k wrote:
] Decius wrote:
] ] ] a conversation between Clay Shirky, Stewart "Flickr"
] ] Butterfield,
] ] ] Joshua "Delicious" Schachter and Jimmy "Wikipedia" Wales
] at
] ] the
] ] ] O'Reilly Emerging Technology Conference in San Diego.
] ] Folksonomy
] ] ] is the process of letting users generate and apply their
] own
] ] tags
] ] ] to shared items and then discovering which tegs users
] share
] ] with
] ] ] one another. Unlike previous ventures into this field, the
]
] ] tags
] ] ] aren't "controlled"
] ]
] ] What are people's thoughts about replacing the topics system
]
] ] in MemeStreams with a folksonomy like Del.icio.us?
]
] [ I'm with Rattle. I think it's a fine idea. Strict
] hierarchies are either too complex, and thus unmanageable, or
] not complex enough, and therefore not useful for real
] categorization. If the category can't be determined
] automatically via semantic indexing (which is the most
] difficult, but still, i think, the best) then total
] flexibility is better. Of course, even the best semantic
] analysis isn't always going to mark a document with all the
] tags you might consider relevant, so this would even be useful
] as an addition to that level of categorization. There's no
] reason something shouldn't get multiple tags, in fact, though
] memestreams I guess only sets that on initial submission...
] -k]

I'm forced to agree. This type of free-flowing organization offers a much wider ability in the area of usage. I'd like to suggest however that if it were considered, so then might be, a core or initial tagset from which branches might grow. Making them vanilla enough to be useful and unconstrictive, both giving some internal engine the ability to make relevant high-end references to any user created tag. And perhaps even a method for users to update or grow the coreset, with obvious limitations.
This might help handle the initial potential for it growing into chaos and also give any "transition software" a leg up.

RE: Folksonomy conversation