Study on the Path of High Quality Enhancement of International Communication Effectiveness of Rural Revitalization Literature in the Age of Digital Intelligence

This paper identifies 19 influencing factors for the international dissemination of rural revitalization literature. The identified systematic influencing factors are set according to the corresponding dimensions as A1 (cultural and historical themes), A2 (character themes), A3 (social themes), A4 (political and economic themes), A5 (international themes), A6 (local themes), A7 (national themes), A8 (nodal selections), A9 (non-nodal selections), B1 (video dissemination), B2 (audio dissemination), B3 (text dissemination), B4 (single-episode/single-episode style, B5 (series style), and B6 (with obvious emotional bias), B3 (textual dissemination), B4 (single-episode/single-article genre), B5 (series genre), B6 (significant emotional bias), B7 (no significant emotional bias), B8 (endoperspective narrative), B9 (zero-perspective narrative), and B10 (exoperspective narrative). And C1 (first-class spreading heat/awarded rank), C2 (second-class spreading heat/awarded rank), and C3 (third-class spreading heat/awarded rank) were used as outcome variables, respectively.

The Delphi method is used to investigate the problem under study, to determine the mutual logical relationship between the influencing factors, and then by experts to make a quantitative evaluation of the relationship between the factors, the scoring rules of the questionnaire in this paper use a 5-level scale to measure, in accordance with the 0-4 scale for scoring, and this method is also the most common method. 0 indicates that the row of factors has no influence on the column of factors, 1 indicates that the row of factors has a weak influence, 2 indicates that the row factor has a normal influence on the column factor, 3 indicates that the row factor has a strong influence on the column factor, and 4 indicates that the row factor has a strong influence on the column factor, so as to establish the direct influence matrix of rural revitalization literature dissemination X. In this paper, we assumed that the direct influence matrix of the system is: 1X(X=[Xij]n×n)$$X\left( {X = {{\left[ {{X_{ij}}} \right]}_{n \times n}}} \right)$$

Establish the normalized influence matrix G and calculate the normalized direct influence matrix G. The core of the normalization process for the direct matrix X lies in the method of obtaining the maximum value, which is calculated as follows: 2G=1max1≤i≤n∑j=1nxijX$$G = \frac{1}{{\mathop {\max }\limits_{1 \le i \le n} \sum\limits_{j = 1}^n {{x_{ij}}} }}X$$

It is easy to know 0 ≤ g_ij ≤ 1, and max1≤i≤n∑j=1ngij=1$$\mathop {\max }\limits_{1 \le i \le n} \sum\limits_{j = 1}^n {{g_{ij}}} = 1$$, where max1≤i≤n∑j=1n$$\mathop {\max }\limits_{1 \le i \le n} \mathop \sum \limits_{j = 1}^n$$ represents the maximum value of the sum of the rows in the direct influence matrix, and then use the individual values in the matrix in the division of the maximum value of the rows can be derived from the normalized normalized influence matrix.

Establish the combined impact matrix and calculate the combined impact matrix T(T=[tij]n×n)$$T\left( {T = {{\left[ {{t_{ij}}} \right]}_{n \times n}}} \right)$$ between the influencing factors of the system, and the combined impact matrix T is calculated by the formula: 3T=G(I−G)−1$$T = G{(I - G)^{ - 1}}$$

where I is the unit matrix, i.e., the matrix whose diagonal value is 1 and whose values are 0 elsewhere.

The formula for calculating the degree of influence f_i and the degree of being influenced e_i for each element is as follows: 4Influenceei=∑j=1ntij,(i=1,…,n)$${\text{Influence}}\ {e_i} = \sum\limits_{j = 1}^n {{t_{ij}}} ,(i = 1, \ldots ,n)$$ 5Degree of influenceei=∑j=1ntij,(i=1,…,n)$${\text{Degree of influence}}\ {e_i} = \sum\limits_{j = 1}^n {{t_{ij}}} ,(i = 1, \ldots ,n)$$

Calculate the center degree m_i and cause degree n_i of each element with the following formulas: 6Centralitymi=∑j=1ntij,(i=1,…,n)$${\text{Centrality}}\ {m_i} = \sum\limits_{j = 1}^n {{t_{ij}}} ,(i = 1, \ldots ,n)$$ 7Degree of causeni=∑j=1ntij,(i=1,…,n)$${\text{Degree of cause}}\ {n_i} = \sum\limits_{j = 1}^n {{t_{ij}}} ,(i = 1, \ldots ,n)$$

According to its center degree and cause degree in drawing out the corresponding Cartesian coordinate system, with the cause degree as the vertical coordinate, the center degree as the horizontal coordinate, drawing out the position of each factor in the coordinate system, a more intuitive analysis of the importance of each factor and the factor attributes, and in this way to assist in the comprehensive analysis of each factor, screening the key influencing factors.

