Purposes of (archaelogical) linked data(?)

Submitted by Isto Huvila on Wed, 03/28/2012 - 18:50

 One of the interesting issues raised in the discussions of the data track at this year's CAA 2012 conference in Southampton is the question of teh purpose of the Linked Data. The purpose in the sense that whether Linked Data should be opened on the premises of its creators or users. @kidehen was kind enough to underline in Twitter (once I posed this rhetorical question) that "#LinkedData == Open Data Access, Connectivity, and Representation. It delivers all of that at Web scale via Hyperlinks." He is of course perfectly right. This is what Linked Data is supposed to be and with a little bit of luck, so it be.


At the same time, however, this is a sort of a political statement that does not really account for the reality of creating and publishing data. Someone chooses which data to open and transform to a linked data and when opening up a particular piece of data, somebody is always making some assumptions and decisions how to do that, how to document the data and how to inform others of its existence. These are done on the premises of the creator of the data.


The conclusion is then that data is opened and linked on the premises of its creators? Well, yes and no. We can and people also do think of its users. Being well informed of the main findings of user studies in information science, it is clear that there are some limits in making inferences about users and their preferences, and in the case of most linked data, the intentions and preferences of unknown users and uses.


As the somewhat unconclusive discussion at the conference and my own pondering of the topic possibly shows (until someone comes and solves the problem), there is no simple answer to the question. It might not be entirely unproblematic to close open data with too many creator side assumptions but at the same time it may be equally destructive to go too far with assumptions about the preferences of the users. So the basic principle of linked data as phrased by @kidehen holds even in practice. At the same time, however, it is extremely important to remember that even if the principle is a good guideline how to do things, it is not necessarily how things go in practice. A way out might be a combination of the principled outspoken idealism of Linked Data initiative and a healthy reminder that the practicalities and ideals of the purposes play a role when data is actually made available.