University of Glasgow

Roderic Page

::: {.separator style="clear: both;"}
[![](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhBaMrDj0KRPFPHmjSBYzffLbJZhA711GGekWPbiRqzSpSYLvBdef7OLwDrPaarFrxMHrBEQP5iiM_u3hF_AdNODs5whiO0i5cJmoQx3NWDcWUroU9OGF9o99BMcOIkIoCTETHH/s200/Rf7UoXTw_400x400.jpg){border="0"
width="128" original-height="240"
original-width="240"}](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhBaMrDj0KRPFPHmjSBYzffLbJZhA711GGekWPbiRqzSpSYLvBdef7OLwDrPaarFrxMHrBEQP5iiM_u3hF_AdNODs5whiO0i5cJmoQx3NWDcWUroU9OGF9o99BMcOIkIoCTETHH/s240/Rf7UoXTw_400x400.jpg){style="display: block; padding: 1em 0; text-align: center; clear: right; float: right;"}
:::

The [Plazi](http://plazi.org) project has become one of the major
contributors to GBIF with some 36,000 datasets yielding some 500,000
occurrences (see [Plazi\'s GBIF
page](https://www.gbif.org/publisher/7ce8aef0-9e92-11dc-8738-b8a03c50a862)
for details). These occurrences are extracted from taxonomic publication
using automated methods. New data is published almost daily (see [latest
treatments](https://tb.plazi.org/GgServer/static/newToday.html)). The
map below shows the geographic distribution of material citations
provided to GBIF by Plazi, which gives you a sense of the size of the
dataset.

::: {.separator style="clear: both;"}
[![](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhbzyO3NVCA0r2-2bkxUBCZdA9iuZZATf_mlqYDy6CCZjO4NTSEQMC8txdCLGGo_Ynz7ISTEGVLuPJDX3Rx8ZOh1vDZVzseKKa6OsvPsc2qoirL_tGwF4i0JJEWR5Uhyphenhyphen8b1R4Ey/s400/Screenshot+2021-10-24+at+20.43.23.png){border="0"
width="400" original-height="514"
original-width="1030"}](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhbzyO3NVCA0r2-2bkxUBCZdA9iuZZATf_mlqYDy6CCZjO4NTSEQMC8txdCLGGo_Ynz7ISTEGVLuPJDX3Rx8ZOh1vDZVzseKKa6OsvPsc2qoirL_tGwF4i0JJEWR5Uhyphenhyphen8b1R4Ey/s1030/Screenshot+2021-10-24+at+20.43.23.png){style="display: block; padding: 1em 0; text-align: center; "}
:::

By any metric Plazi represents a considerable achievement. But often
when I browse individual records on Plazi I find records that seem
clearly incorrect. Text mining the literature is a challenging problem,
but at the moment Plazi seems something of a \"black box\". PDFs go in,
the content is mined, and data comes up to be displayed on the Plazi web
site and uploaded to GBIF. Nowhere does there seem to be an evaluation
of how accurate this text mining actually is. Anecdotally it seems to
work well in some cases, but in others it produces what can only be
described as bogus records.

## Finding errors

A treatment in Plazi is a block of text (and sometimes illustrations)
that refers to a single taxon. Often that text will include a
description of the taxon, and list one or more specimens that have been
examined. These lists of specimens (\"material citations\") are one of
the key bits of information that Plaza extracts from a treatment as
these citations get fed into GBIF as occurrences.

To help explore treatments I\'ve constructed a simple web site that
takes the Plazi identifier for a treatment and displays that treatment
with the material citations highlighted. For example, for the Plazi
treatment
[03B5A943FFBB6F02FE27EC94FABEEAE7](https://tb.plazi.org/GgServer/html/03B5A943FFBB6F02FE27EC94FABEEAE7)
you can view the marked up version at
<https://plazi-tester.herokuapp.com/?uri=622F7788-F0A4-449D-814A-5B49CD20B228>.
Below is an example of a material citation with its component parts
tagged:

