A Pilot Study to Integrate Mobility Data Collection APPs with Personalized Recommendation Systems

Term Start:

November 1, 2023

Term End:

December 31, 2024

Budget:

$75,000

Keywords:

Smartphone Application, Travel data

Thrust Area(s):

Data Collection Mechanisms, Data Modeling and Analytic Tools

University Lead:

University of Washington

Researcher(s):

Shuai Huang

Recent years have witnessed many efforts to use smartphones to collect travel data. Typical examples include the automatic collection of sensor data such as location, accelerometer, or microphone readings, and personalized recommendation/behavior modification by gamifying travel and providing incentives for particular mode choices or building route choice models for active transportation modes such as bicycling. However, it seems that the two lines of work are usually pursued separately. Our hypothesis is that, to make better personalized recommendation, the data collection could be further optimized as well, with the help of incorporation of latest developments in adaptive sensing, uncertainty quantification, and predictive science that can help the APP prioritize data collection tasks, identify crucial time points for data collection, etc.  On the other hand, the better personalized recommendation the APP can offer, the better user engagement, that will ultimately translate into a long-term adaption of the APP by a wide range of users. The proposed project will take on the following tasks:

  1. Review the existing works to identify and evaluate current open-source software that can collect mobility trajectories, user data, and offer personalized recommendations.
  2. Design a smartphone application to collect travel behavior data and recommend personalized options.
  3. Develop the application and conduct a pilot test on the functions for Android and iOS systems.

The proposed research is expected to launch a new smartphone application on travel behaviors and provide more personalized recommendations based on the collected data, while the end tasks (i.e., the personalized recommendation algorithms) could also help improve data collection as well.

Scroll to Top