Why it matters
Groups a keyword list into topic clusters by shared words and SERP-free similarity, with a pillar per cluster.
Tokenise, Porter/Snowball-lite stemming (English), stopwords, TF-IDF vectors, agglomerative clustering with cosine threshold, plus optional volume column to choose pillars. Honest note: no SERP overlap (that needs network); lexical clustering only. Everything runs in your browser; nothing you paste is sent unless you press Fetch.
How to use it
- Paste or type the input the tool asks for.
- Read the summary and the table or preview.
- Adjust the options if you need a different result.
- Copy or download the output. Nothing leaves this page.
FAQ
Is my input uploaded?
No. The work runs in your browser and nothing you type, paste or upload is sent anywhere.
Will this change my site for me?
No. It shows results and, where it writes rules or files, you copy them and apply them yourself.
Can I rely on this as official guidance?
It follows public documentation and well-known algorithms, and tells you where it is a subset or a simulation. Confirm critical changes in the official tool or a staging environment.
