SNIC
SUPR
SNIC SUPR
Semi supervised named entity tagging with recurrent neural networks
Dnr:

SNIC 2018/5-56

Type:

SNAC Small

Principal Investigator:

Thomas Mejtoft

Affiliation:

UmeƄ universitet

Start Date:

2018-04-10

End Date:

2019-05-01

Primary Classification:

21103: Interaction Technologies

Webpage:

Allocation

Abstract

This research project aims to evaluate the use of a semi supervised learning technique for Named Entity Recognition (NER) on Swedish tweets and image captions from Instagram. Semi supervised techniques has shown promise in previous studies applied on the English language. However there is a lack of studies for the Swedish language as well as on natural language work outside of newswire text like social media posts. This research also looks to answer questions surrounding the #metoo campaign, we see the campaign as an opportunity to map where, how, and when violence against women occurs. The campaign offers an opportunity for an alternative data collection method surrounding sexual violence outside of traditional outlets.