A Multiple Objective Decision Support Tool (MODS)

Software · Sun Dec 14 00:00:00 EST 2003

OSTI ID:1230736

The Multiple Objective Decision Support (MODS) tool is an automated tool used to assist decision makers and policy analysts with multiple-objective decision problems. The classes of problems that this decision support tool addresses have both multiple objectives and multiple stakeholders. Decision problems, which have multiple objectives that in general cannot be maximized simultaneously, and multiple stakeholders, who have different perspectives about the relative importance of the objectives, require analytic approaches and tools that can provide flexible support to decision makers. This tool provides capabilities for the management, analysis, and graphical display for these types of decision problems drawn from diverse problem domains. The MODS tool is a unique integration of analysis algorithms, an information database, and a graphical user interface. This collection of algorithms, the combination of an information database with the analysis into a single tool, and the graphical user interface provides a technically advanced tool to decision makers and policy analysts. There are two main issues when addressing problems of this type: what set of attributes should be used to characterize the tokens in the domain of interest, and how should the values of these attributes and their weights be determined and combined to provide a relative ordering to the tokens. This tool addresses both of these issues. This decision support tool provides a flexible way to derive and use a chosen set of attributes. For example, the tool could be used to first perform a paired comparison of a large set of attributes and from this evaluation select those attributes that have the highest weights. The flexibility of the tool allows experimentation with various attribute sets and this capability, along with domain expertise, addresses the first issue. To address the second issue, several algorithms have been implemented. For example, two algorithms that have been implemented are the geometric mean and the product of the weighted attributes. These, and other algorithms, are described in detail in the following sections. The algorithms can be selected easily and various evaluations can be compared. The tool also provides flexible methods for assigning weights to the attributes. A range of values can be assigned to an attribute when there is uncertainty as to which specific value should be assigned.

• Short Name / Acronym: MODS; 001740IBMPC00
• Site Accession Number: MODS Ver. 1.0a 11/25/03
• Version: 00
• Programming Language(s): Medium: X; OS: Any system with Java Runtime Environment 1.4 or later; Compatibility: PC
• Research Organization: Sandia National Laboratories (SNL), Albuquerque, NM, and Livermore, CA (United States)
• Sponsoring Organization: Not Specified
• Contributing Organization: David L. Harris, David M. Teter
• DOE Contract Number: DE-AC04-94AL85000
• OSTI ID: 1230736
• Country of Origin: United States