Historical Quantitative Reasoning on the Web

The Semantic Web is an extension of the Web through standards by the World Wide Web Consortium (W3C) [4]. These standards promote common data formats and exchange protocols on the Web, most fundamentally the Resource Description Framework (RDF). Its ultimate goal is to make the Web a suitable data space for humans (using the paradigm of Linked Documents) as well as for machines (using the paradigm of Linked Data). The latter has experimented an enormous growth in the last years, mostly due to the adoption of Linked Data publishing practices by institutions, governments and users, giving birth to the Linked Open Data cloud: a Web-graph of 100 billion interconnected facts and schemas (also called ontologies) [3]. Many domains have published their datasets as Linked Open Data, including History [6,12], converting them to RDF and linking them to related historical datasets and concepts on the Web. One of the fundamental problems of historical research are so-called gaps in historical evidence: the non-existence of relevant historical data for a particular matter. The very exercise of historical research has a primary focus on filling these gaps, generating the missing knowledge (to a level of certainty) using diverse methods. When raw historical datasets are published on the Semantic Web as Linked Data, these gaps still exist [2]. In this paper, we investigate whether existing Semantic Web technology is useful to fill these gaps of historical knowledge. We study the way in which social historians derive new knowledge from historical quantitative sources to fill missing gaps [2,13], and we mimic their behavior in a Semantic Web setting, by adapting existing technologies to, first, identify these gaps [14] and, second, to fill them [17] We explore whether these mechanisms can be generalized to be applied to other domains [18].