The aim of the project is to build a model which would predict the opinion of the people in tweets under particular # In this case we have used the omicron variant of COVID to see the world's opnion about it
The dataset was created by fetching the tweets from twitter using tweepy and converted to dataframe using Pandas library
In this project, we compare the sentiments of the tweets either positive,negative, or neutral opinion from vader_lexicon and tried to calculate the score of each sentiments
- Created a twitter developer account and fetched my consumer_key, consumer_secret, access_token, access_token_secret
- Connecting to Twitter with API
- Gathering the tweets regarding OMICRON
- Setting up the dataset
- Then performed Data cleaning :Removing the stopword and handling unnecessary wordse
- Then performed Visualising of the tweets using wordcloud
- Then used vader_lexicon (Valence Aware Dictionary and sEntiment Reasoner) to analyse positive,negative and neutral tweets
- Then Calculated the Sentiment Score
To be specific we have used a tookit Natural Language Toolkit from python to which provides easy-to-use interfaces to over 50 corpora and lexical resources such as vader_lexicon, SentimentIntensityAnalyzer, SnowballStemmer , stopwords removal to analyse the tweets and calculate the Sentiment Score
##LIBRARIES NEEDED
- Pandas
- Matplotlib
- nltk
- Numpy
- tweepy
- wordcloud
From this project we conclude that using Natural Language Toolkit we can analyse the sentiments of tweets