1 Introduction

This document proposes a representation for measures of effect (also called measures of merit) of objectives in air campaign plans. This representation is based on discussions with CHECKMATE. More specifically, its development is part of the ISI work in the continuing development of INSPECT (INtelligent System for air campaign Plans evaluation based on ExpeCT). INSPECT is a knowledge-based system developed using ISI's EXPECT framework. The system checks the consistency of air campaign plans with respect to several criteria. INSPECT provides consistency checking for operational plans developed (or modified) using ACPT. For a detailed description of INSPECT, see [Valente et al., 1996a].

The term measure of effect should not be confused with measure of effectiveness. The former, is a synonym for measure of merit, something that defines when an objective has been achieved -- that is, how to measure the effect of an objective/task. In contrast, a measure of effectiveness relates to the performance in a campaign, i.e. in executing an air campaign plan. Typical measures of effectiveness refer to sortie rates, loss exchange ratios, repair rates, etc. All these elements help defining how well certain forces or units are performing in a campaign (but say nothing about whether or not they are achieving their objectives).

The representation proposed in this document is a continuation of work in the representation of objectives in air campaign plans, initially documented in [Valente et al., 1996b]. The basic idea is to provide a richer and more structured representation for air campaign plan objectives than the one currently used. It must be noted that we do not make any attempt to define methodological guidelines regarding the development of measures of effect, or what constitute ``good'' measures of effect. The proposed representation is basically syntactical in character (even though it hints at the semantics of MOEs).

This document assumes some familiarity with the air campaign planning domain, and the tools currently used to support it, like ACPT and INSPECT. Further, it assumes knowledge of the representation for ACP objectives proposed in [Valente et al., 1996b]. This document is structured as follows. In Section 2 we briefly describe the background and rationale for the proposed representation. In Section 3 we present the representation, and show how a number of examples of MOEs would be expressed using it. In Section 4 we discuss several issues raised in constructing the representation, and how they have been solved through discussions with Checkmate. In section 5 we discuss briefly the role measures of effect can play in connecting Intelligence and Operation groups in air campaign planning. Section 6 contains the specification of the representation as a Loom ontology (definitions). Section 7 contains the specification of an IDL schema to store measures of effect using the structured representation proposed here.

Acknowledgements

This document is the result of many discussions with Major Alred and Major Cunico, of Checkmate, who I would like to thank very much for their invaluable and continuous help and support. The views expressed here do not necessarily agree with theirs; in particular, all mistakes are my sole responsibility.