QCA analysis is an approach based on set theory, which determines set relationships between measured variables by calibrating them. This method usually converts the variables into a set of sets and assigns a degree of affiliation to them.

Arbitrary mapping on a closed interval from the domain X to [0,1]$$\left[ {0,1} \right]$$: 8μA:X→[0,1]$${\mu_A}:X \to [0,1]$$ 9x→μA(x)$$x \to {\mu_A}(x)$$

Determine a fuzzy set A on X, μ_A is called the affiliation function of A, and μ_A(x) is called the degree of affiliation of x to the fuzzy set A, denoted as: 10A={(x,μA(x))/x∈X} $$A = \left\{ {\left( {x,{\mu_A}(x)} \right)/x \in X} \right\}$$

FsQCA is applicable to different types of data, not only is it able to convert different types of data to the [0, 1] range using fuzzy set calibration, but at the same time, FsQCA can be combined with categorical variables that do not need to be converted to fuzzy sets. In this case, some variables are dichotomous or multicategorical values, while others have continuous values in the range of [0, 1] FsQCA has a good compatibility with variable types [20].

The “four-value fuzzy set calibration method” can effectively detect discontinuous values, and according to the membership degree of the case result variable and the condition variable, [0−0.33−0.67−1]$$\left[ {0 - 0.33 - 0.67 - 1} \right]$$ quartile difference calibration is used to ensure the accuracy and reliability of the data, “1” = full membership, “0.67” = partial membership, “0.33” = partial non-membership, “0” = no membership at all. In the case of dichotomous variables, only full and complete affiliation are considered.

For continuous variables, in FsQCA, it is necessary to calibrate the value of a variable in a case to the fuzzy set affiliation degree from 0 to 1, so as to become a set. In this paper, the fuzzy set calibration is carried out by “Mean Anchor Method”, in which the variable data are arranged in descending order, the quartiles are selected as the fully affiliated and fully unaffiliated values, and the mean of the upper and lower quartiles is the intersection value, so as to determine the anchor point. The following is the calculation method for anchor point values:

Fully unaffiliated anchor point value: 11Q1=QUARTILE(x1j,xrj,1)$${Q_1} = QUARTILE\left( {{x_{1j}},{x_{rj}},1} \right)$$

Intersection anchor value: 12Q2=QUARTILE(x1j,xnj,2)$${Q_2} = QUARTILE\left( {{x_{1j}},{x_{nj}},2} \right)$$

Fully subordinate anchor point values: 13Q3=QUARTILE(x1j,xnj,3)$${Q_3} = QUARTILE\left( {{x_{1j}},{x_{nj}},3} \right)$$

Fuzzy set calibration calculation method: 14xj′=Calibrate(xj,Q3,Q2,Q1)$${x'_j} = Calibrate\left( {{x_j},{Q_3},{Q_2},{Q_1}} \right)$$

Qualitative comparative analysis determines whether there is a relationship of necessity and sufficiency between variables through the calculation of consistency and coverage. 1)

Consistency

Consistency refers to the degree of necessity of a case that meets a condition leading to an outcome variable and is calculated as follows: 15Consistency(Xi≤Yi)=∑[min(Xi,Yi)]/Xi$$Consistency\left( {{X_i} \le {Y_i}} \right) = {{\sum {\left[ {\min \left( {{X_i},{Y_i}} \right)} \right]} } \mathord{\left/ {\vphantom {{\sum {\left[ {\min \left( {{X_i},{Y_i}} \right)} \right]} } {{X_i}}}} \right. } {{X_i}}}$$

When the consistency index is greater than 0.9, then X is a necessary condition for Y, i.e., X is always present when Y occurs, and if the consistency index is less than 0.9, then the combined effects of multiple conditioning variables need to be analyzed. 2)

