Information Technology Reference
In-Depth Information
Combine the Objective Features with the Subjective
Feelings in Personal Multi-alternative Decision Making
Modeling
Jiyun Li
1
and Jerome R. Busemeyer
2
1
School of Computer Science and Technology, Donghua University,
No. 1882, West Yan'an Road, Shanghai, the People's Republic of China, 200051
2
Psychology and Brain Science Department, Indiana University, Bloomington, IN 47405
jyli@dhu.edu.cn, jbusemey@indiana.edu
Abstract.
In this paper we propose a computer modeling framework for per-
sonal multi-alternative decision making. It integrates both the objective feature
space searching and evaluation and subjective feeling space deliberating and
evaluation process. A case study on fashion decision making is given out as an
example. It shows that the proposed model outperforms the currently widely
studied ones in terms of prediction accuracy due to the consideration of the sto-
chastic characteristics and the psychological effects that occur quite often in
human multi-alternative decision making.
1 Introduction
The kernel of a computer recommendation system or computer aided decision making
system is the model of decision making and preferences of the decision maker. Most
of the decision making processes in them can be taken as a multi-alternative decision
making process. Economists use preference in rational decision making to interpret
decision making activities in terms of preference, looking at the effects or influences
of preferences in making reasoning faster and introducing new decisions. While the
focus of economists on “nice” preferences satisfying the properties [1] such as Cohe-
rent, Consistent, Complete, Comparatively constant, Concrete, Conveniently cleaved,
completely comparative, etc., many studies in behavioral psychology have suggested
many irrational effects exist in human judgment and decision making practice. These
are important factors for personal decision making modeling. A good computer model
for personal decision making should both help the decision maker make quick and
reasonable decision, and in the mean time, make the decision according to his 'own'
taste. In order to achieve this goal, we should have a better understanding of human
decision making process, especially the personal factors' part.
Generally speaking, decision making process can be divided into three processes,
namely, search, evaluation and deliberation. Both the search and evaluation strategy
and deliberation and evaluation process have been carefully studied separately all
these years. Many models and algorithms have thrived out. In the searching field,
