Yaguang Li

Effective map-matching on the most simplified road network

The effectiveness of map-matching algorithms highly depends on the accuracy and correctness of underlying road networks. In practice, the storage capacity of certain hardware, e.g. mobile devices and embedded systems, is sometimes insufficient to maintain a large digital map for map-matching. Unfortunately, most existing map-matching approaches consider little about this problem. They only apply to environments with information-rich maps, but turn out to be unacceptable for map-matching on simplified road networks. In this paper, we propose a novel map-matching algorithm called Passby to work on most simplified road networks. The storage size of a digital map in disk or memory can be greatly reduced after the simplification. Even under the most simplified situation, i.e., each road segment only consists of a couple of intersection points and omits any other information of it, the experimental results on real dataset show that our Passby algorithm significantly maintains high matching accuracy. Benefiting from the small size of map, simple index structure and heuristic foresight strategy, Passby improves matching accuracy as well as efficiency.

Network-Matched Trajectory-Based Moving-Object Database: Models and Applications

Tracking and managing the locations of moving objects are essential in modern intelligent transportation systems (ITSs). However, a number of limitations in existing methods make them unsuitable for real-world ITS applications. In particular, Euclidean-based methods are not accurate enough in representing locations and in analyzing traffic, unless the locations are frequently updated. Network-based methods require either digital maps to be installed in moving objects or transmission of prediction policies, which inevitably increase the cost. To solve these problems, we propose a network-matched trajectory-based moving-object database (NMTMOD) mechanism and a traffic flow analysis method using the NMTMOD. In the NMTMOD, the locations of moving objects are tracked through a dense sampling and batch uploading strategy, and a novel edge-centric network-matching method, which is running at the server side, is adopted to efficiently match the densely sampled GPS points to the network. In addition, a deviation-based trajectory optimization method is provided to minimize the trajectory size. Empirical studies with large real trajectory data set offer insight into the design properties of the proposed NMTMOD and suggest that the NMTMOD significantly outperforms other mobile-map free-moving-object database models in terms of precision of both location tracking and network-based traffic flow analysis.

Compressing large scale urban trajectory data

With the increasing size of trajectory data generated by location-based services and applications which are built from inexpensive GPS-enabled devices in urban environments, the need for compressing large scale trajectories becomes obvious. This paper proposes a scalable urban trajectory compression scheme (SUTC) that can compress a set of trajectories collectively by exploiting common movement behaviors among the urban moving objects such as vehicles and smartphone users. SUTC exploits that urban objects moving in similar behaviors naturally, especially large-scale of human and vehicle which are moving constrained by some geographic context (e.g., road networks or routes). To exploit redundancy across a large set of trajectories, SUTC first transforms trajectory sequences from Euclidean space to network-constrained space and represents each trajectory with a sequence of symbolic positions in textual domain. Then, SUTC performs compression by encoding the symbolic sequences with general-purpose compression methods. The key challenge in this process is how to transform the trajectory data from spatio-temporal domain to textual domain without introducing unbounded error. We develop two strategies (i.e., velocity-based symbolization, and beacon-based symbolization) to enrich the symbol sequences and achieves high compression ratios by sacrificing a little bit the decoding accuracy. Besides, we also optimize the organization of trajectory data in order to adapt it to practical compression algorithms, and increase the efficiency of compressing processes. Our experiments on real large-scale trajectory datasets demonstrate the superiority and feasibility of the our proposed algorithms.

On Efficient Map-Matching According to Intersections You Pass By

Map-matching is a hot research topic as it is essential for Moving Object Database and Intelligent Transport Systems. However, existing map-matching techniques cannot satisfy the increasing requirement of applications with massive trajectory data, e.g., traffic flow analysis and route planning. To handle this problem, we propose an efficient map-matching algorithm called Passby. Instead of matching every single GPS point, we concentrate on those close to intersections and avoid the computation of map-matching on intermediate GPS points. Meanwhile, this efficient method also increases the uncertainty for determining the real route of the moving object due to less availability of trajectory information. To provide accurate matching results in ambiguous situations, e.g., road intersections and parallel paths, we further propose Passby*. It is based on the multi-hypothesis technique and manages to maintain a small but complete set of possible solutions and eventually choose the one with the highest probability. The results of experiments performed on real datasets demonstrate that Passby* is efficient while maintaining the high accuracy.

