News channels word sentiment analysis

keywords "migrant", "refugee" and "asylum seeker"

A NLP project, enables you to scrape news channels Al Jazeera, BBC(uk), PBS, Positive-news, RTV and The Local (all 9 branches).

Preprocess the text, build word2vec models and analyze the sentiment.

Installation

Prerequisites

Python 3.6 (tested on ubuntu 18.04)

Scraping

beautifulsoup4
selenium
requests
lxml
pyqt
qt

Processing

nltk
gensim
googletrans # pip3 install git+https://github.com/BoseCorp/py-googletrans.git --upgrade
pandas

Visualizations

seaborn

Installing (conda)

Simple way to install all the packages is to simply import the environement (onj) into your conda installation.

Use conda environement file (.yml) to create environement from file. The file is located in environment/conda-env/ To use, run:

conda env create -f onj-env.yml

Scraping

Data modeling

Analysis

License

This project is licensed under the MIT License - see the LICENSE.md file for details