HN has a dupe checker based on URLs. If you're asking for a dupe checker based on content, not URLs, that's harder than it sounds. I don't know how to do it and I don't know of anyone who's solved the problem.

To solve it, you don't just have to identify when two articles contain the same story—you also have to distinguish when they don't contain the same story, even if the content is similar.

Since we don't know how to do that, we rely on a combination of moderation and community feedback (i.e., human judgment) to identify when a story has already had significant attention on HN, and merge or downweight the follow-ups. Since it isn't automated, this approach misses things. It does catch most of them though.