EmotioNL. Emotion detection for Dutch

Start - End 
2019 - 2022 (completed)
Type 
Department(s) 
Department of Translation, Interpreting and Communication
Research Focus 
Research Period 

Tabgroup

Abstract

This research project is situated in the field of NLP and focuses on automatic emotion detection in Dutch texts. Supervised machine learning techniques are currently considered as a state-of-the-art methodology to identify emotional states in text. These systems, however, rely on large quantities of annotated data, so progress is mainly made for majority languages such as English. For Dutch, emotion detection has not yet been studied properly, since no high-quality annotated data sets are available and acquiring such large amounts of annotated data is arduous. Therefore, we explore a (cross-lingual) transfer learning approach to investigate how data sets and systems that were developed for English can be used to bootstrap emotion detection for Dutch. Moreover, this transfer learning methodology will be used to explore cross-domain (different genres) and cross-task (different emotion frameworks) transfer, which will result in a flexible methodology for Dutch emotion detection that can be applied to various use cases.

People

Supervisor(s)

Co-supervisor(s)

Phd Student(s)

Subprojects