Perceptual Hashes, Persistent Predictable Identifiers and more

[Maarten Zeinstra]

Hi all,

Great to have met some of you yesterday and congratulations on the many interesting project your doing I saw during this mornings lightning talks. Unfortunately I had previous engagement, I could not stay the entire three days. I promised some of you to share some links about my lighting talk about Perceptual Hashing on IIIF-discuss.

I’m a developer, and law scholar at Kennisland, a Dutch independent Think Tank. We for example represent Creative Commons Netherlands, work with Europeana on their licensing framework, Last year we developed, together with Klokan Technologies, Embedr.eu. Embedr.eu was a means to embed IIIF player via an iFrame on other sites. The platform was hosted on AWS and used Dockers. Unfortunately we could not find a good business case for the platform and it is now offline. All source code is open source and can be found here: github.com/embedr/

Our other project is a means to (re)connect the attribution chain when that chain is broken. The attribution chain is what we call the availability of provenance information (metadata) at every copy of a media file. often when a file is re-encoded, resized or uploaded to a mediaplatform, embedded metadata is stripped (I don’t know if this is the case with the IIIF Image API). Externally placed metadata, like manifest, is also often lost when people copy a work from a source. Without this provenance information the file cannot be properly attributed, permission to reuse cannot be requested, etc.

To recover the link back to this provenance information I raised the need for persistent predictable identifiers. I’ve written a call to action here: https://www.kl.nl/opinie/need-shared-persistent-reproducible-identifiers/. We did the underlying research with our partner CommonsMachinery.se at a project called videorooter.eu that tries to do perceptual hashing on video. This was a natural follow up of our partner's  work on image perceptual hashing (elog.io and blockhash.io). Blockhash.io is also one of the most popular perceptual hashing algorithms at the moment.

After a workshop on standardising hashing (http://videorooter.eu/2016/05/21/announcement-workshop-on-standardising-hashing/) we created a WhitePaper on this subject that I want to bring to your attention: https://docs.google.com/document/d/1V96B6SwSxS3SDhqeWlkS07ZYSluMXG3KexqQ0yWVNwo/edit. We would love your feedback on that document. We believe adding perceptual hashing to the IIIF project would raise the value of the series of APIs by being able to get into the network of IIIF implementations and for relatively cheaply checking if a media file exists in other IIIF repo’s. 

What should we do to bring our project to the next level?

Kind regards,

Maarten Zeinstra

--

Kennisland | www.kl.nl | t +31205756720 | m +31643053919 | @mzeinstra