Printer Friendly

Research on cross border e-commerce logistics service based on improved AHP algorithm.

1. Introduction

The rising and the development of the cross-border e-commerce was the product of the financial crisis and the popularity of the Internet technology. With the continuous expansion scale of the foreign trade in China, the cross-border e-commerce developed rapidly. At the same time, the factors of the appreciation of the RMB, the rising labor costs and the serious situation for the foreign trade also promoted the development of the economic commerce. However, the cross-border e-commerce was facing many questions. The most outstanding question was the cross-border e-commerce logistics service (Gomez-Herrera et al., 2014).

The logistics service was always the important question in the electronic commerce. Because the longer distance and the higher logistics cost of the cross-border e-commerce, the cross-border e-commerce logistics had been concerned widely by many scholars (Grubert, 2003). Therefore, evaluating the logistics service of the cross-border e-commerce can help the enterprises to find their own problems and enhance the international competitiveness (Gessner and Snodgrass, 2015).

With the continuous process of the science, there were many evaluation methods (Lai et al., 2015). For example, we use the AHP method (Dong and Cooper, 2016; Hu and Yi, 2016), ANP method (Golcuk and Baykasoglu, 2016), TOPSIS method (Chiu and Hsieh, 2016) and DEA method (Song and Wang, 2014) etc. however, these methods had some defects. Therefore, many scholars combined the traditional methods with other methods and proposed the improved evaluation methods. For example, there were Fuzzy-AHP, AHP-Entory, Grey-AHP and AHP-TOPSIS etc.

In this paper, we establish the evaluation system of the cross-border e-commerce logistics service. Then, this paper also proposes the improved AHP method to evaluate the logistics service of the economic commerce. In the final simulation results, the simulation results of the improved method have more error. It shows that the evaluation index that we establish has the rationality and the effect is ideal.

2. Evaluation index of the cross border e-commerce logistics service

Establishing the evaluation index of the cross border e-commerce logistics service has some principles. These principles are the scientific principle, legality principle, integrity principle, operability principle, quantitative principle and qualitative principle.

Based on the above principle, we establish the evaluation index of the cross border e-commerce logistics service. The indexes are as follows.

3. AHP method

AHP method is a simple, flexible and practical multi criteria decision making method which aiming at the qualitative question for the quantitative analysis. Since 1982, this method was introduced into our country. Then, it has applied in some fields widely. These fields are the energy system analysis, the urban planning, the economic management and scientific research evaluation etc.

The steps of AHP method are as follows.

1. Structuring the analytic hierarchy structure

According to the relationship among the various factors, we divide all the factors into different levels and determine the relation among them. Each superior level has a dominant effect on the lower level. At the same time, the lower level is the refinement for the superior level. The general level can be divided into three categories. They are the target layer, the middle layer and the lowest layer. The target layer usually has only one element. In general, it is the intended target of the analysis problem. The middle layer is the intermediate link for achieving the goals. It may include several levels. The lowest level is the program or decisionmaking rules. It is also the most direct evaluation factors.

2. Structuring the judgment matrix

A = {[a.sub.ij]} (1)

3. Standardizing judgment matrix

[a.sub.ij] = [a.sub.ij]/ [n.summation over (k=1)] [a.sub.kj] (i = 1,2, *** ,n) (2)

4. Calculating the sum of each row for the matrix

[[omega].sub.i] = [n.summation over (j=1)] [a.sub.ij] (i = 1,2, ***, n) (3)

5. Standardizing the sum of each row for the matrix

[[omega].sub.i] = [[omega].sub.i]/[n.summation over (i=1)] [[omega].sub.i] (i = 1,2, ***, n) (4)

6. Calculating the largest eigenvalue of the judgment matrix

A[omega]= [[lambda].sub.max] [omega] (5)

7. Random consistency test of the judgment matrix

The consistency index of the judgment matrix is,

CI = [[lambda].sub.max] - n/n - 1 (6)

Where, [[lambda].sub.max] is the largest eigenvalue of the judgment matrix. Consistency ratio is as follows.

CR = CI/RI (7)

When CR<0.1, we think that the consistency of the matrix is acceptable. The value of the RI is as follows.

4. The improved AHP method

The shortcoming of the AHP method is that the evaluation results are influenced by the subjective factors and the objectivity is poor. At the same time, the operation repeatability of the evaluation is larger. In addition, AHP evaluation method has no accumulation for the work experience. The historical data is difficult to have a positive impact on the subsequent assessment. Therefore, in order to improve the characteristics of the AHP method, we combined the BP neural network model with the AHP method and propose an improved AHP method. The BP network model is composed of the input layer, hidden layer, the output layer and the link connection weight of each layer. The results are shown as follows.

