Product Sentiment Analysis

Description

This project is a Logistic Regression model trained and fine tuned using pscikit-learn and python, the frontend is a react server connected to a backend flask server

Features

• Input a product review and the ML model will determine whether the review is positive or negative

Table of Contents

• Technologies Used
• Getting Started
• Installation
• Running the Application

Technologies Used

• React - For building the user interface.
• Flask - For handling routes and connecting to the frontend.
• Python - For training the ai prediction model on the data
• CSS3 - For styling the layout.

Getting Started

Prerequisites

• Node.js, npm, and python installed on your system.

Installation

• To get started, clone the repository and install dependencies

  git clone git@github.com:dellis2cs/HaiSentimentAnalysis.git

  cd aiSentimentAnalysis

  npm install

  pip install flask
  pip install pandas
  pip install scikit-learn
  pip install flask_cors
  pip install nltk
  pip install pickle

Usage

Run the application

cd frontend
npm run dev

cd backend
python3 app.py

Ways to improve the model

• The main issue I ran into when training this model was that when encountering phrases, not individual words the AI struggled
• For example if I said "this product is not fun", it would see the word fun and mark it positive
• I have trained and fine tuned to keep these discrepancies to a minimum, however you could transition to using a more power and context aware model such as bert