https://github.com/srashti13/twitter-sentiment-analysis

Categorize tweets as a positive or negative
https://github.com/srashti13/twitter-sentiment-analysis

jupyter-notebook python python-3-6

Last synced: 1 day ago
JSON representation

Categorize tweets as a positive or negative

• Host: GitHub
• URL: https://github.com/srashti13/twitter-sentiment-analysis
• Owner: Srashti13
• Created: 2019-08-30T16:45:04.000Z (over 5 years ago)
• Default Branch: master
• Last Pushed: 2019-09-05T02:47:47.000Z (over 5 years ago)
• Last Synced: 2025-01-19T07:42:22.525Z (6 days ago)
• Topics: jupyter-notebook, python, python-3-6
• Language: Jupyter Notebook
• Size: 97.7 KB
• Stars: 1
• Watchers: 2
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Twitter-sentiment-Analysis

This project was built as a part of my NLP assignment in my "introduction to Big Data" coursework at George Mason University. The code categorizes incoming tweets as either a positive or negative emotion based on words used.  

The project is built in jupyter notebook and utilizes Scikit-Learn & NLTK library to perform all the heavy lifting steps of feature extraction and preprocessing of tweets.  

Steps include:

* Remove punctuation

* Stemming

* Lemmatizing

* Tokenization

* Feature extraction

* Modelling

The code is built in function form, different steps are performed by different functions. This increases the readability of the code and its easy to re-use different functions later with minimal efforts

The workflow then includes building classification models (such as Naive Bayes used here).

### Results:

I was able to **achieve 91.79% accuracy** on test set with minimal efforts (which in turn also beats the class average of 88%).

To utilize the functions just download the .ipynb file