Identification of an individual flat in a building using Address Validate

Nigel Light

Hi,

I've checked the following address (which is valid with an appropriate postcode) using Address Validate

Addr 1 :16 Wythburn Court

Addr2 : 34 Seymour Place

Addr3 : London

Postcode : W1H 7NS

However, because the "Flat" is missing from the address, the 16 at the front is ignored and I get a "3-tentative full match" returned with the address amended to

Wythburn Court 34 Seymour Place London W1H 7NS

ie missing the precision of the flat number and therefore a deliverable address has been amended to an undeliverable one. I have tried with both standard and detailed components, individual address elements and a concatenation.

Checking the Royal Mail website gives me this as a format

Is there any way that I can get the flat number recognised?

Thanks

Nige

Pavan Rao

Thanks @Nigel Light . I am going to work with the team to get this working in V1 too. I will keep you posted on the progress!

Clinton Jones

hi @Nigel Light I believe that @Valentin Tchergarsky may be looking into this.

As per normal, we would usually handle this via a support ticket which you can open, per the below link.

However I thought it would be an interesting exercise to test out this particular address in a couple of different ways

In this instance I am using a single block address in a number of formats - this can be easily done with a concatenate

This particular address comes back from the batch integration engine thus:

This wasn't very satisfactory as you rightly point out - even with flat numbers, so I tested it against a couple of other approaches too!

For that address the Royal Mail correctly provide the requirement of selecting a Flat number.

I tested this also with a couple of competitors and they also didnt quite say it was a perfect address,

The EDQ real time api similarly didnt prompt for Flat Number.

I also couldn't get very granular with What3words on the actual address beyond the generalized postcode

Below is the link to the support contact info, it could be that there has been an update to the data to provide additional data.

https://www.edq.com/documentation/contact-support/uk/#:~:text=Customer%20Support%20Hub-,Call%20us,England%20and%20Wales%20public%20holidays).

Nigel Light

Hi @Clinton Jones

Thanks for getting back to me (and support ticket has been raised as suggested).

Further information; I am informed by @Valentin Tchergarsky that we are missing a Multi-Residential Building file GBRMRF which should correct our issue (I was unaware of this, as others may be too). @Pavan Rao informs me that this isn't straight forward to apply to Aperture 2 and some reconfiguring is required. We are currently working on this - any tips that you can share here?

Crossing fingers (and everything else) that this will meet our needs

Regards

Nige

Pavan Rao

Hi @Nigel Light , I have managed to get this working in Aperture V2. There were some configuration changes and updates to json file to get this working as Additional dataset in Aperture Data Studio.

Nigel Light

Great work @Pavan Rao - it sounds fiddly though (and hopefully this can also be applied to V1 too until we are in a position to migrate successfully too)

Jenna Leiper

Hi @Pavan Rao We are currently having a similar issue to this, can you advise how you got this to work as an additional dataset? We are also hoping to match UPRN but can only see Parent UPRN, is it possible to match UPRN? Thanks

Sueann See

@Jenna Leiperplease log a case with support so we can track this. For the multi-residence dataset, it may be a similar issue as experienced by other clients that we can provide a workaround for, but we should confirm if that is the case.

For the UPRN, there may be an issue we need to investigate.

In any case, We will need you to supply the following data, so it is best to contact support who can work through this with you.

• Address examples
• Aperture Data Studio version
• qawserve/qaworld (address validation configuration files) from the problematic instance
• version of dataset /data vintage used.

Pavan Rao

Hello @Jenna Leiper , As Sue Ann mentioned, we can log a case to investigate this further, one is related to GBR Additional dataset (GBRMRF) which returns multi-residence and the other is related to APR (Address Base Premium) dataset which returns UPRN.

But here is what I did to update Multi-residence (sub-building info) in Aperture Data Studio:

Following are the key changes that are required to configure Multi-residency (Sub-building information) in Aperture:

• Firstly, we will need a license key for Multi-residence datasets to be updated. This would have been provided by Experian as part of the purchase of this additional dataset.
• To update the license key in Aperture Data Studio - Select the user icon on top right corner and select “Update License” :

• In addition to this, Multi-residence related datasets licensed (gbrmr*)are required to be added to the location specified in qawserv.ini file (where the standard 5 datasets (top 5 in the screengrab below) from the same vintage are also included):

• We also need to configure qawserve.ini file (located normally in C:\ProgramData\Experian\addressValidate\runtime) to include GBRMRF, something like this: DataMappings=GBR,Great Britain,GBR+GBRMRF
•  And finally, an addition of GBRMRF in GBR.json file (attached) found in C:\ProgramData\Experian\addressValidate\dataPlusMappings, which will bring this Additional dataset to life.

GBR.json

Hope this helps.

Thanks,

Pavan

Nigel Light

Nb Probably fair to say, this isn't straightforward and might involve 2 passes of the address dataset - with and without the multi-residency buildings file for the best results.

Nigel @Nigel Light @Nigel Light

PS It works with Aperture 1.x too (thanks @Pavan Rao )

Sueann See

@Jenna Leiper i'm hoping this will help you to surface the UPRN attribute you are looking for in the APR (AddressBase Premium dataset). You will need to define a custom layout.

The documentation about custom layouts is available here.

A sample custom layout specific to the APR dataset is as follows:

[CustomAPR]

CountryBase=APR

 

CleaningAction=Address

APRAddressLineCount=8

APRAddressLine1=W60

APRAddressLine2=W60

APRAddressLine3=W60

APRAddressLine4=W60,L21

APRAddressLine5=W0

APRAddressLine6=W60,C11

APRAddressLine7=W60,X11

APRAddressLine8=W60,A11

APRCapitaliseItem=X11

APRExcludeItem=L21 C11 X11

APRSeparateElements=Yes

APRElementSeparator={ } C11{ ^-}

Jenna Leiper

Hi @Sueann See

We have added this custom layout, we are now pulling out UPRN, but have noticed some of the UPRN's are actually the Building UPRN rather than the flat UPRN, they are all in the same column, is there a way to separate these out?

Thanks

Jenna

Sueann See

@Jenna Leiper it sounds like we will need to investigate the data itself and will need you to provide a sample. Are you able to Contact Support to log a case?

@Valentin Tchergarsky anything you want to add here?

Jenna Leiper

Thanks @Sueann See I have raised one :-)