Towards an e-commerce recommendation system based on MCDM methods

The expanding assortment of products and devices available on the market is associated with dynamic technological development. Consequently, the number of parameters taken into account during purchasing decisions is also growing. It causes the choice of the product to be purchased, a multidimensional problem that is difficult to solve and requires a consumer to consider many criteria. The presented paper proposes an approach incorporating multi-criteria decision-making (MCDM) methods, which may serve as a helpful instrument in consumer choice support systems during purchases made on e-commerce websites. The research includes applying an approach to solve a multi-criteria monitor selection problem based on four MCDM meth-ods. Two of them, Characteristic Objects METhod (COMET) and Stable Preference Ordering Towards Ideal Solution (SPOTIS) method, are new methods that are rank reversal free. Technique for Order of Preference by Similarity to the Ideal Solution (TOPSIS) and Vlse Kriterijumska Optimizacija I Kompromisno Resenje (VIKOR), on the other hand, are methods that are well known and have been widely used for many years and were used as reference methods. The obtained results prove that the final rankings depend on the algorithm of the MCDM method used and confirm the need for benchmarking in investigations of multi-criteria problems. The received outcomes indicate that the proposed approach seems to be a promising solution for consumer decision support systems.