INFORMATICA-ONLINE-TRAINING: September 2018

Sunday 30 September 2018

Saturday 29 September 2018

Saturday 22 September 2018

PYTHON BASIC SENTIMENT ANALYSIS USING ALEXA DATA

Basic sentiment analysis using python pandas

#python #pandas #dwbiadda #analysis #repeatedwords

import pandas as pd

alexa_data=pd.read_csv('C:\Data_Set\conv_data\Alexa_rating.csv')

alexa_data.head(6)

alexa_data['desc_conver'] = alexa_data.verified_reviews.str.lower().str.strip().str.split()

alexa_data.head()

stop_words = frozenset(['the', 'as', 'is','i','it','to','and','my','this','for','have','a','but','of','on','in','that','was','so','can','you'])

conv_list_reviews = list()

for data in alexa_data[['variation', 'desc_conver']].iterrows() :

row_data = data[1]

for final_word in row_data.desc_conver:

if final_word not in stop_words:

conv_list_reviews.append((row_data.variation, final_word))

verified_reviews = pd.DataFrame(conv_list_reviews, columns=['produc_type', 'Repeated_word'])

data_set=verified_reviews[verified_reviews.Repeated_word.str.len()>0]

data_set=data_set.groupby('produc_type').Repeated_word.value_counts().to_frame().rename(columns={'Repeated_word':'Cnt_of_word'})

data_set.head()

def display_top_rec(records, index_lvl=0):

final_result = records.groupby(level=index_lvl).apply(lambda records: records.nlargest(3,keep='first')).reset_index(level=index_lvl, drop=True)

return final_result.to_frame()

display_top_rec(data_set['Cnt_of_word'])