Archive for the ‘Umlaut’ Category

What is Umlaut anyway?

Monday, April 30th, 2012

What is Umlaut anyway?

From the webpage:

Umlaut is software for libraries (you know the kind with books), which deals with advertising services for specific known citations. It runs as Ruby on Rails application via an engine gem.

Umlaut could be called an ‘open source front-end for a link resolver’ — Umlaut accepts requests in OpenURL format, but has no knowledge base of it’s own, it can be used as a front-end for an existing knowledge base. (Currently SFX, but other plugins can be written).

And that describes Umlaut’s historical origin and one of it’s prime use cases. But in using and further developing Umlaut, I’ve come to realize that it has a more general purpose, as a new kind of infrastructural component.

Better, although a bit buzzword laden:

Umlaut is a just-in-time aggregator of “last mile” specific citation services, taking input as OpenURL, and providing an HTML UI as well as an api suite for embedding Umlaut services in other applications.

(In truth, that’s just a generalization of what your OpenURL Link Resolver does now, but considered from a different more flexible vantage).

Reading under Last Mile, Specific Citation I find:

Umlaut is not concerned with the search/discovery part of user research. Umlaut’s role begins when a particular item has been identified, with a citation in machine-accessible form (ie, title, author, journal, page number, etc., all in seperate elements).

Umlaut’s role is to provide the user with services that apply to the item of interest. Services provided by the hosting institution, licensed by the hosting institution, or free services the hosting institution wishes to advertise/recommend to it’s users.

Umlaut strives to supply links that take the user in as few clicks as possible to the service listed, without ever listing ‘blind links’ that you first have to click on to find out whether they are available. Umlaut pre-checks things when neccesary to only list services, with any needed contextual info, such that the user knows what they get when they click on it. Save the time of the user.

Starts with a particular subject (nee item) and maps known services to it.

Although links to subscriber services are unlikely to be interchangeable, links to public domain resources or those with public identifiers would be interchangeable. Potential for a mapping syntax? Or transmission of the “discovery” of such resources?