KnowLedge bAse enrichment from Uncertain medieval pRosopographical dAta – LAURA

One of today's issues is the construction of knowledge bases from large collections of documents. To do this, it is necessary to master a complete and complex process that encompasses several areas of computer science, such as digitization, extraction of named entities, data fusion/alignment, etc. The enrichment of knowledge bases is essential in history. Indeed, these databases are the foundation on which historians base their hypotheses. By regularly adding new information to these databases, we can not only add to our knowledge, but also, where necessary, reinforce the credibility of a piece of information, or, conversely, diminish it, or even correct existing knowledge. This enrichment offers new opportunities for discovery and significant advances, by improving the quality and quantity of data that can be used to build historical hypotheses. This is the ambition of the LAURA project, which focuses on prosopographic databases for the medieval period. Prosopography is a social science method that seeks to analyze a group through a systematic study of the singular itineraries of its constituent individuals. To do this, researchers collect all possible facts (factoids) about each individual. In history, these data are scarce, discontinuous, uncertain and often of mediocre quality. For example, people are referred to by several names, places change their names and boundaries over time, and a course of graduation may change according to the person's time, place or social class. Because of this complexity, many rules remain opaque to historians.
There are many areas of joint research between historians and computer scientists on these prosopographical data. The populating and enrichment of these data deposits could be achieved using a protocol that would build on the experience and results capitalized on in the ANR DAPHNE project, which resulted in several major contributions (4 journals, 16 conferences, 1 software and 2 related projects). The objectives of the LAURA project capitalize on the results of the DAPHNE project, in order to carry out the complete processing chain: from digitization and data extraction to enrichment and exploitation of the knowledge base, while taking uncertainty into account. This knowledge base will be made available to the community via an open platform.