Passive Air Cooling System and Solar Water Heater.ppt
Sentiment Analysis on Demonetization Tweets
1. Project: Sentiment Analysis on Demonetization Tweets
Student-Name: Amit Phaujdar
Eckovation
Course-Name: Machine Learning in Python
2. PROBLEM STATEMENT:
•The purpose of the project is to find out the
views of different people on the
demonetization by analysing the tweets
from twitter. Here the Twitter API is used to
extract all tweets into a demonetization-
tweets.csv related to demonetisation in
India by using appropriate filters.
3. • For the project we divide each tweet text into words
to calculate the sentiment of the whole tweet and
rate the word as per its meaning from +5 to -5 using
the dictionary AFINN. The AFINN is a dictionary which
consists of 2500 words which are rated from +5 to -5
depending on their meaning.Then a sentiment rating
for each tweet is generated. After that, we perform
the sentiment analysis by filtering out positive
sentiment tweets from negative sentiment tweets.
4. CRITICAL STEPS/ALGORITHM(STEPS CARRIED OUT):
• The project is divided into 3 parts:-
• Sentiment Analysis.py - in this file we perform the actual sentiment
analysis as mentioned above to determine the sentiment rating for
each tweet of a dataset containing 14940 tweets. For this purpose we
use natural language processing to break up the tweets into words
and as said above taking reference from the AFINN dictionary each
word is allotted a rating and the sum of all the ratings for the words
becomes the sentiment rating for the tweet. Based on if the overall
rating is positive or negative we can derive our respective insights.
5. •EssentialData.py - Contains extracting and modifying
the data to be used for plotting and used for drawing
conclusions from it , filtering only the data required to
explain the insights derived to best explain the results.
•Plotting.py - Here 5 plots are made that help us
answer questions like Is the general sentiment
positive or negative and by how much? what is the
maximum positive sentiment and maximum negative
sentiment, the maximum number of retweets
obtained, average sentiment for the group and
likewise.
6. • data set for Total Sentiment:
Positive/Negative :
• [2081.8304363029592, -
1481.106017300224]
Conclusions Drawn:-
1. The general sentiment as
can be seen is positive and
is greater than the negative
sentiment by
600.7244190027632.
7. • data set for Sentiment
Extremities:
Positive/Negative:
• [1.299867367239363,
0.04020913112468294, -
2.2283440581246223]
Conclusions drawn:-
1. The maximum positive
sentiment is around
1.299(greater by 3.527)
while the maximum
negative sentiment is
around -2.228
8. • data set for Top 5 Negative
Influences:
• ['CPIMBadli'
'HASHTAGFARZIWAL'
'PRAMODKAUSHIK9'
'deeptiyvd'
• 'rahulja13034944']
9. • data set for Top Positive Influences:
• ['1DharMMA' 'AniruddhasT' 'Hello_Guddi'
'JusteyAlex' 'JustinSandefur'
• 'Mridul_Sharma1' 'MuzzammilAap'
'Nischaytweets' 'Reaganite8'
• 'Saquibanwargift' 'Stupidosaur'
'SukhvinderShahi' 'golgappewala'
• 'hemant_shiyal' 'iridemach5' 'pavan_sethi']
• [1.29986737 1.29986737 1.29986737
1.29986737 1.29986737 1.29986737
• 1.29986737 1.29986737 1.29986737
1.29986737 1.29986737 1.29986737
• 1.29986737 1.29986737 1.29986737
1.29986737]
Conclusions drawn:-
1. The top(16) positive
influencers all share the same
sentiment value 1.2998
10. • data set for Top Re-tweeted
Users' sentiment Values:
• ['1SunnyElias' 'apoliceshanigm2'
'sxP6DbxfufguCc0' 'rayyat9tfoi'
• 'subhashjsr' 'Krishna20977027']
• [0. 0. 0. 0.18569534
0.37139068 0.37139068]
Conclusions drawn:-
1. Out of the 6 top
retweeted Users half have
a neutral sentiment value
as can be seen from graph.