1 (edited by Darwyn Sumner 04-02-2014 11:34:44)

Topic: Using Gateway to refuse verification offers from National Schemes

That seems to be what most of you are doing.
Today I received all the responses made by LRCs through the NBN Gateway response system to a 2 person NSS which had made requests for improved access to your Diptera datasets.
There are worldwide only around 10 experts able to verify records from this specialist group (Chloropidae) and many LRCs have just turned down the offer from 2 of them to verify the records you've posted onto the NBN Gateway.
If the Gateway was being set up afresh today then it would be totally open access and you either uploaded it for all to see (restrictions only based on species status and sensitivity) or you kept it to yourselves (that was what Stuart Ball said to me; makes sense).
It was LRCs who asked for these controls, it's the price of conserving business models.
(I know, I know, an hour spent chatting to a JNCC friend and I'm all anti-LRCish and then an hour spent chatting to LRC friends and I'm all anti-JNCCish - my compromise is to give everyone earache)

So when people talk at NBN conferences about NSSs complaining about junk data uploaded by LRCs then they do rather have a point if LRCs are refusing to allow experts to make attempts to fix this. Paula Lightfoot just gave a Gateway presentation to the Dipterists Forum schemes (there are 18 of them so the other 17 will follow eventually) telling them all how the Gateway controls work - so they're all well informed about how to set about it now, BRC set up this meeting and they've already done the Riverflies and Coleoptera and presumably will be meeting up with the same sort of approach to all the other NSSs - so expect this trickle to turn into a flood. All the NSSs should be listed on the CEH website.

Small NSSs struggle, 60 LRCs cannot expect such NSSs to approach them all directly with requests for sufficient detail to be able to verify records on the NBN Gateway, as the Chloropidae schemes organisers said "we can only verify data if it is in bulk from the NBNG and gets centrally distributed by others, which is what we tried out - and it looks as if we failed".
If the fact that they are also consultants throws you, most of them have to earn a crust as well as spending a vast amount of voluntary time on training others, taxonomy, managing records, verification and so on.
Tom Hunt and I are in possession of all 23 responses that this NSS have received and are looking at a fix.

Darwyn Sumner

Re: Using Gateway to refuse verification offers from National Schemes

Hi Darwyn

Having been one of the LRCs that 'refused' I think it worth responding. Our repsonse to the request was that it was better for us to send the data direct as we don't at the moment have all diptera records on the Gateway so they would get more data via a direct enquiry with us. The text of the request that we received did not specify that they needed or preferred to do this via the gateway. We thought we were providing a better and more complete service by offering the solution that we did. The data request also specifically mentioned an RSPB dataset which again we don't necessarily have permission to make fully accessible via the Gateway,but can offer access directly from NBIS. We do have to consider our data providers wishes.

The text of the request was this:

We organise the Chloropidae (1 family of True Flies) Study Group and when we downloaded the Chloropidae data set we noted that the RSPB dataset did not have recorder / determiner included. These are most likely specimens identified by us, but without seeing the recorder / determiner we cannot verify the records. We ask for access to all or as many terrestrial invertebrate groups as possible since it is often useful to be able to compare our data with that of other groups, e.g. how well surveyed was an area / site? However, there is no filter for this, so we used Diptera - the data set we are most interested in.

Our response:

Barbara, Due to the type of request, this would be better placed to come through to us as a standard data request directly, rather than through the NBN. As we may have more records and further information not on the NBN. Please visit this page on our website, as it gives details on how you go about this

I don't think that was an unreasonable response, and is actually better than just granting the access instantly. We are happy to revisit this and grant the permission, however it will be on an incomplete dataset.

Regards
Martin

Martin Horlock
Chair of ALERC
Senior Biodiversity Officer
Norfolk County Council

Re: Using Gateway to refuse verification offers from National Schemes

Generalisations don't really help here. The impersonal self-service nature of the NBNG doesn't help either. It would be useful to look at the reasons each LRC gave for 'refusal' and address the problem on a case by case basis. I feel for anyone putting in a request for data only to be confronted with having to approach multiple LRCs directly. I've just filled in my tax return and so feel particularly empathetic towards fellow online form fillers. That said, the reasons Martin gave for 'refusal' are valid, reasonable and conscientious. I don't know what the solution is, but we can and must do better at working together to build trust and build better, more efficient systems predicated on that trust.

