Welcome to the Emotion Classifier project! This repository contains a comprehensive solution for emotion classification using Natural Language Processing (NLP) and deep learning techniques.
This project aims to build a model capable of accurately categorizing text samples into three emotion classes: anger, joy, and fear. The provided dataset, "Emotion_classify_Data.csv," forms the basis for training and evaluating our model.
- Introduction
- Dataset Overview
- Data Preprocessing
- Tokenization and Padding
- Model Architecture
- Model Training
- Model Evaluation
- Visualizations
- Actual vs. Predicted Examples
- Conclusion
Welcome to the Emotion Classifier journey! Our mission? Unleash the power of NLP and deep learning to decode emotions from text.
Our dataset, named "Emotion_classify_Data.csv," is a treasure trove of emotions—anger, joy, and fear—spread across 5937 entries.
Let’s open the treasure chest and explore the dataset structure.
Transforming text into numbers—our magical preprocessing step.
Balance is key! We split the dataset into realms of training and testing.
Turning text into sequences and ensuring a uniform length for our magical machine.
Our emotion classification castle: an embedding layer, an LSTM tower, and a dense layer with softmax magic.
Witness the model’s evolution over 10 enchanting epochs.
The grand reveal! Classification metrics paint a vivid picture of our model’s emotional insight.
A visual symphony of accuracy and validation accuracy dancing over epochs.
A heatmap spectacle revealing the model’s performance in vivid colors.
The stage is set! Ten examples showcasing the model’s prowess in predicting emotions.
In this magical journey, we’ve crafted a robust solution for emotion classification. The model, a beacon of performance, awaits those delving into the enchanting world of NLP.
- LinkedIn: Vidhi Waghela
- Kaggle: Vidhi Kishor Waghela
- GitHub: Vidhi1290
- Email: vidhiwaghela99@gmail.com
🚀 Open to Collaborations and Tech Discussions!