Reccommantion system based on sentiment analysis

Reccommantion system based on sentiment analysis

Project Overview

The core idea is to provide recommendations based on sentiment analysis of reviews. Here's how it works:

Users provide multiple inputs.

The system scrapes reviews from Twitter and other sources.

It then classifies these reviews as positive, negative, or neutral.

A pie chart visually represents the sentiment distribution.

The system recommends the option with the highest positive-to-negative ratio.

Key Components

Data Collection

Collected Egyptian language datasets from various sources:

Google Maps

Twitter

Google Play

Data Preprocessing

Implemented standard preprocessing techniques to clean and prepare the data for analysis.

Machine Learning Models

Utilized a range of models to classify sentiments:

SVM (Support Vector Machine)

LSVM (Linear SVM)

KNN (K-Nearest Neighbors)

Naive Bayes

Decision Tree

Random Forest

Logistic Regression

Deep Learning Models

Explored advanced deep learning architectures:

Forward Neural Network

CNN (Convolutional Neural Network)

Sequential Models

Applied models that are effective in processing sequential data:

RNN (Recurrent Neural Network)

GRU (Gated Recurrent Unit)

LSTM (Long Short-Term Memory)