I'm certainly interested in learning more about what NSSs need and in what form so that we are able to provide the best possible service we can. I think the will is there - we just need to get better at communication. This is partly the role of ALERC, no? Would it not be better for NSSs to kick off  dialogue through Tom and/or the directors? First thing to tackle is expectation-management  that mass requests for LRC data through the NBN website will often meet with 'refusal' for many and good reasons. That being the case, how can ALERC assist with making a mass requests manageable for both parties? I'm sure good old-fashioned liason between ALERC and NSSs is what's needed here.

Re: Using Gateway to refuse verification offers from National Schemes

ALERC will take it up with the NSSs in question.  The fact that NSSs are not aware that they can contact ALERC as a point of contact is the first point to address.  After that, it will be a case of making sure everyone is aware that requests are likely to come through and what they will contain.  I'm confident all issues can be resolved by appropriate communication.

One of the side issues that I have spotted is that it took Darwyn to raise this issue with the directors before it was discussed.  If LRCs receive unusual or difficulty data access requests, it would be good if they could mention them on the forum somewhere, so we can try and sort them out immediately and avoid any bad feeling.  I know this is happening with some other requests.

Thanks,

Tom Hunt - ALERC National Coordinator

Re: Using Gateway to refuse verification offers from National Schemes

Thanks Tom.

6 (edited by Darwyn Sumner 05-02-2014 09:05:45)

Re: Using Gateway to refuse verification offers from National Schemes

I think it useful here to give a little background to the story.
At the NBN Conference back in November, Stuart Ball was elected an honorary member. A brief comment made by him at the time concerning "rubbish data from LRCs" made some LRCs cross but caused me to look at ways to address the problem.
One of the methods was already in progress, a joint LRC/Dipterists Forum initiative, started in this forum, to hunt down County Recorders. This has been a success, resulting in a pdf with a map on one side and a detailed listing of all the Diptera Recording Schemes on the other (distributed to ALERC Directors a little while ago) - I'll  put all the details on that thread.
The other method was related to a BRC initiative. Supported by Martin Harvey and Chris Raper (again DF members), Helen Roy put together a meeting of all DF Recording Schemes (it attracted about 8 of the 18, the others will have to await our write-up in the DF Bulletin) it was mainly about BRC's online initiatives. I thought the NBN Gateway perspective would be useful to have at such a rare gathering of so many DF Recording Schemes so I asked Paula Lightfoot if she could turn up and tell them about NBN Gateway (Helen said - oh yes, I was going to ask Paula). So we caught Paula on one of her last few days at NBN and she gave a very stimulating presentation, good enough to fire the enthusiasm of a few Recording Scheme organisers who attended and certainly enough to stimulate them into attempts to use what she'd told us to begin work on turning the tide of the alleged "rubbish data"
The Chloropidae scheme has been the first one to try, after experiencing the difficulties detailed above, they asked me as an ALERC Director to help out.

Darwyn Sumner

7

Re: Using Gateway to refuse verification offers from National Schemes

HI - I think we were the other 'refusal' here in Cambs

I responded that I was confused over the mention of an RSPB dataset.... we don't have ANY rspb data anywhere, let alone on the gateway.. (hence I assumed it was yet another example of poor geographic filtering which is currently rife on gateway5)
Additionally, I stated that we had not put any recorder names on the records on the gateway, and that it was far better to contact us directly to resolve any verification issues.
This is pretty similar to the NBIS response,
for the same sort of reasons, and would be happy to deal with the request directly, although I'm fairly sure all of our fly data comes from superior fly-folk such as Perry, Cole, Stubbs, etc......


mhorlock wrote:

Hi Darwyn

Having been one of the LRCs that 'refused' I think it worth responding. Our repsonse to the request was that it was better for us to send the data direct as we don't at the moment have all diptera records on the Gateway so they would get more data via a direct enquiry with us. The text of the request that we received did not specify that they needed or preferred to do this via the gateway. We thought we were providing a better and more complete service by offering the solution that we did. The data request also specifically mentioned an RSPB dataset which again we don't necessarily have permission to make fully accessible via the Gateway,but can offer access directly from NBIS. We do have to consider our data providers wishes.

