Tool for journalists: FOIA Predictor, for estimating the success rate of a Freedom of Information request in the US

Credit: Image from pexels.com

What is it? A free, open source tool that uses an algorithm to predict the likelihood of a Freedom of Information request being accepted by an organisation in the US.

Cost: Free

How is it of use to journalists: Freedom of Information requests are at the core of many investigations and allow journalists to solicit material that is in the public interest, such as how budgets are spent or communications between public figures.

However, not all FOIA requests are granted by government bodies and other organisations, which can reject them based on their content – whether or not they include all the relevant details – or argue that they do not hold the records a journalist is asking for.

FOIA Predictor is a free and open-source tool launched by data.world in June with the aim of predicting how likely a request is to be accepted based on similar ones made in the past.

The tool's algorithm is based on more than 9,000 FOIA requests tracked by non-profit website MuckRock in the US. It takes into account the number of words used, whether the document references fees, and the organisation's rate of granting requests, among others, to calculate the rate of success.

How it works: To get started, simply copy and paste the text of your request into the box on the FOIA Predictor website, and choose the US organisation you are aiming to send it to from the drop down menu. The list includes organisations specific to each state, but also the US Department of State and the US Census Bureau for example.

Click the 'get prediction' button and the tool will return a percentage of success and a breakdown of the attributes the algorithm looks at. For example if you haven't included your email, or referenced fees, FOIA Predictor will flag this so that you can consider making some adjustments.

"After we opened [data.world's] public beta last July, we very quickly noticed that journalists on not only wanted to find and use data that was already out there, but were interested in creating and sharing new sets of their own – often sourced from federal agencies and accessed via FOIA request," Ian Greenleigh, data.world's head of brand, told Journalism.co.uk via email.

"It's a way to shed a bit of light on what is often a fairly complicated and lengthy process. However, even if the success probability the app displays seems low, it's meant only to inform, not to be a factor as to whether someone submits a FOIA request or not," he pointed out.

The code for the tool is also available on GitHub so anyone with an idea for improving or expanding the FOIA Predictor can get in touch with the data.world team and request to be a contributor.

For more resources on how to craft and submit a FOIA request in the US, check out the Freedom of Information Center from Investigative Reporters and Editors, and the FOI step-by-step guide from the Society of Professional Journalists.

For the UK, use WhatDoTheyKnow, the Information Commissioner's Office website and our guide to submitting a FOI request as a starting point.

Free daily newsletter

If you like our news and feature articles, you can sign up to receive our free daily (Mon-Fri) email newsletter (mobile friendly).