1

Topic: Species Search still very confusing

Can someone check where we are on making the Species Search work in a more intuitive way? The use of 'fuzzy logic' matching is completely unnecessary for the majority of searches and should only be used if what the user searched for doesn't return anything. Also, the way that the list is being sorted is still bizarre.

Take "Dreissena bugensis" as an example - it returns 13 results but none are an identical match. But because the sort order is so unreliable I have to check both pages to make sure that the species isn't being listed on the second page. *  The issue of why Quagga Mussel doesn't return anything is a separate issue, which I have emailed you about.

[* BUG: If you go to page-2 and then click "Species" to refine the results it doesn't reset the page count ("offset=10") so you get a blank page. Any refinement needs the offset resetting to 0.]

Searching for "Sparrow" still lists "House Sparrow" on the second page, third from last. How is the search ordering the results? A search like "sparrow" shouldn't return the commonest UK species on page-2 and 8 rare vagrants on page-1.

Searching for "tachina magnicornis" returns the junior synonym of Phryxe magnicornis before the accepted name Tachina magnicornis. Surely taxa that match identically should be listed above junior synonyms?

Ideally, I would want the species search to search for what you type "LIKE %search-text%" and if that returns nothing then run a fuzzy logic search. With any returned taxa we could list then in one of 2 ways:

1. Taxonomically: So matches where the name is the accepted name listed above junior synonyms; well-formed names above ill-formed names etc.

2. Rarity: Matching common species listed above matching rarities

I'm very happy to help with this if necessary because I think that having a simple and intuitive Species Search is vital to the usability and credibility of the NBN Atlas.

Chris Raper, Manager of the UK Species Inventory, Angela Marmont Centre for UK Biodiversity,
Natural History Museum, Cromwell Road, London, SW7 5BD.  (tel: 020 7942 5894)
also Tachinid Recording Scheme (http://tachinidae.org.uk/)

2

Re: Species Search still very confusing

Agreed, this needs sorting out.

Relatedly, on the search boxes at nbnatlas.org when you start typing it comes up with a list of suggestions quite quickly. e.g. type vulpes. I can click on vulpes vulpes and it fills it it, but then I have to click on the search icon to bring up the full search results (and then on the appropriate result in the list).

Although it would be satisfying to go straight to a species page from here, I can see that could cause issues where different taxa have the same name (whether common or scientific) because you don't have the full info, so not a good idea. But it would be more user friendly if selecting from the autofill list automatically took you to the search page, so you only have to click twice and not three times.

The autofill options don't always work as expected, think it is a bit buggy. E.g. typing "reds" three redshanks appear in the list, but typing the next letter they disappear for some reason ("redsh"). (And of course the same problems Chris outlines about ordering for things like sparrow are repeated here.)

-----------------
Teresa Frost | Wetland Bird Survey National Organiser | BTO
Other hat  | National Forum for Biological Recording Council
(Old hats  | NBN Board, ALERC Board, CBDC, KMBRC)

3

Re: Species Search still very confusing

We're also still getting locations appearing in the species search when vernacular names are used e.g. Pyramidal Orchid produced 423 results including 143 localities. This is nonsense, I thought that we had removed place names from the species search.

Problems with the search were identified as soon as the Scotland Atlas went on-line( 12 months ago) and have been highlighted as a priority issue since the Atlas became operational in April.

We don't seem to be making very much progress on this or related issues - perhaps it's time we had a progress report.

4

Re: Species Search still very confusing

chrisjohnson wrote:

We're also still getting locations appearing in the species search when vernacular names are used e.g. Pyramidal Orchid produced 423 results including 143 localities.

That happens when you use the "Search the NBN Atlas" box on the homepage. If you click on "Species" in the menu and then enter the name on that page you should get better results but I'd argue that the NBN Atlas is mainly searched by species name and so the default should be just species. By all means have a "Search the whole NBN Atlas" option but by default the search boxes should assume the user is looking for a species.

Better still, search by species and if no results are returned then do a broader, fuzzy-logic, search.

Chris Raper, Manager of the UK Species Inventory, Angela Marmont Centre for UK Biodiversity,
Natural History Museum, Cromwell Road, London, SW7 5BD.  (tel: 020 7942 5894)
also Tachinid Recording Scheme (http://tachinidae.org.uk/)

5 (edited by chrisjohnson 05-07-2017 07:56:01)

Re: Species Search still very confusing

Hi Chris,
The result quoted was from the "species" search on Atlas Scotland and not the general search. If you use the species name rather than the vernacular the locations do not appear. However, many recorders, search on vernacular rather than the species name in the species box.

Christine

PS Apologies I tried to replicate the search without success so I must have inadvertently used the wrong box.

6

Re: Species Search still very confusing

Now I'm even more confused!.
This time I made sure I used the species search box. For "ferret"  the search returned, four location names before two species names and then four vernacular names.

7

Re: Species Search still very confusing

It just seems crazy ... the search facility has been designed in a very powerful way but the implementation in the NBN Atlas seems to fly in the face of user requirements. How many users were asked what they wanted from the search functionality and how many said that they wanted the default to return a mixed search across all types of object in the database? Very few, I'd guess. I'd go so far as to suggest that, if the default search options are not taxon based, then this is making the Atlas look out of touch with its core audience. Fixing it should be a matter of urgency.

Chris Raper, Manager of the UK Species Inventory, Angela Marmont Centre for UK Biodiversity,
Natural History Museum, Cromwell Road, London, SW7 5BD.  (tel: 020 7942 5894)
also Tachinid Recording Scheme (http://tachinidae.org.uk/)

8

Re: Species Search still very confusing

Hi Chris,

Improving the results produced by the search functions is on our list of to-do items. We currently have limited access to developer time and, therefore, have to prioritise fixing bugs and any issues that are having a major impact on the usability of the NBN Atlas. While the search function may not be optimal, it does not appear to be having a significant impact on use of the NBN Atlas.

With regard to the general search function, it has always been the intention for this to produce more than just species returns. The NBN Atlas is designed to appeal to a wide audience and make use of the very powerful search engine within the site. While your focus is understandably on species and taxonomy, not all users will have the same requirements and as you state, using the search box under the species tab will produce species specific results.

chrisjohnson wrote:

Hi Chris,
The result quoted was from the "species" search on Atlas Scotland and not the general search. If you use the species name rather than the vernacular the locations do not appear. However, many recorders, search on vernacular rather than the species name in the species box.

Christine

PS Apologies I tried to replicate the search without success so I must have inadvertently used the wrong box.

Christine, there does indeed appear to be a bug in the species search on the NBN Atlas Scotland, which means it is returning all results not just species specific ones. I have raised this with the developer and this will be addressed.

I am sorry I can't give you a timescale yet for enhancement of the search results, but rest assured it has not been forgotten.

Jo