This study aims to develop a framework for data source analysis in multicriteria decision models to support the choice of the best data source for a given context and use this definition to adequately deal with uncertainty. The proposed framework was applied in conjunction with the PROMETHEE II multicriteria method to solve a decision problem of selecting a steel supplier in the Brazilian construction company.
Data source analysis is used in the modeling phase. The PROMETHEE II multicriteria method was used to evaluate the suppliers. Then, a sensitivity analysis based on a combination of a Monte Carlo simulation and scenario analysis was conducted, with scenarios created by randomly varying the criteria weights and/or performances of alternatives within a determined percentage.
This paper presents a structured framework to help decision-makers track down some interesting data sources and relevant experts and decide how to deal with uncertainty based on the available data source in the company’s context, applied to supplier selection in a small construction company.
The proposed framework is based on a data source analysis that investigates, for each criterion, the possibility of experts participating in the decision-making process by providing information or the possibility of using other data sources. It analyzes the relationship between the context of the company and data sources. Based on this analysis, dealing with uncertainty can be directed toward approaches at the beginning of the decision-making process with the experts and/or after the recommendation provided by the model.
