Global Investigative Journalism Network (https://archive.gijn.org/tag/data-mining/)
Converting text documents like PDFs to spreadsheets is tedious and expensive work. To see how well AI tool ChatGPT can extract data from PDFs, data journalist Brandon Roberts wrote a Python script to convert two document sets to spreadsheets.
In 2022, many of GIJN’s original stories focused on reporting techniques relevant to global threats that grew or emerged this year — including Russia’s invasion of Ukraine, democratic decline, growth of far-right populism, the challenge of accountability journalism in the Arabian Gulf, abuse of migrants and minorities, and the exiling, assault, and legal harassment of independent media.
In this edition of GIJN Toolbox, we examine the latest advancements from the IRE22 conference on data extraction and optical character recognition (OCR) tools for turning unwieldy documents into searchable spreadsheets.
In the latest webinar in GIJN’s series on Investigating the Pandemic, a leading data journalism trainer and a Pulitzer Prize winning data scientist shared insights on data sources that can illuminate COVID-19 impacts beyond health.
In Malaysia, the Sinar Project has taken up the challenge of collating patchy government statistics to provide the public, and journalists, with useable data.
As our governments and businesses become increasingly flush with information, more and bigger data are becoming available from across the globe. Increasingly, investigative reporters need to know how to obtain, clean, and analyze “structured information” in this digital world.
What’s the global data journalism community tweeting about this week? Our NodeXL #ddj mapping from February 4 to 10 finds exciting releases with a document analysis tool by @ICIJorg and an audio-driven visualization tool by @f_l_o_u_r_i_s_h, while @kloopnews highlights domestic labour imbalance in Kyrgyzstan and @davidottewell emphasizes that data journalism is not just for nerds in a corner.