Modeling human interactions with learning under the extended belief-desire-intention framework

Sojung Kim, Hui Xi, Santosh Mungle, Young Jun Son

Research output: Contribution to conferencePaperpeer-review

4 Scopus citations

Abstract

While the extended Belief-Desire-Intention (BDI) framework has been successfully applied to mimic human behaviors under various scenarios, detailed human interactions have not been addressed even though human behaviors are affected by the interactions with other people. This paper is to address various types of human interactions together with learning under the extended BDI framework. To this end, we first adopt five major human interactions from social science literature, such as avoidance, accommodation, compromise, collaboration and competition. We then develop models to realize each of the interaction types. The proposed method is then illustrated for drivers' route choice behaviors involving a real traffic data of Albany, NY. Three types of drivers (commuters, explorers, and travelers) are considered based on the level of experience (or knowledge). The proposed models have been implemented in Repast agent-based simulation software. Experimental results reveal that travelers with no experiences and explorers with partial experiences can reach their destinations via interactions with other drivers and their own self-learning behaviors. Moreover, the learning behavior of each driver is differentiated based on the interaction types. This demonstrates that the proposed approach provides a new opportunity for modeling a more realistic human that consists of various types of human interrelationships.

Original languageEnglish
Pages2861-2870
Number of pages10
StatePublished - 2012
Event62nd IIE Annual Conference and Expo 2012 - Orlando, FL, United States
Duration: 19 May 201223 May 2012

Conference

Conference62nd IIE Annual Conference and Expo 2012
Country/TerritoryUnited States
CityOrlando, FL
Period19/05/1223/05/12

Keywords

  • Agent-based simulation
  • Belief-desire-intention
  • Human interaction
  • Learning

Fingerprint

Dive into the research topics of 'Modeling human interactions with learning under the extended belief-desire-intention framework'. Together they form a unique fingerprint.

Cite this