Towards Fast and Accurate Solutions to Vehicle Routing in a Large-Scale and Dynamic Environment

The delivery and courier services are entering a period of rapid change due to the recent technological advancements, E-commerce competition and crowdsourcing business models. These revolutions impose new challenges to the well studied vehicle routing problem by demanding (a) more ad-hoc and near real time computation - as opposed to nightly batch jobs - of delivery routes for large number of delivery locations, and (b) the ability to deal with the dynamism due to the changing traffic conditions on road networks. In this paper, we study the Time-Dependent Vehicle Routing Problem (TDVRP) that enables both efficient and accurate solutions for large number of delivery locations on real world road network. Previous Operation Research (OR) approaches are not suitable to address the aforementioned new challenges in delivery business because they all rely on a time-consuming a priori data-preparation phase (i.e., the computation of a cost matrix between every pair of delivery locations at each time interval). Instead, we propose a spatial-search-based framework that utilizes an on-the-fly shortest path computation eliminating the OR data-preparation phase. To further improve the efficiency, we adaptively choose the more promising delivery locations and operators to reduce unnecessary search of the solution space. Our experiments with real road networks and real traffic data and delivery locations show that our algorithm can solve a TDVRP instance with 1000 delivery locations within 20 min, which is 8 times faster than the state-of-the-art approach, while achieving similar accuracy.

Benchmarking big data for trip recommendation

The availability of massive trajectory data collected from GPS devices has received significant attentions in recent years. A hot topic is trip recommendation, which focuses on searching trajectories that connect (or are close to) a set of query locations, e.g., several sightseeing places specified by a traveller, from a collection of historic trajectories made by other travellers. However, if we know little about the sample coverage of trajectory data when developing an application of trip recommendation, it is difficult for us to answer many practical questions, such as 1) how many (future) queries can be supported with a given set of raw trajectories? 2) how many trajectories are required to achieve a good-enough result? 3) how frequent the update operations need to be performed on trajectory data to keep it long-term effective? In this paper, we focus on studying the overall quality of trajectory data from both spatial and temporal domains and evaluate proposed methods with a real big trajectory dataset. Our results should be useful for both the development of trip recommendation systems and the improvement of trajectory-searching algorithms.

Multi-task Representation Learning for Travel Time Estimation

One crucial task in intelligent transportation systems is estimating the duration of a potential trip given the origin location, destination location as well as the departure time. Most existing approaches for travel time estimation assume that the route of the trip is given, which does not hold in real-world applications since the route can be dynamically changed due to traffic conditions, user preferences, etc. As inferring the path from the origin and the destination can be time-consuming and nevertheless error-prone, it is desirable to perform origin-destination travel time estimation, which aims to predict the travel time without online route information. This problem is challenging mainly due to its limited amount of information available and the complicated spatiotemporal dependency. In this paper, we propose a MUlti-task Representation learning model for Arrival Time estimation (MURAT). This model produces meaningful representation that preserves various trip properties in the real-world and at the same time leverages the underlying road network and the spatiotemporal prior knowledge. Further-more, we propose a multi-task learning framework to utilize the path information of historical trips during the training phase which boosts the performance. Experimental results on two large-scale real-world datasets show that the proposed approach achieves clear improvements over state-of-the-art methods

A brief overview of machine learning methods for short-term traffic forecasting and future directions

Short-term traffic forecasting is a vital part of intelligent transportation systems. Recently, the combination of unprecedented data availability and the repaid development of machine learning techniques have brought on immense advancement in this field. In this paper, we aim to provide a brief overview of machine learning approaches for short-term traffic forecasting to facilitate research in related fields. We first introduce traffic forecasting and the challenges, and then introduce different approaches for modeling the temporal and/or spatial dependencies. Finally, we discuss several important directions for the future research.

An Efficient Map-Matching Mechanism for Emergency Scheduling and Commanding

Efficient vehicle tracking and trajectory analyzing are important to emergency scheduling and commanding as they are essential for assessing and understanding the current situation. One of the fundamental techniques is map matching which aligns the trajectory points of moving objects to the underlying traffic network. In this paper, we propose an efficient map matching algorithm called EM3 to meet the requirement of high efficiency and accuracy posed by emergency management. Instead of matching every single GPS point, the algorithm concentrates on those close to intersections and infers the matching results of intermediated ones, which makes the algorithm quite efficient and robust to edge simplification. To provide accurate matching results in ambiguous situations, e.g., road intersections and parallel paths, we further propose EM3*, which is based on the multi-hypothesis technique with novel candidate generation and management methods. The results of experiments performed on real datasets demonstrate that EM3* is efficient while maintaining the high accuracy.