Selecting products through text reviews: An MCDM method incorporating personalized heuristic judgments in the prospect theory

Online reviews have become an increasingly popular information source in consumer’s decision making process. To help consumers make informed decisions, how to select products through online reviews is a valuable research topic. This work deals with a personized product selection problem with review sentiments under probabilistic linguistic circumstances. To this end, we propose a multi-criteria decision making (MCDM) method incorporating personalized heuristic judgments in the prospect theory (PT). We focus on the role of personalized heuristic judgments on review helpfulness in the final decision outcomes. We demonstrate the consistency between the three common heuristic judgments (with respect to review valence, sentiment extremity, and aspiration levels) and the three behavioral principles of the PT. Then, the products are ranked with the probabilistic linguistic term set (PLTS) input, based on the proposed adjustable PT framework, in which the coefficients of negativity bias are derived from the consumer’s heuristic judgments. Finally, a real case on TripAdvisor.com and two simulation experiments are given to illustrate the validity of the proposed method.