In Representation of Data Perspectives, we investigate long-term changes of perspectives and perception through studying concepts in cultural texts (fiction, biographies, pedagogical texts, journals, newspapers and encyclopedias), from 1600-2000. By elaborating on datasets and NLP techniques that have recently been developed in other projects (e.g. “The Riddle of Literary Quality” (KNAW/UvA), NWO’s Zwaartekrachtproject “Language in Interaction” (UvA), VU/Escience “Embodied Emotions” and BiographyNet (VU)), we cluster, classify and interpret different perspectives on topics that feature in currently available machine-readable Dutch texts from 1600-2000.
Changes in these long-term perspectives/perceptions are collected and analyzed. Such a long-term search of changes in perspectives in texts has never been accomplished before and is challenging from both a Computer Science perspective and a humanities perspective.
The project will deal with complex questions like: ‘How can we track changing perspectives on youth, health, nature, emotions, finance, or poverty during the last four centuries?’ Our goal is to bring these long term perspectives in dialogue with data presented by other scientific fields, e.g. life science, youth studies, business analytics.
As a first use case we are tracking the changes in concepts in literary magazine De Gids.