edit page / share feedback / contact
Search the database/ Introduction / Methods / Prerequisites / Results / Discussion
Works published in 1994 mentioning “Koopman”:
Bat researchers rely on access to a vast corpus of bat literature to help advance our understanding of bats and the ecosystems they live in. Many researchers build and organize their personal literature collections using mainstream digital tools like Zotero and EndNote, whereas others use homegrown digital methods or even manage their collections manually. However, all researchers routinely encounter roadblocks to literature access including paywalls and older literature resources that have not yet been digitized. To help provide access to bat research literature for all, Plazi and the GBatNet Bat Eco-Interactions Working Group are compiling the Bat Literature Corpus (BatLit). BatLit is an actively managed, digital, versioned, and citable collection of bat research literature and associated metadata compiled from existing literature contributed by bat researchers. BatLit is designed to be used in manual (e.g., point-and-click) as well as automated workflows (e.g., text mining, language model training), and can be accessed in many ways, including, but not limited to, external storage media, Zenodo, and GitHub. As BatLit continues to improve and grow, we aim to continue to democratize access to bat literature, accelerate research, and help reduce the barrier to knowledge for bat researchers around the world. We invite you to contribute your reference library, especially the PDFs, to BatLit and thereby help increase information access for all.
The Bat Literature Project (BatLit) facilitates discovery of scientific literature on bats (Chiroptera).
by Aja C. Sherman (curator, batbase.org), Jorrit H. Poelen (reviewer, archivist), Donat Agosti (reviewer, Plazi), Nathan Upham (reviewer, mammaldiversity.org), Cullen K. Geiselman (reviewer, batbase.org).
Cite as:
Sherman, A.C., …[add more names here], Simmons, N.B., Zijlstra, J.S., Geiselman C.K. 2024. Bat Literature Project (BatLit, https://batlit.org) v0.6 hash://md5/db73e659e8cf16ef50e82bb5e72ae97b .
with contributions from DeeAnn Reeder, Nancy Simmons, Kendra Phelps, …
⚠️ this is a work in progress⚠️
If you’d like to explore BatLit, please review the examples below. Alternatively, if you’d like to learn more about why BatLit exists, and how it was built, please skip to the Introduction
.
One way BatLit can be searched and discovered is through Zenodo at https://zenodo.org/communities/batlit. Zenodo offers search features through a point-and-click web interface, as well as a Web Application Programming Interface (Web API, https://developers.zenodo.org).
For instance, let’s say that Carla would like to search for literature that mention “Rhinolophus sinicus” in their associated metadata (e.g., title, abstract). She uses Zenodo and enters “Rhinolophus sinicus” in the search box of the BatLit community producing the following query: https://zenodo.org/communities/batlit/records?q=rhinolophus%20sinicus . On inspecting the initial results, she limits to see only publications that are openly available by selecting Access Status > Open
in the Zenodo web search results.
BatLit is managed in Zotero. Each Zotero record has a group id and an item id. These ids uniquely identify a record. These ids are combined to create a life science identifier for each Zotero literature record with the following syntax:
urn:lsid:zotero.org:groups:[group id]:items:[item id]
For example, the Zotero record with group id 5435545
(the batlit group) and item id C4GD9Y6Q
has a lsid:
urn:lsid:zotero.org:groups:5435545:items:C4GD9Y6Q
Now, you can search Zenodo for deposits (publications) related to this particular Zotero record by putting the lsid (in double quotes) in the search box. After hitting “enter”, you are directed to the following search results page .
The example above shows that BatLit records in Zotero are linked to their derived records in Zenodo by their record identifiers.
Because BatLit is integrated with Zenodo, you can benefit from the powerful features of the Zenodo Web Search and Zenodo Web API for discovery. See https://developers.zenodo.org for more information.
