def api():
# Check for request type, content type
if request.method == 'GET':
return make_response({'error': 'Method Not Allowed'}, 405)
if request.content_type != 'application/json':
return make_response({'error': 'Unsupported Media Type'}, 415)
else:
# Check for request body in correct format
try:
txtInput = request.json['text']
except:
return make_response({'error': 'Bad Request'}, 400)
# Create API response
scores = SentimentIntensityAnalyzer().polarity_scores(txtInput)
sentiment = getSentiment(scores.get('compound'))
if sentiment == 0:
sentiment = 'ERROR'
wordCount = getWordCount(txtInput)
mostFrequent = getMostFrequent(txtInput, 5)
return jsonify(sentiment=sentiment,
compound=scores.get('compound'),
pos=scores.get('pos'),
neu=scores.get('neu'),
neg=scores.get('neg'),
wordCount=wordCount,
mostFrequent=mostFrequent,
)
def on_data(
self,
data): # pulling the text out of the JSON file each tweet comes in
json_load = json.loads(data)
try:
texts = json_load['text']
coded = texts.encode('utf-8')
s = str(coded)
ss = SentimentIntensityAnalyzer().polarity_scores(s)
datetime = json.loads(data)[
'created_at'] # print the datetime of the tweet
text = s[1:] # print the text of the tweet
compound = ss.get('compound', '0')
negative = ss.get('neg', '0')
positive = ss.get('pos', '0')
neutral = ss.get(
'neu', '0') # print the VADER sentiment scores for the tweet
user_id = json.loads(data)['user'].get('id', 'NA')
user_name = json.loads(data)['user'].get('name', 'NA')
user_screen_name = json.loads(data)['user'].get(
'screen_name', 'NA')
location = json.loads(data)['user'].get('location', 'NA')
num_of_followers = json.loads(data)['user'].get(
'followers_count', 'NA')
time_zone = json.loads(data)['user'].get('time_zone', 'NA')
cur.execute('insert into twitter_stream_netflix (datetime_created, user_id, user_name, user_screen_name,\
location, num_of_followers, time_zone, tweet_text, vader_compound, vader_negative,\
vader_positive, vader_neutral) values (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)' , datetime, user_id, user_name,\
user_screen_name, location, num_of_followers, time_zone, text, compound, negative, positive, neutral)
conn.commit() # insert values into SQL database
except:
cur.execute('insert into twitter_stream_netflix (datetime_created, user_id, user_name, user_screen_name,\
location, num_of_followers, time_zone, tweet_text, vader_compound, vader_negative,\
vader_positive, vader_neutral) values (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)' , 'NA', 'NA', 'NA', \
'NA', 'NA', 'NA', 'NA', 'NA', 0, 0, 0, 0)
conn.commit() # accounting for blank values
err = sys.exc_info()[2]
print(err)
print('YEEEEEEEEEEEEH BWOI SORTED MATE')
def on_error(self, status):
print(status)
def on_error(self, status_code):
if status_code == 420:
# returning False in on_data disconnects the stream
return False
def service():
form = serviceForm()
descriptions = {
'neg': 'Proportion of the text that has negative sentiment',
'neu': 'Proportion of the text that is neutral',
'pos': 'Proportion of the text that has positive sentiment',
'compound': 'A normalized, weighted composite score for the text overall. Ranges \
between -1 (most negative) to 1 (most positive)',
'sentiment': 'Overall sentiment, based on compound score. Below -0.05 is negative',
'word_count': 'Word Count',
'most_frequent': 'Most frequently used words and number of appearances in the text'
}
displayResults = False
txtInput = ''
scores = {
'neg': '-',
'neu': '-',
'pos': '-',
'compound': '-'
}
sentiment = '-'
wordCount = '-'
mostFrequent = '-'
if request.method == 'POST':
displayResults = True
txtInput = form.inputText.data
scores = SentimentIntensityAnalyzer().polarity_scores(txtInput)
sentiment = getSentiment(scores.get('compound'))
if sentiment == 0:
sentiment = 'ERROR'
wordCount = getWordCount(txtInput)
mostFrequent = getMostFrequent(txtInput, 5)
return render_template(
'service.html',
nav=nav,
serviceUrl=serviceUrl,
form=form,
displayResults=displayResults,
txtInput=txtInput,
scores=scores,
sentiment=sentiment,
wordCount=wordCount,
mostFrequent=mostFrequent,
descriptions=descriptions
)
def print_sample_sentiments(samples):
"""*Processes sentiment analysis for a list of samples and prints the results*
samples (list): A list of strings to process
"""
sample_sentiments = []
counter = 0
for sample in samples:
score = SentimentIntensityAnalyzer().polarity_scores(sample)
counter += 1
sample_sentiments.append([
counter,
score.get('neg'),
score.get('neu'),
score.get('pos'),
score.get('compound'), sample
])
print(counter, score.get('neg'), score.get('neu'), score.get('pos'),
score.get('compound'), sample)
def analyze():
"""
This function will analyze the top 5 best reviews for sentiment. The imdb library already lists the reviews according to their filter::"Best". So we do not have to write a function for that. Also, instead of treating these as individual strings, we combine them into one large string. This allows for us to use a single natural language processing to determine the overall sentiment of the reviews.
"""
search()
value = 0
analyizetext = ""
for element in range(0, 5):
analyizetext = analyizetext + imdb._get_reviews_data(
identi)[value]['text']
value += 1
score = SentimentIntensityAnalyzer().polarity_scores(analyizetext)
if score.get("pos") > score.get("neg"):
print("The top 5 reviews for the movie:", title,
". Has been generally positive.")
elif score.get("pos") < score.get("neg"):
print("The top 5 reviews for the movie:", title,
". Has been generally negative.")
elif score.get("pos") == score.get("neg"):
print("The top 5 reviews for the movie:", title,
". Has been basically neutral.")
againfun()
# Top 10 Most Frequent Words in the Wiki Page
print(
f'Top 10 Most Frequent Words and its Frequency in Wiki page \'{page_1.title}\' are:'
)
word_freq.top_10_most_freq(word_freq.clean(page_1.content))
# Naturalization
import nltk
from nltk.sentiment.vader import SentimentIntensityAnalyzer
score_1 = SentimentIntensityAnalyzer().polarity_scores(
page_1.content) # AI Emotions
print(score_1)
neu_1 = score_1.get('neu')
def is_neutral(neu):
"""
Identify if the inputed neutrality measurement is greater or equal to .80
"""
if neu >= 0.80:
return True
else:
return False
if is_neutral(neu_1) is True:
print(f'The {page_1.title} Wikipedia Page is Neutral')
else: