[Talk-GB] Post-processing shop values (was mechanical edit)

David Woolley forums at david-woolley.me.uk
Mon Nov 3 13:07:30 UTC 2014

On 03/11/14 12:19, SK53 wrote:

> Bringing together bookmaker tags is harmless, but does little to gather

Except, possibly that one of the big political issues about such shops 
is that they now get a lot of their money from fixed odds machines, 
which means there is no element of bookmaking involved.

> The key to getting shop data better is actually to map many more of
> them. There will be no serious interest in the shop data in aggregate
> (rather than individual POIs) until it is much more complete. Improving

And more than just collecting the initial data, unless you can actually 
maintain it is going to be very unreliable. Not only are a very small 
proportion of shops mapped, but many of them have been mapped by 
non-local visitors (one in particular) who are only interested in the 
first time mapping and have no plan in place to regularly re-survey 
them.  This information rapidly becomes stale.

I'd guess you would need almost two orders of magnitude more active 
mappers before you had any hope of reasonably complete and up to date 
information on retail outlets, particularly those which are not 
fashionable in youth culture.  I'd estimate the amount of effort 
required as similar to that needed to map every shop in the country 
every one to two years.  Although we are talking about maybe 20% 
turnover a year, you still need to go out to the street and compare the 
details against what is recorded far more regularly than that. Each 
visit still requires careful checking of the details of everything.

Incidentally, if anyone knows of a pre-written tool that will construct 
a schedule of premises on a street in strict walking order, I might be 
interested, as it might make auditing streets mapped by outsiders 
easier, and therefore encourage me to maintain over a slightly wider 
area.  I prefer paper for this sort of thing, both because it is less 
likely to attract unwanted attention, and because it doesn't put 
expensive gadgets at risk.

At the moment, I'm only really tracking two streets, which I visit often 
enough to notice changes, and for which I can remember what was there 
the last time.

More information about the Talk-GB mailing list