City-scalable Destination Recommender System for On-demand Senior Mobility

As one of the direct beneficiaries of advances in information & communications technologies, mobility on-demand (MOD) fleet operations have grown increasingly relevant to the operation of smart cities. Many seniors face mobility issues and differences in life-style from the rest of the population; as a result, recommended destinations for the general population may not be preferable to this specific subpopulation. This project is developing a new inference tool that fleet operators can use over time to efficiently learn users’ preferences for different destinations. A “routing-constrained recommender system” is being developed with a component to select a subset of destinations that are route-constrained to recommend to the user to select, and sequential learning component that takes different users’ feedback (e.g. a 5-star rating and selection of destination in each case) to update the knowledge base.

Language

  • English

Project

Subject/Index Terms

Filing Info

  • Accession Number: 01648397
  • Record Type: Research project
  • Source Agency: Connected Cities for Smart Mobility towards Accessible and Resilient Transportation
  • Files: UTC, RiP
  • Created Date: Oct 12 2017 10:11AM