The text of the request was this:

We organise the Chloropidae (1 family of True Flies) Study Group and when we downloaded the Chloropidae data set we noted that the RSPB dataset did not have recorder / determiner included. These are most likely specimens identified by us, but without seeing the recorder / determiner we cannot verify the records. We ask for access to all or as many terrestrial invertebrate groups as possible since it is often useful to be able to compare our data with that of other groups, e.g. how well surveyed was an area / site? However, there is no filter for this, so we used Diptera - the data set we are most interested in.

Our response:

Barbara, Due to the type of request, this would be better placed to come through to us as a standard data request directly, rather than through the NBN. As we may have more records and further information not on the NBN. Please visit this page on our website, as it gives details on how you go about this

I don't think that was an unreasonable response, and is actually better than just granting the access instantly. We are happy to revisit this and grant the permission, however it will be on an incomplete dataset.

Regards
Martin

Re: Using Gateway to refuse verification offers from National Schemes

Cofnod have not been asked for improved access in this instance (we only have a small proportion of our data holdings on the Gateway and it may not contain any relevant species). But if we had been asked our response would have been the same as everyone else who

Aisling May
Data Manager - Cofnod   www.cofnod.org.uk

Re: Using Gateway to refuse verification offers from National Schemes

I could do with seeing a transcript of what LRCs were originally approached with f that's possible.  Discussions here, and with the directors, has progressed to data exchange, but I am not certain that was what was originally asked for in this case (although I realise that lack of certainty was actually a major part of the problem in the first place).

Thanks,

Tom Hunt - ALERC National Coordinator

10

Re: Using Gateway to refuse verification offers from National Schemes

I've seen it now elsewhere and can understand why it received a negative response.

Regards,

Tom Hunt - ALERC National Coordinator

Re: Using Gateway to refuse verification offers from National Schemes

BRERC also have all our data on the Gateway but still denied the request, particularly because the request stated personal use (BRERC charge discounted rates for personal use).

I just wanted to say that for BRERC it was not considered a difficult request or one we should have tip-toed around. I am happy with our response and dont feel the need to address it with ALERC.  If there was such a request where we thought there would be a reasonable expectation of controversy then I have no doubt LRC managers would raise it with ALERC at the time.

I appreciate Tom's request and offer to assist but sometimes people will submit poor requests.

ALERC has already set up a data flow group that are looking at many of these issues.

Tim Corner

Re: Using Gateway to refuse verification offers from National Schemes

Perhaps not a good choice, someone whose native language isn't English, but that's just chance and I can think of several others who would express the request less effectively. It was also made pretty clear that this was an NSS making a request, making them instantly recognisable as the "gold" that Steve Wilkinson tells us is what BRC and NBNG are seeking out to un"rubbish" the Gateway and iRecord data; "rubbish" being a criticism levelled at LRCs at the NBN conference (which makes my initiative here more of a support for LRCs than NSSs and not vice versa - it's not at all helpful for LRCs to be telling me I have a bias in either direction)

Those choice boxes on the NBN Gateway aren't adequate in any case, I've tried them and there are several that cannot be used:
1. Educational purposes - that excludes use in peer reviewed papers, you can't put yourself forward as an expert or verify without peer review
2. Research and scientific analysis - well that one just gives a big red notice saying you have to contact all data suppliers individually
3. Media and most of the remainder - ditto
4. Conservation/NGO work - not applicable
The only logical choice for an NSS wanting to verify records is "Personal interest" because all the others are inappropriate. This isn't what the requester said, it's the corner that NBNG forced them into via the restricted set of choices on the dropdown box on the NBN Gateway.
A simple solution would be an additional one that said "Accredited NSS verifier" which could only be used by people on an accredited list - you'd still have the problem of how best to obtain the single datasets rather than one from each LRC but it would be a move in the right direction.