If you have any comments, suggestions, or questions, please open an issue.
name | version | date | size | # pages | # references | # attachments | fingerprint |
---|---|---|---|---|---|---|---|
Bat Literature Corpus | v0.1 | 2024-04-26 | 7.9 GiB | 2929 | 5055 | hash://sha256/6ba…189 | |
Bat Literature Corpus | v0.2 | 2024-05-16/2024-05-17 | 11.6 GiB | 3310 | 5471 | hash://md5/be6…1d7 | |
Bat Literature Corpus | v0.3 | 2024-06-25/2024-06-26 | 13.6 GiB | 5501 | 7229 | hash://md5/350…77d | |
Bat Literature Corpus | v0.4 | 2024-08-01/2024-08-02 | 50.9 GiB | 20146 | 29860 | hash://md5/b39…72a | |
Bat Literature Corpus | v0.5 | 2024-08-16/2024-08-17 | 50.9 GiB | 20145 | 29850 | hash://md5/26f…b20 | |
Bat Literature Corpus | v0.6 | 2024-09-19/2024-09-20 | 44.7GiB | 427105 | 19038 | 22590 | hash://md5/db7…97b |
We use Zotero for managing our literature corpus, and Preston for tracking their associated content in a versioned corpus. This versioned corpus is designed to be published through various channels such as local storage media (e.g., external harddisk), GitHub pages and Zenodo.
(Bat Literature Zotero Group) -[:track]-> (Bat Literature Corpus)
(Bat Literature Corpus) -[:publish]-> (https://bat-literature.github.io)
(Bat Literature Corpus) -[:publish]-> (Zenodo BLR)
To help keep BatLit current (e.g., add new references) and accurate (e.g., update existing records), we’ve implemented the following curation workflows:
{
"...": "...",
"key": "YWNCWPYJ",
"...": "...",
"relations": {
"dc:replaces": "http://zotero.org/groups/5435545/items/2PWXAVQL"
}
"...": "...",
}
and translates this into an action to annotate any existing Zenodo record associated with http://zotero.org/groups/5435545/items/2PWXAVQL (or urn:lsid:zotero.org:groups:5435545:items:2PWXAVQL) as deprecated and being replaced by https://www.zotero.org/groups/5435545/items/YWNCWPYJ (or urn:lsid:zotero.org:groups:5435545:items:YWNCWPYJ), or
(urn:lsid:zotero.org:groups:5435545:items:2PWXAVQL)
-[:replaced_by]->
(urn:lsid:zotero.org:groups:5435545:items:YWNCWPYJ)
For context see notes related to approach curating duplicate literature entries
.
In the following sections, some examples are listed that uses a notation commonly used in the Unix shell, also known as the “commandline” or “terminal”. And, at the time of writing, these examples can be executed/run provided the following programs are available: preston, jq, mlr as well as more commonly available unix/posix/linux programs like grep, sort, and uniq.
To run these programs, please use some Linux distribution, MacOS, or Windows Subsystem for Linux (WSL) available on Windows 10 and higher.
Also, please make sure to install preston, jq, and mlr.
The tools used in the example workflows are designed to handle lots of data quickly using so-called “Standard Streams”. In addition, they have the ability to run offline after an initial caching (or cloning) of remote resources. If you are unfamiliar with these tools or processing methods and would like to learn more, you may benefit from a Carpentries Lesson like https://librarycarpentry.org/lc-shell/ or many of the other educational materials. Note that some of these tools have been around since the 1970s, are well documented and are likely to stick around a little while longer.
To track the Zotero group and compile a version of the bat literature corpus, the following command is used (in bash/linux):
ZOTERO_TOKEN=[SECRET] preston track --algo md5 https://www.zotero.org/groups/5435545/bat_literature_project
Note that this group has access restrictions for copyright reasons. This is why you need to replace the “[SECRET]” with your personal access token.
To publish the batlit metadata only (not pdfs), use the following commands
# first copy provenance index
preston cp --algo md5 --type provindex [target dir]/data
# then copy the provenance
preston cp --algo md5 --type prov [target dir]/data
cd [target dir]
# and get the associated zotero metadata
preston ls --algo md5\
| grep -v "file/view"\
| grep hasVersion\
| preston cat --algo md5 --remote file://[source dir]/data\
> /dev/null
Estimating number of references in a corpus version -
preston cat --remote https://linker.bio hash://md5/db73e659e8cf16ef50e82bb5e72ae97b\
| grep "items[?]"\
| grep hasVersion\
| preston cat --remote https://linker.bio\
| jq -c '.[]'\
| jq --raw-output -c '.data | select(has("creators"))'\
| wc -l
Estimating number of associated corpus pdfs -
preston cat --remote https://linker.bio hash://md5/db73e659e8cf16ef50e82bb5e72ae97b\
| grep "file/view"\
| grep hasVersion\
| grep hash\
| wc -l
Estimating the total volume of data for the most recent (i.e. “head”) version
preston cat --remote https://linker.bio hash://md5/db73e659e8cf16ef50e82bb5e72ae97b\
| grep hasVersion\
| grep -oE "hash://md5/[a-f0-9]{32}"\
| sort\
| uniq\
| preston cat\
| pv > /dev/null
An example of a tracked Zotero record generated using
preston cat --remote https://linker.bio hash://md5/db73e659e8cf16ef50e82bb5e72ae97b\
| grep "items[?]"\
| grep hasVersion\
| preston cat --remote https://linker.bio\
| jq -c '.[]'\
| head -n1\
| jq .
