This project explores different classification algorithms to classify Twitter data based on Sentiment and Sarcasm. Following steps are performed step-by-step :
- Twitter Data is being imported using the Twitter api python module. The dataset consists of tweet_ids, tweet_texts, usernames and date.
- Perform Exploratory Data Analysis on the dataset, cleaning out unnecessary values, visually plotting our results.
- Perform Sentiment Analysis using Textblob's inbuilt sentiment function and also by fitting a model on a labeled dataset and then predicting the sentiment using the model.
- Detect Sarcasm by training our model based on a labeled dataset, then predicting the sarcasm in our tweets.
- Compare F1 scores of classification models.