::: {.separator style="clear: both;"}
[![](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjinptnav7_CNmsrnCzCArizBz19_srW3AyQ8LL5b5PHqFw-m3zjXDDqxd4PMCacnTQ2xyJh61Wo_0omEbwQ5AGJRsHteeqa7-btTty4QGcVsdTVp0CpOxZrw7hFkpwBGWN_lea/s400/Screenshot+2021-10-24+at+20.59.56.png){border="0"
width="400" original-height="94"
original-width="693"}](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjinptnav7_CNmsrnCzCArizBz19_srW3AyQ8LL5b5PHqFw-m3zjXDDqxd4PMCacnTQ2xyJh61Wo_0omEbwQ5AGJRsHteeqa7-btTty4QGcVsdTVp0CpOxZrw7hFkpwBGWN_lea/s693/Screenshot+2021-10-24+at+20.59.56.png){style="display: block; padding: 1em 0; text-align: center; "}
:::

This is an example where Plazi has successfully parsed the specimen. But
I keep coming across cases where specimens have not been parsed
correctly, resulting in issues such as single specimens being split into
multiple records (e.g.,
<https://plazi-tester.herokuapp.com/?uri=5244B05EFFC8E20F7BC32056C178F496>),
geographical coordinates being misinterpreted (e.g.,
<https://plazi-tester.herokuapp.com/?uri=0D228E6AFFC2FFEFFF4DE8118C4EE6B9>),
or collector\'s initials being confused with codes for natural history
collections (e.g.,
<https://plazi-tester.herokuapp.com/?uri=252C87918B362C05FF20F8C5BFCB3D4E>).

Parsing specimens is a hard problem so it\'s not unexpected to find
errors. But they do seem common enough to be easily found, which raises
the question of just what percentage of these material citations are
correct? How much of the data Plazi feeds to GBIF is correct? How would
we know?

## Systemic problems

Some of the errors I\'ve found concern the interpretation of the parsed
data. For example, it is striking that despite including marine taxa
**no** Plazi record has a value for depth below sea level (see [GBIF
search on depth range 0-9999 for
Plazi](https://www.gbif.org/occurrence/map?depth=0,9999&publishing_org=7ce8aef0-9e92-11dc-8738-b8a03c50a862&advanced=1)).
But [many records do have an
elevation](https://www.gbif.org/occurrence/map?elevation=0,9999&publishing_org=7ce8aef0-9e92-11dc-8738-b8a03c50a862&advanced=1),
including records from marine environments. Any record that has a depth
value is interpreted by Plazi as being elevation, so we have aerial
crustacea and fish.

### Map of Plazi records with depth 0-9999m

::: {.separator style="clear: both;"}
[![](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhnYHyMsNeElQaa-WjfVe6PHQfPx5O4sbAt6Spxn2Cp8FPPK50JbuKCg0o188uAg8ryVM_aZtYvBSl1tpAdAn25yIvmSh97wf1fxa6nb4C9ijNy_hOTtvK3k1aA3HbWJ6xpSma2/s320/Screenshot+2021-10-25+at+12.03.51.png){border="0"
width="400" original-height="258"
original-width="673"}](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhnYHyMsNeElQaa-WjfVe6PHQfPx5O4sbAt6Spxn2Cp8FPPK50JbuKCg0o188uAg8ryVM_aZtYvBSl1tpAdAn25yIvmSh97wf1fxa6nb4C9ijNy_hOTtvK3k1aA3HbWJ6xpSma2/s673/Screenshot+2021-10-25+at+12.03.51.png){style="display: block; padding: 1em 0; text-align: center; "}
:::

### Map of Plazi records with elevation 0-9999m

::: {.separator style="clear: both;"}
[![](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEh4kucSwi2YxsVsjDboeeZM6Nyy9W0REaqnXEsqgdk2tn_oqzMflKMzqpWnXJCwXvpKD1EOhhtN4CA9B20dJFcMycFpQQJGSlRD-BaWQMv6hApysjFWHkReaaeqYnvwrG0-yjQi/s320/Screenshot+2021-10-25+at+12.04.06.png){border="0"
width="400" original-height="256"
original-width="675"}](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEh4kucSwi2YxsVsjDboeeZM6Nyy9W0REaqnXEsqgdk2tn_oqzMflKMzqpWnXJCwXvpKD1EOhhtN4CA9B20dJFcMycFpQQJGSlRD-BaWQMv6hApysjFWHkReaaeqYnvwrG0-yjQi/s675/Screenshot+2021-10-25+at+12.04.06.png){style="display: block; padding: 1em 0; text-align: center; "}
:::

Anecdotally I\'ve also noticed that Plazi seems to do well on zoological
data, especially journals like *Zootaxa*, but it often struggles with
botanical specimens. Botanists tend to cite specimens rather differently
to zoologists (botanists emphasise collector numbers rather than
specimen codes). Hence data quality in Plazi is likely to taxonomic
biased.