is shown below:
{
"key": "6TEYS6CT",
"version": 60055,
"library": {
"type": "group",
"id": 5435545,
"name": "Bat Literature Project",
"links": {
"alternate": {
"href": "https://www.zotero.org/groups/bat_literature_project",
"type": "text/html"
}
}
},
"links": {
"self": {
"href": "https://api.zotero.org/groups/5435545/items/6TEYS6CT",
"type": "application/json"
},
"alternate": {
"href": "https://www.zotero.org/groups/bat_literature_project/items/6TEYS6CT",
"type": "text/html"
},
"up": {
"href": "https://api.zotero.org/groups/5435545/items/ARXB2UDJ",
"type": "application/json"
},
"enclosure": {
"type": "application/pdf",
"href": "https://api.zotero.org/groups/5435545/items/6TEYS6CT/file/view",
"title": "1-s2.0-S2589004222010513-main.pdf",
"length": 2286303
}
},
"meta": {
"createdByUser": {
"id": 13229919,
"username": "acsherman",
"name": "",
"links": {
"alternate": {
"href": "https://www.zotero.org/acsherman",
"type": "text/html"
}
}
},
"numChildren": 0
},
"data": {
"key": "6TEYS6CT",
"version": 60055,
"parentItem": "ARXB2UDJ",
"itemType": "attachment",
"linkMode": "imported_file",
"title": "1-s2.0-S2589004222010513-main",
"accessDate": "",
"url": "",
"note": "",
"contentType": "application/pdf",
"charset": "",
"filename": "1-s2.0-S2589004222010513-main.pdf",
"md5": "7d24ebd866e9b3b0a74ce9a6577d8677",
"mtime": 1726696006321,
"tags": [],
"relations": {},
"dateAdded": "2024-09-18T21:46:56Z",
"dateModified": "2024-09-18T21:46:56Z"
}
}
The tracked metadata was used to list the kinds of content included in the Bat Literature Corpus.
The follow bash script was used to generated the content type frequency table below.
cat\
<(echo count contentType)\
<(preston cat --remote https://linker.bio hash://md5/db73e659e8cf16ef50e82bb5e72ae97b | grep items? | grep hasVersion | preston cat --remote https://linker.bio | jq --raw-output '.[].data.itemType' | sort | uniq -c | sort -nr)\
| mlr --ipprint --omd cat
Note that there’s roughly two kinds of content: top level content like journal articles, books, reports and conference papers. These top level content may have one of more association with associated content like attachments, notes, and annotations.
count | contentType |
---|---|
23831 | attachment |
18176 | journalArticle |
637 | note |
358 | book |
269 | bookSection |
95 | annotation |
93 | report |
53 | thesis |
33 | conferencePaper |
21 | preprint |
15 | dataset |
11 | magazineArticle |
6 | webpage |
2 | newspaperArticle |
1 | presentation |
Literature records can be extracted from this corpus in various ways. As an example, we show the output of an executed script in bin/list-refs.sh again a recent version of the BatLit Corpus. For ease of processing, we’ve included a sample of 3 records in the table below, as well as files in tsv/csv formats include 100 records and all records.
filenames | description |
---|---|
refs-100.tsv / refs-100.csv | author/date/title/journal of first 100 records |
refs.tsv / refs.csv | author/date/title/journal of all records |
First 3 records shown below as an example:
authors | date | title | journal | doi |
---|---|---|---|---|
Vriesendorp | Schulenberg | Alverson | Moskovits | Moscoso | 2006 | Rapid biological inventories: Sierra del Divisor | ||
Carus | 1896 | Wissenschaftliche Mittheilungen. | Zoologischer Anzeiger | |
Handley, Jr. | 1996 | New species of mammals from northern South America: bats of the genera Histiotus Gervais and Lasiurus Gray (Chiroptera: Vespertilionidae). | PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON |