Publication:
A multiagent framework for consumer behavior and purchase intentions in electronic commerce

Thumbnail Image
Authors
Gil-Herrera, Eleazar
Embargoed Until
Advisor
Cesaní-Vázquez, Viviana I.
College
College of Engineering
Department
Department of Industrial Engineering
Degree Level
M.S.
Publisher
Date
2009
Abstract
An agent based model (ABM) to investigate the dynamics of human purchasing behavior within an online environment is presented. Based on empirical research, both extrinsic and intrinsic influencing factors such as disposition to trust, impulsive behavior, security perception and others were identi fied to describe the characteristics of a population comprised mainly of college students. Classifi cation tree analysis was used to generate the decision rules and parameters serving as input data for the model. As a result, this study demonstrates how ABM can be used to analyze consumer behavior to cope with the dynamic changes and complexities in a real-world environment. Experimental fi ndings suggest that the adaptation process to this environment is mainly based on both the increment on the level of trust in online review sites visited, along with an increment in the knowledge about the procedures and security mechanisms available for online transactions. Both factors reduce the rate of non-buyers.

Esta investigaci on presenta el desarrollo de un modelo basado en agentes (ABM) para investigar la din amica del comportamiento humano en las transacciones va internet. Basado en un estudio emp rico se identi caron factores internos y externos tales como la disposici on a la con anza, el comportamiento impulsivo, la percepci on de seguridad entre otros; los cuales, describen las caracter sticas de una poblaci on compuesta mayoritariamente por estudiantes universitarios. Se utilizo una t ecnica de miner a de datos para generar las reglas de comportamiento y los par ametros que sirvieron de entrada al modelo. Como resultado, se muestra la utilizaci on de un ABM para analizar el comportamiento del consumidor y manejar los cambios din amicos y complejos de un entorno real. Los resultados sugieren que el proceso de adaptaci on a este medio se basa principalmente en el incremento del grado de con anza en las revisiones presentadas en un sitio web conjuntamente con el incremento en el conocimiento sobre los procedimientos y mecanismos de seguridad ofrecidos para realizar transacciones en l nea, los cuales, reducen la tasa de no-compradores.
Keywords
Agent based model,
Purchasing behavior,
Online environment
Cite
Gil-Herrera, E. (2009). A multiagent framework for consumer behavior and purchase intentions in electronic commerce [Thesis]. Retrieved from https://hdl.handle.net/20.500.11801/1508