Coverage

Coverage is the extent to which these given conditions explain the occurrence of the outcome. When consistency is satisfied, the researcher can assess the ability of X to explain Y by calculating the coverage rate, where the higher the coverage indicator, the better the ability of X to explain empirically, with the following formula: 16Consistency(Xi≤Yi)=∑[min(Xi,Yi)]/∑Yi$$Consistency\left( {{X_i} \le {Y_i}} \right) = {{\sum {\left[ {\min \left( {{X_i},{Y_i}} \right)} \right]} } \mathord{\left/ {\vphantom {{\sum {\left[ {\min \left( {{X_i},{Y_i}} \right)} \right]} } {\sum {{Y_i}} }}} \right. } {\sum {{Y_i}} }}$$

QCA is a set theory configuration analysis method based on Boolean algebra, which uses Boolean operation to simplify the conditional configuration and obtain the key elements. The threshold of sufficient consistency of configuration is set to 0.8 and the threshold of PRI consistency is set to 0.75, and a truth table with the exclusion of logical remainders is constructed to obtain the behavior configuration path of the high (low) network cluster. The symbol “*” in the conditional combination path means “and”, “+” means “or”, and “~” means “not”, that is, the opposite value.

Boolean arithmetic is the logical deduction method of numerical symbolization, including union, intersection, and subtraction. Boolean operations are performed by combining, differing, and intersecting two or more objects to obtain a new object form. Logical deduction and thus simplification of conditional combinations by Boolean operation methods are commonly used as follows:

The method of concatenation: 17Y=ABC+A′BC=(A+A′)BC=BC$$Y = ABC + A^\prime BC = \left( {A + A^\prime } \right)BC = BC$$

Absorption method: 18Y=AB+ABC=AB$$Y = AB + ABC = AB$$

Elimination method: 19Y = AB+A′C+B′C=AB+(A′+B′)C = AB+(AB)′C=AB+C$$\begin{array}{rcl} Y &=& AB + A^\prime C + B^\prime C = AB + \left( {A^\prime + B^\prime } \right)C \\ &=& AB + (AB)^\prime C = AB + C \\ \end{array}$$

Matching term method: 20Y = AB+BC+AC′=AB(C+C′)+BC+AC′ = (A+1)BC+AC′(B+1)=BC+AC′$$\begin{array}{rcl} Y &=& AB + BC + AC^\prime = AB\left( {C + C^\prime } \right) + BC + AC^\prime \\ &=& (A + 1)BC + AC^\prime (B + 1) = BC + AC^\prime \\ \end{array}$$

Checking the robustness of the results is a key step in QCA research, and there are many different ways to check the robustness of QCA, the most common of which is to adjust the parameters, re-perform the analysis, and revisit the changes to assess the robustness of the results based on the changes. It is common to adjust the calibration affiliation, consistency threshold, frequency of cases, etc. If the parameter changes do not lead to substantial changes in the histogram results, the findings will be highly reliable.

This paper adopts the robustness test method applicable to set theory, and adjusts the consistency threshold to test the robustness of the grouping results by adjusting the original consistency threshold from 0.8 to 0.85, and other treatments remain unchanged. The analysis reveals the number of group states, whether the neutral arrangement with the same core conditions but different edge conditions has undergone a large change, and whether the change can support meaningful and distinct substantive explanations.

Based on the evaluation results generated by the experts after analyzing and evaluating the communication factors, 19 initial judgment matrices were obtained, and then the resulting matrices were normalized to obtain the direct impact matrix. The tabular data specifically reflect the extent to which A1 has a direct influence on A1-C3 respectively, and the extent to which A2 has a direct influence on A1-C3 respectively ...... and so on. For example, A1 has no impact on A1 itself, and the direct impact of A1 on A2 is higher than the impact of A1 on A3. The larger the value shown in the table, the greater the degree of correlation of the mutual impact, and vice versa, the smaller it is. Then according to the normalized influence matrix G and formula (3) to get the comprehensive influence matrix T between the factors.

Using the formula to get the degree of influence Di, the degree of being influenced Ci the degree of centrality Mi the degree of cause Ri of each factor, the specific total influence relationship of all the indicators is shown in Table 1. Example analysis: the cultural and historical theme (A1) has the highest degree of influence of 4.506, which indicates that the rural revitalization literature on this theme has a higher likelihood of causing international dissemination. The non-nodal selection (A9) has the lowest influence degree of 2.047, indicating that this factor has the least influence on the dissemination of literature. Cultural-historical themes (A1) were affected by the highest influence degree value of 4.269, indicating that they were influenced by other factors. There is a significant emotional bias (B6) was shown to have the lowest influence value of 2.743, indicating that it is not easily influenced by other factors.

