Overview

The concept of preferential or context aware Location-based Search (LBS) came into light in the recent past with the inclusion of local businesses into many social networks. To better the user experience of LBS, personalized recommendations are being generated. These recommendations are often produced by utilizing the user’s travel history, points of Interest, and search history. This problem is termed as a individual mobility prediction. However, the idea of integrating multi-dimensional data from various sources including but not limited to user's social network profile, smart devices that communicate the context of travel is also helpful is providing better user experience. Such a inclusion of data aggregation from various sources provides a rich dataset that helps us in truly understanding the user's personal choices. However, there is a huge gap between utilizing such data and providing Mobility-as-a-service (MaaS) while preserving user's privacy has not been studied extensively. Therefore, our goal in this project is to understand the problems of mobility prediction and provide a framework to context-aware and personalized user's mobility prediction. To be more specific, our goals are:

• Creating a digital twin to simulate social network and user patterns
• Privacy and context-aware frequent route pattern mining to predict transportation mode for the user
• Generating privacy preserved similar travelers information
• Predicting the personalized route with Points of Interest

Intellectual Merit

Data privacy has to be given utmost importance due to the data availability to any adversary. A large amount of data is often public which includes user’s social network profile, his demographic information that is published every year by the government, his work and income information, and so on. This allows adversaries to fuse information to find out about the user and his behavior. This is a leading issue for behavioral and psychological attacks. Therefore, we focus on developing a smart transportation system that combines information from various sensors and providing a personalized route choice to the user while preserving data privacy. Research in this area is limited and does not address all the concerns. Therefore, our research will significantly advance the state-of-the-art system in personalized route/destination prediction based on the user behavior.

Broader Impacts

The proposed study creates a futuristic location based search that aims at providing a personal prediction of routes and places to visit based on user's history and profile. The work can later be combined with the smart city technology to truly understand how to improve the quality of life in city planning with this knowledge.