shahil98 / opinion_mining_and_analysis

This repository has the code for an application that performs different types of analysis on tweets.

Opinion Mining And Analysis

This is a NLP, Deep Learning and Data Analysis Project.

Types of analysis performed in this project on tweets fetched using a keyword or sentence searched

1. Displaying most frequently used hashtags.
2. Sentiment Analysis
3. Visibility Analysis i.e. calculating the reach of positive tweets vs negative tweets.

Sentiment Classification Architecture

Getting started

1. Clone this repository.

git clone https://github.com/Shahil98/Opinion_Mining_And_Analysis.git

2. Create a twitter developer account.
3. Execute pip install -r requirements.txt to install necessary libraries.
4. Download dataset from https://www.kaggle.com/kazanova/sentiment140 and place the csv file inside Classification network folder.
5. Execute python generate_data.py inside Classification network folder to generate data for training the sentiment classification network.
6. Execute python network.py to train the netwotk. After execution is completed a model.h5 and model.json file will be created.
7. Move to webApp folder and execute set FLASK_APP=main.py.
8. Add twitter developer credentials in main.py file.
9. Execute python main.py to start the application at http://localhost:5000.