Identifying Crime-prone Areas Based on Tweet Sentiments

Abstract

Social media’s big data can be used to infer real-time sentiments of people about notable events, topics, and places. The goal of this paper is to use social media data for route navigation, a popular everyday application. Whereas existing navigation systems are optimized for the shortest distance or the fastest time, social media sentiments can be used to explore a new dimension such as safety and happiness. We propose a system called SocRoutes that aims to find a safer, friendlier, and more enjoyable route based on sentiments inferred from real-time, geo-tagged messages from Twitter. SocRoutes tailors routes by avoiding places with extremely negative sentiments, thereby potentially avoiding crime-prone areas at a marginal cost in total distance compared to the shortest path. The system supports three types of traveling modes: walking, bicycling, and driving. Based on the real crime-history data published by the City of Chicago Data Portal, we demonstrate a significant correlation between regional Twitter sentiments and crime rates and that SocRoutes can successfully avoid crime hotspots by using social media sentiments.