In order to be able to analyze the influencing factors more clearly, the causality diagram is drawn according to the calculation results in Table 1. The specific content includes: establish a Cartesian coordinate system with the degree of influence as the horizontal axis and the degree of being influenced as the vertical axis, and correspond the values of the degree of influence and the degree of being influenced of each influencing factor on the coordinate system. The Cartesian coordinate system diagram obtained for the degree of influence - the degree of being influenced is shown in Figure 1.

Figure 1.

The degree of influence - being influenced diagram

In the graph, the horizontal coordinate is the value of the influence degree D, the vertical coordinate is the value of the influenced degree C, and the two lines are the average values of D or C, respectively. The first quadrant represents a high influence degree value and a high influenced degree value, i.e., a high elemental influence degree and a high influenced degree. The second quadrant represents low and high influence values, i.e., low elemental influence and high influence. The third quadrant represents low impact and low influence values, i.e., low elemental influence and low influence. The fourth quadrant represents high impact and low influence values, i.e., high elemental influence and low influence.

In terms of the degree of influence, the degree of influence of the cultural-historical theme (A1) is relatively significant compared to other factors, mainly because the cultural-historical theme influences the local theme (A6), international theme (A5), etc., thus affecting the whole process of international dissemination of rural revitalization literature. Compared with the cultural-historical theme (A1), the character theme (A2), the political-economic theme (A4) and the national theme (A7), which have a slightly weaker degree of influence, affect the direction of the dissemination of rural revitalization literature to a large extent. The character theme emphasizes the degree of importance of rural revitalization literature in the portrayal of rural characters, while the political-economic theme and the national theme emphasize the influence of reflecting China’s national conditions on the dissemination of rural revitalization literature on the international dissemination.

In terms of the degree of being influenced, the cultural and historical theme (A1), the character theme (A2), the international theme (A5), and the local theme (A6) are more likely to be influenced by other factors, and changes in these factors need to be paid attention to in the process of promoting the international dissemination of literature on rural revitalization. It can also be seen that the three factors of political and economic themes (A4), international themes (A5), and local themes (A6) have a great degree of influence and being influenced, and their changes will have a greater impact on the dissemination of rural revitalization literature.

At the same time, in order to more effectively analyze the relationship between the degree of centrality and the degree of cause among the factors, the causality diagram is drawn according to the calculation results in Table 1. The specific content includes: take the center degree as the horizontal axis, the cause degree as the vertical axis to establish the Cartesian coordinate system, the center degree and cause degree value of each influencing factor corresponds to the coordinate system, and the obtained center degree-cause degree Cartesian coordinate system diagram is shown in Figure 2.

Figure 2.

The degree of center - reason diagram

The centrality degree reflects the importance of each influential factor in the dissemination of literary works. According to the analysis of the order of the centrality degree, cultural and historical themes (A1), character themes (A2), international themes (A5), political and economic themes (A4), nodal selections (A8), national themes (A7), and video dissemination (B1) are ranked in the first eight places, and the results obtained are verified by the relevant experts in the field of literary dissemination, which basically conform to the current situation of international dissemination of literature on China’s rural revitalization. After the results were verified by relevant experts in the field of literature dissemination, they were basically in line with the current international situation of China’s rural revitalization literature. These communication factors largely influence the international communication process of Chinese rural revitalization literature.

The degree of causality reflects the relationship between the factors in the international dissemination of rural revitalization literature. According to the analysis of the order of the degree of cause, national theme (A7), obvious emotional bias (B6), audio communication (B2), single episode/single story style (A4), nodal selection (A8), character theme (A2), cultural-historical theme (A1), series style (B5), single episode/single story style (B4), social theme (A3), international theme (A5), no A positive value for significant affective bias (B7) indicates a causal influence factor. In particular, the national theme (A7) has the largest value of causality, which significantly influences the other indicators, while the values of the remaining indicators will be influenced by other factors, which should focus on literary themes with Chinese characteristics and vision in the process of international dissemination of rural revitalization literature.

By analyzing the influencing factors of the international dissemination of rural revitalization literature, this paper selects content theme (CT), topic nature (SM), time node (TN), writing form (WF), emotional bias (ET), and narrative perspective (NP) as the antecedent variables, and international dissemination effect (ISE) as the outcome variable, and applies FsQCA to conduct a cohort analysis to further analyze whether certain antecedent variables in the combinations whether they have an effect on the international dissemination of rural revitalization literature.

