On-Line Shopping Evaluation Based On The Extension Theory Of Type-2 Fuzzy Sets

With the increase in the number of domestic Internet use, the use of the Internet for online shopping and bank card payments are gradually popular, and the market share is also growing rapidly, thus e-commerce sites are endless. E-commerce has penetrated into all areas of our lives. On-line shopping evaluation is a key technology in e-commerce. As consumers’evaluation on commodity is full of subjectivity and uncertainty, type-2 fuzzy logic system based on type-2 fuzzy sets is an effective way to deal with uncertain information. It applies a model based on type-2 fuzzy logic system to on-line shopping recommended system.Type-reduction is an important part in type-2 fuzzy logic system. It is important to solve type-reduction of interval type-2 fuzzy sets. To improve computational efficiency, many algorithms have been proposed one after another. KM algorithm is widely used for the centroid computation of interval type-2 fuzzy sets. Although studies have shown that the iterative process of KM algorithm is monotonic convergence and surplus exponential convergence, KM algorithm still needs to be able to reach the convergence by a number of iterations, and each iteration requires a large amount of numerical calculations. Firstly, it expands the KM algorithm in the matter of large computation amount and long computation time. The proposed algorithm (N-EIASC) improves in starting condition, calculation process and ending condition on the basis of EIASC. And the numerical simulations show that the new algorithm’s calculation efficiency are higher than several existing algorithms. Secondly, for type-2 fuzzy evaluation, an improved type-2 TOPSIS method is proposed, it covers the traditional TOPSIS method’ shortage that cannot deal with uncertain data and introduces type-2 fuzzy sets into TOPSIS method. As existing type-2 TOPSIS method cannot keep type-2 fuzzy sets due to Rank formula, it gives "positive ideal solution" and "negative ideal solution" a new definition, thus it can avoid information lost. And the calculation process is also optimized, it can improve computational efficiency by N-EIASC algorithm It shows its advantage over existing type-2 TOPSIS method by an example. Lastly, type reduction of interval type-2 fuzzy sets and the improved type-2 TOPSIS method are combined and applied to on-line shopping evaluation. It shows that type-2 fuzzy logic system is reasonable to deal with shopping evaluation problems.Applying type-2 fuzzy logic system to dealing with on-line shopping evaluation problems, it not only promotes further theory research but also provides decision support for e-commerce individual service.