Spatio-Temporal Sentiment Hotspot Detection Using Geotagged Photos

TL;DR

This paper introduces a deep learning approach for analyzing public sentiment through geotagged photos, identifying spatial and temporal hotspots of emotions, and correlating them with real-world events.

Contribution

It develops a novel deep learning classifier for emotion prediction from images and applies spatio-temporal hotspot detection to analyze emotional patterns over space and time.

Findings

Different emotions have distinct spatial distributions.

Temporal analysis reveals correlations with known events.

The method accurately detects emerging emotional hotspots.

Abstract

We perform spatio-temporal analysis of public sentiment using geotagged photo collections. We develop a deep learning-based classifier that predicts the emotion conveyed by an image. This allows us to associate sentiment with place. We perform spatial hotspot detection and show that different emotions have distinct spatial distributions that match expectations. We also perform temporal analysis using the capture time of the photos. Our spatio-temporal hotspot detection correctly identifies emerging concentrations of specific emotions and year-by-year analyses of select locations show there are strong temporal correlations between the predicted emotions and known events.