miahelpt/mia_twitter_factchecker

An end to end factchecker for Covid-Vaccine related tweets

1

stars

26

commits

Python

primary language

Dec 12, 2022

updated

README

MIA Twitter Factchecker

This repository contains the source code for the MIA Twitter Factchecker.

The factchecker has been created by Stichting MIA Helpt, with financial support from the SIDN Fund

Installation (local)

To get started, clone the repo to your local machine, in a new folder.

Prerequisites

Running the code depends on:

  1. A local installation of SQLite (in case of using a local database) or access to an AWS-hosted MySQL or Aurora database, with credentials stored in AWS SecretManager
  2. Access to the Google Factcheck API (requires a developer key)
  3. Access to the twitter API, requires developer key and credentials.

Getting started

  1. Rename config_ini to config.ini and fill the required api keys (twitter / google)
  2. If using a local database, set use_local to 1, this will create a SQLite database in the main directory; if using AWS, fill the secret manager secret (and if using a tunnel, set tunnel to 1, this will assume a tunnel to the database at localhost:3001)
  3. Install the requirements (pip install -r requirements.txt)
  4. Download the spacy dutch tokenizer: python -m spacy download nl_core_news_sm
  5. Download the models: ./models/download.sh

Filling the database and start matching facts.

  1. Sync the Google factchecks: python load_factchecks.py

Contributors

rsips

26 commits

miahelpt/mia_twitter_factchecker

An end to end factchecker for Covid-Vaccine related tweets

1

stars

26

commits

Python

primary language

Dec 12, 2022

updated

README

MIA Twitter Factchecker

This repository contains the source code for the MIA Twitter Factchecker.

The factchecker has been created by Stichting MIA Helpt, with financial support from the SIDN Fund

Installation (local)

To get started, clone the repo to your local machine, in a new folder.

Prerequisites

Running the code depends on:

  1. A local installation of SQLite (in case of using a local database) or access to an AWS-hosted MySQL or Aurora database, with credentials stored in AWS SecretManager
  2. Access to the Google Factcheck API (requires a developer key)
  3. Access to the twitter API, requires developer key and credentials.

Getting started

  1. Rename config_ini to config.ini and fill the required api keys (twitter / google)
  2. If using a local database, set use_local to 1, this will create a SQLite database in the main directory; if using AWS, fill the secret manager secret (and if using a tunnel, set tunnel to 1, this will assume a tunnel to the database at localhost:3001)
  3. Install the requirements (pip install -r requirements.txt)
  4. Download the spacy dutch tokenizer: python -m spacy download nl_core_news_sm
  5. Download the models: ./models/download.sh

Filling the database and start matching facts.

  1. Sync the Google factchecks: python load_factchecks.py

Contributors

rsips

26 commits

Languages

Python

90.6%

JavaScript

4.9%

HTML

2.1%

Dockerfile

1.2%

CSS

1.1%