Personalized Decision Support Systems

Introduction

Decision support systems (DSS) are computerized systems that assist humans to make decisions. Early versions were designed for executives, but over time DSSs were designed for workers at any level in the organization (Keen & Morton, 1978; Rockart, 1979). Due to increasing costs in providing benefits and services, organizations are forcing workers and consumers to take increasing responsibility for insurance, health care, and financial planning decisions. Extreme events, such as terrorism, pandemics, and natural disasters will swamp the capacity of governmental agencies to serve their citizenry. Individuals in affected communities must turn to local agencies or ad hoc groups for assistance. Personal decision support systems (PDSS), consisting of databases, model-based expertise, and intelligent interfaces, along with wireless communications, Internet resources, and personal computing, provide sufficient resources to assist informed individuals and groups in solving problems.

This article reviews the typical components of a DSS and the different types of systems that have evolved. The article poses three types of problems facing individuals, including routine problem solving, immediate survival needs, and long-term evolutionary growth. Personal decision support issues of acquiring information, processing information, and dissemination are outlined. Future trends and research opportunities are discussed.

Background

DSS aid human thinking by accessing information, integrating this information in some way, structuring decisions, and optimizing decisions (Sprague & Carlson, 1982). These benefits are obtained using three major system features of a DSS, which include a database, which records knowledge; a model base, which models or represents expertise and problem-solving; and an interface, which provides a user with a means to interact with the other system components (Sprague, 1980).

Powers (2007) characterized DSS in terms of how the system provides assistance. Model-driven DSSs for individuals include spreadsheets. Data-driven DSS, such as Executive Information Systems (EIS), are used by organizations and institutions for strategic and tactical decisions. Communication-driven DSSs can be seen in groupware, video conferencing, and bulletin boards. A document-driven DSS, such as provided by search engines, facilitates document retrieval. A knowledge-driven DSS would be used to solve specialized problems and consist of knowledge represented in terms of rules, procedures, hierarchical frames, or networks. Most recently, web-based DSSs are found in browser searching, intranets, and portal use.

Decision support systems are based on the notion that human reasoning is a rational process, although this is not always the case particularly when humans are faced with complexity and stress (Druzdzel & Flynn, 2000). Experts’ decisions in real settings have been shown to demonstrate less quality than linear models (Hastie & Dawes, 2001). Judgmental heuristics reduce cognitive load but decrease the quality of decisions. Characteristics of the DSS components vary in a PDSS in order to compensate for the type of problems faced by individuals. In general for a PDSS the data bases are customized, the model bases are organized along preferential outcomes (e.g., more or less, quantitative), decisions (e.g., lists and value ordering), and uncertainty (specific actions resulting in gain considering constraints and price).

Personalized Decision Support

This article summarizes three problem types facing individuals, including routine problem solving, extreme survival needs, and long-term change. The article outlines system architecture requirements in terms of acquiring and processing of information, interacting with this information, and the dissemination of information and recommendations.