Provide Data Analysis Ideas for Small Travel Agencies

2019 2nd International Conference on Informatics, Control and Automation (ICA 2019) ISBN: 978-1-60595-637-4

Provide Data Analysis Ideas for Small Travel Agencies

Wen-rong JIANG

and Yu-lei YANG

Shanghai Second Polytechnic University, China

2_{SANDA UNIVERSITY, China}

*Corresponding author

Keywords: Correlation analysis, Factor analysis, Easy data processing methods, Small travel agency.

Abstract. This paper mainly studies Chinese outbound tourists' preferences in three potential factors (sunch as quarters, source area and travel modes). The data source is provided by a data company, which includes the passenger flow information for the thole China and the Yangtze River Delta region of China in the four quarters of 2017. The purpose of this paper is to show small travel agencies that they can improve their personalized travel products with simple data processing. The SPSS software is a simple operation software. Therefore, this paper apply SPSS software to first determine the correlation between variables by correlation analysis, and then use factor analysis to find the correlation between the original variables.

Introduction

According to the China Outbound Tourism Development Annual Report 2017, we can see that Chinese outbound tourism market continues to grow. The destination consumption behavior in this report shows the incereasing demand for the personalized experience of destinations and the quality of travel services. The satisfaction requirements are gradually increasing. Therefore, for small travel agencies, the importance of launching customized customer-specific customized outbound travel products is self-evident. Furtheremore, Japan, Thailand, and Taiwan China are relatively popular destinations. So this paper uses the passenger flow data to Japan, Thailand and Taiwan China in 2017. Table 1 shows the original data.

𝑋₁- Number of Chinese group travellers

𝑋2- Number of Chinese free travellers

𝑋3- Number of group travellers in the Yangtze River Delta 𝑋4- Number of free travellers in the Yangtze River Delta 𝑋5- Quarter

Table 1. Original data.

Desternation X₁ X₂ X₃ X₄ X₅

Japan 18898 563091 6930 238964 1

Japan 21288 695948 7938 312988 2

Japan 23599 813400 8705 354600 3

Japan 14731 513176 6351 261364 4

Thailand 19981 521253 4174 102727 1

Thailand 17233 436362 3688 93608 2

Thailand 16640 436718 4557 108270 3

Thailand 7721 198643 1771 45877 4

Taiwan China 24684 481874 5250 100952 1

Taiwan China 16871 291160 4662 72954 2

Taiwan China 12269 183045 2787 41165 3

Correlation Analysis

Correlation analysis involves bivariate correlation analysis, partial distance correlation analysis, and distance-related analytical variables. Here is the bivariate correlation analysis. The Calculation method of correlation coefficient includes Person's simple correlation coefficient, Spearman's rank correlation coefficient, and Kendall's tau-b grade correlation coefficient. Because the variables here are not ordered variables, use Person's simple correlation coefficient, as shown in Table 2.

Table 2. Correlation analysis.

X1 X2 X3 X4 X5

X1

Pearson correlation 1 .818** .767** .579* -.714**

Significant (bilateral) .001 .004 .048 .009

N 12 12 12 12 12

X2

Pearson correlation .818** 1 .928** .906** -.384

Significant (bilateral) .001 .000 .000 .217

N 12 12 12 12 12

X3

Pearson correlation .767** .928** 1 .944** -.345

Significant (bilateral) .004 .000 .000 .273

N 12 12 12 12 12

X4

Pearson correlation .579* .906** .944** 1 -.103

Significant (bilateral) .048 .000 .000 .750

N 12 12 12 12 12

X5

Pearson correlation -.714** -.384 -.345 -.103 1

Significant (bilateral) .009 .217 .273 .750

N 12 12 12 12 12

**. Significantly correlated at the .01 level (both sides). *. Significant correlation at the 0.05 level (both sides).

Table 2 shows that the four variables X1, X2, X3, and X4 have strong correlations. Because the Pearson correlation (correlation grade factor, r) of X1, X2, X3, and X4 is greater than 0.5. The significance of X5 and X1, X2, X3, and X4 (p>0.05) indicates that X5 has no correlation with X1, X2, X3, and X4.Therefore, factor analysis is performed on X1, X2, X3, and X4 below.

Factor Analysis

Factor analysis is a method of counting the quantitative relationship between related variables. First of all, we should determine whether it is suitable for the extraction factor. Next, the factor will be extracted. After that which variable is effective in characterizing the factor will be determined. Finally, the score ranking of the factor is calculated so as to the ranking of the destinations can be showed.

Table 3 below shows that the extracted factor has a KMO value greater than 0.5 but less than 0.6, although it is not suitable for extraction, but can still be extracted. When the KMO value is greater than 0.6, it indicates that the factor extraction is generally suitable, and when KMO is greater than 0.7, it indicates that it is suitable.

Table 3. KMO and Bartlett's inspection.

Sampling enough Kaiser-Meyer-Olkin metrics .503

Bartlett's sphericity test

Approximate chi square 131.129

df 6

Sig. .000

Table 4. Total variance of interpretation.

