The Research of a Reputation Assistant Decision Model in Mobile Commerce Based on Optimal Combination Determining Weights Method

Reputation evaluation in mobile commerce (MC) is a complicated system issue. After analyzing deficiencies of decision model in existing reputation management system, and combining with characteristic of MC, the paper presents a new reputation evaluation model (REM) used in MC. The model improves the reputation fraud prevention ability, for example, the cahoots fraud, the cumulating the reputation fraud which is used in large amount dealing after getting enough reputation from small-value transactions, and so on. AHP method is used to compute pre-transaction user's preference weight, objective weight is determined by using the entropy-weight coefficient method, and an optimal combination weights model is proposed based on maximal deviations. Finally, a MC reputation evaluation case is given to illustrate the availability of the proposed model.