Here's a question I've been saving up for a suitable Friday!
I regularly use online collections, whether via a web interface or
increasingly through APIs, but typically end up with a wide range of
results that I have to scour through to extract what I really want. Yes,
typically I might not have worded my query very well and many collections
provide tools where I can refine my search or drill down by facets like
date, subject or keyword. Often I just want images, and again systems
typically allow you to filter to just records where digitised version are
available. But what about the less tangible measures around quality and
interest - in other words the 'wow factor' or what I would very
subjectively call the good stuff?
So my question is, have any of you with online collections used qualitative
tools like sentiment analysis or actually just simple, easily available
quantitative metrics like web stats to expose certain records? To explain
that last part, I am thinking of things like basic web analytics around
item views and referrers, and social mentions of specific items e.g. on
Twitter, Pinterest etc. The closest thing I can think of is Flickr's
'interestingness' measure which uses a (secret) algorithm based seemingly
on Views, Favourites, Comments and so on.
I've wrote a short piece a few weeks back with a bit more detail and some
sample links -
http://www.catchingtherain.com/2014/02/sentiment-analysis-for-cultural-collection-objects-aka-how-to-identify-the-good-stuffwhich
at the time genrated a short discussion with a few ideas -
https://twitter.com/jamesinealing/status/435813116258304000
Cheers, James
****************************************************************
website: http://museumscomputergroup.org.uk/
Twitter: http://www.twitter.com/ukmcg
Facebook: http://www.facebook.com/museumscomputergroup
[un]subscribe: http://museumscomputergroup.org.uk/email-list/
****************************************************************
|