Analyzing Global and Pairwise Collective Spatial Attention for Geo-social Event Detection in Microblogs

Shoko Wakamiya, Adam Jatowt, Yukiko Kawai, Toyokazu Akiyama

Abstract: Microblogging has been recently used for detecting common opinions of users at different geographic places. In this paper we propose a novel spatial visualization system for uncovering collective spatial attention and interest of users not at but rather towards different locations. In other words, we aim to answer questions of the type: what do users collectively talk about when they refer to certain geographical places? In addition, we analyze relations between geographical locations from where Twitter users issue messages and the locations they tweet about. This allows answering questions such as: what do users at a certain place commonly talk about when they refer to another geographical place? We demonstrate an online visualization system that supports the interactive analysis of collective spatial attention over time using 4 months' long collection of tweets in USA.