Reaching a decision when multiple, possibly conflicting, criteria are taken into account is often a difficult task. This normally requires the intervention of an analyst to aid the decision maker in following a clear methodology with respect to the steps that need to be taken, as well as the use of different algorithms and software tools. Most of these tools focus on one or a small number of algorithms, some are difficult to adapt and interface with other tools, while only a few belong to dynamic communities of contributors allowing them to expand in use and functionality. In this paper, we address these issues by proposing to use the R statistical environment and the MCDA package of decision aiding algorithms and tools. This package is meant to provide a wide range of MCDA algorithms that may be used by an analyst to tailor a decision aiding process to their needs, while the choice of R takes advantage of the yet poorly explored opportunity to interface data analysis and decision aiding. We additionally demonstrate the use of this tool on a practical application following a well-defined decision aiding process.

International audienceIn this paper we describe diviz, a software for modeling, processing and sharing Multiple Criteria Decision Aid (MCDA) techniques and experiments. One of its main features is the possibility to design complex MCDA algorithms by combining elementary calculation components via an intuitive graphical user interface. The target user community of diviz reaches from researchers to teachers, students and MCDA practitioners. The diviz software also eases the comparison of results produced by different algorithms on the same decision problem instance and facilitates the dissemination of MCDA methods and experiments