Re: Understanding keywords

< Next Message | Back to archived message list | Previous Message >

Note: This message is from the outliners.com archive kindly provided by Dave Winer.

Outliners.com Message ID: 2673

Posted by chris
2005-02-01 01:37:24

I suppose the method you describe would be useful if you can’t remember all the related keywords; for example, if you are working with very technical or dense information that you are not familiar with.

My actual usage of information managers suggests that I am never really in the situation where I don’t know the general scope of what is living in my data. I never struggle to come up with the search terms which will lead me to the items I desire to see. Perhaps there is a different usage pattern for someone studying new material from someone like me who is mostly managing domains I already know well.

I think the efficacy of locating items based on multiple search terms depends somewhat on how well an application searches. As an example, if I have clipped some articles about Freud and I search for “Freud dream analysis free association” (to build on your example), most of the apps I’ve used (Ultra Recall, Zoot, Info Select) that support implicit “and” searches would pull together all of the articles that had all of these terms - and the search results are essentially a “folder” or “keyword hoist” of the matching articles. True, I knew what I was searching for beforehand, as opposed to exploring my data in some more conceptual way, but it seems cumbersome to me to have to enter in concepts that become search terms via some separate form. Again, though, this may just be based on my personal usage patterns.

One thing I do think is interesting is the idea of semantic relationships based on word frequency. I recently read this article <http://www.stevenberlinjohnson.com/movabletype/archives/000230.html> about DevonThink (Mac only), which has this feature. Haven’t seen any tools for Windows that do this.

I am interested in exploring these concepts to improve how I can structure my data, with the goal of cutting out extraneous steps that don’t offer much benefit. If categories/keywords/folders are essentially word tags, it seems the best way to manage things is through a very strong text manipulation tool (or set of tools). I have been playing around with using plain text files and trying to identify where the benefits kick in with full-fledged information managers.

Back to archived message list