News Sift Bot

Application that classifies news into positive and negative and gives the results to the user via telegrams of the bot.

It consists of four main parts:

receives the text of articles from the site https://www.bbc.com/.

prepares data, creates features, fits Logistic Regression and gives predicts for the new articles.

provides connections to the database.

helps users receive news classified into positive and negative, depending on which user prefers.

Getting started

To start using this application you need to run the following commands:

python3 main.py migrate - initialize tables,

python3 main.py train - train model and save it

python3 main.py parse - parse new articles and make predict for them

python3 main.py run-bot - run telegram bot

Name		Name	Last commit message	Last commit date
Latest commit History 1 Commit
kerasenv/lib/python3.8/site-packages		kerasenv/lib/python3.8/site-packages
model		model
parser		parser
storage		storage
telegram_bot		telegram_bot
.env_dist		.env_dist
.gitignore		.gitignore
Procfile		Procfile
README.md		README.md
__init__.py		__init__.py
config.py		config.py
main.py		main.py
nltk.txt		nltk.txt
parse_new_data.py		parse_new_data.py
requirements.txt		requirements.txt
run_model.py		run_model.py
runtime.txt		runtime.txt