[Board] separate discussion on the value of video archiving

["Peter Löwe"]

Jody, all,

just a few thoughts about the potential of video archiving in general and what TIB Hannover (german national library for science & tech, strategic OSGeo partner for video archiving) can enable the community to achieve.

Over the last years, OSGeo hast established workflows to record and archive conference presentations. A few minor glitches (-> loss of some vids following FOSS4G Portland ) made clear that it makes sense to work with a partner who's professional about video archiving. That's where TIB comes in. As a national infrastructure they have a clear mandate (and budget!) to collect geoinformatics-related content, currently in English and German.

I want to emphasize that TIB does much more in the background than just providing a long term dumping place for OSGeo-related videos. For starters, each video receives a Digital Object Identifier (DOI), which serves as an *unbreakbale* web link, which allows to link to and cite the video for many years to come (e.g.: https://doi.org/10.5446/20420). This is very useful as it allows 1.) scientific citation: OSGeo videos can be cited by scientists just like they cite an article in a scientific journal (which is also done by a DOI) (just throw the DOI of a particular OSgeo video into the Datacite search engine and it will generate proper citations in a variety of established standards like Harvard, IEEE, BibTex, etc. eg: https://search.datacite.org/works?query=10.5446%2F20420 ), 2.) the capability to reference just sections from a lengthy video (like, when the presenting person says something important. eg: https://doi.org/10.5446/20420#t=21:59,22:03). 3.) For the OSGeo members who work in science (or are still stundents), this allows them to receive scientific merit for presentations, since all DOI-referenced material becomes part of the scientific track record (-> and can be linked to a particular person by an ORCID ID: https://orcid.org/ or DataCite (e.g.: https://search.datacite.org/works?query=Jody+garnett)).

So all this infrastructure is already in place and can be used to access, cite and reuse OSGeo-related video content.

FWIW, a next step for OSGeo could be to adopt DOI-links also to reference software, both on the scale of OSGeo projects, but also for individual functionalities, like a particular module within GRASS, GIS, etc. This would allow to give scientific merit (by citation) to the code devlopers and maintainers. This will require a different discussion. 

When a conference video is uploaded to TIB, it is not published right away: TIB applies voice-to-text conversion and OCR to automatically create a lot of additonal metadata (in additon to publisher/conference, recording date, speaker, conference title, presentation title, etc.). The potential value of this feature remains to be explored by OSGeo together with TIB.

What I noticed at FOSS4G conferences is, that speakers really make an effort to cram as much interesting, funny and visionary content into their presentations, which usually does not reflect the abstract texts. Also, presentations touch on different levels of technology or different topics. It is not unusual to have a presentation which talks about advanced server sided OGC services, geological data volumes and an particular region of interest. With standard meta data it would be next to impossible search the trove of OSGeo videos for vague queries like "give me all videos which mention the geology of Berlin", "give me all videos which make references to the movie Wargames, Star Trek or Amy Winehouse songs" (all this can be found in the video archives !), etc. The advanced search capabilities by TIB will enable just that, but OSGeo needs to become an active sparring partner for TIB to push forward with these developing technologies and services. It would be interesting to consider a joint research programme / research project with them, but I am not sure how to wing this. Ideas, anyone ?

best,
Peter  
 


<peter.loewe at gmx.de>