Enzymind Labs

While I see the intent of the current "type and content" selecction tools, ultimately I think this should be automated. There's no reason a parser can't extract the "type", and I think the "content" list is very restrictive. I imagine you want to constrain content by forcing this selection, and use the selections for graphical node views.

But I think it would be better to use the text itself for tagging identifiers. I imagine being able to select a set of words from the text that would be used as tags. This would result in a richer (more possibilities) semantic network. Perhaps once particular keywords were used several times, the user could assign them a group, corresponding to the current topic list. The del.icio.us tag cloud works this way, and is very efective.