27 oktober 2020
25 minuten - 25 minutes
When travelling around, people’s mobile phone periodically communicates with the cellular network towers (even there are not call/SMS/internet activities), which leaves lots of “footprints” at these towers. Due to its high coverage of the whole
population and relatively low cost, this kind of cellular signaling data, which is available at the mobile telecommunication operators, has attracted strong interests from urban planning, transportation and human geography.
This work aims to develop methods to detect transport modes using mobile phone cellular signaling data. The transport modes of interests include walking, cycling, bus, tram, subway, train, and private cars. Several methods are developed, including two supervised methods - one combining rule-based heuristics (RBH) with random forest (RF), and the other combining RBH with a fuzzy logic system - and a third, unsupervised method with RBH and k-medoids clustering.
Evaluation with a labeled ground truth dataset shows that the best performing method is the hybrid one with RBH and RF, where a classification accuracy of 73% is achieved when differentiating these modes. The proposed unsupervised method also achieved a promising result, with an accuracy of 68%. To our knowledge, this is the first study that distinguishes fine-grained transport modes in CSD and validates results with ground truth data.