A Machine Learning Recommender Model for Ride Sharing Based on Rider Characteristics and User Threshold Time
Date of Award
College of Information Technology and Engineering
Type of Degree
Dr. Wook-Sung Yoo, Committee Chairperson
Dr. Husnu S. Narman
Dr. Haroon Malik
In the present age, human life is prospering incredibly due to the 4th Industrial Revolution or The Age of Digitization and Computing. The ubiquitous availability of the Internet and advanced computing systems have resulted in the rapid development of smart cities. From connected devices to live vehicle tracking, technology is taking the field of transportation to a new level. An essential part of the transportation domain in smart cities is Ride Sharing. It is an excellent solution to issues like pollution, traffic, and the rapid consumption of fuel. Even though Ride Sharing has several benefits, the current usage is significantly low due to limitations like social barriers and long rider waiting times. The thesis proposes a novel Ride Sharing model with two matching layers to eliminate most of the observed issues in the existing Ride Sharing applications like UberPool and LyftLine. The first matching layer matches riders based on specific human characteristics, and the second matching layer provides riders the option to restrict the waiting time by using personalized threshold time. At the end of trips, the system collects user feedback according to five characteristics. Then, at most, two main characteristics that are the most important to riders are determined based on the collected feedback. The registered characteristics and the two main determined characteristics are fed as the inputs to a Machine Learning classification module. For newly registering users, the module predicts the two main characteristics of riders, and that assists in matching with other riders having similar determined characteristics. The thesis includes subjecting the proposed model to an extensive simulation for measuring system efficiency. The model simulations have utilized the real-time New York City Cab traffic data with real-traffic conditions using Google Maps Application Programming Interface (API). Results indicate that the proposed Ride Sharing model is feasible, and efficient as the number of riders increases while maintaining the rider threshold time. The expected outcome of the thesis is to help service providers increase the usage of Ride Sharing, complete the pool for the maximum number of trips in minimal time and perform maximum rider matches based on similar characteristics, thus providing an energy-efficient and a social platform for Ride Sharing.
Machine learning -- Development.
Machine learning -- Technological innovations.
Artificial intelligence -- Computer programs.
Yatnalkar, Govind Pramod, "A Machine Learning Recommender Model for Ride Sharing Based on Rider Characteristics and User Threshold Time" (2019). Theses, Dissertations and Capstones. 1259.