This project aims to use the sentiment pipeline to trace historical shifts in awareness. Many debates - whether about climate change, genetically modified foodstuffs, or #metoo - hint at a high form of awareness in our current global society. It is, however, far less evident where these sentiments are rooted in and how they have evolved over time. We aim to investigate this for the Dutch case by focusing on a genealogical study of central modifiers of awareness - (un)healthy, (not) harmful, etc. We are particularly interested in the roles of multinationals like Shell and Unilever as agents of change in these debates.

To do so, this project develops a historical sentiment analysis pipeline that is based on machine learning. The use of this text mining approach in historical scholarship has been hampered by the manner in which "sentiments" are usually implemented in thesauri and binary systems of sentiment qualifications (positive - negative; based on static lists). With this project we aim to make sentiment analysis more historically dynamic and context-specific.

• dr. Gertjan Plets
• dr. Pim Huijnen

Marie Sklodowska-Curie Innovative Training Network grant for the program CHEurope: Critical Heritage Studies and the Future of Europe - towards an integrated, interdisciplinary and transnational training model in cultural heritage research and management.

The consortium consists of 21 European academies and museums, led by the University of Gothenburg. Utrecht University participates in the € 3,9 million project with one of the 15 PhDs in the subprojects project “Mining transnational reference cultures in multilingual and multimedia repositories” that is developed with the Netherlands Institute for Sound and Vision in Hilversum. 

CHEurope aims to develop a new integrated theoretical and methodological framework to enhance the academic and professional training and open future job opportunities in cultural heritage preservation, management and promotion. Heritage has commonly been perceived through its contingent relationship to other areas, preventing it to be considered as a “legitimate” scientific discipline. Moreover, research and practice in this field are still too often seen as separate dimensions. Thus, there is an increasing need to address these diverging trends in the expanding heritage industry with a critical approach that situates cultural heritage in its social, economic and political frameworks, as well as in professional practice.

Bringing together a network of key European academic and non-academic organisations, the project will explore the processes by which heritage is “assembled” through practice-based research in partner institutions that connect students to their future job markets and publics. Our aim is to inform more conventional aspects of cultural heritage designation, care and management with a strong focus on present and future consumers. The project will develop an advanced learning strategy based on the emerging field of Critical Heritage Studies, which combines theoretical and instrumental knowledge at a transnational and interdisciplinary level, in a series of research seminars, summer schools and secondments.

The program is based on themes where cultural heritage is undergoing profound change, such as Heritage Futures, Curating the City, Digital Heritage, Heritage and Wellbeing and Management and Citizen Participation. In so doing, this research will have a direct impact on future heritage policies and be linked explicitly to new modes of training. These will enable future practitioners to facilitate a more democratic and informed dialogue between and across various heritage industries and their users, promoting entrepreneurship and innovation in this field.

• prof. dr. Joris van Eijnatten
• dr. Gertjan Plets

The goal of this project is to develop a repurposable tool that enables humanities researchers to mine the historical development of concepts and the vocabulary with which they are expressed in big textual data repositories. Recent research suggests that vector representations derived by neural network language models offer new possibilities for obtaining high quality semantic representations from huge data sets. We aim to use this methodology to mine changing vocabularies that are used to articulate historical concepts in public discourse (such as citizenship, democracy, evolution, health, liberty, security, etc). This does not only allow us to assess conceptual stability, change and replacement, but also offers us a new way to distinguish between core concepts and the marginal associations that are attached to them in certain historical contexts. Mining Shifting Concepts through Time (ShiCo) uses forementioned innovative word vector algorithms to address two critical challenges for humanities scholars: 1. How can we use text mining methodologies to trace the continuities and changes in historical concepts in big data repositories of digitized periodicals; and 2. How can we present the historical dimension of these cultural formations to humanities researchers in a way that is intuitive and interactive, but at the same time quantitative and reproducible.

• Maarten de Rijke (UvA)
• Charles van den Heuvel (UvA)
• Tom Kenter (UvA)

The purpose of this project is to have a stable, reliable IT environment for the innovative generic text mining application “Texcavator” which is an integral part of the NWO-funded program “Translantis: Digital Humanities Approaches to Reference Cultures.” One of the aims of the Translantis program is to test and further develop Texcavator to research “big data” sets. This will allow a multidisciplinary team of humanities scholars to address conceptual historical questions by mining large textual repositories such as the collection of digitized historical newspapers as provided by National Library of the Netherlands (KB). Texcavator is implemented as a set of modules. It combines xTAS – a scalable open source text analysis service developed by ISLA (UvA) – with Elasticsearch – a big data search and analytics platform. We ask support from the eScience centre in dealing with the challenges of creating a stable IT environment for Texcavator that can be used on a day-to-day basis by a user group of 15-30 historians. The supply of eScience infrastructure (server and computing capacity) will guarantee stable access for users to a prototype of Texcavator within an OTAP-structure. The development capacity of the eScience engineer will be essential for expanding Texcavator’s text mining functionalities.

