Week 1: Automated Factchecking

Giacomo, a Mathematics student from Imperial College London, tells us about his first week at Full Fact. He's working on a pioneering project: automated factchecking. We're grateful to Imperial's Charity Insights programme for making this happen. 

My internship is at Full Fact, the UK's independent factchecking organisation. Ahead of this year's referendum, they worked with ITV and Sky News to correct factual errors made in live debates, and they have asked for and got corrections in all the national newspapers. They play an ever-growing role in the effort to hold the media and politicians accountable to their claims.

Many assertions made in public debate come up again and again, they call them “zombie claims” at Full Fact (because they just don’t die). Claims like ‘poverty increased in the past six months’ or ‘unemployment decreased last year’. Factcheckers spend valuable time finding and interpreting government data for poverty or unemployment every time new datasets are released. In order for Full Fact to spend more time getting into deeper questions, and for journalists to have faster access to the truth, the charity is aiming to automate the most repetitive parts of their work.

This is what I’m working on: automated factchecking. The code I write will hopefully lead to the first ever factcheck carried out by Full Fact with a computer program. My project has quickly taken shape, and is divided into two parts. The first is natural language processing (NLP), that interprets, for example, Jeremy Corbyn's claim at Prime Minister’s Questions that more people than ever live on the streets. The second does the statistical work, checking the claim against government data for homelessness.

So far I have been using NLP and discovered that the problems involved in teaching a computer how to read are both interesting to solve, and also a massive challenge. Sentences are constructed with linguistic rules, but they are no way near as logical as  instructions a computer understands. Every rule that I come up with that can interpret text seems to require a hundred exceptions.

More broadly, the office experience is great. Having been provided with a good starting point, I am working on the code itself on my own. I am even helping shape what direction to take it in, so there is a certain amount of responsibility I have never felt before in the workplace. I am doing this because I like politics and this is something that can make a difference to democracy. I love having a TV on the wall showing the BBC news channel, and receiving emails with information on the day’s political happenings! To top it off, sunny lunchtimes with a falafel wrap in Gray’s Inn Gardens are beautiful.