Language Density Driven Route Navigation System for Pedestrians based on Twitter Data

Navigation systems often aid tourists traveling in unfamiliar cities. This work presents a novel navigation system for suggesting routes based on traveler density estimated from Twitter. We aim to recommend two types of routing: one is for tourist spots on a route frequently visited by large groups of tourists, and the other one is for suggesting a route with a small number of expected tourists. Our method is based on shallow processing of tweets and can be used in online applications. We demonstrate online route navigation system that recommends routes based on diversity of foreign languages spoken in two touristically popular cites: San Francisco and Kyoto.