An evaluation study of rural scientific and technological talents based on TODIM method with hybrid indicator

Abstract

A scientific evaluation model can be effectively used for the evaluation of regional talent development level. This paper proposes a set of scientific index systems for evaluating rural science and technology talents, which can be used for understanding the development status and level of rural science and technology talents in various regions; putting forward the corresponding talent cultivation and introduction policies, and; promoting the development of rural economic construction. Moreover, in order to avoid the shortcoming of over-subjective indicator weight in analytic hierarchy process (AHP), this paper uses the entropy weight method to determine indicator weight. Furthermore, giving the fact that the evaluation experts may have individual personal preferences, this paper proposes an extended TODIM method based on hybrid index values, for achieving more scientific and effective evaluation results of rural science and technology talents. Finally, the proposed methods are evaluated on an actual case, where relevant analysis and suggestions are given.

Keywords

Rural scientific and technological talents TODIM method entropy weight method hybrid indicator multi-attribute decision-making

1 Introduction

To implement the strategy of rural revitalization, talents are the key, while farmers are the mainstay. Rural revitalization requires not only cadres at all levels who are willing to work hard but also rural scientific and technological talents. As an important part of rural talents, rural scientific and technological talents play an important role in many aspects, such as the driving force of agricultural science and technology progress, the backing of agricultural modernization development, the guarantee for accelerating the construction of new rural areas in each country, and the foundation of agricultural economy development. However, through an investigation of the current status of agriculture, rural areas, and farmers in China, we have found that the country’s rural science and technology penetration rate is relatively low and farmers’ overall scientific and technological quality is generally poor. In order to solve these problems, there are 5 aspects we need to consider: understanding the current status and quality level of rural scientific and technological talents in various regions; purposefully strengthening the cultivation and development of rural scientific and technological talents; proposing corresponding talent introduction policies and evaluation incentive mechanisms; accelerating the construction of rural scientific and technological talents, and; enhancing rural economic development level.

In the construction process of rural scientific and technological talents, it is crucial to manage and evaluate the talents. A sound rural talent evaluation mechanism can give full play to the role of talents, fully mobilize their enthusiasm, and enable them to serve rural areas in a targeted and planned manner. There are many research works on the evaluation theory of rural talents. Guo [1] believed that the incomplete talent evaluation and incentive mechanism was an important factor for restricting the construction of rural talent teams. At present, China has not really established and perfected a rural practical talent evaluation system with specific classification and scientific indicators. Li [2] pointed out that the establishment of a talent evaluation mechanism was the prerequisite of both the talent competition mechanism and talent supervision mechanism; specifically, in order to establish a solid evaluation mechanism for rural scientific and technological talents, there are several evaluation measures needing to be considered, such as clear evaluation focus, high-quality professional evaluation teams, and scientific evaluation methods. Lee and Cho [3] evaluated the importance of education programs to the development of rural tourism human resources, and believed that the content of education practice was the most important evaluation factor for improving the practical operation ability of rural tourism talents.

In order to evaluate rural scientific and technological talents, it is crucial to determine the evaluation indicator system. For example, Chen [4] believed that the scientific rural talent evaluation mechanism should adhere to the principle of both integrity and ability. That is, it is necessary to focus on evaluation of talents through practice, and establish a variety of rural talent evaluation indicator systems that focus on performance and consist of elements such as character, knowledge, and ability. Wu [5] believed that rural talent evaluation needed to combine qualitative evaluation and quantitative evaluation, to further highlight the evaluation of character, ability, and performance, and to establish a talent classification evaluation mechanism and system oriented to the quality, contribution, and performance of scientific and technological innovation. In order to evaluate the ability of talents, Mahdi et al. [6] divided the indicators into three main indicators of professional ability, innovation ability and social ability, where each primary indicator was specifically divided into four secondary indicators, and each secondary indicator was further divided into four tertiary indicators. However, this evaluation indicator system was far from the rural talent evaluation system. Wu et al. [7] built a more scientific and reasonable evaluation indicator system (combining the competency model) for rural science and technology talents, which included three first-level indicators, personal literacy, work performance, and performance effectiveness, as well as multiple second-level and third-level indicators. Chen and Sun [8] provided suggestions on the construction of local engineering talent evaluation systems in three aspects, which included the conditional dimension, the talent manifestation dimension, and the talent potential dimension. Through the analysis of the above literatures, it is found that although the scholars propose some indicators to evaluate rural talents, there are still some problems and limitations in above existing works, such as unclear evaluation standards, non-scientific evaluation indicators, and invalid evaluation results (cannot be calculated from mathematics). Thus, the existing evaluation methods are too objective cannot achieve the scientific and fair evaluation results due to above problems/limitations.

Talent evaluation is a typical multi-attribute decision-making problem [9 –11], which evaluates or sorts the alternatives based on multiple index information. At present, the research of the evaluation methods of talents mainly focuses on aggregate operators, TOPSIS and PROMETHEE. For example, Boran et al. [12] proposed a talent evaluation model based on intuitionistic fuzzy sets and TOPSIS method, and applied this model to the evaluation of manufacturing management talents. Afshari [13] proposed a talent evaluation model based on fuzzy linguistic terms and fuzzy integrals that considered the dependencies between evaluation indicators. Through this model, the fuzzy integral values of the talents can be calculated, based on which the talents can be evaluated. Yu et al. [14] explored the prioritized aggregation method based on hesitant fuzzy elements, and proposed a new hesitant fuzzy group decision method to perform talent evaluation for overcoming the shortcomings of traditional talent evaluation methods. Dejan and Slavica [15] argued that talent evaluation was a multi-attribute decision problem; therefore, they proposed an integrated approach to solve the talent evaluation problem, by combining the TOPSIS method and the PROMETHEE method. In addition, it should be noted that the method of determining the weights of evaluation indicators is also worthy of attention. But the existing AHP method [16] focused on the use of subjective weights, and the determination method was relatively simple, which made the AHP method not suitable for complex evaluation indicator systems. For the weights of quantitative and qualitative evaluation indicators, we can determine them by further considering the ambiguity of the evaluation indicators, and by adding some objective weight calculation methods. In addition, there are still some shortcomings in the above methods. For example, the TOPSIS method is difficult to give a clear explanation for data with relatively small differences, and may obtain different ranking results with different closeness formulas. Though the PROMETHEE method has a certain advantage in dealing with the evaluation problem with a small number of indicators, its computation complexity will be very high, when the number of indicators is large. Thanks to the continuous development of decision-making science, some multi-attribute decision-making methods can be applied to the evaluation of rural scientific and technological talents. Specifically, the TODIM method [17, 18] is more suitable for decision-making evaluation in reality, and it has some advantages as follows. First, the TODIM method is a mature evaluation method that considers the psychological behavior of decision makers based on prospect theory; uses the descriptions provided by people based on experience, and; makes evaluation that fits with the decision makers’ decision psychology according to the risk attitudes of different decision makers. Moreover, the TODIM method is more flexible, since it can enable the decision makers to choose different decision parameters according to different decision situations. Furthermore, the TODIM method can achieve more scientific and reasonable decisions, by avoiding the problems of TOPSIS and PROMETHEE methods.

