Site Search
Computer Science


Christie Ezeife, Ph.D.Dr. Christie Ezeife
Dr. Christie Ezeife
Windsor WaterfrontWindsor Waterfront Park
Windsor Waterfront Park
Arunita Jaekel, Ph.D.Dr. Arunita Jaekel
Dr. Arunita Jaekel
Dr. Luis RuedaDr. Luis Rueda
Dr. Luis Rueda
Dr. Robert KentDr. Robert Kent
Dr. Robert Kent
Robin Gras, Ph.D.Dr. Robin Gras
Dr. Robin Gras
Alioune Ngom, Ph.D.Dr. Alioune Ngom
Dr. Alioune Ngom
Jessica Chen, Ph.D.Dr. Jessica Chen
Dr. Jessica Chen
Xiaobu Yuan, Ph.D.Dr. Xiaobu Yuan
Dr. Xiaobu Yuan
Dr. Scott GoodwinDr. Scott Goodwin
Dr. Scott Goodwin
Imran Ahmad, Ph.D.Dr. Imran Ahmad
Dr. Imran Ahmad
Lambton TowerLambton Tower
Lambton Tower
Dr. Ziad Kobti lecturingDr. Ziad Kobti
Dr. Ziad Kobti

Big Data Mining To Construct Truck Tours

Add this event into your calendar using the iCAL format
  • Wed, 12/12/2018 - 1:00pm - 3:00pm

The School of Computer Science at the University of Windsor is pleased to present …

MSc Thesis Defense by:

Vidhi Patel

 Cross-Border shipping of goods among different distributors is an essential part of transportation across Canada and U.S. This research considers the identification of the international tours accomplishing the shipping of goods. A truck tour is a round trip where a truck starts its journey from its firm or an industry, performing stops for different purposes that include taking a rest, fuel refilling, and transferring goods to multiple locations, and returns back to its initial firm location. In this thesis, we present a three step method on mining the GPS truck data to identify all possible truck tours belonging to different carriers. In the first step, clustering technique is applied on the stop locations to discover the firm for each carrier. A modified DBSCAN algorithm is proposed to achieve this task by automatically determining the two input parameters based on the data points provided. Various statistical measures like count of unique trucks and count of truck visits are applied on the resulting clusters to identify the firms of the respective carriers. In the second step, we tackle the problem of classifying the stop locations into two types: primary stops, where goods are transferred, and secondary stops like rest stations, where vehicle and driver needs are met. This problem is solved using one of the trade indicator called Specialization Index. Moreover, several set of features are explored to build the classification model to identify the purpose of stop locations. In the third step, an automated path finder is developed to identify the truck tours starting from each firm. The results of the proposed classification model are compared with the entropy index from the previous work. Experimental results show that the proposed set of cluster features significantly add classification power to our model giving 98.79% accuracy which in turn helps in discovering accurate tours.



Melissa Robinet
(519)253-3000 ext.3773