I keep searching for software that will enhance this blog. Two days ago, Bill Dimm, President of Hot Neuron, showed me one that impressed me.
Dimm demonstrated as I watched how his Clustify software organized 354 of my posts on productivity. The algorithm looks for and brings together posts that share words and concepts.
In the five-minute setup I picked a similarity cutoff, which is a threshold for when the software will identify a concept group. I could have also played around with words that are significant or words that should be ignored.
The software took about two seconds to sort my collection. The screen shows the clusters, the number of posts in each cluster, and “key words” extracted from the cluster. You can drill down into each key word and see the other clusters where it surfaced. The lower right window shows the actual document (blog post) and puts the key words in bold. I did not do it, but it is easy to “tag” related clusters so that they are combined. There is an automatic tagging function that makes this even easier to do.
Clearly a potent tool for making sense out of masses of documents in discovery, I was enthralled by the speed and accuracy of Clustify’s ability to assemble related posts. It would do the same for hundreds of thousands or millions of e-mails or other documents. I only wish I had ten times as many posts so I could let the software do for me what I painstakingly do with my metaposts.