Information Technology Reference
In-Depth Information
genetic algorithm [2], ant colony algorithm [3], etc. are just a few. For the delibera-
tion part, ever since Von Neumann and Morgenstern's [4] classic expected utility
theory, much development has popped out from both rational theorist and behavioral
scientists [5, 6, 7, 8]. They tackle the decision making problem from the representa-
tion of the preference of an ideal decision maker and identification of the behavioral
principles that human preferences actually obey, respectively. Busemeyer et al. [9]
study decision making from a dynamic cognitive perspective. They paid much more
attention to the dynamic motivational and cognitive process during decision making,
thus their models are good candidates for both modeling and predicating personalized
preference in decision making process especially under risky or uncertainty. Besides,
the computational attribute of their model makes it a good candidate to be incorpo-
rated into other computer systems than the descriptive ones.
From computer decision making system's perspective, scientists have realized al-
most all the searching algorithms, and have incorporated the rational expected utility
theory into their implemented decision support systems. This makes it possible to
provide the general decision support by taking advantage of the large capacity and
huge memory of the computer system. But for personal decision making, in which the
behavioral research plays an important role, computer systems till now have done
quite less except some tracking and simple statistics of the user usage information.
In this paper, we will propose a personal decision making model which incorporate
both the objective feature space searching and subjective feeling deliberating into the
decision making process. Section 2 is an overview of the architecture of the model
and details the different components of the architecture. Section 3 gives a case study
of the proposed architecture in personal fashion decision making system. Section 4
includes some discussions and points out further research directions.
2 Proposed Architecture
The many factors that affect the process and result of a decision making can be rough-
ly divided into two categories: objective factors and subjective factors. For the objec-
tive part, we refer to those factors that are definite and easy to be defined. Due to its
objective nature, they are easy to be formalized and be accepted by computer systems.
We have a relatively better understanding of this part. Models based on these factors
are widely studied and implemented in different areas; but for the subjective parts,
which are relatively less studied, and due to the stochasticity and randomness exist
in them, as far as our knowledge is concerned, less computer decision making systems
have taken much consideration in this aspect. But we would like to argue that the
subjective experiences that accompany our thinking process are informative in their
own right in decision making processes. We therefore can't understand and model
human judgment and decision making without taking the interplay of these subjective
and experiential information into account. But there are many subjective factors like
motivation, feeling, etc. which will be effective in decision making, it is not easy or
almost impossible for us to define which one will play a major role in a specific deci-
sion making process, let alone consider all of them in one decision making model.