Data calibration is a key step in performing FsQCA analysis in order to convert the data into a fuzzy affiliation degree, which takes values between 0 and 1 and represents whether and to what extent a case belongs to the fuzzy set formed by this variable. When the value is close to or equal to 1, the case (sample) is considered to be fully affiliated to this set. With a value close to or equal to 0, it is considered that the case (sample) is not affiliated with this set at all. With a value of 0.5, the case (sample) is considered to belong to both fuzzy and non-fuzzy sets, also known as fuzzy affiliation.

After the initial processing of the data, the values corresponding to the 95% quartile, 50% quartile, and 5% quartile in the data of each latent variable were first calculated and they were used as the full affiliation, fuzzy affiliation, and full non-affiliation points, respectively, for that latent variable, and the data calibration anchors for the variables are shown in Table 2.

The preliminarily processed data were imported into FsQCA3.0 software, and at the same time, the calibrate function function that comes with FsQCA3.0 software was utilized to complete the data calibration of each variable by inputting the complete affiliation point, the fuzzy affiliation point, and the complete unaffiliated point, respectively. Considering that FsQCA3.0 software automatically ignores the sample (case) with data calibrated to 0.5, the value calibrated to 0.5 is manually adjusted to 0.501 to ensure the validity of all the data as well as the completeness of the results of the grouping analysis, and the specific data calibration results are shown in Table 3.

Before conducting FsQCA analysis on the calibrated data, it is also necessary to conduct necessity analysis on each antecedent variable as a way to understand whether there exists a certain antecedent variable that always exists or always does not exist in the process of influencing the international dissemination of rural revitalization literature. In the necessity analysis, consistency and coverage are used as key indicators. Consistency represents the extent to which the antecedent variable (independent variable) is a necessary condition for the production of a particular outcome (dependent variable), reflecting the likelihood of the independent variable as a necessary condition. Coverage is the proportion of sample cases that cover the necessary condition variable and reflects the explanatory power of the variable. Coverage and consistency need to be considered together when determining necessity, with high consistency and high coverage implying that the number of cases presenting the condition is proportional to the number of cases forming the outcome, representing a high degree of importance of the necessary condition. High consistency and low coverage indicate that the number of cases presenting the condition is not proportional to the number of cases presenting the result, representing that the necessary condition is not very important. The antecedent variable is determined to be a necessary condition when the threshold of consistency reaches 0.9 or more. In this paper, through the Necessary Conditions analysis function in FsQCA3.0 software, the calibrated data are analyzed for necessity, and the specific results are shown in Table 4.

From the results, it can be seen that the consistency values of the presence and absence of the six antecedent variables are lower than 0.9, and the content theme (CT), nature of the topic (SM), time node (TN), writing form (WF), emotional bias (ET), and narrative perspective (NP) are not the necessary conditions affecting the international dissemination of the rural revitalization literature, and they cannot individually explain the reasons for the formation of the results. Therefore, this paper will further analyze the role of the combination of conditions on the international dissemination of rural revitalization literature.

1)

Constructing the truth table

Constructing a truth table is the basis of FsQCA analysis, which aims to judge and assess the adequacy of the grouping on the outcome variables, so it is necessary to construct a truth table to categorize and summarize the sample cases into different combinations of conditions (groupings) before grouping analysis. In this study, six influencing factors in the model of influencing factors for international dissemination of rural revitalization literature were selected as antecedent variables, which may exist in 26=64 configurations. The 556 sample cases were categorized and generalized into these 64 configurations (there are cases where there are no cases in some groupings) to get the truth table used for grouping analysis.

After the initial generalization, there are some groupings in the truth table that have low consistency and cover a small number of sample cases. Low consistency indicates that the probability of the grouping resulting in an outcome is low, so a threshold needs to be set to encode the outcome variable corresponding to the grouping with low consistency as 0, representing that the outcome does not exist. In this study, the consistency threshold is set to 0.8, and the case frequency threshold is set to 6. At the same time, based on the truth table initially constructed by the FsQCA3.0 software, this paper manually assigns the result codes with a consistency value of PRI of more than 0.75 as 1, and those with a consistency value of less than 0.75 as 0, to obtain the final truth table. Table 5 shows part of the truth table results, where: ● represents the existence of the core condition, ● represents the existence of the edge condition, and blank represents that the condition may or may not exist.