Plazi is [using GitHub to track issues with
treatments](https://github.com/plazi/community/issues) so feedback on
erroneous records is possible, but this seems inadequate to the task.
There are tens of thousands of data sets, with more being released
daily, and hundreds of thousands of occurrences, and relying on GitHub
issues devolves the responsibility for error checking onto the data
users. I don\'t have a measure of how many records in Plazi have
problems, but because I suspect it is a significant fraction because for
any given day\'s output I can typically find errors.

## What to do?

Faced with a process that generates noisy data there are several of
things we could do:

1.  Have tools to detect and flag errors made in generating the data.
2.  Have the data generator give estimates the confidence of its
    results.
3.  Improve the data generator.

I think a comparison with the problem of parsing bibliographic
references might be instructive here. There is a long history of people
developing tools to parse references ([I\'ve even had a
go](https://iphylo.blogspot.com/2021/05/finding-citations-of-specimens.html)).
State-of-the art tools such as [AnyStyle](https://anystyle.io) feature
machine learning, and are tested against [human curated
datasets](https://github.com/inukshuk/anystyle/blob/master/res/parser/core.xml)
of tagged bibliographic records. This means we can evaluate the
performance of a method (how well does it retrieve the same results as
human experts?) and also improve the method by expanding the corpus of
training data. Some of these tools can provide a measures of how
confident they are when classifying a string as, say, a person\'s name,
which means we could flag potential issues for anyone wanting to use
that record.

We don\'t have equivalent tools for parsing specimens in the literature,
and hence have no easy way to quantify how good existing methods are,
nor do we have a public corpus of material citations that we can use as
training data. I [blogged about
this](https://iphylo.blogspot.com/2021/05/finding-citations-of-specimens.html)
a few months ago and was considering using Plazi as a source of marked
up specimen data to use for training. However based on what I\'ve looked
at so far Plazi\'s data would need to be carefully scrutinised before it
could be used as training data.

Going forward, I think it would be desirable to have a set of records
that can be used to benchmark specimen parsers, and ideally have the
parsers themselves available as web services so that anyone can evaluate
them. Even better would be a way to contribute to the training data so
that these tools improve over time.

Plazi\'s data extraction tools are mostly desktop-based, that is, you
need to download software to use their methods. However, there are
experimental web services available as well. I\'ve created a simple
wrapper around the material citation parser, you can try it at
<https://plazi-tester.herokuapp.com/parser.php>. It takes a single
material citation and returns a version with elements such as specimen
code and collector name tagged in different colours.

## Summary

Text mining the taxonomic literature is clearly a gold mine of data, but
at the same time it is potentially fraught as we try and extract
structured data from semi-structured text. Plazi has demonstrated that
it is possible to extract a lot of data from the literature, but at the
same time the quality of that data seems highly variable. Even minor
issues in parsing text can have big implications for data quality (e.g.,
marine organisms apparently living above sea level). Historically in
biodiversity informatics we have favoured data quantity over data
quality. Quantity has an obvious metric, and has milestones we can
celebrate (e.g., [one billion
specimens](GBIF%20at%201%20billion%20-%20what's%20next?)). There aren\'t
really any equivalent metrics for data quality.

Adding new types of data can sometimes initially result in a new set of
quality issues (e.g., [GBIF metagenomics and
metacrap](https://iphylo.blogspot.com/2019/12/gbif-metagenomics-and-metacrap.html))
that take time to resolve. In the case of Plazi, I think it would be
worthwhile to quantify just how many records have errors, and develop
benchmarks that we can use to test methods for extracting specimen data
from text. If we don\'t do this then there will remain uncertainty as to
how much trust we can place in data mined from the taxonomic literature.

## Update

Plazi has responded, see [Liberating material citations as a first step
to more better
data](http://plazi.org/posts/2021/10/liberation-first-step-toward-quality/).
My reading of their repsonse is that it essentially just reiterates
Plazi\'s approach and doesn\'t tackle the underlying issue: their method
for extracting material citations is error prone, and many of those
errors end up in GBIF.


The Plazi project has become one of the major contributors to GBIF with some 36,000 datasets yielding some 500,000 occurrences (see Plazi's GBIF page for details). These occurrences are extracted from taxonomic publication using automated methods.

Problems with Plazi parsing: how reliable are automated methods for extracting specimens from the literature?

