What are you using for keyword clustering?
- Thread starter CPANathan
- Start date
- May 17, 2010
- 1,655
- 634
CPANathan
Elite Member
- Jul 9, 2019
- 2,850
- 10,255
Charly is the most overpowered LinkedIn poster I've ever come across
Purely licking his *ss these days, thanks ksha mate
Repsolution
BANNED
- Sep 27, 2025
- 15
- 20
Haha, been there keyword chaos in full swing.
You can try pandas + scikit-learn’s KMeans or DBSCAN for clustering; solid combo for multilingual sets like Swedish. If you want semantic grouping, toss in SentenceTransformers for embeddings first works way better than raw keyword matching.
CPANathan
Elite Member
- Jul 9, 2019
- 2,850
- 10,255
Maaaaaan that's a response and a half, I'll happily take heat speaking for the forum; we need more responses like that.
After giving me two or three weeks of study in that message hahaha, can't remember the last time I seen scikit popping up.
Really appreciate it dude
chloesanders
Newbie
- Oct 11, 2025
- 9
- 2
Yellowbrick has some neat clustering visualization tools if you want to inspect results.
- Oct 3, 2019
- 457
- 148
Try using scikit-learn or sentence transformers for clustering. Both handle big keyword sets if you batch them right.
- Sep 25, 2025
- 504
- 297
I use sentence-transformers (multilingual MiniLM) for embeddings and HDBSCAN to cluster. For Swedish, clean with stopwords and dedupe with rapidfuzz. Works fine in big batches.