Postal codes as alternate names

[Ryan Levering-2]

Is this ok?  Several of the European countries have postal codes as
alternate names for most cities (France/Spain).  This seems wrong to
me, even if they are given "post" language identifiers.  Postal codes,
if they are integrated, should be a separate feature class since there
is rarely a 1:1 postal:city pairing.

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "GeoNames" group.
To post to this group, send email to geonames@...
To unsubscribe from this group, send email to geonames+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/geonames?hl=en
-~----------~----~----~----~------~----~------~--~---

[Marc Wick]

Most users are actually directly asking for a postalcode-city relation.
At the moment we don't have enough information about postal codes to add
them as separate feature class.

Marc

Ryan wrote:
> Is this ok?  Several of the European countries have postal codes as
> alternate names for most cities (France/Spain).  This seems wrong to
> me, even if they are given "post" language identifiers.  Postal codes,
> if they are integrated, should be a separate feature class since there
> is rarely a 1:1 postal:city pairing.
>
> >
>
>  


--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "GeoNames" group.
To post to this group, send email to geonames@...
To unsubscribe from this group, send email to geonames+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/geonames?hl=en
-~----------~----~----~----~------~----~------~--~---

[Ryan Levering-2]

Yes and postal codes are very valuable, so I understand the desire.  
I'd love to see them integrated in the data.  However, I have a  
semantic and practical problem with doing it via alternate names:

Semantic - postal codes and cities, at least in most countries that I  
know of, don't have a 1:1 mapping with cities and generally have  
fairly different boundaries.  To say that a postal code is an  
alternate name of a city is just wrong.  The city is not defined by  
it's zip code nor is it used in common language as an identifier for  
the city.

Practical - I have an address parser that uses geonames data.  I have  
to filter out these postal alternate names or the parser gets very  
confused at why there are two cities, possibly conflicting, in the  
same address.

Is the eventual plan to add them as separate feature classes?  What  
information are you lacking?

On Jun 22, 2009, at 12:47 AM, Marc Wick wrote:

>
> Most users are actually directly asking for a postalcode-city  
> relation.
> At the moment we don't have enough information about postal codes to  
> add
> them as separate feature class.
>
> Marc
>
> Ryan wrote:
>> Is this ok?  Several of the European countries have postal codes as
>> alternate names for most cities (France/Spain).  This seems wrong to
>> me, even if they are given "post" language identifiers.  Postal  
>> codes,
>> if they are integrated, should be a separate feature class since  
>> there
>> is rarely a 1:1 postal:city pairing.
>>
>>>
>>
>>
>
>
> >

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "GeoNames" group.
To post to this group, send email to geonames@...
To unsubscribe from this group, send email to geonames+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/geonames?hl=en
-~----------~----~----~----~------~----~------~--~---

[Bernard Vatant]

Hi all

Completely tuned with Ryan on this, I have to cope those days with data
where places (in france) are "identified" by Name + Postal code, that
have to be attached to cities identified by insee codes or geonames id
(we have the mapping). What happens in France is roughly as following

- Big cities have several postal codes.
- Some small-to-middle cities can be uniquely identified by a single
postal code.
- Many small villages share a postal code with their neighbours and a
local city.

Only in the second case the use of postal code as alternate name can
make sense, but I agree it should be stored in a specific field.

Bernard

> Yes and postal codes are very valuable, so I understand the desire.  
> I'd love to see them integrated in the data.  However, I have a  
> semantic and practical problem with doing it via alternate names:
>
> Semantic - postal codes and cities, at least in most countries that I  
> know of, don't have a 1:1 mapping with cities and generally have  
> fairly different boundaries.  To say that a postal code is an  
> alternate name of a city is just wrong.  The city is not defined by  
> it's zip code nor is it used in common language as an identifier for  
> the city.
>
> Practical - I have an address parser that uses geonames data.  I have  
> to filter out these postal alternate names or the parser gets very  
> confused at why there are two cities, possibly conflicting, in the  
> same address.
>
> Is the eventual plan to add them as separate feature classes?  What  
> information are you lacking?
>
> On Jun 22, 2009, at 12:47 AM, Marc Wick wrote:
>
>  
>> Most users are actually directly asking for a postalcode-city  
>> relation.
>> At the moment we don't have enough information about postal codes to  
>> add
>> them as separate feature class.
>>
>> Marc
>>
>> Ryan wrote:
>>    
>>> Is this ok?  Several of the European countries have postal codes as
>>> alternate names for most cities (France/Spain).  This seems wrong to
>>> me, even if they are given "post" language identifiers.  Postal  
>>> codes,
>>> if they are integrated, should be a separate feature class since  
>>> there
>>> is rarely a 1:1 postal:city pair

