The evaluation of mobile health apps: A psychological perception-based probabilistic linguistic belief thermodynamic multiple attribute decision making method

In this paper, we propose a probabilistic linguistic belief thermodynamic multiple attribute decision making method for evaluating mobile health Apps. To extract the evaluation information for mobile health Apps, the quantification transformation of linguistic evaluation of decision makers is critical. However, decision makers may have different perceived values for linguistic terms in their minds. Thus, we deeply investigate the quantification process of linguistic terms based on the psychological perception by introducing prospect theory which can describe the psychological influences of decision makers. Meanwhile, we extend the operational laws of probabilistic linguistic term set (PLTS). In order to efficiently use probabilistic linguistic evaluation information, we further introduce the thermodynamic method to measure the quantity and quality of evaluation information. To process the qualitative rankings for attributes, we propose a non-linear programming model for the determination of attribute weights by improving Borda score which considers the different preferences of decision makers for the rankings. Then, to address the uncertainty situation of probabilistic linguistic evaluation information during the aggregation process, we use the Dempsters' combination rule of the evidence theory. Considering inconsistent of the ranking results of alternatives based on the exergy indicator and the entropy indicator, we construct a comprehensive score by combining the two indicators. Finally, we apply our proposed method to evaluate existing mobile health Apps and validate it by comparison analysis and sensitivity analysis.

Automated summary

This paper proposes a probabilistic linguistic belief thermodynamic method for evaluating mobile health Apps. The method considers decision makers' perceived values for linguistic terms, quantifies linguistic terms based on prospect theory, and uses entropy and exergy indicators for comprehensive evaluation. Through a non-linear programming model, attribute weights are determined, and Dempsters' combination rule is used for uncertain linguistic evaluation information aggregation.