Personalized Recommender System based on Taxi Driver’s Behavior to Optimize Supply and Demand

Please enable javascript in your browser.

Pouyabahar, Ardalan | 2017

582 Viewed

Type of Document: M.Sc. Thesis
Language: Farsi
Document No: 50503 (19)
University: Sharif University of Technology
Department: Computer Engineering
Advisor(s): Heydarnoori, Abbas; Habibi, Jafar
Abstract:
Taxis provide a flexible and indispensable service to satisfy the urban travel demand of public commuters. Balancing supply and demand and minimizing the driving time before finding a customer would directly increase taxi drivers and system income. Availability of GPS traces and easy access to the Internet has enabled taxi companies to be aware of supply-demand level in every region and all taxi states in real time. In this research, we propose a novel fair recommender system to maximize the sum of taxi drivers’ income by recommending regions with the highest profitability due to each driver’s acceptance rate and each regions’ minimum supply availability. Experiment results on TAP30’s data in Tehran shown we are able to increase driver’s income by 6%, which means company’s income can be increased by 40%
Keywords:
Recommender System ; Intelligent Transportation System (ITS) ; Data Mining ; Big Data Proccessing ; Fleet Optimization ; Supply and Demand

No TOC