2019/Grants/fakenewsguard
Title:
fakenewsguard
Name:
Mattia Spinelli
Wikimedia username:
E-mail address:
matt
fakenewsguard.com
Resume:
https://www.linkedin.com/in/mattia-spinelli-029044a8/ https://www.linkedin.com/in/mattia-trino-8a9ba8114/ https://www.linkedin.com/in/giuseppe-ianniello-7120a5112/ https://www.linkedin.com/in/igor-shelvinskyi-96055249/
Geographical impact:
global
Type of project:
Technology
What is your idea?
Use currently available Machine Learning techniques to help speed up the factchecking process. We have already built a search engine dedicated to fake news. The idea is to use ML to compare semantically a piece of text with our whole database of fake news (10.000+ at the moment). If a statement agrees semantically with one or more fake news, it's probably fake as well.
Why is it important?
It would speed up researches about fake news, and partially automate it.
Is your project already in progress?
Yes. www.fakenewsguard.com info.fakenewsgaurd.com We won a competition organised by Singapore press holding for 20k$ to develop this prototype.
How is it relevant to credibility and Wikipedia? (max 500 words)
Wikipedia could integrate the tool/API to help user verify intormation. We could also benefit from Wikipedia data.
What is the ultimate impact of this project?
Our goal is to provide a tool to help people stay informed.
Could it scale?
Yes, we already tested the scalability with the Singaporeans.
Why are you the people to do it?
In our team, we have the technical knowledge and the domain knowledge.
What is the impact of your idea on diversity and inclusiveness of the Wikimedia movement?
I am not sure my answer will be meaningful, but our project help surface the work of many fact-checkers from around the word. They are lesser known voices in the world of journalism, but they deserve a bit more recognition.
What are the challenges associated with this project and how you will overcome them?
There are some technical challenges, some of which we already resolved and some that still need to be addressed. We design this solution starting from readily available technologies, so we are confident this can be done. Moreover our last prototype has passed validation from our customer, Singapore Press Holding. Some more steps need to be taken to improve the data and the reliability of the results.
How much money are you requesting?
10.000 UK Pound
How will you spend the money?
To pay for development and related expenses.
How long will your project take?
We are already live. We aim to deploy a publicly available API in 3 months.
Have you worked on projects for previous grants before?