Thanks for the article.

My concern is that we have to be careful not to assume web-based use cases are representative. That may lead us to misidentify which datasets and categories of open data have the widest potential for re-use.

Open data lends itself more to web-based applications than does closed data, because the licensing terms are so much more flexible. However many firms re-use open data for their internal business purposes, or incorporate it in products or services delivered to customers via channels other than the web. Attribution for that re-use may appear only in offline documentation.

Publishers’ download statistics will also usually understate the scale of re-use. Firms may procure a dataset once centrally and then cascade it to internal users or forward it under sub-licence to end users.

I think tracking re-use comprehensively will always be an intractable problem for open data publishers. It’s rather inherent in the nature of open data, because there’s no exchange of contracts and no obligation on re-users to report back to the publisher.

Just started another on musical improvisation. This should be more of a challenge for me. I’ll probably do some other courses in future on other topics that interest me.