Mae’r prosiect treftadaeth ddigidol cydweithiol Making Sense of Illustrated Handwritten Archives yn datblygu amgylchedd digidol uwch-dechnolegol cyfeillgar i’r defnyddiwr a fydd yn hwyluso gwaith haneswyr, biolegwyr a churaduron sydd â diddordeb mewn treftadaeth byd natur wedi’i digido a threftadaeth llawysgrifenedig ddarluniedig. Mae’r prosiect Making Sense yn rhoi sylw arbennig i archif Pwyllgor Byd Natur India’r Iseldiroedd, menter gasglu ar raddfa fawr a ariannwyd gan y brenin Isalmaenig Willem I. O 1820 hyd 1850 bu aelodau’r Pwyllgor yn gwneud teithiau helaeth drwy Ynysfor Indonesia, gan greu casgliad unigryw o ddogfennau llawysgrifenedig, sbesimenau a darluniau. Yn ogystal â bwrw golwg cyffredinol dros y prosiect, bydd fy narlith yn trafod y cyfleoedd, peryglon a goblygiadau ehangach sydd ynghlwm wrth gymhwyso system adnabod delweddau (geiriau) a thechnegau digidol eraill yng nghyd-destun casgliadau llawysgrifenedig darluniedig wedi’u digido. The collaborative digital heritage project Making Sense of Illustrated Handwritten Archives develops a user-friendly and technologically advanced digital environment which is meant to facilitate the work of historians, biologists and curators interested in digitized natural historical and other illustrated handwritten heritage. Core use case of the Making Sense project is the archive of the Committee of Natural History of the Netherlands Indies, a large scale collecting endeavour financed by the Dutch king Willem I. From 1820 to 1850, members of the Committee made extensive tours through the Indonesian Archipelago and brought together a unique set of handwritten documents, specimens and visuals. Next to a project overview, my lecture discusses opportunities, pitfalls, and wider implications which the application of an (word) image recognition system and other digital techniques in the context of digitized illustrated handwritten collections entail.

1. Misschien plaatje bijNatural Historical Archives as Digital
Challenge and Opportunity
Andreas Weber, PhD
Department of Science, Technology, and Policy Studies (STePS)
University of Twente

2. USE CASE
PROJECT: MAKING SENSE OF ILLUSTRATED HANDWRITTEN ARCHIVES (2016-2020)
• Archive of 17,000+ digitzed handwritten
and illustrated documents.
• 18 naturalists travelling in insular
Southeast Asia in the early 19th century.
• never fully studied and interlinked
(stored in Naturalis Biodiversity Center)
• Consortium project: NWO Creative
Industries, Brill publishers private
partner.
• Aims: 1] infrastructure (for researchers,
offered by Brill) and 2] a searchable
digital repository of the NC archive (for
historians/biologists and general public)
For more information on “Making Sense” project
and consortium see: www.brill.com/makingsense

3. SOME DESCRIBED ANIMALS CRITICALLY ENDANGERED

4. ‘NATUURKUNDIGE COMMISSIE VAN NEDERLANDS-INDIË, 1820-1850’
THE ARCHIVE
MARITIME SOUTHEAST ASIA
AS AREA OF COLLECTION

5. A GARDEN AS HUB IN SOUTHEAST ASIA
harbour
garden
Botanical garden in Bogor:
• storage of notes and specimens
• processing of observations and collected items
(preparations, drawings)
• point of departure for expeditions
• local and global distribution of specimens and
knowledge

6. Field notes / traval diaries
Drawings
Publications
Links weak or
missing
… necessary for
humanities
and biodiversity
scholarship
THE PROBLEM
Specimens

7. VISUAL CHALLENGES
I. Mixes of French, Latin,
Dutch, Greek, Malay,
German and French.
II. Visual and textual
elements often intertwined.
III. Different writers and styles
often on one page.
Latin
German

8. SEMANTIC CHALLENGES (1) – SPECIES AND PLACE NAMES
1820s: Rhinolopi crumniferi Peron
insignis Horsf.
Present day: Hipposideros larvatus
Complex
taxonomical shifts

9. LINKING CHALLENGE (1) – REFERENCES
Peron, pl. 35
Voyage de découvertes aux terres
australes (1807-1816), co-edited by
M. F. Péron, pl. 35

10. LINKING CHALLENGE (2) – VISUAL REFERENCES
Inserted in field notes;
visual guide for fieldwork
Seba’s Cabinet of Couriosities
facsimile edition

11. HOW TO REALIZE ‘LINKING’ IN A DIGITAL ENVIRONMENT (1)
Wordzone labels
SEMI-AUTOMATED (SEMANTIC)
SYSTEM THAT ESTABLISHES LINKS
Links to sources
(external and
internal)

12. SEMI-AUTOMATED (SEMANTIC)
SYSTEM THAT ESTABLISHES LINKS
REALIZATION IN DIGITAL ENVIRONMENT (2)
SEMANTIC FIELDBOOK
ANNOTATOR
by LISE STORK
PhD Student,
Leiden University
Visual search engine for handwritten
material developed by L. Schomaker
MONK
AI handwriting recognition system,
developed by prof. L. Schomaker
Groningen University
Phd student: Mahya Ameryan

13. MONK AS HANDWRITING RECOGNITION SYSTEM
EXPERIENCES WITH MONK
• Tabula rasa approach ideal for NC
material
• Language independent
(currently MONK processes different forms of
Arabic, Chinese, Western handwriting of all ages
and kinds, as well as Dead Sea scrolls)
• Delivers sufficient precision and speed
for making illustrated handwritten
collections searchable!
• Scalable and reliable
Dead sea scrolls Chronicon Bohemorum
Chinese manuscripts Arabic manuscripts

14. LABEL WORDS (LINK)
HOW DOES IT WORK IN PRACTICE?
For the screencast see:
https://sites.google.com/naturali
s.nl/makingsenseproject/get-
involved/tutorial

15. HOW DOES IT WORK IN PRACTICE?
TRANSCRIBE LINES (LINK)
For the screencast see:
https://sites.google.com/naturali
s.nl/makingsenseproject/get-
involved/tutorial

16. HOW DOES IT WORK IN PRACTICE?
SELECT CORRECT LABELS IN HITLIST
For the screencast see:
https://sites.google.com/naturali
s.nl/makingsenseproject/get-
involved/tutorial

17. MONK OUTPUT
A. INDEX B. GROUNDWORK FOR FULL TEXT
TRANSCRIPTION
C. ENRICHED SCANS
-
Zunge

18. DESIRED LAYOUT (WITHIN MAKING SENSE PROJECT)
SPECIES
NAMES
PLACE
DATE
PERSON
WHICH BAT SPECIES
WERE COLLECTED AND
DRAWN IN JAVA
BETWEEN 1820 AND
1833?
VISUAL
FEATURES

19. 24/02/2018
MORE INFORMATION
For more papers and posters see:
www.makingsenseproject.org
Springer Lecture Notes in
Computer Science (LNCS)
10605 (2018, in press)

20.  Handwritten digital biodiversity heritage collections create opportunities
for cutting-edge computer science research as well as for taxonomic,
historical ecology and history of science research.
 Challenge:
 Linking with technology is only starting point for contextualization of
archives (historians and biologists needed for interpretation)
CONCLUSION