MERIT connects research in text analytics, contemporary history, cultural heritage, as well as memory studies. From a technological perspective, the project will help to develop existing tools in information extraction of unstructured and semi-structured texts. We move beyond the state of the art; aside from being able to identify names of persons, locations, organizations, and time expressions, we develop new methods for event identification in multiple texts, i.e. discovering that differently phrased descriptions of an event actually refer to the same event. As we work with English, Dutch, and German texts, we will need to combine the use of pre-existing tools with new cross-lingual tools. The lack of proper multi-lingual domain ontologies furthermore suggests a data-driven, bottom-up approach.
This bottom-up approach is notably similar to the practice of historical research: It offers both historians and cultural heritage centres help in the often laborious task of retrieving and combining specific factual information from large corpora of sources which are typically spread across several archives. Historians typically need to rely on their memories or notes to piece together bits of information spread across sources. MERIT will provide a user-friendly working environment in which historians can make use of the results of the semi-automatic content analysis, edit them and manually enter additional information. This approach allows historians to store and quickly retrieve factual information. Yet still, it ensures full control over the interpretation and verification of the retrieved information.
With regard to the history of the Battle of the Rhineland, MERIT offers a multi-faceted view of the past by linking the experiences of Allied and German soldiers and Dutch civilians who witnessed the events and/or participated in them e.g. in resistance activities. This strong commitment to multiperspectivity not only stands in accordance with good practice in historical research but also with our partners’ concepts for commemorating the Second World War in the Netherlands and in Germany.#
MERIT’s method not only helps to view past events from more than one perspective, it also allows the study of the development of these views. MERIT will be able to semi-automatically identify the key components in the narratives of the Battle of the Rhineland and analyze their distribution over time since 1944.
Prof. dr. Antal van den Bosch (MA, computational linguistics, 1992, Tilburg University; Ph.D., computer science, 1997, Universiteit Maastricht) held research positions at Tilburg University, the Netherlands and the Université Libre de Bruxelles, Belgium (1993-1994), Universiteit Maastricht, the Netherlands (1994-1997) and Tilburg University (1997-2011), where he was appointed full professor in computational linguistics and AI in 2008. He is also guest professor of the University of Antwerp. In 2011 he took on a full professorship in language and speech technology at Radboud University Nijmegen, The Netherlands. His research interests include memory-based natural language processing and modeling, text analytics and applications thereof in cultural heritage and digital humanities, machine translation, and proofing tools. Prof. Van den Bosch has led several cross-disciplinary projects on improving access to cultural heritage with innovative ICT methods: with Naturalis, the National Natural History Museum in Leiden he worked on enriching digital versions of historical fieldbooks and registry books in the NWO CATCH ‘MITCH’ project. With the International Institute for Social History (IISH, Amsterdam) he carries out the NWO CATCH project ‘HiTiME’ (Historical Timeline Mining and Extraction) and the NWO Digging Into Data project ‘ISHER’ (Integrated Social History Environment for Research) on exploiting, linking and enriching primary and secondary historical texts on the history of the social movement.
Marten Düring (BA European Cultural History, 2001, MA War & Memory Studies, 2006, Ph.D. Contemporary History, 2012) worked as Junior Research Fellow and Research Fellow at Kulturwissenschaftliches Institut Essen, Germany in the project Frames of Reference of Help. In July 2012 he defended his PhD thesis on support networks for persecuted Jews during National Socialism. In 2012 he joined Radboud Universiteit Nijmegen as eHumanities guest Post Doc. His research interests revolve around software applications in historical research, he has hands-on experience with Social Network Analysis, qualitative data analysis and computer simulations.