This document summarizes Martijn Kleppe's dissertation research from 2007-2012 on tracing the afterlife of iconic photographs using IPTC metadata. The research involved analyzing a dataset of around 5,000 photographs from 400 Dutch history textbooks to identify which photos were most frequently published. Kleppe used Fotostation Pro software to code images with IPTC fields and export the metadata for analysis. While IPTC was useful, manual work was still needed and errors occurred. Kleppe's dataset and metadata are available online but have seen little use. The research demonstrated both benefits and limitations of using commercial software and standards for scholarly image analysis.
7. Outline
7
• How to find images that are often published?
> Distant Reading
• How to analyse images to study their changing
meaning over time?
> Close Reading
• Lessons learned
8. 1. Often published
8
• Which identical photos are published most often?
• 400 Dutch History textbooks
• Dataset of ~ 5.000 photographs
• Analyzed on 41 variables (f.e. name person & topic)
• How to organise & annotate my image dataset?
9. 1. Often published
My wishlist
• Create my ‘own metadata’ of each image
• Copy my metadata into other images
• (Quick) Search through dataset based on my metadata
• Export my metadata for further analyses
• Allow other researchers to use my images & metadata
• Find similar images in my dataset
14. Advantages for photographers:
- Code multiple images
at once
- All information in 1 file
- All software can read
information in the file
- Easy to transfer
Advantages for academics:
- Code multiple images
at once
- All information in 1 file
- All software can read
information in the file
- Easy to transfer (& share)
- Find images easier
- Export all data
International Press Telecommunications Council
15. 1. Often published – My case
15
• Which photos are published most often in 400 Dutch
History textbooks?
• Locate photos in History textbooks
• Digitize them & add them to Fotostation Pro
• Code all images & add values in IPTC fields
• Using (commercial) software: Fotostation Pro
18. 18
• Export data variables (f.e. topics) to SPSS
• Create frequency lists topics
• Which topics are most frequent?
• Manually go over most frequent topics to find most
published photographs
1. Often published – My case
22. 2. Changing Meaning
22
• Close reading of the captions provided in the history
textbooks
1912?
or
1918?
23. 3. Lessons Learned
23
• IPTC was useful & efficient
• BUT: Manual handwork remained necessary &
typos will happen
• Does the most frequent topic also contain the
most published identical photo?
(Spoiler alert: NO! )
• My dataset & metadata is available for other
researchers. (But hardly used)
25. 3. Lessons Learned
25
• IPTC was useful & efficient
• BUT: Manual labor remained necessary & typos will
happen
• Does the most frequent topic also contain the
most published identical photo?
(Spoiler alert: NO! )
• My dataset & metadata is available for other
researchers. (But hardly used)
• Commercial software was okay for time being, but
sucks in terms of sustainability
27. My wishlist
• Create my ‘own metadata’ of each image
• Copy my metadata into other images
• (Quick) Search through dataset based on my metadata
• Export my metadata for further analyses
• Allow other researchers to use my images & metadata
• Find similar images in my dataset
3. Lessons Learned