During the iterative process, the BP neural network algorithm is easy to fall into the local minimum and the convergence rate is slow. Therefore, based on the back propagation method, we produce the new weight change. The specific method is that we add the previous weight adjustment amount into the current weight adjustment amount. And we take it as the actual weight adjustment amount.

[DELTA][w.sub.ij](n) = [alpha][w.sub.ij] (n + 1) + (1 - [alpha])[eta][nabla]f ([w.sub.ij] (n - 1)) (8)

Where, [eta] is the learning rate. n is the training times. [nabla][??]([sub.ij](-1)) is the gradient of the error function.

After we adopt the additional momentum method, the adjustment of weight value trend to the average direction of the error surface bottom. When the network weights are into the flat area, it can avoid the occurrence of [DELTA]w=0. It also help it to jump the local minimum of the error surface.

The basic idea of the adaptive learning rate method is that the learning rate can adjusted adaptively according to the chance of the error. During the training process, we adopt the following function of the learning rate adaptive adjustment mechanism. It can increase the stability and improve the convergence speed.


Where, E(n) is the error of the n step.

Then, we establish the improved AHP model. The steps are as follows.

1. We establish the evaluation index.

2. We evaluate the sample data.

3. Standardized data

4. We take the standardized data as the input value of the BP neural network. The evaluation data that obtained by the AHP method is as the target data.

5. Determining the number of the nodes for the input layer, output layer and the hidden layer. Then, we initialize the weight and the thresholds of the neural network nodes.

6. We input the sample data. Then we calculate the errors of the output for each node.

7. Back propagation. According to the size of the set of the momentum coefficient, we modify the weight.

8. Calculating the error. When the error is smaller than the given target error, the neural network training ends. Otherwise, we go back to the step 6.

9. We obtain the training results of the part sample test.

5. Numerical test

We apply the improved AHP method to study the logistics service evaluation for one t cross-border electronic commerce. We take the haven sample data as the initial data. Among them, the first 25 data is as the training data and the last data is as the testing data. We score the logistics service level of the cross-border e-commerce enterprise. Among them, [80, 100] is excellent. [60, 80] is the good. [40, 60] is normal. [10, 40] is poor. [0, 20] is very poor. Firstly, we calculate the weight of each index and the results are shown as follows.

After that, according to the data that obtain by the AHP method, we make the BP neural network training. We set the error precision as [e.sup.-10]. The maximum iterations number is 3000. The step is 0.02. The momentum coefficient is 0.2. The simulation results are shown as follows.

According to the above figure, we can see the evaluation results intuitively. The simulation curve is similar to the actual curve. It shows that the simulation results are ideal.

The specific data are shown in the following table.

From the above table, the maximum relative error of the improved AHP simulation algorithm is 2.30%. It is lower than the general accuracy 5%. It shows that the evaluation of the improved AHP algorithm is good.

6. Conclusion

The normal cloud model is introduced to the distributed data aggregating system to perform data analysis. Inspired by the idea that the quantitative-to-qualitative conversion can be done using the cloud model, the shape of the cloud picture is used to determine the right computing node that should be fed with the data.

At present, all over the world attach the great attention to the electronic commerce of international trade. The cross-border e-commerce is as an important branch of e-commerce. It has become the important means for developing the international trade of the enterprises in our country. In addition, it has become the driving force of adjusting the structure for our country' foreign trade. The cross-border e-commerce reduces the cost of the enterprises and enhances the enterprise's price advantage. At the same time, the cross border e-commerce also enhances the international competitiveness of Chinese enterprises. During the development of the e-commerce, the logistics service has been an important part of e- commerce service and even affects the development of the whole enterprise. In this paper, firstly, we establish the evaluation system of the cross-border e-commerce logistics service. After that, we propose the improved AHP method to evaluate the cross-border e-commerce logistics service. The main work of this paper is as follows. Firstly, this paper establishes the evaluation system of the cross-border e-commerce logistics service. Secondly, this paper introduces simply the AHP method. Thirdly, this paper proposes the improved AHP method. Then it uses the method to evaluate the cross-border e-commerce logistics service. The higher precision of the simulation results show that the method has good operability.

Recebido/Submission: 09/05/2016

Aceitacao/Acceptance: 19/07/2016


Chiu M.C., Hsieh M.C. (2016). Latent human error analysis and efficient improvement strategies by fuzzy TOPSIS in aviation maintenance tasks, Applied Ergonomics, 54: 136-147.

Dong Q.X., Cooper O. (2016). A peer-to-peer dynamic adaptive consensus reaching model for the group AHP decision making, European Journal of Operational Research, 250: 521-530, DOI: 10.1016Zj.ejor.2015.09.016.

Gessner G.Y., Snodgrass C.R. (2015). Designing e-commerce cross-border distribution networks for small and medium-size enterprises incorporating Canadian and U.S. trade incentive programs, Research in Transportation Business & Management, 16: 84-94, DOI: 10.1016/j.rtbm.2015.07.005.

