Abstract
An automated way to measure a text's perspective is essential to a computational analysis of literary style and narrative. However, no such measure of narrative perspective exists to date. We aim to provide such a measure by taking two approaches to operationalise narrative perspective. The first measure is a machine learning approach, the second is based on computing the ratio of combinations of pronouns. Our research demonstrates that narrative perspective can be predicted with high accuracy and provides a reflection on the advantages and pitfalls of both methods.
Original language | English |
---|---|
Publication status | Published - 2020 |
Event | DH Benelux 2020 - Cloud Duration: 03 Jun 2020 → 05 Jun 2020 https://www.universiteitleiden.nl/en/events/2020/06/dhbenelux-leiden-3-5-june-2020 |
Conference
Conference | DH Benelux 2020 |
---|---|
Period | 03/06/2020 → 05/06/2020 |
Internet address |