We do something similar at Newslines, a crowdsourced news search engine that is a curated database of news events, kind of like a database-driven Wikipedia meet

We do something similar at Newslines, a crowdsourced news search engine that is a curated database of news events, kind of like a database-driven Wikipedia meets daily news. We generally source each event to a single news report, but we may add more sources programmatically. By sorting the news into discrete events we can eliminate a huge amount of manual searches on Google and other sites, compiling news from multiple sources into a crosslinked timeline format that saves readers days of work. That said, I think the drive to fact-check every statement in an article is a red herring. Other sites have tried this (Infobitt, Grasswire) and it doesn't work well for general reader.