> Cool! @[alaskamuseum](https://twitter.com/alaskamuseum) has DOIs for
> specimens, e.g.
> [dx.doi.org/10.7299/X78052...](http://t.co/YHkMl6hmt5 "http://dx.doi.org/10.7299/X78052RB")
> (minted by @[ezidcdl](https://twitter.com/ezidcdl) )
>
> --- Roderic Page (@rdmpage) [May 22,
> 2013](https://twitter.com/rdmpage/status/337284270605348864)

  
I\'ve been banging on about having citable, persistent identifiers for
specimens, so was suitably impressed when Derek Sikes posted a [comment
on
iPhylo](http://iphylo.blogspot.co.uk/2013/04/a-decadal-view-of-biodiversity.html#comment-904886746)
that [Arctos](http://arctos.database.museum/) already does this. For
example, here is a DOI for a specimen:
<http://dx.doi.org/10.7299/X7VQ32SJ>.  
  
![Uam](http://lh4.ggpht.com/-DR3GNpJ5gYk/UZ3cHtCslHI/AAAAAAAACyc/ObXUQzvXLQ8/uam.png?imgmax=800 "uam.png"){style="display:block; margin-left:auto; margin-right:auto;"
border="0" width="460" height="358"}  
  
So, we\'re all done, right? Not quite. DOIs by themselves don\'t get us
where we (OK, where I think we) want to be. The DOI identifies a
specimen, which is great (see discussion on [iDigBio: You are putting
identifiers on the wrong
thing](http://iphylo.blogspot.com/2013/01/idigbio-you-are-putting-identifiers-on.html)
for why this matters). We can also get machine-readable metadata using
the DOI (by using the URL <http://data.datacite.org/10.7299/X7VQ32SJ> ).
The metadata is limited (ideally we\'d want something like Darwin Core),
but it is a start. It\'s not clear how we get from the DOI to Darwin
Core.  
  
There are at least two issues that remain to be tackled. The first is
that we now have a bunch of identifiers for the same thing, e.g.:  

- <http://dx.doi.org/10.7299/X7VQ32SJ>
- <http://arctos.database.museum/guid/UAM:Ento:230092>
- <http://data.gbif.org/occurrences/773259491/>
- UAM Ento 230092 (GBIF)
- UAM Insects 230092 (Arctos)
- BYU_gb32.1 (see
  [DQ457300](http://www.ncbi.nlm.nih.gov/nuccore/DQ457300)

  
Most of these identifiers don\'t know about each other (for example,
GBIF doesn\'t know about the DOI, nor does Arctos link to GBIF). So we
have disconnected pieces of information about the same thing.  
  
The second issue is how do we discover a specimen DOI? CrossRef supports
services where you can take a bibliographic citation, e.g. [Phylogeny
and biogeography of ice crawlers (Insecta: Grylloblattodea) based on six
molecular loci: designating conservation status for Grylloblattodea
species](http://search.crossref.org/?q=Phylogeny+and+biogeography+of+ice+crawlers+%28Insecta%3A+++++++++++++Grylloblattodea%29+based+on+six+molecular+loci%3A+designating+++++++++++++conservation+status+for+Grylloblattodea+species)
and get back a DOI (in this case,
<http://dx.doi.org/10.1016/j.ympev.2006.04.013>). This makes it possible
for publishers to take lists of literature cited in authors\'
manuscripts and quickly add DOIs to those citations. We don\'t have an
equivalent service for specimens, which is going to make our task of
linking specimens to sequences and the literature something of a
challenge.  
  
We are making progress, but there is some way to go. Identifiers are
only part of the solution, we also need services.  


I've been banging on about having citable, persistent identifiers for specimens, so was suitably impressed when Derek Sikes posted a comment on iPhylo that Arctos already does this. For example, here is a DOI for a specimen: http://dx.doi.org/10.7299/X7VQ32SJ. So, we're all done, right? Not quite.

DOIs for specimens are here, but we're not quite there yet

Continuing the theme of trying to map specimens cited in the literature
to the equivalent GBIF records, consider the GBIF record
<http://data.gbif.org/occurrences/685591320>, which according to GBIF is
specimen \"ZFMK 188762\" (a \[sic\] holotype of *Praomys hartwigi*).  
  
