Sentiment analysis under temporal shift

Publikation: Bidrag til bog/antologi/rapport › Konferencebidrag i proceedings › Forskning › fagfællebedømt

Sentiment analysis models often rely ontraining data that is several years old. Inthis paper, we show that lexical featureschange polarity over time, leading to degradingperformance. This effect is particularlystrong in sparse models relyingonly on highly predictive features. Usingpredictive feature selection, we are able tosignificantly improve the accuracy of suchmodels over time.
OriginalsprogEngelsk
TitelProceedings of the 9th Workshop on Computational Approaches to Subjectivity, Sentiment and Social Media Analysis
ForlagAssociation for Computational Linguistics
Publikationsdato2018
Sider65–71
StatusUdgivet - 2018
Begivenhed9th Workshop on Computational Approaches to Subjectivity, Sentiment and Social Media Analysis - Brussels, Belgien
Varighed: 31 okt. 2018 → 31 okt. 2018

Workshop

Workshop9th Workshop on Computational Approaches to Subjectivity, Sentiment and Social Media Analysis
LandBelgien
ByBrussels
Periode31/10/2018 → 31/10/2018

ID: 214758960