3 ms·
StackOverflow has this feature that when you write a new question, it tries to fuzzy-match that up against existing questions. I wonder if an approach similar t
by kleiba 1y ago
StackOverflow has this feature that when you write a new question, it tries to fuzzy-match that up against existing questions. I wonder if an approach similar to yours, using search, could be employed on HN as well to reduce the number of dupes?!
This would be helpful especially for those cases where the same story gets covered on multiple places on the internet and so URL matching doesn't help.
- nielsbot 1y agoWonder if ML could help automate merging of duplicate (or similar topic) threads? (I don't know much about ML)