--

*Bernard Vatant
*Senior Consultant
Vocabulary & Data Engineering
Tel:       [phone]
Mail:     bernard.vatant@... <mailto:bernard.vatant@...>
----------------------------------------------------
*Mondeca**
*3, cité Nollez 75018 Paris France
Web:    www.mondeca.com <http://www.mondeca.com>
Blog:    Leçons de Choses <http://mondeca.wordpress.com/>
----------------------------------------------------**


--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "GeoNames" group.
To post to this group, send email to geonames@...
To unsubscribe from this group, send email to geonames+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/geonames?hl=en
-~----------~----~----~----~------~----~------~--~---

[Marc Wick]

Hi Bernard

Do you have concrete examples where it does not work? For Paris as far
as I know each postal codes also has a proper name and if it is only the
number of the Arrondissement. (Ex 75013= 13e Arrondissement,13eme
Arr,13ème Arr,Arrondissement des Gobelins,Gobelins,Paris 13e,XIIIe)

In the third case where several villages share the same postal code the
code is just added to each of the villages sharing the code.

Bernard Vatant wrote:

> Hi all
>
> Completely tuned with Ryan on this, I have to cope those days with data
> where places (in france) are "identified" by Name + Postal code, that
> have to be attached to cities identified by insee codes or geonames id
> (we have the mapping). What happens in France is roughly as following
>
> - Big cities have several postal codes.
> - Some small-to-middle cities can be uniquely identified by a single
> postal code.
> - Many small villages share a postal code with their neighbours and a
> local city.
>
> Only in the second case the use of postal code as alternate name can
> make sense, but I agree it should be stored in a specific field.
>
> Bernard
>> Yes and postal codes are very valuable, so I understand the desire.  
>> I'd love to see them integrated in the data.  However, I have a  
>> semantic and practical problem with doing it via alternate names:
>>
>> Semantic - postal codes and cities, at least in most countries that I  
>> know of, don't have a 1:1 mapping with cities and generally have  
>> fairly different boundaries.  To say that a postal code is an  
>> alternate name of a city is just wrong.  The city is not defined by  
>> it's zip code nor is it used in common language as an identifier for  
>> the city.
>>
>> Practical - I have an address parser that uses geonames data.  I have  
>> to filter out these postal alternate names or the parser gets very  
>> confused at why there are two cities, possibly conflicting, in the  
>> same address.
>>
>> Is the eventual plan to add them as separate feature classes?  What  
>> information are you lacking?
>>
>> On Jun 22, 2009, at 12:47 AM, Marc Wick wrote:
>>
>>  
>>> Most users are actually directly asking for a postalcode-city  
>>> relation.
>>> At the moment we don't have enough information about postal codes to  
>>> add
>>> them as separate feature class.
>>>
>>> Marc
>>>
>>> Ryan wrote:
>>>    
>>>> Is this ok?  Several of the European countries have postal codes as
>>>> alternate names for most cities (France/Spain).  This seems wrong to
>>>> me, even if they are given "post" language identifiers.  Postal  
>>>> codes,
>>>> if they are integrated, should be a separate feature class since  
>>>> there
>>>> is rarely a 1:1 postal:city pair
>
>

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "GeoNames" group.
To post to this group, send email to geonames@...
To unsubscribe from this group, send email to geonames+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/geonames?hl=en
-~----------~----~----~----~------~----~------~--~---

[Marc Wick]

There is a separate dump with postal codes only:
http://download.geonames.org/export/zip/

Marc

Ryan Levering wrote:

> Yes and postal codes are very valuable, so I understand the desire.  
> I'd love to see them integrated in the data.  However, I have a  
> semantic and practical problem with doing it via alternate names:
>
> Semantic - postal codes and cities, at least in most countries that I  
> know of, don't have a 1:1 mapping with cities and generally have  
> fairly different boundaries.  To say that a postal code is an  
> alternate name of a city is just wrong.  The city is not defined by  
> it's zip code nor is it used in common language as an identifier for  
> the city.
>
> Practical - I have an address parser that uses geonames data.  I have  
> to filter out these postal alternate names or the parser gets very  
> confused at why there are two cities, possibly conflicting, in the  
> same address.
>
> Is the eventual plan to add them as separate feature classes?  What  
> information are you lacking?
>
> On Jun 22, 2009, at 12:47 AM, Marc Wick wrote:
>
>> Most users are actually directly asking for a postalcode-city  
>> relation.
>> At the moment we don't have enough information about postal codes to  
>> add
>> them as separate feature class.
>>
>> Marc
>>
>> Ryan wrote:
>>> Is this ok?  Several of the European countries have postal codes as
>>> alternate names for most cities (France/Spain).  This seems wrong to
>>> me, even if they are given "post" language identifiers.  Postal  
>>> codes,
>>> if they are integrated, should be a separate feature class since  
>>> there
>>> is rarely a 1:1 postal:city pairing.
>>>
>>>
>>
>
>
> >
>

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "GeoNames" group.
To post to this group, send email to geonames@...
To unsubscribe from this group, send email to geonames+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/geonames?hl=en
-~----------~----~----~----~------~----~------~--~---