Eye candy by Eric
The 2012 Spring/Summer issue of Information Standards Quarterly includes a number of LODLAM articles, and of particular interest to me was the article authored by Jane Stevenson entitled “Linking Lives: Creating And End-User Interface Using Linked Data”. In short, Stevenson reported on a project called Linking Lives to make archival information about individuals available as linked open data.
As describes in a previous posting, there were a number of issues to address before the project could be called a perfect success. For example, after aggregating data created by a divergent set of people over a long period of time makes for inconsistent data — data of a great “variety” as Dorthea Salo would say. How to create a level of consistency is a problem to solve. This was even a greater problem as DBPedia data was pulled in the environment because some (Wikipedia) articles contain images, and some don’t. One of the biggest problems creating URIs for individuals because the individuals were not always listed in exactly the same manner from finding aid to finding aid, thus there are multiple identifiers for multiple people. Moreover, sometimes individuals are creators and sometimes individuals are index terms — persons versus agents. Because of all of these issues, a lot of data clean up is necessary before finding aids — even if they do come from EAD files — can be made accurately available as linked data.