (word) > 3 ) ] result = ' ' .join(result) wordcount = {} for word in result.split(): if word not in wordcount: wordcount[word] = 1 else : wordcount[word] + = 1 word = WordCloud(max_font_size = 40 ).generate(result) plt.figure() plt.imshow(word, interpolation = "bilinear") plt.axis("off") plt.show() |
Output: From titles, we found out that the most concern with Merkel is Ardogan, turkey president.