John Robb has been spotted here since he left UserLand. John is a champion of K-Logs, and he has elaborated some of his ideas on a recent post.
As a computer scientist, I take exception to the use of text documents (or blog posts) to be referred to as knowledge. Knowledge essentially has to have intelligence, like rules, decision making or learning ability -- hence a database can never be a knowledgebase.
Trackbacks/Backflips: 1 , 2
Having pinpointed my pet peeve, I have some notes from customer sites that are actually using my SimplyBlog software for capture of human knowledge.
- A K-Log system must have categorization -- this can be human specified (like in Movable-Type), or automated via a content classification engine (CCE)
- Full-text search is also essential to dissimilate the knowledge.
- Chronological organization of posts is quite useless in a K-Log. Rather, organization by importance (most important), relevancy (post popular), author, and whether peer-reviewed or not, are more important.
- At all the installations, the customers asked a search by Last-Updated feature. This tells many things (more below)
- Knowledge is not equal -- some posts must be available only to some eyes.
- As absurd as it might sound, nicely printing the blog post is very important. I have seen executives go to meetings with thick stack of blog printouts!
More on "Last Updated"
A recent experiment called "Winer Watcher" by Mark Pilgrim that reported changes made to a blog post was very popular (it has been since withdran, hence no link), so I am thinking there might be a lot of interest in micro-following the emergence of a blog-post (or a company policy, or a business decision). Definitely something for the designers to consider...
