It's possible by analyzing the title and subtext (and the article snippet, if it exists). I tried to have an AI model estimate the likeness of articles. Worked relatively well, but I lack the motivation to build it out into a usable app.