Intelligent model for contemporary supply chain barriers in manufacturing sectors under the impact of the COVID-19 pandemic.

The COVID-19 pandemic has cast a shadow on the global economy. Since the beginning of 2020, the pandemic has contributed significantly to the global recession. In addition to the health damages of the pandemic, the economic impacts are also severe. The consequences of such effects have pushed global supply chains toward their breaking point. Industries have faced multiple obstacles, threatening the fragile flow of raw materials, spare parts, and consumer goods. Previous studies showed that supply chain barriers have multi-faceted impacts on industries and supply chains, which demand appropriate measures. In this regard, seven major barriers that directly impact industries have been identified to determine which industry is most affected by the COVID-19 pandemic. This paper utilized a hybrid multi-criteria decision-making (MCDM) approach under a neutrosophic environment using trapezoidal neutrosophic numbers to rank those barriers. The Analytical Network Process (ANP) quantifies the effects and considers the interrelationships between the determined barriers (criteria) involved in decision-making. Subsequently, the Measurement Alternatives and Ranking according to the COmpromise Solution (MARCOS) method was adopted to rank six industries according to the impact of those barriers. Results show that the lack of inventory is the largest barrier to influencing industries, followed by the lack of manpower. Sensitivity analysis is performed to detect the change in the rank of industries according to the change in the relative importance of the barriers.

Introduction

The world has faced major challenges since December 2019 when the Coronavirus disease 2019 (COVID-19) emerged (Lu et al., 2020). On March 11, 2020, the World Health Organization declared the COVID-19 outbreak a pandemic (García-Basteiro et al., 2020, Cayvaz et al., 2022). The COVID-19 pandemic has forced factories worldwide to slow or halt production (Karmaker et al., 2021). As a result, every aspect of the operation in the global supply chain has been disrupted.

Logistics and supply chains worldwide became mired with the onset of the COVID-19 pandemic. The situation worsened with the continuation of the pandemic, affecting many sensitive strategic sectors such as health care, transportation, spare parts, defense, and security (De Vito & Gómez, 2020). Events in 2019, which has become known as the “year of the epidemic,” showed that the pace of disruption in supply chains accelerated, with governments around the world turning to the domestic protection front, the risk was great (Chenarides et al., 2021). Each country sought to secure the necessary reserve of essential consumer goods and commodities and products outside the direct consumer range. The disruption inevitably reached the associated logistic sectors with the affected supply chains. Thus, the main problem is that despite all expectations for the growth of global commodity trade in the next two years, the risks arising from the pandemic continue to impact supply chains (Ivanov, 2021).

COVID-19 has affected the fundamental elements in the global supply chains on which the globalization of industrial production depends (Grida et al., 2020). Since the outbreak of COVID-19, the industrial sector has faced many new difficulties in addition to existing challenges (Ivanov, 2020). The effects were apparent in industries that depend on production requirements imported from abroad. Moreover, partial closures led to a decline in the activity and productivity of factories, thereby reducing sales and revenues (Belhadi et al., 2021). As a result, supply chains have been affected by multiple policies and various barriers that have involved many industries to avoid the spread of the disease, whether in supply, demand, or logistics (Biswas & Das, 2020). In this study, the impact of various barriers identified (i.e., lack of inventory, lack of transportation, local law enforcement, scarcity of raw materials, fluctuation of demand, deficiency in cash flow in the market, and lack of manpower) will be evaluated on a number of major industries in the world under an uncertain environment.

