Majority OWA operator for opinion rating in social media

Abstract

In recent years, the acquisition of products and services via Internet has grown exponentially. Today, a person goes through different websites in search of better alternatives and prices. In this searching process, he/she uses social media applications to examine the opinions of others who have purchased such services or products. Also, the seller offers summarized information of buyers’ valuation by a satisfaction value on a predefined scale (expressed with a system of stars or linguistic labels). This number is calculated using an arithmetic mean and, in many cases, does not adequately reflect all users’ opinions. In this article, a new consensus operator called selective aggregated majority ordered weighted averaging (SAM-OWA) is proposed. It is based on the opinions of the majority though also considering the opinion of the minority. SAM-OWA calculates the individual weights of each value of the satisfaction scale, using the number of votes obtained (cardinality). The mathematical foundation of SAM-OWA operator and some application examples to improve the results obtained by the arithmetic mean are presented.