Affiliation: University of Waterloo
Sector: Sciences, Humanities, Academic, Research
Rob Warren is a postdoctoral fellow working on databases, information retrieval and data interchange systems. His current interest is the re-use of historical data-sets for modern research projects and using linked open data for computing in the public interest.
The aim of Muninn is to take archives of the first world war, extract the written data using massive amounts of computing power and turn the resulting information into semi-structured databases for further research. I'd like to swap ideas with a few of you on: * Using some of the current tools to ensure the link-ability of some of the archival sources used. * How do we make Dublin Core work properly for complex documents? * Neat strategies to separate layers of evidence from belief as well as really novel data-quality problems. * Some tags that we use in production to speed up reasoning, encode ambiguity and some common-sense linkages.