[Tagging] General tagging system problems

Daniel Koć daniel at xn--ko-wla.pl
Tue May 12 23:47:43 UTC 2015


W dniu 13.05.2015 0:56, Bryce Nesbitt napisał(a):
> Something that gets proposed from time to time is a tree hierarchy:
> shop=food:bakery:muffins+sweets:cookie

So what are the reasons it does not catch up?

I think the downside of this example is that it would be tedious and too 
detailed for people to enter and thus more error prone - we should rely 
more on editors with validation rules then. In a "bricked", lightweight 
version it would be rather:

shop + muffins + cookies

because if we have the general category tree curated on the wiki, we 
don't have to explicitly repeat it inside each item's tagging.

> That said the google approach would be to infer everything from text,
> social and web linking analysis:
> name=Fred's Bakery
> website=http://freds.example.org/ [1]

As we already have these informations, we could just ignore the rest and 
make a big software effort to recognize the meaning. But that would be 
hard problem, involving parsing the websites. So it is for Google, but 
their strength is automatic big data analysis, and ours is multiple 
users with their own analysis powers. =}

I would also say Google has much lower expectations - they want the map 
to be just a part of their services ecosystem, especially for 
advertising locations, while we want the map to have all the items one 
can think of.

-- 
"The train is always on time / The trick is to be ready to put your bags 
down" [A. Cohen]



More information about the Tagging mailing list