A New Proposal to Represent the Linguistic Conditional Preference Networks

In this article, we focus on group decision making when preferences are expressed through the LCP-nets (Linguistic Conditional Preference networks). There is a need for complementing efforts and research on mixing LCP-nets together to be able to compute group preferences. First steps are the comparison between two LCP-nets and the aggregation of LCP-nets. In order to compare then aggregate two LCP-nets, we propose a matrix representation that takes into account the number of nodes, the directions and three kinds of arcs. With this simple modeling, computations become very easy and fast. In particular, we develop an adjunction operator called + that permit to aggregate two LCP-nets under certain conditions.