Although TODIM method has been widely used in real applications, such as logistics evaluation [19], ecosystem health evaluation [20] [21], product selection [22], opinion polls [23], and CCUS storage site selection [24], to the best of our knowledge, there is no research finding(s) on the practical application of talent assessment. In addition, the existing TODIM methods can only deal with a single type of fuzzy information [25 –29], and cannot handle decision situations where the attribute values are hybrid fuzzy information. Given that the evaluation of rural scientific and technological talents is a complex multi-attribute evaluation problem with national or regional characteristics, this study is mainly engaged in the following aspects of the original work: (1) constructing a set of evaluation indicator systems of rural scientific and technological talents through literature review and field investigation, which contains various types of hybrid indicators; (2) introducing an entropy weight method to determine indicator weights based on hybrid indicator information; (3) proposing a TODIM model for solving the evaluation problems of hybrid indicator system, and; (4) employing the proposed method to solve the evaluation problem of actual regional rural scientific and technological talents, to analyze the regional talent development problems, and to put forward relevant suggestions.

In this paper, we make the following contributions. We builds an evaluation indicator system for rural science and technology talent based on the hybrid evaluation indicators in Section 2. We introduce some relevant theoretical knowledge and an entropy weight method for the proposed evaluation indicator system to determine the indicator weight is introduced in Section 3, and propose a TODIM model to evaluate rural science and technology talents in Section 4. In Section 5, we present an actual case study and some corresponding policy recommendations. Section 6 concludes this paper with some suggestions.

2 The evaluation system of rural scientific and technological talents

A scientific and reasonable indicator system is the premise and foundation for the effective evaluation of our China’s rural scientific and technological talents. In order to construct the evaluation indicator system for rural science and technology talents, we should conform to the basic characteristics of rural science and technology talents such as scalability, high value, and creativity, follow the regularity of talent evaluation, and adhere to the combination of principle and flexibility. Moreover, the indicator system should aim at practicability and operability; be guided by ability and performance; transfer the leader role from the government to society; combine subjective evaluation and objective evaluation; scientifically and effectively evaluate the rural scientific and technological talents, and; provide talent guarantee and intellectual support for rural talent promotion strategies. Based on the existing literature, this paper conducts questionnaire surveys of talent management departments and experts in many regions, analyzes the common quality of rural scientific and technological talents, and builds an evaluation indicator system based on three elements, which are basic condition, ability performance, and moral character, respectively. Specifically, the basic condition has three evaluation indicators: talent quantity, cultural level, and moral quality; the ability performance also consists of three evaluation indicators: work ability, cognitive ability, and innovation ability; and the moral character constitutes three evaluation indicators: the amount of research achievements, and social and economic benefits created by scientific and technological talents for the region. For quantitative indicators, the source data mainly refers to “China Rural Statistical Yearbook”, “China County Statistical Yearbook (Township Volume)”; for qualitative indicators, the source data is provided by relevant department leaders and experts. The specific system framework is shown in Table 1.

Table 1
Comprehensive evaluation indicator system of rural scientific and technological talents

Target Elements Evaluation indicators Indicator types Indicator properties

evaluation indicator system of rural scientific and technological talents Basic condition Talent quantity benefit quantitative

Cultural level benefit qualitative

Moral quality benefit qualitative

Ability condition Work ability benefit qualitative

Cognitive ability benefit qualitative

Innovation ability benefit qualitative

Performance Research achievement benefit quantitative

Social benefit benefit qualitative

Economic benefit benefit quantitative

Target	Elements	Evaluation indicators	Indicator types	Indicator properties
evaluation indicator system of rural scientific and technological talents	Basic condition	Talent quantity	benefit	quantitative
		Cultural level	benefit	qualitative
		Moral quality	benefit	qualitative
	Ability condition	Work ability	benefit	qualitative
		Cognitive ability	benefit	qualitative
		Innovation ability	benefit	qualitative
	Performance	Research achievement	benefit	quantitative
		Social benefit	benefit	qualitative
		Economic benefit	benefit	quantitative

3 The basis of relevant theory knowledge

3.1. Definition of different indicators

Different types of evaluation indicators can be expressed by different data types. For example, quantitative indicators can be expressed by data types such as real numbers, interval numbers, and 6666 etc.; qualitative indicators can be expressed by data types such as linguistic variables, hesitant fuzzy linguistic elements (HFLEs), and etc. We briefly introduce the relevant definitions of interval numbers and HFLEs as follows.

(1) Interval Numbers

Definition 1: given a real number set R, for arbitrary a^l, a^u ∈ R, if a^l ⩽ a^u, we call the closed interval A = [a^l, a^u] as an interval number. If we have its membership function a (x) : R → [0, 1] as follows: $a_{A} (x) = {\begin{matrix} 1, x \in (a^{l}, a^{u}) \\ 0, x \in (- \infty, a^{l}) \cup (a^{u}, \infty) \end{matrix}$ (1) we set I (R) as the set of all interval numbers. Intuitively, for any A ∈ I (R), if a^l = a^u, we can obtain that A is an ordinary real number. Therefore, we have R ⊂ I (R). That is, the interval numbers are a generalization of real numbers.

(2) Hesitant Fuzzy Linguistic Term Sets (HFLTSs)

Definition 2 [31]: let S = {s_t|t = - τ, ⋯ , -1, 0, 1, ⋯ , τ} be the set of linguistic terms, x_α ∈ X, α = 1, 2, ⋯ , N, the hesitant fuzzy linguistic set can be represented in a math manner as: $H_{S} = {〈 x_{α}, h_{S} (x_{α}) 〉 | x_{α} \in X}$ (2) where h_S (x_α) : X → S is a mapping function of the membership degree from linguistic variable x_α ∈ X to linguistic term set S, h_S (x_α) ={ s_{φ
_l} (x_α) |s_{φ
_l} (x_α) ∈ S, l = 1, ⋯ , L } is a set of potential values of linguistic term set S, and L is the number of terms in h_S (x_α). The terms in the HFLTS are sorted in an ascending order. For the sake of simplicity, we use h_S (x_α) to denote the HFLE, and H_S to represent the set of all possible HFLNs in linguistic term set S.

According to the definition of the HFLEs, different HFLEs may contain different numbers of linguistic terms. In order to eliminate the influence caused by the difference in the number of linguistic terms and to better compare the two HFLEs, Zhu and Xu [32] proposed a corresponding method to add linguistic terms to HFLEs.