This is odd, because the original publication of this name (Eisentraut,
M. 1968 .Beitrag zur Saugetierfauna von Kamerun. Bonner Zoologische
Beitraege, 19:1-14, see PDF below) gives the type (p. 11) as \"Museum A.
Koenig, Kat. Nr. 68. 7\").  
  

::: {style="text-align:center"}
::: {.iframe}
::: {.ndfHFb-c4YZDc .ndfHFb-c4YZDc-AHmuwe-Hr88gd-OWB6Me .dif24c .XV0XSd .LgGVmb .bvmRsc .ndfHFb-c4YZDc-TSZdd .ndfHFb-c4YZDc-TJEFFc .ndfHFb-c4YZDc-vyDMJf-aZ2wEe .ndfHFb-c4YZDc-i5oIFb .ndfHFb-c4YZDc-e1YmVc aria-label="Showing viewer."}
::: {.ndfHFb-c4YZDc-bnBfGc .ndfHFb-c4YZDc-zTETae}
:::

::: {.ndfHFb-c4YZDc-JNEHMb}
:::

::: {.ndfHFb-c4YZDc-K9a4Re}
::: {.ndfHFb-c4YZDc-E7ORLb-LgbsSe .ndfHFb-c4YZDc-LgbsSe-OWB6Me aria-label="Previous"}
::: {.ndfHFb-c4YZDc-DH6Rkf-AHe6Kc}
::: {.ndfHFb-c4YZDc-Bz112c .ndfHFb-c4YZDc-DH6Rkf-Bz112c}
:::
:::
:::

::: {.ndfHFb-c4YZDc-tJiF1e-LgbsSe .ndfHFb-c4YZDc-LgbsSe-OWB6Me aria-label="Next"}
::: {.ndfHFb-c4YZDc-DH6Rkf-AHe6Kc}
::: {.ndfHFb-c4YZDc-Bz112c .ndfHFb-c4YZDc-DH6Rkf-Bz112c}
:::
:::
:::

::: {.ndfHFb-c4YZDc-q77wGc}
:::

::: {.ndfHFb-c4YZDc-K9a4Re-nKQ6qf .ndfHFb-c4YZDc-TvD9Pc-qnnXGd role="main"}
::: {.ndfHFb-c4YZDc-EglORb-ge6pde .ndfHFb-c4YZDc-K9a4Re-ge6pde-Ne3sFf role="status" tabindex="-1" aria-label="Loading"}
::: {.ndfHFb-c4YZDc-EglORb-ge6pde-RJLb9c .ndfHFb-c4YZDc-AHmuwe-wcotoc-zTETae}
::: {.ndfHFb-aZ2wEe dir="ltr"}
::: {.ndfHFb-vyDMJf-aZ2wEe .auswjd}
::: {.aZ2wEe-pbTTYe .aZ2wEe-v3pZbf}
::: {.aZ2wEe-LkdAo-e9ayKc .aZ2wEe-LK5yu}
::: {.aZ2wEe-LkdAo .aZ2wEe-hj4D6d}
:::
:::

::: {.aZ2wEe-pehrl-TpMipd}
::: {.aZ2wEe-LkdAo .aZ2wEe-hj4D6d}
:::
:::

::: {.aZ2wEe-LkdAo-e9ayKc .aZ2wEe-qwU8Me}
::: {.aZ2wEe-LkdAo .aZ2wEe-hj4D6d}
:::
:::
:::

::: {.aZ2wEe-pbTTYe .aZ2wEe-oq6NAc}
::: {.aZ2wEe-LkdAo-e9ayKc .aZ2wEe-LK5yu}
::: {.aZ2wEe-LkdAo .aZ2wEe-hj4D6d}
:::
:::

::: {.aZ2wEe-pehrl-TpMipd}
::: {.aZ2wEe-LkdAo .aZ2wEe-hj4D6d}
:::
:::

::: {.aZ2wEe-LkdAo-e9ayKc .aZ2wEe-qwU8Me}
::: {.aZ2wEe-LkdAo .aZ2wEe-hj4D6d}
:::
:::
:::

::: {.aZ2wEe-pbTTYe .aZ2wEe-gS7Ybc}
::: {.aZ2wEe-LkdAo-e9ayKc .aZ2wEe-LK5yu}
::: {.aZ2wEe-LkdAo .aZ2wEe-hj4D6d}
:::
:::

::: {.aZ2wEe-pehrl-TpMipd}
::: {.aZ2wEe-LkdAo .aZ2wEe-hj4D6d}
:::
:::

