Analysis of Diurnal and Seasonal Mood using Twitter Data

Last updated on 2019-10-01
PDF
  • Collected over 60GB of twitter data of year 2016
  • Used VADER Sentiment Analysis to extract positive and negative affects from tweets
  • Analyzed hourly diurnal mood change by day of the week
  • Analyzed the relationship between mood change and number of friends/followers and device impact
  • Calculated the top PMI words for morning/night
natural language processing
Avatar
Xuewen Yao
PhD student

My research interests include deep learning, wearable computing, and activity/affect detection.