Ingredients Initial eigenvalue Extract square sum loading Rotation square sum loading

Total Variance

%

Accumulatio n %

Total Variance %

Accumulatio n %

Total Variance %

Accumulation %

1 3.483 87.069 87.069 3.483 87.069 87.069 2.359 58.985 58.985

2 .442 11.051 98.120 .442 11.051 98.120 1.565 39.134 98.120

3 .067 1.686 99.805

4 .008 .195 100.000

Extraction method: principal component analysis.

From Figure 1, we can also see that the curve tends to be stable from the third factor, which also means that it is reasonable to extract two factors. This can also be verified from Table 5 below. More than 90% of X4 can be reflected on the first factor, and more than 50% of the X1 element can be reflected in the second factor. At the same time, these two variables are in the respective factors. The proportion is the largest. However, the first factor is better interpreted than the second factor. This can also be verified from Table 5 below. More than 80% of the variables can be reflected on the first factor.

Figure 1. Gravel Map.

Table 5. Composition matrix𝑎.

Ingredients

1 2

𝑋1 .841 .539

𝑋2 .980 .001

𝑋3 .978 -.111

𝑋4 .926 -.373

Extraction method: main component. a. 2 components have been extracted.

Table 6 below is a table of rotational component matrices. This table shows that the first factor characterizes the variable X4, and the second factor is used to characterize the variable X1.

Table 6. Rotation component matrixa.

Ingredients

1 2

𝑋1 .277 .960

𝑋2 .796 .571

𝑋3 .879 .435

𝑋4 .980 .180

Extraction method: main component.

Rotation method: Orthogonal rotation method with Kaiser standardization. a. The rotation converges after 3 iterations.

Figure 2. Composition diagram in rotating space.

Table 7 shows that the first factor 𝐹₁ = −0.546𝑋₁+ 0.223𝑋₂+ 0.376𝑋₃+ 0.724𝑋₄and the second factor 𝐹₂ = 1.115𝑋₁+ 0.172𝑋₂− 0.0.29𝑋₃− 0.509𝑋₄ .Plus, we can calculate F =

0.87069𝐹₁+ 0.9812𝐹₂.And the coefficients of 𝐹₁ and 𝐹₂ are their variance contribution rates (in Table 5).Thus, we can see the ranking in the table 8.

Table 7. Component score coefficient matrix.

Ingredients

1 2

X1 -.549 1.115

X2 .223 .172

X3 .376 -.029

X4 .724 -.509

Extraction method: main component.

Rotation method: Orthogonal rotation method with Kaiser standardization. Composition score

Table 8. Ranking.

Destination Quarter Rank

Japan 3 1

Japan 2 2

Taiwan China 1 3

Japan 1 4

Thailand 1 5

Japan 4 6

Thailand 3 7

Thailand 2 8

Summary

Although the factor of quarter was not considered in the factor analysis, the rankings of destinations and travel modes were taken to the rankings of Japan, Thailand, and Taiwan. For small trips, the following improvements can be made to the three-purpose products: 1. Multi-organizational products for Japanese tourists, regardless of the quarter, Japan is a popular destination choice. 2. It is possible to launch products targeting Thailand in the first, second and third quarters. 3. For the travel portfolio to Taiwan, you can focus on the first quarter. However, compared with Japan and Thailand, the proportion of Taiwan as a destination can be less. From Table 7, we can see that China, Taiwan in the second, third and fourth quarters are in a relatively backward ranking.

Acknowledgements

This work is supported by the Key Disciplines of Computer Science and Technology of Shanghai Polytechnic University under Grant No. XXKZD1604.

References

[1] Wenrong Jiang, Wenting Shao. Analysis and Evaluation of Development Degree of the Special Township in a Prefecture-level City[J]. 2018 2nd International Conference on Applied Mathematics, Modelling and Statistics Application (AMMSA2018), Sanya, China, ATLANTIS Press. May 28, 2018, 265-270.

[2] Cao Yuru, Research on Applicability of Regression Analysis Conditions Based on SPSS Weighted Regression, [J].Statistics & Decision, 2019, 35(04):89-92.

[3] Huang Wenxia, Li Min. Analysis of the main factors affecting the development of tourism areas based on SPSS data analysis [J]. Software, 2019, 40 (01): 144-149.

[4] LI Yan.Influencing factors and countermeasures of marine tourism development in Beibu Gulf based on grey correlation analysis[J].Journal of Southwest China Normal University(Natural Science), 2019,44(01):56-61.

[5] Ren Yujie, Ma Kun, Tang Xiaotong, Liu Cao. Selecting points and heat analysis of tourism destination clusters based on LBSN big data[J].Bulletin of Science and Technology, 2019, 35(01):94-100.

[6] Xiao Jie. Application Analysis of Big Data in Tourism Management [J]. Vacation Travel, 2018 (12): 104-105.

[7] Outbound tourism report: tourism consumption tends to be rational. The number of people going along the Belt and Road is rising [J]. Air Freight Business, 2017 (11): 20-22.

[8] Liu Qiuyan, Wu Xinxin. Review on the Method of Determining Index Weights in Multi-element Evaluation[J].Knowledge Management Forum, 2017, 2(06):500-510.

[9] Maryse Boivin, Georges A. Tanguay. Analysis of the determinants of urban tourism attractiveness: The case of Québec City and Bordeaux[J]. Journal of Destination Marketing & Management, 2019, 11.