Customizing Your Own Route with QQIP. A Quantitative and Qualitative Itinerary Planner for New Transportation Routes

Abstract

Public transportation route planning is crucial for both traffic management authority and residents. Current procedures for deciding new routes are time-consuming and ineffective due to the complicate simulation process or overwhelming numbers of opinions from stockholders. In this paper, we propose a novel decision supporting tool, Quantitative and Qualitative Itinerary Planner (QQIP), to help governments pre-evaluate new route services in the city in a timely manner. The function of QQIP is three-fold: visualization of urban informatics, a flexible interface for sketching designate routes, and passenger flows estimation in certain time intervals. With acquired relevant urban information, user can pre-estimate the effectiveness of designed routes using QQIP. To capture the spatial-temporal factors correlated with passenger flows, we propose route-affecting region (RAR) and adopt Deep Neural Network (DNN) framework to combine several dynamic and static features. According to our experimental results on bus-ticket data of Tainan city, the proposed RAR-based feature engineering methods are effective for handling and combining high-correlated dynamic and static data; meanwhile, QQIP can help decision makers infer the passenger flow effectively and efficiently for given designated routes.