::: {.aZ2wEe-LkdAo-e9ayKc .aZ2wEe-qwU8Me}
::: {.aZ2wEe-LkdAo .aZ2wEe-hj4D6d}
:::
:::
:::

::: {.aZ2wEe-pbTTYe .aZ2wEe-nllRtd}
::: {.aZ2wEe-LkdAo-e9ayKc .aZ2wEe-LK5yu}
::: {.aZ2wEe-LkdAo .aZ2wEe-hj4D6d}
:::
:::

::: {.aZ2wEe-pehrl-TpMipd}
::: {.aZ2wEe-LkdAo .aZ2wEe-hj4D6d}
:::
:::

::: {.aZ2wEe-LkdAo-e9ayKc .aZ2wEe-qwU8Me}
::: {.aZ2wEe-LkdAo .aZ2wEe-hj4D6d}
:::
:::
:::
:::
:::
:::

[Loading...]{.ndfHFb-c4YZDc-EglORb-ge6pde-fmcmS
.ndfHFb-c4YZDc-AHmuwe-wcotoc-zTETae aria-hidden="true"}
:::

::: {.ndfHFb-c4YZDc-ujibv-nUpftc}
![](/viewerng/thumb?ds=AON1mFzONkB9Regfl1E9U5Zd5ET12atPZR-fC0gWP_-01hiMK7ho72jL08hMMR-l1ZAhUKfGTqjMJ0eVXAx_F8t4KV2lTWrEy7yJ9cilFWRHZMSdgSVefNPMP69yPIapR5-JDjf8CMVUElIeEUSThKx7hQAJiN-MTb9MHqGVWHbaDGi-pPaQ3VYuzHYwQTfsNdpC0ss9flvIv79welB6HmzClHHKUtPTlhUgsA7VqV6VS9qMQ0MVrRjBCBBjClmULD873NOjtGbtu553QdYD2k7-WRYg9UKEHUygUZ9p1MHqpq_8-eEkLEktVqeKuZrqGHJ4YuLpeDIECFM-FZAcOSVTex5GJ-8WJK3qIFsVbfyX9krrvh2k2ILTdrpPOMvQYLqQygOpm8AL&ck=lantern&authuser&w=800&p=proj){.ndfHFb-c4YZDc-ujibv-JUCs7e}
:::
:::
:::
:::
:::
:::

  
  
The GBIF record includes links to images of ZFMK 188762, such as
<http://www.biologie.uni-ulm.de/cgi-bin/imgobj.pl?sid=T&lang=e&id=102323>.  
  
![Bild
pl](http://lh3.ggpht.com/-8PzhTwrTJOE/UPl5VLupJ9I/AAAAAAAACaw/Zv027kXF4uY/bild.pl.jpg?imgmax=800 "bild.pl.jpg"){style="display:block; margin-left:auto; margin-right:auto;"
border="0" width="460" height="345"}  
  
If we open this link we see that specimen is listed as \"ZFMK-68.7\",
which matches the original description. \"ZFMK-68.7\" is a link to
<http://www.biologie.uni-ulm.de/cgi-bin/herbar.pl?herbid=188762&sid=T&lang=e>,
which is the record for this specimen in the SysTax database.  
  
Note that this URL includes the number **188762**, which is treated as
the catalogue number by GBIF (i.e., \"ZFMK 188762\"). So, it seems that
in the data provided by SysTax the primary key in that database (188762)
has become the catalogue number in GBIF (I tried to verify this by
clicking on the original provider message on the GBIF page but it failed
to produce anything). This means any naive attempt to locate the
specimen \"ZFMK-68.7\" in GBIF is going to fail because the harvesting
and indexing as conflated a local primary key with the catalogue number
that appears in publications that refer to this specimen.  
  
Sometimes I think we are doing our level best to make retrieving data as
hard as possible\...


Continuing the theme of trying to map specimens cited in the literature to the equivalent GBIF records, consider the GBIF record http://data.gbif.org/occurrences/685591320, which according to GBIF is specimen "ZFMK 188762" (a [sic] holotype of
<i>
 Praomys hartwigi
</i>
).This is odd, because the original publication of this name (Eisentraut, M. 1968 .Beitrag zur Saugetierfauna von Kamerun.

Rogue Scholar Beiträge

Problems with Plazi parsing: how reliable are automated methods for extracting specimens from the literature?

DOIs for specimens are here, but we're not quite there yet

More GBIF specimen identifier strangeness