Crowdsourcing public transport data via livemobile tracking : Feasibility study of a system capable of collecting mobile data to build a database ofpublic transit routes, stops and timetables, using machine learning techniques andgraph theory

Abstract: Millions of people use public transport systems everywhere in the world, while the number of smartphones connected to the Internet is dramatically increasing. The aim of this thesis is to study, design, and prototype a system to collect data from the mobile devices of regular public transport users and analyse them in order to provide useful information to travellers all over the world. To study the feasibility of this project at a large scale, data will be created to simulate what will be collected via mobile phone applications. The main objective of this paper is to study the effect of data mining techniques on mapping original transport routes with associated timetables.