Twitter Sentiment Analysis on GPT-related Tweets This project performs sentiment analysis on tweets related to GPTs using R. The analysis uses two different sentiment lexicons: NRC and AFINN.

Data Collection

• A Kaggle dataset that scraped tweets from Twitter (currently known as X) with hashtags related to GPTs was used

Data Preprocessing

• Firstly, all individual words were extracted from the tweets dataset.
• Secondly, the words were converted to lowercase.
• Afterwards, the extracted word list was cleaned by removing the "stop words" and words with non-alphabetical characters.

Data Analysis and Visualisation

• Sentiment analysis was condicted using the NRC and AFINN lexicon
• NRC lexicon categorizes words into emotions and AFINN lexicon assigns a numerical sentiment score based on the positivity or negativity of the word.
• For NRC, a bar chart of the frequency of the sentiments was plotted and the top 3 sentiments were recorded.
• For AFINN, a bar chart of the frequency of the sentiments was plotted and the top 3 positive words were recorded. The top word from each sentiment score was also noted.
• The above recorded words were used to speculate people's opinions on GPTs (found in the Powerpoint).
• Finally, ChatGPT's opinion on this results was recorded (found in the Powerpoint).

If you made it this far, thank you! Have a great day :)