• prof. dr. Joris van Eijnatten

• prof.dr. Toine Pieters

This project proposes, firstly, that the emergence of trans-national reference cultures was one of the most significant consequences of cultural encounters within the European framework between 1815 and 1992, and, secondly, that these encounters were asymmetrical. ASYMENC investigates how these reference cultures, defined as spatially and temporally identifiable cultures that offer a model to other cultures, have been established in public debates. The availability of large digital data collections enables us for the first time to study long-term developments and transformations of such cultural imaginaries in a systematic, longitudinal, and quantifiable way. This enables us to chart the regional and national public discourses in which collective frames of orientation and cultural cross-referencing to European "others" have been established. ASYMENC explores how public discourses about reference cultures contributed to the formation of communities, from local to trans-national, and thus foster adherence or resistance to particular versions of “Europe”.

ASYMENC strives to accomplish three related objectives: a. explore the concept of reference cultures as a way to understand asymmetrical cultural encounters within Europe; b. develop a digital humanities demonstrator to map and analyse the vectors of asymmetrical encounters in European public debates; and c. build an interdisciplinary consortium of three European research centres around the crosscultural study of printed media in Europe by means of multilingual text mining.

ASYMENC uses the innovative digital humanities methodology of multi-lingual text mining to map
the dynamics, intensity, and direction of intercultural references within European public discourse. Innovative semantic-text analytics will be applied to large corpora of digitized newspapers in the Benelux countries and the surrounding countries that served as reference cultures. The demonstrator will be applied to case studies in which regional, national, and European dimensions of cultural encounters are visible, such as references to European urban centres, mass media, and consumer products.

• prof. dr. A.H.L.M. (Toine) Pieters
• prof. dr. Joris van Eijnatten

• Prof.dr. Caroline Sporleder (Digital Humanities Centre / Trier University)
• Dr. Ulrich Tiedau (Centre for Digital Humanities / University College London)

Searching public discourse. Research into our culture is about understanding conversations and debates as forms of public discourse. https://www.esciencecenter.nl/project/spudisc

• prof. dr. A.H.L.M. (Toine) Pieters

• prof.dr Maarten de Rijke
• prof.dr Ed Tan

Op welke manieren hebben de Verenigde Staten in de afgelopen eeuw als cultureel model gefungeerd voor Nederland? De term referentiecultuur verwijst naar het feit dat sommige culturen een dominante rol spelen in de internationale uitwisseling van ideeën, producten en praktijken. Zij fungeren als een cultureel model dat geïmiteerd, aangepast of juist verworpen kan worden. Met behulp van digitale technologieën zoals text mining gaat dit HORIZON-project de rol van de VS als ‘referentiecultuur’ analyseren in Nederlandse debatten over sociale veranderingen en collectieve identiteiten. De Universiteit Utrecht spreekt in dit project de ambitie uit om in de toekomst een leidende rol te spelen bij de verdere ontwikkeling van het veelbelovende gebied van de digitale humaniora.

Dynamiek van referentieculturen
De historische dynamiek van referentieculturen is nog nooit systematisch onderzocht en eigenlijk weten we er te weinig van af. Om er meer inzicht in te krijgen, richt dit project zich op drie onderling samenhangende vragen:
1. Hoe waardeerde het Nederlandse publieke debat tussen 1890 en 1990 de ideeën, producten en praktijken die het associeerde met de Verenigde Staten?
2. Hoe kan text mining software worden gebruikt om trends en verschuivingen in het publieke debat te traceren, vooral op het gebied van economie, cultuur, en wetenschap en technologie?
3. Hoe weerspiegelt en beïnvloedt een publiek debat de opkomst en inwerking van referentieculturen?

Om de ontwikkeling van referentieculturen in kaart te kunnen brengen, is het nodig om het publieke debat over een langere periode te volgen. Enorme gedigitaliseerde krantenbestanden in de Koninklijke Bibliotheek (KB) maken het voor het eerst mogelijk om grootschalige ontwikkelingen in het nationale discours systematisch te analyseren. Dat gebeurt met innovatieve en krachtige ‘text-mining’ software, waarmee patronen en trends kwantificeerbaar uit de teksten kunnen worden gedestilleerd. Deze gegevens kunnen vervolgens worden gekoppeld aan cijfers over bijvoorbeeld economische en maatschappelijke trends, en aan resultaten van meer traditioneel kwalitatief onderzoek.

• dr. José de Kruif
• prof. dr. Joris van Eijnatten
• dr. Jesper Verhoef

• prof.dr. Toine Pieters
• dr. Charles van den Heuvel
• prof.dr. Maarten de Rijke