Utility Mill

The idea and some of the code comes from chapter three of "Programming collective intelligence".

Each line should contain a phrase or entry of text you wish to cluster. For example it could be search phrases from your web site (one per line), or whole blog entries (one entry per line so remove newlines before).

NOTE: This utility only accepts thirty lines of input currently so it probably won't be too useful except as a novelty.

Try copying the code and running it locally to handle larger input.

Here is an example I made (locally) clustering 500 search phrases coming in Answer my searches(for a one week period)

Full Size Image

TODO:

• Allow more inputs some how, or clean up code to be more easily run locally
• Allow optional title for each data line (comma seperated)
• If no title, show only first N characters of each input line