Gomez-Herrera E., Martens B., Turlea G. (2014). The drivers and impediments for cross-border e-commerce in the EU, Information Economics and Policy, 28: 83-96, DOI: 10.1016/j.infoecopol.2014.05.002.

Golcuk I., Baykasoglu A. (2016). Analysis of DEMATEL approaches for criteria interaction handling within ANP, Expert Systems with Applications, 46: 346-366, DOI: 10.1016/j.eswa.2015.10.04i.

Grubert H. (2003). Source Rules, Trade and Income Taxes, and Electronic Commerce: Designing Rules for the Taxation of Cross-Border Income, Economic Analysis and Policy, 33: 237-249, DOI: 10.1016/S03i3-5926(03)500i9-4.

Hu W., & Yi M. (2016). Network reliability study based on analytic hierarchy process (AHP), RISTI - Revista Iberica de Sistemas e Tecnologias de Informacao, 18, 165-175.

Lai P.L., Potter A., Beynon M., Beresford A. (2015). Evaluating the efficiency performance of airports using an integrated AHP/DEA-AR technique, Transport Policy, 42: 75-85, DOI: 10.1016/j.tranpol.2015.04.008.

Song M.L., Wang S.H. (2014). DEA decomposition of China's environmental efficiency based on search algorithm, Applied Mathematics and Computation, 247: 562-572.

Ying Liu

Department of Logistics Technology, Zhejiang Technical Institute of Economics, Hangzhou, China.

Table 1 - The evaluation index of the cross
border e-commerce logistics service

First order      Second          Third level index
index            level index

                                 Cross border e-commerce
                                 logistics model

                                 Cross border e-commerce
                                 logistics strategy

                 Logistics       Cross border e-commerce logistics
                 capability of   function planning
                 cross border
                 e-commerce      Cross border e-commerce
                                 logistics technology

                                 Cross border e-commerce
                                 shop coverage rate

                                 Customs clearance ability

                                 Overseas storage capacity

                                 Online communication ability

                                 Order processing time
The evaluation   quality of      Order tracking
index of the     cross border
cross border     e-commerce      The cross-border
e-commerce       logistics       insurance service provider
service                          The cross-border e-commerce
                                 customer service ability

                                 The cross-border e-commerce
                                 platform registration information

                 Transaction     The cross-border e-commerce
                 security of     platform authentication information
                 cross border
                 e-commerce      The cross-border e-commerce
                                 user privacy protection

                                 The payment of cross-border

                                 The cross-border payment
                                 service provider

                                 Technical support for
                                 cross-border payments

                                 Legal factors

                 Other factors   Regulatory factors

                                 Policy factors

Table 2 - Average random consistency test

Matrix   1   2    3      4      5      6

RI       0   0   0.52   0.89   1.12   1.26

Matrix    7      8      9      10     11

RI       1.36   1.41   1.46   1.49   1.52

Table 3 - Index weigh

Second          Weight   Third level index                   Weight
level index
                         Cross border e-commerce             0.17
                         logistics model

                         Cross border e-commerce             0.17
                         logistics strategy

                         Cross border e-commerce             0.17

Logistics       0.34     logistics function planning
capability of
cross border             Cross border e-commerce             0.15
e-commerce               logistics technology

                         Cross border e-commerce             0.13
                         shop coverage rate

                         Customs clearance ability           0.105

                         Overseas storage capacity           0.105

                         Online communication ability        0.25

                         Order processing time               0.21

Service                  Order tracking                      0.11
quality of
cross border    0.31     The cross-border insurance          0.14
e-commerce               service provider
                         The cross-border e-commerce         0.29
                         customer service ability

                         The cross-border e-commerce         0.12
                         platform registration information

                         The cross-border e-commerce         0.12
                         platform authentication

Transaction     0.28     The cross-border e-commerce user    0.27
security of              privacy protection
cross border
e-commerce               The payment of cross-border         0.20

                         The cross-border payment            0.13
                         service provider

                         Technical support for               0.16
                         crossborder payments

                         Legal factors                       0.33

Other factors   0.07     Regulatory factors                  0.33

                         Policy factors                      0.33

Table 4 - Simulation result

Indicators          26      27      28      29      30

Simulation value    81.38   76.21   54.87   66.50   79.05
Actual value        80.60   76.82   56.13   67.15   80.31
error               0.96%   0.80%   2.30%   0.98%   1.59%
COPYRIGHT 2016 AISTI (Iberian Association for Information Systems and Technologies)
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2016 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Liu, Ying
Publication:RISTI (Revista Iberica de Sistemas e Tecnologias de Informacao)
Date:Sep 1, 2016
Previous Article:A performance analysis model of the ideological and political education in colleges and universities based on the extension classification method.
Next Article:Research on the multi-dimension ideological and political education of college students based on computer platform.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters