We used the internet source of http://auto.consumerguide.com as it is a "trusted resource helping smart consumers make buying decisions with confidence." Our target user is the typical owner of a new vehicle. The ratings used in the system come from the evaluations by experts looking at how the new vehicle meets everyday needs - from riding comfortably over potholes to how difficult it is to adjust the controls while driving. The questions posed, therefore, to the system user are preference questions with no technical specifications given.
The decision making process is a difficult one. It would make life easier if all problems were highly structured and an optimal solution was known. But life is not that simple. The problem of choosing a new car to purchase is a semi-structured problem, which allows a decision support system, such as ours, to aid a user in his/her final decision. It does not simply give the best solution, but it gives a range of alternatives using the information available, leaving the final decision making to the user. The problem is also a form of management control, meaning user wants to obtain the best vehicle according to his/her preferences, within a suitable price range.
Given the complexity and constraints of this problem, a decision support system is an excellent means to an end. It offers better decision quality, time savings and cost reduction. It has higher decision quality as it supports the entire decision making process from providing all the raw data available, to analyzing preferences, to suggesting solutions. It also reduces cost and time by presenting everything in one place. This collection of information intended to eventually give a solution to the car choice problem is our system.
Every system can be divided into three parts, and our system is no exception. The inputs are the user preferences. There are 15 different user preferences all designed to get an accurate picture of the user and his/her needs when it comes to a new vehicle.