Yes, completely. If you want some fun on holiday sometime set the
option on your sat nav to 'shortest distance' and then see where it
takes you. You get to go down all the narrow wiggly lanes through
places you would normally never discover. This gives a rather
beautiful illustration of the fallacy that people might navigate
according to distance minimisation (pre-GPS of course :-)
**A word of warning though - don't try it in those Italian hill towns
where the roads get narrower and narrower until nothing larger than a
cinquecento will fit through...
Alan
On 14 Jun 2008, at 12:10, Alasdair Turner wrote:
> There is a problem with GPS tracks from cars however, in that you need
> to know that you are not simply discovering the GPS manufacturer's
> algorithm for routing.
>
> Alasdair
>
> On Fri, Jun 13, 2008 at 5:21 PM, Rui Carvalho <[log in to unmask]>
> wrote:
>> Hi Bin, Martin,
>>
>> Yes, the use of GPS data is definitely a step in the right direction.
>>
>> Any chance someone might be able to summarize the findings with these
>> datasets in a couple of paragraphs -sorry, time is limited these
>> days :)
>>
>> Rui
>>
>>
>>
>> On Fri, 13 Jun 2008 18:07:37 +0200, Martin Tomko <[log in to unmask]
>> >
>> wrote:
>>
>>> Hi Bin, Rui,
>>> very interesting paper indeed. As for what the use may be for
>>> ranking
>>> streets according to some other criteria then administrative/
>>> functional
>>> rankings designed by road network planners, we have been looking
>>> at how a
>>> similar process may underline people's inference of familiarity
>>> with a
>>> street. We argued that developping experiential rankings may support
>>> generation of better route directions. See Tomko, Claramunt, Winter
>>> (2008): Experiential hierarchies of streets , Computers,
>>> Environment and
>>> Urban Systems 32(1): 41-52
>>> http://linkinghub.elsevier.com/retrieve/pii/S0198971507000221
>>>
>>> and some consecutive papers. I am reading Bin's paper with great
>>> interest,
>>> as he was able to look at actual correlation with traffic data. In
>>> our
>>> approach, we had to use named streets as basic network elements,
>>> as common
>>> labels are relied upon in route directions. I will see if the two
>>> approaches could be brought together somehow.
>>>
>>> cheers
>>> Martin
>>> =
>>> =
>>> =
>>> =
>>> =
>>> ====================================================================
>>
>>
>
>
>
> --
> Course Director
> MSc Adaptive Architecture and Computation
> UCL Bartlett School of Graduate Studies
>
> http://www.vr.ucl.ac.uk/people/alasdair
>
|