Tweet Sentiment Analysis with Machine Learning and Deep Learning

Overview

The goal of this project is to build a machine learning model to accurately predict tweet sentiment. The project utilizes various machine learning models and deep learning techniques, with an emphasis on preprocessing, feature extraction, and model evaluation.

Results

The following visualizations and tables provide insights into the model's performance:

Model Performance

Model	Accuracy
Logistic Regression	70.99%
K-Nearest Neighbors (KNN)	48.50%
Random Forest Classifier	70.83%
Multinomial Naive Bayes	64.12%
Neural Network	98.36%

Features

• Data Processing: Clean and preprocess tweet text by removing noise, stopwords, and applying tokenization and stemming.

• Text Vectorization: Converts tweet text into numerical features using TF-IDF (Term Frequency-Inverse Document Frequency).

• Modeling: Evaluates several machine learning models including:

  • Logistic Regression
  • K-Nearest Neighbors (KNN)
  • Random Forest
  • Naive Bayes
  • Artificial Neural Network (ANN)

• Evaluation: Model performance is measured using accuracy, classification reports, and confusion matrices.

Sprint Features

Sprint 1: Data Collection and Preprocessing

• Collect and preprocess the tweet dataset.
• Clean the data by removing unnecessary symbols, stopwords, and applying text normalization techniques like stemming and lemmatization.
• Deliverable: Cleaned dataset ready for training.

Sprint 2: Feature Engineering

• Convert text data into numerical features using TF-IDF vectorization.
• Explore and experiment with different feature extraction techniques to improve model accuracy.
• Deliverable: Feature matrix ready for model training.

Sprint 3: Model Training and Evaluation

• Train several machine learning models (Logistic Regression, KNN, Random Forest, Naive Bayes) and evaluate their performance using accuracy, precision, recall, and F1 score.
• Compare the performance of different models to determine the best one for the task.
• Deliverable: Best-performing model selected and evaluated.

Sprint 4: Deep Learning Model Implementation

• Implement an artificial neural network (ANN) for tweet sentiment classification.
• Tune hyperparameters and evaluate the ANN's performance.
• Deliverable: Trained neural network model with high accuracy.

Conclusion

This project demonstrates how machine learning and deep learning models can be applied to sentiment analysis tasks. The neural network model performed the best, achieving an accuracy of 98.36%. Future improvements could involve exploring more advanced deep learning models or using more complex feature extraction methods to further enhance model performance.