Browse interface for BBC Creative Archive clips

Last week the BBC made available the first video content under the terms of the Creative Archive initiative. The footage includes material covering natural history, wildlife, science, locations. I heard there was a hundred hours of footage, though it’s hard to check this; the clips appear to vary in length from under 30 seconds to 10 or 15 minutes.
The positioning of this material as video resources for VJs makes sense for at least two reasons. First it emphasises the re-mix, re-purposing scope and ambitions of the Creative Archive. Second, more prosaically and practically, it gets round the problem that little or none of the footage is the kind anyone would just sit down and watch (the Tomorrow’s World clip is almost comical in the way the audio drops out — sometimes mid-sentence — presumably when some uncleared music is mixed in with the voice-track).
I may be proved wrong here, but I suspect the BBC will be hoping to discourage anyone drawing direct comparisons between use of this video material and the recent runaway figures for downloads of Beethoven symphonies. The Creative Archive is a very different proposition from music downloads, and the ‘public value’ test in a year’s time will no doubt reflect that — including the different uses to which the material is put. Obviously this trial is just a start, and there has been correspondingly little fanfare about it: the blog coverage is fairly neutral so far, mostly just noting the existence of the new material.
The element of the trial that gives me pause is not the material, the formats or the (UK only) licence, but the user interface for browsing and finding clips.

Long-standing readers of this site may remember that I’ve been concerned about the issue of making large archives accessible to ‘informal’ users for a while: see earlier comments on the usability of archives, 6 Music as a learning resource and the limitations of search for supporting learning.
The Creative Archive trial is not what I described as a “massive slew of undigested resources dumped on the web with just a fancy search engine to help you find what you want” — in fact there is no search facility — but nevertheless I found it quite hard to re-find the Tomorrow’s World clip referred to above, having first located it a couple of days previously. Try it yourself: go to the index of clips and see how long it takes you to find this clip (scroll to bottom of this text for clues).
Hopefully this illustrates the trickiness of designing a browse interface for a disparate collection of material. (If the material were more homogeneous — all natural history, say — then browsing by animal, environment or weather could be a reasonably reliable and powerful means of finding clips.) I’m guessing that the designers did test the classifications with a sample of VJs, and if the top-level categories make sense for their purposes, that is a sound start. But the browse approach still has limitations for material that doesn’t correspond intuitively to the categories.
And this approach clearly will not scale up. Assuming the Creative Archive is a success and the trial is extended, there could presumably be fifty or a hundred times as many clips in the archive in the future. The current browse interface would buckle under that volume.
Guessing again, perhaps the approach will be to design a different type of user interface for each expected user group: a browse interface for the short clips used by VJs, a different approach for health groups, say, and another one for clips of news and sport? That would make more sense than a single user interface to the Creative Archive that tried to be all things to all people — though a ‘distributed’ approach should try not to miss out on the scope for serendipity in people using clips in different ways than the BBC anticipated.
If you’re having difficulty finding the Tomorrow’s World clip, you first need to select Historical Periods and then 1980s. Intuitive?

Leave a Reply

Your email address will not be published. Required fields are marked *