AOS 2020 Meeting: Bird Data Harmonization Workshop

[Carla Cicero]

Curators/managers of ornithological collections: Please see the invitation below for a data workshop to be held at the North American Ornithological Conference in Puerto Rico. If you actively manage collection data that are published to VertNet/GBIF/iDigBio, and want to see data standardized to make them more discoverable for research and education, this workshop will be of interest to you. Please fill out the form with your intent to attend, if you have not done so already.

Title: Bird Data Harmonization Workshop invitation
Date and time: Tuesday, 11 August 2020
Convenors: John Bates, Carla Cicero, Town Peterson
Invitees: Bird curators and collections managers

Rationale: The natural history museum collections community in ornithology has made major strides in recent decades towards sharing its data resources openly to scientists and interested citizens. As a consequence, 7.69 million bird specimen records are now available and searchable online in DarwinCore format, which is a remarkable achievement. This massive storehouse of bird data is making possible many exciting scientific analyses that are teaching important new lessons about bird diversity and biology.

The DarwinCore provides an ISO standard set of fields for aggregating data from diverse institutions and sources. In theory, these data can be integrated seamlessly once structured into DarwinCore fields. However, DarwinCore describes only the nature of each field and what it should contain; it does not control the values entered into those fields. Thus, aggregated data often include diverse, confusing, and near-random content in DarwinCore fields, which detracts from their utility and hinders discoverability.

As one example, we recently analyzed the terms served by different natural history museums under the field PreparationType. Instead of a small number of expected terms such as “study skin,” “skeleton,” “pickle,” “wing,” etc., we found over 22,000 distinct terms including the names of people, field preparation numbers, and many other values. Inspections of other DarwinCore fields indicate a similar lack of control in data content.

NAOC 2020 plans: This workshop is designed as a meeting of minds among those who work with and manage avian specimen data. We will review the overall situation, and lay out a plan to create standardized “vocabularies” for key data fields. In 2020, we will start by working with three important fields (PreparationType, Sex, and Age) with the goal of creating a standard vocabulary for these fields. Participants will be expected to implement those three vocabularies in their respective collections databases over succeeding months, resulting in a qualitative improvement in the quality and utility of bird specimen data. Once we have a list of the participating collections, we will produce summaries of the data that each is contributing to VertNet, and we will provide summaries of the data “situation” to each curator/collections manager in advance of the meeting, to facilitate productive discussions.

Future years will involve more complex data unification/standardization challenges, such as creating a taxonomy field or fields that would offer an interpretation of bird specimen identifications under one or several global-scale authority lists for bird names. The final outcome will be the possibility of creating modern “bird specimen inventories” … updating data from the 1980s… such as an inventory of bird study skins by age and sex for each species, or a single searchable catalog of avian tissue resources. But that is a vision into the future; for 2020, we will work with the relatively simple fields listed above.

PLEASE fill out a brief survey regarding your potential participation.