Assume h_S ={ s_{φ
₁}, ⋯ , s_{φ
_L} } as a HFLE, where L is the number of linguistic terms, and s_{φ
^*} and s_{φ
^-} are the best and worst linguistic terms in the HFLE h_S, respectively. We can add linguistic terms to the relatively short HFLEs as follows. ${\bar{h}}_{S} = ζ φ^{*} + (1 - ζ) φ^{-}$ (3) where φ^* and φ^- are the subscripts of linguistic terms s_{φ
^*} and s_{φ
^-}, respectively, and ζ ∈ [0, 1] is an optimization parameter related to the decision-markers’ attitude towards risk (indicating the preference of decision markers).

In order to compare the size of two HFLEs, Liao et al. [33] defined a scoring function of HFLEs as follows: $S (h_{S}) = \frac{\sum_{l = 1}^{L} φ_{l}}{L}$ (4) where φ_l is the subscript of linguistic term s_{φ
_l} in HFLE, and L is the number of linguistic terms in HFLE. Intuitively, larger function values of the HFLE indicate the better HFLE.

3.2 Standardization of hybrid indicators

As shown in Table 1, evaluation indicators can be divided into two categories, quantitative indicators and qualitative indicators, each of which has corresponding individual metrics. Meanwhile, for indicators with the same nature (qualitative or quantitative), they may have quite different measurement forms. Taking quantitative indicator as an example, the talents and scientific research achievements is measured by “number”, while the economic benefit of the region is measured by “hundred million yuan”. Thus, it is important to consider different indicator types in the process of evaluating rural scientific and technological talents. Moreover, indicators themselves have incommensurability and contradictions, as well as different dimensions and orders of magnitude. Therefore, we first need to standardize the original evaluation data, and then analyze and sort the evaluation objects. Assume that there are m evaluation objects and n evaluation indicators, where y_ij represents the evaluation value of the i-th evaluation object under the j-th evaluation indicator. Thus, the initial evaluation matrix can be written as Y = [y_ij] _m×n. Different types of indicator data have different standardization methods. In the sequel, we mainly introduce the standardization methods of three types of indicator data, which are real numbers, interval numbers, and HFLEs, respectively. The normalized matrix is represented by Z = [z_ij] _m×n.

(1) Standardization of Real Numbers [20]: $\begin{matrix} z_{ij} = y_{ij} / \sqrt{\sum_{i = 1}^{m} y_{ij}^{2}}, j = I_{1}; \\ z_{ij} = \frac{1}{y_{ij}} / \sqrt{\sum_{i = 1}^{m} {(1 / y_{ij})}^{2}}, j = I_{2} \end{matrix}$ (5) where z_ij is the standardized real number, y_ij is the original real number, I₁ is the benefit indicator set, and I₂ is the cost indicator set.

(2) Standardization of Interval Numbers [20]: $\begin{matrix} {\begin{matrix} z_{ij}^{l} = a_{ij}^{l} / \sqrt{\sum_{i = 1}^{m} [{(a_{ij}^{l})}^{2} + {(a_{ij}^{u})}^{2}]} \\ z_{ij}^{u} = a_{ij}^{u} / \sqrt{\sum_{i = 1}^{m} [{(a_{ij}^{l})}^{2} + {(a_{ij}^{u})}^{2}]} \end{matrix} j \in I_{1} \\ {\begin{matrix} z_{ij}^{l} = \frac{1}{a_{ij}^{l}} / \sqrt{\sum_{i = 1}^{m} [{(1 / a_{ij}^{l})}^{2} + {(1 / a_{ij}^{u})}^{2}]} \\ z_{ij}^{u} = a_{ij}^{u} / \sqrt{\sum_{i = 1}^{m} [{(1 / a_{ij}^{l})}^{2} + {(1 / a_{ij}^{u})}^{2}]} \end{matrix} j \in I_{2} \end{matrix}$ (6) where $a_{ij}^{l}$ and $a_{ij}^{u}$ are maximum and minimum of original interval numbers, respectively, $z_{ij}^{l}$ and $z_{ij}^{u}$ is the maximum and minimum of standardized interval numbers, respectively, I₁ is the benefit indicator set, and I₂ is the cost indicator set.

(3) Standardization of HELEs

Given an initial linguistic term set S ={ s_t|t = - τ, ⋯ , -1, 0, 1, ⋯ , τ } and the HFLEs y_ij ={ s_{φ
_l}|s_{φ
_l} ∈ S, l = 1, ⋯ , L } defined over the set S, we have the standardized method as follows [34]: ${\begin{matrix} z_{ij} = {s_{φ_{l}} | s_{φ_{l}} \in y_{ij}} j \in I_{1} \\ z_{ij} = {s_{τ - φ_{l}} | s_{φ_{l}} \in y_{ij}} j \in I_{2} \end{matrix}$ (7) where φ_l is the subscript of term s_{φ
_l}, z_ij is the HFLE, y_ij is the initial HFLEs, I₁ is the benefit indicator set, and I₂ is the cost indicator set.

3.3 Distance measurement of hybrid indicator evaluation value

Based on the standardized values, we calculate the distance measurement d (z_ij, z_kj) of different objects under the indicators, where d (z_ij, z_kj) indicates the distance measurement between the evaluation values of the i-th and k-th objects under the j-th indicator.

(1) Distance Measurement between Real Numbers [20]: $d (z_{ij}, z_{kj}) = | z_{ij} - z_{kj} |$ (8) where d (z_ij, z_kj) indicates the distance between real numbers z_ij and z_kj.

(2) Distance Measurement between Interval Numbers [20]: $d (z_{ij}, z_{kj}) = \frac{\sqrt{2}}{2} \sqrt{{(z_{ij}^{l} - z_{kj}^{l})}^{2} + {(z_{ij}^{u} - z_{kj}^{u})}^{2}}$ (9) where d (z_ij, z_kj) indicates the distance between interval numbers z_ij and z_kj, $z_{ij}^{l}$ and $z_{ij}^{u}$ are the lower and upper bounds of the interval number z_ij, respectively, and $z_{kj}^{l}$ and $z_{kj}^{u}$ are the lower and upper bounds of interval values z_kj, respectively.

(3) Distance Measurement between HFLEs [33]: $d (z_{ij}, z_{kj}) = \frac{1}{2} (\frac{1}{L} \sum_{l = 1}^{L} \frac{| φ_{l}^{ij} - φ_{l}^{kj} |}{2 τ} + \underset{l = 1, \dots, L}{MAX} \frac{| φ_{l}^{ij} - φ_{l}^{kj} |}{2 τ})$ (10) where $φ_{l}^{ij}$ is the subscript of linguistic term $s_{φ_{l}^{ij}}$ , $z_{ij} = ⋃_{s_{φ_{l}^{ij}} \in z_{ij}} {s_{φ_{l}^{ij}} | l = 1, \dots, L_{1}}$ and $z_{kj} = ⋃_{s_{φ_{l}^{kj}} \in z_{kj}} {s_{φ_{l}^{kj}} | l = 1, \dots, L_{2}}$ are HFLEs, L₁ and L₂ are the numbers of the linguistic terms in z_ij and z_kj, respectively, and L = MAX{ L₁, L₂ }. If L₁ is not equal to L₂, based on Eq. (3), we add linguistic terms into the HFLE with smaller size.

