[talk-au] AllThePlaces data in an Australian context?

[Daniel O'Connor]

For maintaining POI, specifically businesses I see a lot of the value in
the following use cases:

1) Phone, hours, website, maybe street address for existing nodes

1b) One of those details has changed, and the prior source was ATP, so one
click update.

1c) Certain points make claims about extra attributes they have

2) Is there an X here, please survey or prove via street level imagery

3) Is there an X still here, flagged as explicitly closed

There's a very large amount of activity in the project and weekly data
aggregation, so it's likely this is not a "one off", but ongoing feed of
information.

On Mon, Jul 29, 2024, 11:09 AM Graeme Fitzpatrick <graemefitz1 at gmail.com>
wrote:

> I've also wondered about using this data?
>
> Just had a look at it though, clicked on a couple at random, & wondering
> if it'll be worth it?
>
> e.g. 99 Bikes:3 missing, 5 mismatch, 4 tags to add but the missing &
> mismatch results seem to be that it's searching for <99_bikes_au>, but
> they're already named as <99 Bikes>. It does also show opening hours &
> phone numbers for some of them so they could be added.
>
>
>
> Thanks
>
> Graeme
>
>
> On Sun, 28 Jul 2024 at 18:19, Daniel O'Connor <daniel.oconnor at gmail.com>
> wrote:
>
>> Hi all,
>> Wanted to open up a discussion thread about https://www.alltheplaces.xyz/
>> and what we're comfortable with ingesting into OSM from it, if anything.
>>
>> *What is it?*
>> https://www.alltheplaces.xyz/ is a set of scrapers that crawl various
>> business' sites; and extracts data of interest to OSM (address, phone, etc).
>>
>> Often, the businesses are publishing structured data under no particular
>> licence (strong indicator they want it consumed), sometimes just presenting
>> it in HTML etc.
>>
>>
>> *Can we actually use it?*
>> https://osmfoundation.org/wiki/Licensing_Working_Group/Minutes/2023-08-14#Ticket#2023081110000064_%E2%80%94_First_party_websites_as_sources
>>
>> <https://osmfoundation.org/wiki/Licensing_Working_Group/Minutes/2023-08-14#Ticket%232023081110000064_%E2%80%94_First_party_websites_as_sources>
>>
>> Probably?
>> Facts themselves aren't protected by copyright, there needs to be an
>> expression of originality, etc.
>> https://alacc.org.au/requirements-to-gain-copyright-protection/
>> https://eprints.qut.edu.au/50425/
>>
>> https://www.mondaq.com/australia/copyright/94168/no-copyright-protection-for-telephone-directories-telstra-v-phone-directories-company
>> Telstra Corp Ltd v Phone Directories Co Pty Ltd (2010) 264 ALR 617 [340].
>>
>> (If this has already been settled, specifically in an Australian context,
>> great! I didn't see anything from a cursory search though)
>>
>> *If we decide to use it; how do we use it?*
>> Obviously, wholesale import of data with no human review is more likely
>> to cause a mess than not.
>>
>> However, the comparison tools between OSM and ATP are being built out by
>> various people; ie
>>
>> https://matkoniecz.codeberg.page/improving_openstreetmap_using_alltheplaces_dataset/#Australia
>> It's likely we'll end up with a way to do maproulette style tasks - one
>> click add phone numbers or websites; maybe something a bit more involved
>> for viewing if there is street view imagery available we can validate a
>> particular store location or similar that is generally user friendly.
>>
>> *How much data is available?*
>> Rough analysis for Australia; ~63782 facts, ranging from "missing place"
>> to "here's a phone number"
>>
>> _______________________________________________
>> Talk-au mailing list
>> Talk-au at openstreetmap.org
>> https://lists.openstreetmap.org/listinfo/talk-au
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-au/attachments/20240729/dcd17fdd/attachment.htm>