Twitter sentiment analysis

This project is a part of a School project.

-- Project Status:

Project Objective

The objective of this project is to build a model capable of predicting the sentiment of a tweet from a labeled dataset using Machine learning and natural language processing techniques.

Methods Used

• Natural Nalguage processing
• Machine Learning
• Data Visualization
• Predictive Modeling
• Data Analysis
• Data cleaning
• etc.

Technologies

• Python
• NLTK
• RegEX
• Pandas
• Jupyter Notebook
• Scikit Learn
• WordCloud
• etc.

Project Description

The project uses the data: tweets and their labels and supervised machine learning techniques in order to predict a binary Sentiment state (positive or negative). The learning phase is implemented after the preprocessing: cleaning + tokeninzing + vectorizing

Getting Started

1. Clone this repo.
2. Raw Data is being kept here within this link.
3. Extract the csv file training.1600000.processed.noemoticon.csv in the Data folder.
4. Imports are being kept here
5. The notebook contains the whole project useing ressources from the Data folder and .py files.

Credits

1. https://github.com/tthustla for the clear steps he used through the project which helped me succeed my first NLP project
2. Data credits by Stanford university

Full Report