3.4 Indicator weight determination

In the process of evaluating rural scientific and technological talents, different evaluation indicators may have different weights. Therefore, we need to determine reasonable weights for each indicator, based on the indicator evaluation system constructed above for rural scientific and technological talent. There are three elements in the evaluation indicator system, where each element has three secondary indicators. It is unreasonable to rely solely on the subjective judgment of experts for assigning weight to each indicator, since the experts may have the personal preferences.

Therefore, we adopt the entropy weight method to calculate the weight value of each indicator as follows:

(1) For indicator j, we calculate the deviation between the i-th evaluation object and other evaluation objects. $D_{ij} = \sum_{k = 1}^{m} d (z_{ij}, z_{kj}) (i = 1, 2, \dots, m; j = 1, 2, \dots, 9)$ (11) where z_ij and z_kj are the standard values of the i-th and k-th evaluation objects under the j-th indicator, respectively, and d (r_ij, r_kj) indicates the distance between z_ij and z_kj.

(2) For indicator j, we calculate the total deviation between all evaluation objects and other evaluation objects. $D_{j} = \sum_{i = 1}^{m} D_{ij} = \sum_{i = 1}^{m} \sum_{k = 1}^{m} d (z_{ij}, z_{kj}) (j = 1, 2, \dots, 9)$ (12)

(3) Calculate the entropy value of each evaluation indicator.

Entropy is a measure of information uncertainty using probability theory. It shows that the more dispersed data leads to the greater uncertainty. The decision information of each evaluation indicator can be expressed by its entropy value E_j: $E_{j} = - K \sum_{i = 1}^{m} \frac{D_{ij}}{D_{j}} ln \frac{D_{ij}}{D_{j}} (1 ⩽ j ⩽ 9)$ (13) where m is the number of evaluation objects, K = 1/ln m, and specifically 01n0 ≡ 0.

(4) Calculate the weight of each evaluation indicator. $w_{j} = (1 - E_{j}) / \sum_{j = 1}^{n} (1 - E_{j}) (1 ⩽ j ⩽ 9)$ (14)

Through the above method, we can obtain the indicator weight vector of the evaluation indicator system of rural science and technology talents: $\begin{matrix} W = {(ω_{1}, ω_{2}, ω_{3}, ω_{4}, ω_{5}, ω_{6}, ω_{7}, ω_{8}, ω_{9})}^{T}, \\ 0 ⩽ ω_{j} ⩽ 1, \sum_{j = 1}^{9} ω_{j} = 1 . \end{matrix}$

4 The evaluation model of rural scientific and technological talents

4.1 Problem definition of evaluating rural scientific and technological talents

There are multiple indicators in the evaluation indicator system of rural science and technology talents. Specifically, the data types of indicators are diverse, and the evaluation value of indicators is ambiguous. Therefore, the evaluation of rural science and technology talents is actually an instance of the complex multi-attribute decision-making problem, which is described as follows:

Suppose that there are m evaluation objects (from different districts) A = (A₁, A₂, ⋯ , A_m) in the evaluation of rural scientific and technological talents. Based on the nine evaluation indicators C = (C₁, C₂, ⋯ , C₉) in the evaluation indicator system of rural scientific and technological talents constructed above, we evaluate the development status of rural scientific and technological talents in m regions. The weight of the evaluation indicator C_j is w_j and meets the requirement $0 ⩽ w_{j} ⩽ 1, \sum_{j = 1}^{9} w_{j} = 1$ . Through methods such as data collection, expert evaluation, and etc., we obtain the initial evaluation decision matrix of each area under each indicator, denoted as Y = [y_ij] _m×9, where y_ij indicates the evaluation value of the i-th district under the j-th evaluation indicator, and its data contains the types such as real numbers, interval numbers, and hesitation fuzzy numbers. Based on above information, it is necessary to determine the ranking order of the evaluation results of rural science and technology talents in various regions.

4.2 Evaluation model of rural scientific and technological talents based on hybrid-indicator TODIM method

Given that the indicator evaluation value contains hybrid information, we propose an evaluation model of rural science and technology talents based on the hybrid-indicator TODIM method. In order to understand the development status of rural scientific and technological talents in various regions since the implementation of the rural revitalization policy, we can evaluate the development level of scientific and technological talents in different regions based on the evaluation model proposed in this paper. Moreover, through the analysis of the evaluation results, we can refine the evaluation model of the rural scientific and technological talents in various regions. The specific decision-making steps of the evaluation model are as follows.

Step 1: Normalize initial evaluation decision matrix.

Based on the normalization method of various data types in Section 2.2, we can obtain the normalized evaluation matrix Z = [z_ij] _m×n of rural scientific and technological talents.

Step 2: Calculate the weight of each evaluation indicator

Based on the entropy weight method, we can calculate the weight vector of evaluation indicators: $W = {(ω_{1}, ω_{2}, ω_{3}, ω_{4}, ω_{5}, ω_{6}, ω_{7}, ω_{8}, ω_{9})}^{T}$

Step 3: Calculate the relative weight $ω_{j}^{'}$ of the reference point indicator j^* corresponding to each evaluation indicator $ω_{j}^{'} = \frac{ω j}{ω^{*}} (j = 1, 2, \dots, 9)$ (15) where $ω^{*} = max_{j} {ω_{j}}$ is the weight of reference point indicator j^*.

Step 4: Calculate the dominance of district i over district k in terms of evaluation indicators $φ_{j}^{ik} = {\begin{matrix} \sqrt{\frac{ω_{j}^{'} d (z_{ij}, z_{kj})}{\sum_{j = 1}^{9} ω_{j}^{'}}} if z_{ij} ≻ z_{kj} \\ 0 if z_{ij} = z_{kj} \\ \frac{- 1}{θ} \sqrt{\frac{(\sum_{j = 1}^{n} ω_{j}^{'}) d (z_{ij}, z_{kj})}{ω_{j}^{'}}} if z_{ij} ≺ z_{kj} \end{matrix}$ (16) where θ is the loss attenuation coefficient and d (r_ij, r_kj) is the distance between evaluation values z_ij and z_kj. Note that, for coefficient θ, the decision-maker can make numerical adjustments based on the preferences and actual conditions. Moreover, the smaller θ indicates the greater loss-avoidance degree of the decision-maker.

