Re: Automating Search Interfaces

Don Park (donpark@quake.net)
Fri, 20 Feb 1998 14:33:31 -0800


This is a multi-part message in MIME format.

------=_NextPart_000_008E_01BD3E0C.844A6680
Content-Type: text/plain;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

Pierre,

>I would like to talk about the location of the person making the search =
versus the location of the product or service provider. If I search for =
a product and I want it now, I only want a list of provider in a =
distance applicable for my request. And if I go to Europe this summer =
and want to make reservation or search for activities occuring at that =
time, the 'where I am' specification change. If I have a secondary house =
and make request on the week-end, I want the restaurant in that region =
and not the one near my primary house. An identity profile should be =
include in the query and give the chance to the search engine to make a =
better choice in regard of my age, sex, etc...

Interesting. Some of the issues with product location are:

1. How to indicate location?

Address or map coordinates? How does one find map coordinates? What =
happens when he moves?

2. How to associate location with products?

If a vendor has all inventory at a single location then the location can =
be #FIXED in his DTD. If inventory is distributed around the globe, =
each product or inventory group will have to be marked. The problem is =
that now it makes no sense to indicate physical location. It will have =
to be a store code which causes problem with search services since store =
codes will have to be converted into location format used by the search =
service.

As far as time constraints go, each product will probably be marked with =
time. The problem is that some time constraints are relative in nature.

*Ouch* I just thought of another painful problem with prices. What =
happens when a store wants to put on a sale? His database of products =
will have to map to different pricing schemes constrained by time, =
location, or association.

All this hurts my head a bit but it is very interesting indeed...

Regards,

Don Park
http://www.quake.net/~donpark/index.html

------=_NextPart_000_008E_01BD3E0C.844A6680
Content-Type: text/html;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD W3 HTML//EN">

Pierre,
 
>I would like to talk about the location of the = person=20 making the search versus the location of the product or service = provider. If I=20 search for a product and I want it now, I only want a list of provider = in a=20 distance applicable for my request. And if I go to Europe this summer = and want=20 to make reservation or search for activities occuring at that time, the = 'where I=20 am' specification change. If I have a secondary house and make request = on the=20 week-end, I want the restaurant in that region and not the one near my = primary=20 house. An identity profile should be include in the query and give the = chance to=20 the search engine to make a better choice in regard of my age, sex,=20 etc...
 
Interesting.  Some of the = issues with=20 product location are:
 
1. How to indicate location?
 
Address or map coordinates?  How does one find = map=20 coordinates?  What happens when he moves?
 
2. How to associate location with = products?
 
If a vendor has all inventory at a single location = then the=20 location can be #FIXED in his DTD.  If inventory is distributed = around the=20 globe, each product or inventory group will have to be marked.  The = problem=20 is that now it makes no sense to indicate physical location.  It = will have=20 to be a store code which causes problem with search services since store = codes=20 will have to be converted into location format used by the search=20 service.
 
As far as time constraints go, each = product will=20 probably be marked with time.  The problem is that some time = constraints=20 are relative in nature.
 
*Ouch*  I just thought of = another painful=20 problem with prices.  What happens when a store wants to put on a=20 sale?  His database of products will have to map to different = pricing=20 schemes constrained by time, location, or association.
 
All this hurts my head a bit but it = is very=20 interesting indeed...
 
Regards,
 
Don Park
http://www.quake.net/~d= onpark/index.html
------=_NextPart_000_008E_01BD3E0C.844A6680-- xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev@ic.ac.uk Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ To (un)subscribe, mailto:majordomo@ic.ac.uk the following message; (un)subscribe xml-dev To subscribe to the digests, mailto:majordomo@ic.ac.uk the following message; subscribe xml-dev-digest List coordinator, Henry Rzepa (mailto:rzepa@ic.ac.uk)