Selecting Visualization Alternative Based on Uncertain Theory

Multiple attribute decision making (MADM) is an efficient way to solve complex systems, and has wide application. This research develops MCDM model based on uncertain theory, used for selecting a suitable visualization alternative for tourism. First, in order to achieve desirable decision making, a new concept is proposed, which is called the best and the worst reference uncertain linguistic variable as a datum uncertain linguistic variable. At the same time, a new method for ranking uncertain linguistic variable is also presented. Second, based on the preference order relation of attributes given by the experts, a new score function is introduced to get the weight vector of attributes. Finally, the evaluation system of tourism big data visualization alternatives is constructed and the order of those alternatives is acquired by the decision method.


Introduction
Big data [1] has been risen as national strategic resources.In recent years, after many developed countries such as Europe, the United States, Japan and South Korea took large data as a national level strategy, China have brought the construction of large data into the national strategic choice.Large datawas written in the government work report in 2014.In January 2015, the National Tourism Administration in the "Guidance on the promotion of the development of intellectual tourism," said: "By 2020, large data mining and intelligent marketing capabilities will be improved significantly, mobile e-commerce, tourism data system analysis, artificial intelligence technology will be applied in the tourism industry more widely.And the main task of building tourism data has been clear many times in the views.
Tourism big data visualization [2] decision-making is an emerging field of the internet age.In the internet context, tourism big data has a large scale and has a high complexity.It is a huge challenge to search, analysis and understand the large number of unstructured data and semi structured data.The value of the rich information behind the hidden is reflected only through the collection, analysis, interpretation and expression.And visualization is the most effective way for people to understand easily the value of complex information data.Therefore, it is significant to study the decision making of tourism big data visualization.It can not only promote the fusion and innovation of data mining, analysis techniques and methods, computer graphics technology and decision theory and methods, but also have a DOI: 10.1051/ , 020 ( 2017) 710002023

2016
MATEC Web of Conferences 100 GCMM matecconf/201 23 transformative influence on thinking and methods of government departments, tourism enterprises and tourists.Therefore, it can provide a more rapid, effective and scientific decision-making protection.
In the process of decision, there are some difficulties for experts to express their preference degrees with crisp numerical values.So, it is another possible way to use linguistic labels [3], which represents qualitative aspects values.We will consider a finite and totally ordered label set " in the usual sense and with odd cardinality, where each label i s represents a possible value for a linguistic real variable.The number of linguistic terms in the set S is called the cardinality of S .In the symbolic computation process, the discrete linguistic set S is extended to a continuous interval , where s is a linguistic label and the numerical value and D represents the value of the symbolic translation [4, 5].
In this paper, the focus is on the extension of discrete support model for MADM [6,7] in which the experts express their opinions by means of uncertain linguistic setting in stead of precise numerical values.In order to get the weight vector, a new method by the definition of score function of attribute and the preference order of attribute presented by different experts is presented.Moreover a new ranking method is presented to rank uncertain linguistic variable, which is based on the best and the worst reference uncertain linguistic variable.

Preliminaries
In this section, the basicconcepts and their extensions of the this paper are briefly introduced.
The deviation measures have been discussed between two uncertain linguistic terms.The following deviation measure for two uncertain linguistic variables is first introduced.
where U is the number of linguistic terms in the set S .then the relative correlation coefficient of i a s is defined as:

Clearly
Obviously, (2) If In the same way, .
is the set of attributes, the score function of attribute

The solution approach
Let } , , , { represents the performance of the alternative i A with respect to the attribute j c .Now, we will give the process of decision based on the mentioned method before.
Step 1.Based on Eq .(4), compute the weight vector of attributes as: where k j s denote the score of j c given by expert Step 2. Based on the decision matrix and Eq.( 1 Step 4.Then, compute the relative correlation coefficient then the bigger the i D , the better the alternative i A , so the overall ranking is acquired.

The establishment of the evaluation
At present, the tourism big data visualization has gradually been widely used in industry.The characteristics of tourism big data visualization alternative can be describes by a group of factor attributes, and experts can evaluate it with these attributes.Based on the National Tourism Administration website and the latest research results, experts summarized the evaluation rules and related factors generally considered in the process of evaluating the tourism big data visualization alternative, then construct the evaluation system, which will be presented as follows: Let be the set of alternatives, where  Suppose there are four experts who give the preference order of attributes as follows:

.
S is called an extended linguistic term set associated with .Let S s , ) (s I is denoted as the position index of s and called the gradation of s in S .For example, of attributes, where 1 c is for the government, 2 c is for the enterprise, 3 c is for the tourism practitioners, 4 c is for the broad masses of the visitor.
After many years 'research analysis by many experts, we can construct the decision matrix as follows: ), for every alternative ,