Extending Concordance and Discordance Relations to Hierarchical Sets of Criteria in ELECTRE-III Method

Visual interactive knowledge management for multicriteria decision making and ranking in linked open data environments

“…Hence, it is important to be able to capture and express these profiles. Del Vasto-Terrientes et al proposed an ELECTRE III hierarchical method, in which a problem was broken down into subproblems and permitted the designation of a preference model at each node of the hierarchy [130]. In the before mentioned method, all the criteria except from the root criterion are pseudocriteria, which are structured in a hierarchy composed of many generalization levels.…”

“…Apart from combining the before mentioned disciplines, we propose a novel technique for multifaceted ranking that seamlessly integrates MCDM, visual analytics and Semantic Web (Figure 1). Though there are several methods that successfully rank the alternatives based on multiple criteria, little work has been done in multifaceted ranking [130,131]. Bearing in mind that all the efforts for multifaceted ranking correspond to multiple levels of hierarchy [130,131], there is none approach that clusters the criteria into groups based on their similarity.…”

“…Though there are several methods that successfully rank the alternatives based on multiple criteria, little work has been done in multifaceted ranking [130,131]. Bearing in mind that all the efforts for multifaceted ranking correspond to multiple levels of hierarchy [130,131], there is none approach that clusters the criteria into groups based on their similarity. In an approach that adopts multiple levels of hierarchy in MCDM, a problem is divided into sub-problems that facilitate the specification of a preference model at each node of the hierarchy [130], while another method decomposes a problem into individual sub-problems in different levels of hierarchy [131].…”

“…Bearing in mind that all the efforts for multifaceted ranking correspond to multiple levels of hierarchy [130,131], there is none approach that clusters the criteria into groups based on their similarity. In an approach that adopts multiple levels of hierarchy in MCDM, a problem is divided into sub-problems that facilitate the specification of a preference model at each node of the hierarchy [130], while another method decomposes a problem into individual sub-problems in different levels of hierarchy [131]. However, treating the criteria in groups with related subjects, allows the generation of individual and combinable ranking sub-profiles that are able to shed light to all the separate missions of an entity.…”

“…There are a few methods that involve important to be able to capture and express these profiles. However, the existing multifaceted MCDM methods consider multiple levels of criteria [130,131], rather than clustering of the criteria that we consider in our method. The literature review will be discussed in the section 2.8.1.3 Literature review on multi-faceted MCDM ranking methods.…”