Step 5: Calculate the comprehensive dominance $δ^{ik} = \sum_{j = 1}^{n} φ_{j}^{ik}$ of district i over the district k

Step 6: Calculate the overall prospect value of district i relative to all other districts $ξ^{i} = \frac{\sum_{k = 1}^{m} δ^{ik} - min_{i} (\sum_{k = 1}^{m} δ^{ik})}{max_{i} (\sum_{k = 1}^{m} δ^{ik}) - min_{i} (\sum_{j = 1}^{m} δ^{ik})}$ (17)

Step 7: Rank (Sort) the development status of rural science and technology talents in various districts

Based on the overall perceived superiority value of each district, we rank the districts based on their development status of rural science and technology talents. Intuitively, the greater overall prospect values indicate the better development level of rural science and technology talents in the district.

5 A case study on the evaluation of rural scientific and technological talents

5.1 Development status evaluation of science and technology talents in four districts

In order to understand the cultivation and development of rural scientific and technological talents since the implementation of the rural revitalization strategy, we selected four districts, A, B, C, and D, as evaluation objects. Based on the proposed TODIM method in this paper, we evaluate the development level of rural scientific and technological talents in these four districts. The initial evaluation data of the four district are shown in Table 2.

Table 2
Initial evaluation data of rural scientific and technologic talents in various districts

Element Evaluation indicator A district B district C district D district

Basic condition (U₁) Talent quantity (η₁) [1500, 1700] [900, 1000] [1300, 1500] [1000, 1200]

Cultural level(η₂) {s₁, s₂} {s₂} {s₁} {s₀, s₁}

Moral quality (η₃) {s₁} {s₂} {s₁, s₂} {s₃}

Ability condition (U₂) Work ability (η₄) {s₂} {s₁, s₂} {s₁} {s₂, s₃}

Cognitive ability (η₅) {s₁} {s₀} {s₂} {s_-1, s₀}

Innovation ability (η₆) {s₀, s₁} {s₁} {s₁, s₂} {s₂, s₃}

Performance (U₃) Research achievement (η₇) 165 135 78 95

Social benefit (η₈) {s₀} {s₀, s₁} {s₂, s₃} {s₂, s₃}

Economic benefit (unit: billion yuan) (η₉) 3.45 1.05 5.37 2.3

Element	Evaluation indicator	A district	B district	C district	D district
Basic condition (U₁)	Talent quantity (η₁)	[1500, 1700]	[900, 1000]	[1300, 1500]	[1000, 1200]
	Cultural level(η₂)	{s₁, s₂}	{s₂}	{s₁}	{s₀, s₁}
	Moral quality (η₃)	{s₁}	{s₂}	{s₁, s₂}	{s₃}
Ability condition (U₂)	Work ability (η₄)	{s₂}	{s₁, s₂}	{s₁}	{s₂, s₃}
	Cognitive ability (η₅)	{s₁}	{s₀}	{s₂}	{s_-1, s₀}
	Innovation ability (η₆)	{s₀, s₁}	{s₁}	{s₁, s₂}	{s₂, s₃}
Performance (U₃)	Research achievement (η₇)	165	135	78	95
	Social benefit (η₈)	{s₀}	{s₀, s₁}	{s₂, s₃}	{s₂, s₃}
	Economic benefit (unit: billion yuan) (η₉)	3.45	1.05	5.37	2.3

We evaluate rural science and technology talents from these four districts in Table 2 as follows:

Step 1: Based on the normalization methods of different data types in the second part of Section 2, we establish a standardized evaluation matrix Z = [z_ij] _m×n as given in Table 3.

Table 3

Normalized evaluation matrix Z = [z_ij] _m×n

	A	B	C	D
η ₁	[0.411, 0.466]	[0.247, 0.274]	[0.356, 0.411]	[0.274, 0.329]
η ₂	{s₁, s₂}	{s₂}	{s₁}	{s₀, s₁}
η ₃	{s₁}	{s₂}	{s₁, s₂}	{s₃}
η ₄	{s₂}	{s₁, s₂}	{s₁}	{s₂, s₃}
η ₅	{s₁}	{s₀}	{s₂}	{s_-1, s₀}
η ₆	{s₀, s₁}	{s₁}	{s₁, s₂}	{s₂, s₃}
η ₇	0.670	0.549	0.317	0.386
η ₈	{s₀}	{s₀, s₁}	{s₂, s₃}	{s₂, s₃}
η ₉	0.523	0.153	0.782	0.335

Step 2: Calculate the weight of each evaluation indicator.

Based on the standardized evaluation data of the four districts, we leverage the entropy weight method to obtain the indicator weight vector for the evaluation indicator system of rural science and technology talents: $\begin{matrix} W = (0.075, 0.084, 0.124, 0.122, 0.112, \\ {0.124, 0.099, 0.122, 0.137)}^{T} . \end{matrix}$

Step 3: Calculate the relative weight of each evaluation indicator $\begin{matrix} W^{'} = (0.549, 0.613, 0.899, 0.887, 0.819, \\ {0.899, 0.721, 0.887, 1)}^{T} . \end{matrix}$

Step 4: Calculate the dominance degree of each district with respect to each evaluation indicator relative to other districts shown in Tables 4–12.

Table 4

The dominance degree $φ_{1}^{ik}$ of several districts relative to other districts in terms of indicator η₁

	A ₁	A ₂	A ₃	A ₄
A ₁	0.000	–1.459	–0.808	–1.277
A ₂	0.007	0.000	0.102	0.060
A ₃	0.068	–1.216	0.000	0.083
A ₄	0.107	–0.718	–0.990	0.000

Table 5

The dominance degree $φ_{2}^{ik}$ of several districts relative to other districts in terms of indicator η₂

	A ₁	A ₂	A ₃	A ₄
A ₁	0.000	–1.154	0.108	0.125
A ₂	0.108	0.000	0.125	0.165
A ₃	–1.154	–1.333	0.000	0.102
A ₄	–1.333	–1.763	–0.990	0.000

Table 6

The dominance degree $φ_{3}^{ik}$ of several districts relative to other districts in terms of indicator η₃

	A ₁	A ₂	A ₃	A ₄
A ₁	0.000	–1.101	–0.954	–1.557
A ₂	0.151	0.000	0.131	–1.101
A ₃	–0.954	0.131	0.000	–1.457
A ₄	0.214	0.151	0.200	0.000

Table 7

The dominance degree $φ_{5}^{ik}$ of several districts relative to other districts in terms of indicator η₅

	A ₁	A ₂	A ₃	A ₄
A ₁	0.000	0.144	–1.154	0.191
A ₂	–1.154	0.000	–1.632	0.125
A ₃	0.144	0.204	0.000	0.240
A ₄	–1.526	–0.999	–1.913	0.000

Table 8

The dominance degree $φ_{4}^{ik}$ of several districts relative to other districts in terms of indicator η₄

	A ₁	A ₂	A ₃	A ₄
A ₁	0.000	0.108	0.125	–1.154
A ₂	–1.154	0.000	0.108	–1.333
A ₃	–1.154	–1.154	0.000	–1.763
A ₄	0.108	0.125	0.165	0.000

Table 9

The dominance degree $φ_{6}^{ik}$ of several districts relative to other districts in terms of indicator η₆

