Understanding Happiness by Using a Crowd-sourced Database with Natural Language Processing

Research output: Contribution to conferencePaperScientificpeer-review

Abstract

We conduct three studies by utilizing a crowd-sourced database called Happy DB Database, which consists of more than 100,000 descriptions of happy moments written by workers in Amazon’s Machine Turk. Firstly, we apply a state-of-art word embedding algorithm BERT to transform all happy moments to context-sensitive representations to learn two critical concepts of happiness, agency and sociality. Our performance is better than that of the existing publications. Next, We study the association between alcohol consumption and happiness and our result suggests that both alcohol consumption and abstaining from drinking can lead to happiness. However, no association is found between gender and different drinking patterns in terms of breeding happiness. We also delve into happiness that results from interpersonal relationships and a significant association is found between gender and interpersonal relationships. Interestingly, men are more likely to get happy from the moments related to interpersonal connections, compared to their female counterparts.
Original languageEnglish
Publication statusPublished - 2020
Event32th Benelux Conference on Artificial Intelligence and the 29th Belgian Dutch Conference on Machine Learning -
Duration: 19 Nov 202020 Nov 2020
https://bnaic.liacs.leidenuniv.nl/

Conference

Conference32th Benelux Conference on Artificial Intelligence and the 29th Belgian Dutch Conference on Machine Learning
Abbreviated titleBNAIC/BENELEARN 2020
Period19/11/2020/11/20
Internet address

Fingerprint Dive into the research topics of 'Understanding Happiness by Using a Crowd-sourced Database with Natural Language Processing'. Together they form a unique fingerprint.

Cite this