Context-aware Multi-data Fusion to Achieve Finest Ride-share Schedule Using Big Data

Abstract

Enabling the taxi rides to share empty seats in their vehicles, to riders, is a win-win situation for both riders and drivers. Existing methods mainly target on minimizing the cost of travel for vehicles. In this work we focus on assigning a new rider to a vehicle considering the real-time context of both riders and drivers. The crux of the problem lies in analysing massive amounts of data in an instance of time to obtain the optimal ride schedule. We further study the scenario, to accommodate best compatible riders with minimal change in the route, to maximize the satisfaction of all possible entities in the business model in several aspects. Our system takes advantage from route recommendation and utility maximization approaches. We also inherit the benefits of collaborative filtering to provide best ride experience. The effectiveness of our approach is shown with experimentation on various datasets. Our work has significant importance, as it redefines the personalization preferences based on realtime context.