Welcome to our GitHub project! Here, we've embarked on an exciting journey to unravel the emotions behind each tweet using Sentiment Analysis. Our project is a digital detective, deciphering whether a tweet is offensive, happy, neutral, and more. Sentiment Analysis is a process that automates mining of attitudes, opinions, views, and emotions from text, in this case, Twitter data.
Our approach is a unique combination of the BERT model and LSTM. BERT, or Bidirectional Encoder Representations from Transformers, is a machine learning model for natural language processing. It reads the entire sequence of words at once, allowing it to learn the context of a word based on all of its surroundings. On the other hand, LSTM, or Long Short-Term Memory, is a type of recurrent neural network designed to remember values over arbitrary time intervals, making it particularly good at processing sequences of data.
The main code for our project is housed in main.py. We invite you to explore, contribute, and join us in this fascinating exploration of emotions in the Twitterverse. Let's dive in and uncover the sentiments behind each tweet!