Supporting Public Decision Making: A Progressive Approach Aided by an Ontology

Introduction

This paper deals with the means of supporting economic development-related territorial public decision making at the French Regional Authorities level. The work presented here has been carried out within the framework of a multidisciplinary research project (the CAVALA project: regional economic development policy monitoring and assessment partnered method) that was conducted in partnership with the Midi-Pyrenees Regional Authority. The aim of the project is to help define indicators that can assess economic development policies and more precisely, financial support for companies. The purpose is to design a cooperative method in the construction of these indicators. The project includes the preparation of an initial set of indicators and their corresponding decision-making tools.

The research undertaken presents two facets that differentiate it from the majority of work that has been carried out on helping the decision-making process. On the one hand, the organisations involved here are territorial ‘collectivités’ (authorities) and not companies, whilst on the other hand, above all else the project aims to assist in the process of policy definition (for which the establishment of assessment indicators represents the final phase) and not in the decision-making for a policy which has already been defined. These two facets will have a strong impact on how the project proceeds (an example is given in the following paragraph).

The indicators reflect the objectives for which their role is to measure the outcome. The implementation of indicators therefore supposes that the objectives have been clearly defined, quantified and that they have milestones to reach over time.

By way of illustration, it would be very difficult (or even impossible) to define assessment indicators for an objective coming under the heading ‘long-term employment development’. It would first prove necessary to redefine the objective in question with a view to making it much more precise. This sort of fine-tuning work supposes that the concepts used have been defined. Does ‘long term’ thus mean permanent and/or letting employees gain new skills to guarantee a more secure professional future and/or giving a lasting competitive advantage to the company, etc? In the same way, what kind of employment are we talking about here: direct, indirect, full-time, qualified, unqualified, and so forth? Or again, what do we mean by ‘development’: creating, maintaining, changing, and so forth?

The preciseness of the objectives therefore depends on the preciseness of the concepts used.

And yet, the way in which territorial authority economic policies are expressed is noteworthy because of the inexactitude of the notions, a general lack of stability in the concepts used and the presence of numerous polysemous terms.

Consequently, the work in defining the indicators must necessarily be accompanied by work on the concepts and the organisation thereof.

It is a work of constructing an ontology in adherence with Gruber’s ‘historical’ definition (1993), “An ontology is a formal, explicit specification of a shared conceptualisation”. Along with Uschold and Gruninger (1996) we would furthermore admit that “an ontology necessarily entails or embodies some sort of world view with respect to a given domain.” The world view is often conceived as a set of concepts (e.g. entities, attributes, processes), their definitions and their inter-relationships; this is referred to as a conceptualisation”. The CAVALA project has produced such a conceptualisation in the form of an ontology of territorial economic development, designed to help in defining economic policies, and more particularly, to help in defining assessment indicators.

This paper describes the development of this ontology by focusing on the specific methodological problems that we encountered, especially during the conceptualisation phase (as used by Gómez-Pérez et al. (2004) in the Methontology method). It should be noted that when this ontology was first applied, no automatic machine reasoning tasks were included. The ontology is used essentially to help index the applications for financial support and to underpin the effort needed by political representatives in defining policies (preciseness of the objectives, choices in the implementation, indicators, etc.) through the clarification of the concepts used.

The paper is organised into five parts, where the introduction makes up the first part.