The Investigative Reporters and Editors’ (IRE) annual data journalism conference – commonly referred to as NICAR, short for National Institute for Computer-Assisted Reporting – brought together over 900 attendees and data experts from seven countries in Minneapolis, Minn. this year. Over four days, attendees immersed themselves in different levels of training sessions, from beginner-friendly tutorials to advanced investigative techniques. 

Kento Tanaka, data journalist from Nikkei, reckoned networking the most valuable because connecting with others was motivating. “The Google Cloud session might be useful for our newsroom,” Tanaka said. “We just started using GCP last year and are exploring how to utilize it effectively.”

Darryl Laiu, freelancer for the BBC, shared with Storybench, “For data projects, I’ve [used] Datawrapper and Illustrator. I particularly [like] the advanced mapping workshop by Ari Sen. He went through how to calculate travel times using python, and I hope I get to use it in a project in the future!”

If you couldn’t attend, there’s still plenty to explore! Storybench has curated a collection of newly introduced and zero-cost tools for those who can benefit even at home.

All about AI

AI topics continue to gain popularity – over 25 out of 217 sessions are related to AI or LLM, discussing data extraction, image analysis and ethical policy. Rosmery Izaguirre brought their session of using AI to extract data from documents. Jonathan Soma from Columbia University introduced AI practice in the newsroom and how to analyze images and videos with AI. Mike Reilley from UIC listed 22 pages’ easy- to-use AI tools for almost all newsrooms. Brandon Roberts from ProPublica demonstrated AI-Python combinations for basic journalism tasks while Alex Garcia focused on JavaScript edition and Matt Waite on R (see tipsheet1 and tipsheet2). Simon Willison posted the latest info in the world of LLMs.

Make your graphics publishable

Check out those impressive charts and maps – you can create them too! Maureen Linke and Stephanie Stamm of The Wall Street Journal taught us how to upscale graphics in R studio in an analytical way. University of Maryland’s Adam Marton introduced a transition tunnel for us to make charts customizable from R to DataWrapper. Another hands-on DataWrapper session by Taylor Johnston and Grace Manthey of CBS is for those who are looking for more ways on visualizations. Jovi Dai of San Jose Mercury News and East Bay Times posted how to make dataviz reusable with JavaScript. Brandon Liu posted different ways to unlock interactive maps.

Have fun with documents

Have you had trouble extracting data from poorly formatted PDFs or spreadsheets? Here are some methods you can try right away. Derek Willis from University of Maryland introduced extracting data from even the worst PDFs, while Jeremy Singer-Vine of the New York Times used pdfplumber to deal with complex PDFs. The Minnesota Star Tribune’s Tom Nehil posted how to handle problem spreadsheets with Python and MaryJo Webster provided a guide to do basic data cleaning with IF and VLOOKUP function in sheets.

Smart scraping

For those looking to scrape data without coding, independent journalist Samantha Sunne introduced some scraping methods without programming. Jonathan Soma of Columbia University shared how to use passive scraping for social media – a method that the scraping process is automated without human monitoring. IRE’s Cody Winchester posted a repo about data scraping from websites with Python. Simon Willison of Datasette shared some cutting-edge techniques particularly for video and image scraping.

For more information on NICAR 2025, IRE members can view information from conference sessions online here.