	A ₁	A ₂	A ₃	A ₄
A ₁	0.000	–0.954	–1.101	–1.557
A ₂	0.131	0.000	–0.954	–1.457
A ₃	0.151	0.131	0.000	–1.101
A ₄	0.214	0.200	0.151	0.000

Table 10

The dominance degree $φ_{7}^{ik}$ of several districts relative to other districts in terms of indicator η₇

	A ₁	A ₂	A ₃	A ₄
A ₁	0.000	0.107	0.181	0.163
A ₂	–1.144	0.000	0.147	0.200
A ₃	–1.948	–1.577	0.000	–0.861
A ₄	–1.747	–1.321	0.080	0.000

Table 11

The dominance degree $φ_{8}^{ik}$ of several districts relative to other districts in terms of indicator η₈

	A ₁	A ₂	A ₃	A ₄
A ₁	0.000	–1.555	–2.979	–2.979
A ₂	0.080	0.000	–2.540	–2.540
A ₃	0.154	0.131	0.000	0.000
A ₄	0.154	0.131	0.000	0.000

Table 12

The dominance degree $φ_{9}^{ik}$ of several districts relative to other districts in terms of indicator η₉

	A ₁	A ₂	A ₃	A ₄
A ₁	0.000	0.253	–1.234	0.175
A ₂	–1.380	0.000	–1.852	–0.955
A ₃	0.227	0.340	0.000	0.041
A ₄	–0.955	–0.996	–1.561	0.000

Step 5: Calculate the comprehensive superiority of each district relative to other districts $\begin{matrix} φ^{11} = 0, φ^{12} = - 5.488, φ^{13} = - 7.582, φ^{14} = - 6.253, \\ φ^{21} = - 6.049, φ^{22} = 0, φ^{23} = - 7.719, φ^{24} = - 4.906, \\ φ^{31} = - 3.058, φ^{32} = - 2.283, φ^{33} = 0, φ^{34} = - 3.056, \\ φ^{41} = - 3.419, φ^{42} = - 4.510, φ^{43} = - 7.456, φ^{44} = 0 . \end{matrix}$

Step 6: Calculate the overall perceived superiority of each district relative to all other districts $ξ^{1} = 0, ξ^{2} = 0.06, ξ^{3} = 1, ξ^{4} = 0.36 .$

Step 7: Rank the development status of rural science and technology talents in various districts

Based on the values of overall perceived superiority of each district, we can rank the four districts in the order of C ≻ D ≻ B ≻ A based on the evaluation result of rural science and technology talents. It can be seen that the C district has the highest development level of rural scientific and technological talents.

5.2 Evaluation result analysis

From the evaluation results of the four districts above, we can see that districts C and A have the best and worst evaluation results of rural scientific and technological talents, respectively. As depicted in Figs. 1–3, we further analyze the impact of each evaluation indicator (in the evaluation indicator system of rural science and technology talents) on the development level of talents in various districts, by demonstrating the overall perceived superiority of each district. Moreover, Fig. 4 illustrates the magnitude of the overall perceived dominance of each district under each element. Specifically, Fig. 1 reflects the changes in the overall perceived superiority of the three evaluation indicators, the number of talents, cultural level, and moral quality, of each district under the basic condition element. Figure 2 shows the changes in the overall perceived dominance on three evaluation indicators, work ability, cognitive ability, and innovation ability, under the ability performance element. Figure 3 demonstrates the changes in the overall perceived dominance on three evaluation indicators, research achievements, social benefits, and economic benefits, in various districts under the benefit performance factor. Figure 4 illustrates the changes in the overall perceived dominance on three elements, basic condition, ability performance, and benefit performance, in various districts.

Fig.1

The overall perceived dominance degree of each district in the corresponding evaluation indicator under the basic condition elements.

Fig.2

The overall perceived dominance degree of each district in its corresponding evaluation indicator under the element of ability performance.

Fig.3

The overall perceived dominance degree of each region on its corresponding evaluation indicator under benefit performance element.

Fig.4

The overall perceived dominance degree of each district under each element.

Based on Figs. 1–4, we conduct a detailed analysis of the quality level of scientific and technological talents and the corresponding influence factors in each district. From the final evaluation result, it can be seen that district A is evaluated as the worst district among the four districts in terms of the evaluation score of science and technology talents. From Fig. 4, we can see that district A has relatively low level of science and technology talents under all three basic elements, which are basic condition, ability performance, and benefit performance. From Fig. 1, district A has the largest number of talents but the worst moral quality among the four districts, which makes district A the lowest level of scientific and technological talents under the basic condition. From Fig. 2, we can see that the work ability and cognitive ability of science and technology talents in district A is not low, but the innovation ability of talents in district A is the lowest in the four districts, which makes district A to be ranked as the second last district among four districts in terms of evaluation scores under the ability performance of scientific and technological talents. From Fig. 3, it can be seen that district A has the largest number of research achievements and good economic benefit, however, district A has the worst economic benefit among the four districts. The above analysis shows that district A blindly pursues the number of talents in the development of science and technology talents, however, ignoring the cultivation of talents’ moral qualities. Consequently, district A has poor innovation ability of science and technology talents, as well as low social benefits brought by scientific and technological talents.

From the final evaluation results, it can be seen that district B is ranked as the third district among the four districts based on the evaluation scores of scientific and technological talents. From Fig. 4, we find that, under the basic condition element, district B has the largest overall perceived superiority. That is, district B has best basic condition quality of science and technology talents among the four districts. However, under the element of ability performance and benefit performance, district B has the worst performance with respect to that of other districts. From Fig. 1, we find that district B has the smallest number of scientific and technical talents among the four districts, but with the relative high cultural level and moral quality of the talents. From Figs. 2 and 3, we find that district B has high the cultural level and moral quality of the science and technology talents, but low level of work ability, cognitive ability and innovation ability, which leads to the poor performance of district B with respect to the social and economic benefit. From above analysis, instead of blindly introducing too many talents, district B paid more attention to the cultural level and moral quality of science and technology talents. However, district B neglects the ability development in the process of cultivating scientific and technological talents, and fails to give full play to the role of scientific and technological talents in actual agricultural production and social activities. As a result, district B cannot benefit from the contribution of scientific and technological talents to regional economic development.

From the final evaluation results, it can be seen that districts C and D are ranked as the first and second districts (among the four districts), respectively, based on the evaluation scores of scientific and technological talents. From Fig. 4, with respect to the overall perceived dominance degree, district C behaves worse than district D under the elements of basic condition and ability performance, but outperforms district D under the element of benefit performance. As a result, district C is evaluated to be inferior to district C in the final evaluation result. From Figs. 2 and 3, districts C and D have similar innovation ability and number of scientific research achievements, and moreover, these two districts achieve best social benefit among the four districts. Meanwhile, it is worth noting that district C significantly outperforms district D in terms of the economic benefits. The above analysis shows that districts C and D have begun to pay attention to the cultivation of moral character and the ability development in the process of cultivating scientific and technological talents. Moreover, technological talents can actively participate in various social activities and create prosperous social profits for the two districts. However, compared with district C, district D only pays attention to scientific research creation, while ignoring the actual transformation from scientific research works to production activities. As a result, the scientific research work of district D has few practical applications and low conversion rate, which greatly reduces the economic benefits created by the scientific research works.