Darwyn Sumner

Re: Using Gateway to refuse verification offers from National Schemes

Just playing devil's advocate for a moment, but has anyone tried putting in a request to all LRCs for a national dataset and determined how much work is involved? It's very easy for us to say "they should just come to us directly" but when we're talking about approaching 60 or 70 LRCs directly in order to obtain a complete national dataset - and then repeating that process on an annual basis - that to me seems wholly unreasonable expectation. The NBNG is supposed to make the aggregation and supply of data on this scale feasible. I don't believe the mantra of "come to us directly" is a good long-term, scalable solution for the end user. I'd be interested in hearing what ideas people have here for reconciling the tension between direct-supply vs. NBNG-supply.

My ideas would be:

* improved access request controls, with faster iteration and updates on those controls in order to fix issues quickly
* More granular access controls; e.g., the ability to grant access to a subset of data within a dataset.

We have uploaded our entire dataset, warts and all, onto the NBNG so we cannot say the data is incomplete. However, we cannot grant access to subsets of that single dataset. We'd have to break it down into groups (which we should do anyway, but that requires a huge admin burden). So generally speaking, although we'd be more than happy to grant full access to data to NSSs, quibble-free, we simply cannot in any practical way and so have to resort to this inadequate, inefficient fallback of direct supply which does not suit us or the end user. So in our case, the constraint is purely a technical limitation of the NBNG.

Re: Using Gateway to refuse verification offers from National Schemes

That issue of breaking down into groups reminds me of a letter in British Wildlife (Vol 24, #3, February 2013) in which Gordon Corbet (responding to a previous article by Roger Morris) says he finds it "cumbersome to submit records to all the numerous national schemes (17 for Diptera alone), often wanting different formats". So he submits all his records to Local Records Centres only. (Solutions via BRC's iRecord initiative perhaps)
Eric Fletcher at rECOrd has a Diptera dataset that he's not uploaded to the Gateway yet and is working with a volunteer (Phil Brighton) to sort them all out - there might be some useful ideas and methodologies emerging from that project, Phil's ex nuclear inspectorate so he's meticulous. Eric's also working with Martin Harvey - an iRecord advocate and NSS organiser.
Sorry to keep harping on about Diptera but it's all I know in detail re NSSs - I hope the issues they have are transferrable to other taxa groups like BWARS, Coleoptera, Riverflies, Fleas.

Darwyn Sumner

Re: Using Gateway to refuse verification offers from National Schemes

Yes, it's definitely cumbersome breaking down our core database into separate taxonomic groups ready for upload to the NBNG. Preparing metadata especially. And then how far do we go? Do we upload a general Diptera dataset, or a dataset for each individual Diptera subgroup? I'd like to see the NBNG allow for ad-hoc selection of data from within a dataset and allow for specification of metadata and a license/agreement based on that subset.

Re: Using Gateway to refuse verification offers from National Schemes

Regarding the cumbersome work flow of national record verification via the NBNG or via individual LRCs: one of the game changing strengths of iRecord, as I see it, is the web-based record verification workflow that has been built into it.

Could iRecord/Indicia be expanded to permit large datasets to be securely uploaded by an LRC or NSSs from their Recorder6 databases, and then the experts use the iRecord verification systems.  The verified records can then be downloaded again, ideally retaining their unique ID numbers, to enable seemless upload back into Recorder6 and eventually onto the NBNG.

Mike Beard

Mike Beard
Natural Course Project Officer
Greater Manchester Local Records Centre

Re: Using Gateway to refuse verification offers from National Schemes

The idea that it is easier to 'request all data from one place and not go to many' is obviously one that all LRCs would recognise and acknowledge.

However the requests have some other bits attached that are never really discussed.  Another way of saying it is 'request all data from one place and not go to many even though I am not willing or have no money to contribute to the costs of enabling that'.

A bit like complaining we have to shop in more than one shop in the high street whilst not offering to contribute to the cost of building a mega store. and then expecting to not actually pay for goods when someone does open such a store.

We need to get away from the idea that it is LRCs who are blocking access, puposely preventing improvements and not making any data avilable to others.  That is patently
untrue.

it is a lack of sufficient investment, funding/income and resources that is the blockage.

Tim

Re: Using Gateway to refuse verification offers from National Schemes

Mike - the answer to your question is yes, datasets could be uploaded with access constrained to verifiers only, and the verification system could be exploited and data could be downloaded again. Uploading large amounts of data would be rather slow, so that would be the biggest challenge in regard to that idea.