Small Area Population Modeling

Authors: Edwin Chow*, Texas State University - San Marcos
Topics: Geographic Information Science and Systems, Population Geography, Spatial Analysis & Modeling
Keywords: population count, web demographics, mobile crowd, geocomputation, agent-based modeling
Session Type: Paper
Day: 4/14/2018
Start / End Time: 10:00 AM / 11:40 AM
Room: Napoleon A3, Sheraton, 3rd Floor
Presentation File: No File Uploaded

Small area population modeling typically involves the estimation of a given population and the modeling of its human dynamics within a small geographic unit at a specific time. Conventional population modeling relies on either field-based questionnaire, remote sensing and/or dasymetric mapping. While these methods are useful, small area population modeling is indeed an intellectual challenge both in theory and in practice. This research will share the findings from exploiting web demographics for small area population estimation and agent-based model for modeling human dynamics. The former involves an alternative demographic data source that leverage personal microdata available on the web, known as web demographics. By searching web demographics by using a list of surnames distinguishable to ethnic groups, a framework has been developed to conduct a “web census” to enumerate demographic records at the street level. The latter component aims to simulate the human movement of a mobile crowd for a specific event (e.g. a rally). Using an agent-based model, each individual is encoded as an agent with associated rules (e.g. move towards a cell based on distance from destination and occupying capacity) to define their crowd behaviors in walking, stopping and personal spacing. The human dynamics was simulated to reconstruct the mobile crowd during the annual July 1st rally in Hong Kong (e.g. density, walking velocities, and entry/departure gateways, etc.). The intellectual merits and research findings sheds useful insights to improve mobile population estimation, and leverage alternative data source to support related scientific applications in small area population modeling.

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login