A consistency-based procedure to estimate missing pairwise preference values

In this paper, we present a procedure to estimate missing preference values when dealing with pairwise comparison and heterogeneous information. The procedure attempts to estimate the missing information in an expert's incomplete fuzzy preference relation using only the preference values provided by that particular expert. Our procedure to estimate missing values can be applied to incomplete fuzzy, multiplicative, interval-valued and linguistic preference relations. Clearly, it would be desirable to maintain experts' consistency levels. We make use of the additive consistency property to measure the level of consistency, and use it to guide the procedure in the estimation of the missing values. Finally, conditions that guarantee the success of our procedure in the estimation of all the missing values of an incomplete preference relation are provided.