5.3 Comparison analysis

There are very few related research works for evaluating the rural talents. In this subsection, we conduct a comparative analysis of several representative multi-criteria decision-making methods for talent evaluation or personnel selection, which mainly includes Boran et al.’ method [12], Afshari et al.’ method [13], Yu et al.’ method [14], Bogdanovic and Miletic’s method [15]. Specifically, we highlight the advantages of the proposed method for the evaluation of rural scientific and technological talents, in three aspects as follows.

(1) Evaluation information processing. The existing methods can only deal with a single type of evaluation information. For example, Boran et al.’ method [12] can only deal with intuitionistic fuzzy numbers, Afshari et al.’ method [13] and Bogdanovic and Miletic’s method [15] can only deal with linguistic variables, and Yu et al.’ method [14] can only deal with linguistic variables hesitant fuzzy numbers. As we all know, the complex rural talent evaluation index system needs a variety of hybrid evaluation indexes, such that the system can conduct convenient and effective evaluation. However, the above methods cannot deal with this kind of evaluation problems (with hybrid evaluation indexes).

(2) Attribute weight selection. The selection of index weight usually has an important impact on the evaluation results. The attribute weights in [12 –15] are directly suggested by experts or indirectly calculated by AHP method based on experts’ subjective assignment of attribute weights. In other words, the weights of these methods are subjective to experts’ knowledge and experience, and the process is complex. In this paper, we use the entropy weight method to determine the attribute weights, and measure the importance of attributes through the evaluation information itself. This process is more convenient, since it effectively considers the objective reflection of the evaluation data on the important relationship between attributes.

(3) Ranking method. The ranking strategy in Yu et al.’ method [14] uses the generalized hesitant fuzzy aggregation operator to integrate the attribute values of each alternative, which is simple to operate but prone to information distortion. The TOPSIS [12] and PROMETHEE [15] rank the alternatives based on closeness coefficient and net flow, respectively, but they do not consider the psychological behavior of decision-makers, which makes them lack some flexibility. In addition, when the number of indicators is large, the computational complexity of the PROMETHEE method will be very high. Instead, the proposed TODIM method in this paper considers the psychological behavior of decision makers based on prospect theory, and thus makes better evaluations according to the risk attitudes of different decision makers. Therefore, the proposed TODIM method is more flexible for evaluating the development level of rural scientific and technological talents in different regions, and can enable the decision makers to choose different decision parameters under different decision situations.

6 Conclusions and suggestions

6.1 Conclusion

Rural scientific and technological talents, as an important part of the construction of rural talents in China, are the representatives of advanced productive forces, and can better promote the development of China’s new rural construction. In order to better construct the rural talent team, we need to scientifically evaluate the current status of rural scientific and technological talents in various districts. Through literature review and field research, this paper builds a complete evaluation indicator system for rural scientific and technological talents, which weight of each evaluation indicator is scientifically determined via the entropy weight method. The TODIM method has two advantages, which are the wide application in the field of multi-attribute decision-making and the consideration of decision markers, respectively. Based on the TODIM method, this paper proposes a hybrid-indicator evaluation model to evaluate the rural scientific and technological talents. The evaluation results indicate that the development level of scientific and technological talents in a certain district is not only affected by the cultural and moral quality of the local scientific and technological talents, but also depends on whether the district can fully develop the innovation ability of the talents. Therefore, it is important to actively guide the scientific and technological talents in social life and production practice, such that they can create economic value.

6.2 Suggestions

In order to continuously strengthen the construction of rural scientific and technological talents in China and improve the development level of rural talents, based on the analysis of evaluation results in Section 5, this article proposes suggestions as follows.

(1) Increase the introduction of rural scientific and technological talents. We should improve the introduction mechanism of rural talents, issue corresponding preferential support policy, continue to increase the introduction of agricultural graduates in colleges and universities through policy publicity and university encouragement, and provide them with superior economic treatment and living conditions to enable them to take root in rural areas. Moreover, we should provide protection of policy, funding, and environment for establishing a reasonable agricultural experimental base, such that the theoretical results can be truly transformed into planting applications.

(2) Establish a solid training mechanism of the rural scientific and technologic talents. First, we must increase the investment for the talents, introduce some agricultural universities, increase some special scientific and technological development funds, and promote the integration of production, education, and research. The government has carried out some measures (e.g., issuing laws, regulations, guarantee policies, and incentive mechanisms) to support the cultivation of rural scientific and technological talents, which promote the better integration of agriculture, science, and technology, education, and promotion. Moreover, we need to improve the training system of the rural science and technology talent. On the one hand, we need to carry out ideological and political education to strengthen their ideological understanding and cultivate their subjective consciousness (dedication) of serving the countryside. On the other hand, we need to strengthen continuing education, such that the rural talents not only master their professional knowledge, but also flexibly master the knowledge and skills in other fields such as society, economy, science, and technology. As a result, the rural talents will have better innovation and research ability for higher benefit value.

(3) Establish a reasonable performance evaluation mechanism for rural scientific and technological talents. The scientific performance evaluation mechanism can stimulate the initiative and motivation of the rural scientific and technological talents, which will lead to a positive incentive effect. The performance evaluation of the talents should not only focus on measures such as personal quality, ability performance, and scientific research achievements, but also pay more attention to the transformation and application of the research achievements, such that we can evaluate the talents through the economic benefits brought by the research achievements. Moreover, we should pay attention to the contributions and efforts made by the talents in terms of social benefits, that is, whether the results are widely recognized by the masses; whether they are universal, and; whether they can bring direct economic benefits and social effects to farmers. Therefore, we need to quantify these indicators, and objectively and fairly evaluate the talents by considering incentive measures such as performance bonus and position promotion. In this way, we can motivate rural scientific and technological talents to create more social value.

(4) Strengthen the transformation of rural science and technology management and service concept. A main reason for current backward development of rural science and technology is the improper management concept, which in turn affects the cultivation and achievement transformation of rural scientific and technological talents (i.e., low talent utilization rate and talent waste). Therefore, it is necessary to transform the ideological awareness and management concepts of technology management and service personnel, due to the urgency and importance of the development of rural science and technology. In this way, we can effectively allocate and manage the rural scientific and technological talents, strengthen the ideological re-education and technical re-training of the talents, and integrate rural talents into the countryside both in ideology and technology. As a result, we can ensure that each talent can be employed effectively and play a main role in serving the countryside.

