Who Wins

Using Twitter API for opinion mining and sentiment analysis to predict Brexit results.

Implementation

Language version : Python 3.6
Operating System : Ubuntu 16.04, Windows 10 and MacOS High Sierra.

Getting Started

Libraries

• pyspark : Apache Spark Library for Python (used for Machine Learning Algorithms) http://spark.apache.org/docs/2.2.0/api/python/index.html
• pandas : open source, BSD-licensed library providing high-performance, easy-to-use data structures and data analysis tool https://pandas.pydata.org/pandas-docs/stable/api.html
• numpy : efficient multi-dimensional container of generic data https://docs.scipy.org/doc/numpy-1.13.0/reference/
• polyglot : natural language pipeline https://github.com/aboSamoor/polyglot

How to install

With pip :

sudo pip install --upgrade <library>

With homebrew :

brew install <library>

With Anaconda :

conda install <library>

Installing polyglot ...

Shell command line:

sudo apt-get install python-numpy libicu-dev

Dataset

• Algorithm used for collecting the data :
  https://github.com/Jefferson-Henrique/GetOldTweets-python

• Download the dataset :
  - directly from the website :
  https://www.kaggle.com/natmonkey/brexit-data-project-bdd/data
  - with the shell command line :
  kaggle datasets download -d natmonkey/brexit-data-project-bdd

• Data cleansing (csv file format):
  

cat fichier | sort | uniq

Machine Learning

The machine learning methods implemented are :

• Naive Bayes
• Logistic Regression
• Decision Tree
• Multi-Layer Perceptron
• Support Vector Machine

To execute ...

Execute the wanted method in the /learning repository.

Authors

Mira Ait Saada, Alexandra Benamar, Cristel Dos Santos Catarino, Joel Oscar Dossa