7 Compliance with ethical standards

(1) Disclosure of potential conflicts of interest

We declare that we do have no commercial or associative interests that represent a conflict of interests in connection with this manuscript. There are no professional or other personal interests that can inappropriately influence our submitted work.

(2) Research involving human participants and/or animals

This article does not contain any studies with human participants or animals performed by any of the authors.

References

Guo

, Research on the construction of rural practical talent team in Weifang City, Doctoral Dissertation of Chinese Academy of Agricultural Sciences, 2009.

, Research on the training mechanism of excellent rural scientific and technological Talents, Doctoral Dissertation of Hunan Agricultural University, 2013.

Lee

S.H.

and Cho

Y.H

, The Evaluation of Importance-Performance on the Education Program for Human Resource Development in Rural Tourism, Journal of Tourism Sciences 38(9) (2014), 265–290.

Chen

, Research on the construction of rural talent team in Qingyuan City, Doctoral Dissertation of Huazhong Agricultural University, 2008.

, Thoughts on innovating agricultural talent evaluation system, Seed Technology 2016(10) (2016), 46–52.

Bohlouli

, et al., Competence assessment as an expert system for human resource management: A mathematical approach, Expert Systems with Applications 70 (2017), 83–102.

, Men

, Wang

, Li

, Shi

and Sun

, Research on evaluation indicator system of science and technology talents in “Three Districts” of Hubei province, Hubei Agricultural Sciences 57(6) (2018), 116–122.

Chen

and Sun

, Research on talent evaluation in local major science and technology introducing projects under the background of innovation drive, Yunnan Social Sciences 2018(4) (2018), 59–64.

Lin

, Xu

, Zhai

and Yao

, Multi-attribute group decision-making under probabilistic uncertain linguistic environment, Journal of the Operational Research Society 69(2) (2018), 157–170.

10.

Lin

, Huang

, Xu

and Chen

, Evaluating IoT platforms using integrated probabilistic linguistic MCDM method, IEEE Internet of Things Journal 7(11) (2020), 11195–11208.

11.

Huang

, Lin

and Xu

, Pythagorean fuzzy MULTIMOORA method based on distance measure and score function: its application in multicriteria decision making process, Knowledge and Information Systems 62(11) (2020), 4373–4406.

12.

Boran

F.E.

, Genç

and Akay

, Personnel selection based on intuitionistic fuzzy sets, Human Factors and Ergonomics in Manufacturing & Service Industries 21(5) (2011), 493–503.

13.

Afshari

A.R.

, Yusuff

R.M.

and Derayatifar

A.R.

, Linguistic extension of fuzzy integral for group personnel selection problem, Arabian Journal for Science and Engineering 38(10) (2013), 2901–2910.

14.

, Zhang

and Xu

, Group decision making under hesitant fuzzy environment with application to personnel evaluation, Knowledge-Based Systems 52 (2013), 1–10.

15.

Bogdanovic

and Miletic

, Personnel evaluation and selection by multicriteria decision making method, Economic Computation and Economic Cybernetics Studies and Research 48(3) (2014), 179–196.

16.

Bozbura

F.T.

, Beskese

and Kahraman

, Prioritization of human capital measurement indicators using fuzzy AHP, Expert Systems with Applications 32(4) (2007), 1100–1112.

17.

Wei

, Ren

and Rodríguez

R.M.

, A hesitant fuzzy linguistic TODIM method based on a score function, International Journal of Computational Intelligence Systems 8(4) (2015), 701–712.

18.

Wang

, Wang

Y.M.

and Martínez

, Fuzzy TODIM method based on alpha-level sets, Expert Systems with Applications 140 (2020), 112899.

19.

Wang

, Wang

and Zhang

, A likelihood-based TODIM approach based on multi-hesitant fuzzy linguistic information for evaluation in logistics outsourcing, Computers & Industrial Engineering 99 (2016), 287–299.

20.

Liu

and Teng

, Ecosystem health evaluation of Dongying City based on mixed indicator value TODIM method, China Population Resources and Environment 27(8) (2017), 167–176.

21.

Liu

, Rong

and Teng

, The evaluation of ecosystem health based on hybrid TODIM method for Chinese case, Technological and Economic Development of Economy 25(3) (2019), 542–570.

22.

Liu

and Teng

, Probabilistic linguistic TODIM method for selecting products through online product reviews, Information Sciences 485 (2019), 441–455.

23.

Arya

and Kumar

, A new picture fuzzy information measure based on shannon entropy with applications in opinion polls using extended VIKOR-TODIM approach, Computational & Applied Mathematics 39(3) (2020), 197.

24.

Guo

, Yin

, Zhang

and Extended

etc.

, TODIM method for CCUS storage site selection under probabilistic hesitant fuzzy environment, Applied Soft Computing 93 (2020), 106381.

25.

Liu

and You

, Improved TODIM Method Based on Linguistic Neutrosophic Numbers for Multicriteria Group Decision-Making, International Journal of Computational Intelligence Systems 12(2) (2019), 544–556.

26.

Liu

and Zhang

, Normal Wiggly Hesitant Fuzzy TODIM Approach for Multiple Attribute Decision Making, Journal of Intelligent & Fuzzy Systems 39(1) (2020), 627–644.

27.

, Pan

and Wang

, A shadowed set-based TODIM method and its application to large-scale group decision making, Information Sciences 544 (2021), 135–154.

28.

Arya

and Kumar

, A novel TODIM-VIKOR approach based on entropy and Jensen-Tsalli divergence measure for picture fuzzy sets in a decision-making problem, International Journal of Intelligent Systems 35(12) (2020), 2140–2180.

29.

Davoudabadi

, Mousavi

S.M.

and Mohagheghi

, A new last aggregation method of multi-attributes group decision making based on concepts of TODIM, WASPAS and TOPSIS under interval-valued intuitionistic fuzzy uncertainty, Knowledge and Information Systems 62(4) (2020), 1371–1391.

30.

Yue

, An extended TOPSIS for determining weights of decision makers with interval numbers, Knowledge-Based Systems 24(1) (2011), 146–153.

31.

Liao

, Gou

and Xu

, A survey of decision making theory and methodologies of hesitant fuzzy linguistic term set, System Engineering-Theory & Practice 37(1) (2017), 35–48.

32.

Zhu

and Xu

Z.S.

, Consistency measures for hesitant fuzzy linguistic preference relations, IEEE Transactions on Fuzzy Systems 22 (2014), 35–45.

33.

Liao

H.C.

, Yang

L.Y.

and Xu

Z.S.

, Two new approaches based on ELECTRE II to solve the multiple criteria decision making problems with hesitant fuzzy linguistic term sets, Applied Soft Computing 63 (2018), 223–234.

34.

Shi

and Xiao

, Optimal Decision-making for Hesitant Fuzzy Linguistic Green Supplier Based on Prospect Theory, Statistics and Decision 513(21) (2018), 46–52.