Given that supply chain barriers and obstacles have many facets and are imprecise, illustrating and estimating the possible significance of those barriers are difficult. Therefore, collecting accurate and reliable data to identify the most affected industry during the COVID-19 pandemic is an important and challenging task (Cayvaz et al., 2022). Thus, neutrosophic theory can be applied to deal with reliable, non-specific, quantitative, and complex data to observe which industries are most affected during the pandemic. Smarandache (1998) introduced neutrosophic theory to deal with ambiguity, indeterminacy, and inaccuracy in everyday problems. The neutrosophic approach has been applied in many scientific studies and research related to real problems. For example, Basha et al. (2021) proposed a neutrosophic rule-based classification model to generate a set of rules to diagnose COVID-19 patients on the basis of their chest X-ray images. Cayvaz et al. (2022) suggested an approach based on neutrosophic fuzzy Decision Making Trial and Evaluation Laboratory to investigate, analyze, and evaluate the significance degrees of various disruption factors. In this regard, the key feature of neutrosophic theory is its ability to deal with indeterminacy as a separate part of the process of assessing uncertainty (Abdel-Basset et al., 2021a). Indeterminacy cannot be dealt with by using fuzzy theory, because the theory always expresses only degrees of truth and falsity (Zadeh, 1965). Moreover, intuitionistic fuzzy theory (Atanassov, 1999) cannot be used to deal with indeterminacy. In this regard, the literature has introduced various neutrosophic scales, such as the bipolar, triangular, interval, trapezoidal numbers, and type-2 neutrosophic numbers (Abdel-Basset et al., 2019). Given that neutrosophic scales consist of three parts, namely, truthness, falseness, and indeterminacy, the appropriate scale is selected according to the nature of the problem. Accordingly, if the study requires to focus on the value of truthness, then the appropriate scale is selected for that purpose; if the study requires focusing on the value of indeterminacy, then the appropriate scale is used; and if the study requires showing the value of falseness, then the appropriate scale is applied (Abdel-Basset et al., 2019). Accordingly, the trapezoidal neutrosophic numbers (TNNs) scale was used to present the value of indeterminacy, as it is expressed as two values within the scale. In this study, the TNNs have been employed to express ambiguity and uncertainty in the problem. Many studies and research have used TNNs. For example, Haque et al. (2021) introduced a novel logarithmic operational law and aggregation operators by using TNN with multi-criteria group decision-making approach to identify the most harmful virus.

Although scholars have actively sought practical solutions to the COVID-19 pandemic, no study has attempted to identify the industries that are most affected by the pandemic. Identifying and prioritizing affected industries require consideration of a large number of often incompatible criteria whose relationship with one another is complex and challenging. Therefore, the multi-criteria decision-making (MCDM) approach is the ideal approach. MCDM has been used to provide an optimal solution to many problems (Campos-Guzmán et al., 2019). Pintelon et al. (2021) presented a new hybrid MCDM methodology for a medical device prototype's risk priority number assessment. Barua et al. (2021) developed a combined MCDM framework to assess barriers to implementing lean, agile, resilient, and sustainable paradigms in the supply chain. Ghorui et al. (2021) developed a hesitant fuzzy MCDM approach for determining the dominant risk element involved in the spread of COVID-19.

In this study, the problem of identifying industries that are affected mainly by the COVID-19 pandemic is formulated as an MCDM problem under a neutrosophic environment. The suggested approach consists of two decision-making methods, namely, the Analytical Network Process (ANP) (Saaty & Vargas, 2013) and the Measurement Alternatives and Ranking according to COmpromise Solution (MARCOS) (Stević et al., 2020) on TNNs. In this regard, using an approach that includes multiple assessment methods helps researchers obtain accurate and realistic results (Amirghodsi et al., 2020). First, the ANP is used to assess seven barriers coupled with TNNs to deal with uncertainties. Saaty developed this method in order to deal with difficult decision-making issues. Interdependencies among selection criteria are created by the network relation of the ANP method, and then the weight of the criteria is computed (Balali et al., 2021). This method is superior in comparison to other MCDM methods as it can create interdependencies among criteria and compute the particular weights of each criterion (Balali et al., 2021). Second, MARCOS is used to rank industries by determining the relationship between alternatives and reference values (ideal and anti-ideal alternatives). This technique is based on measuring alternatives and their rating in relation to a compromise solution (Stanković et al., 2020). The compromise solution entails determining utility functions based on the distance between anti-ideal and ideal solutions and their aggregations (Stević et al., 2020). Furthermore, TNNs have been combined with the MARCOS method to be realistic and accurate in dealing with uncertainty.

This study makes the following contributions to the current literature.

The proposed framework can be used to determine reliable answers in uncertain cases.

The study gives insights into the prospective MCDM methodology.

The study presents an integrated neutrosophic–ANP–MARCOS approach.

The neutrosophic-based ANP framework is utilized to assess the barrier (criteria) weights.

The study clarifies the applicability of the suggested neutrosophic–ANP–MARCOS framework by using an empirical case study to determine the industries most affected by the COVID-19 pandemic within the context of neutrosophic theory.

The study applies sensitivity analysis to show the steadiness of the proposed approach.

The rest of this paper is divided as follows. Section 2 presents the background of the study. Section 3 clarifies the assessment barriers for industries during the COVID-19 pandemic. Section 4 introduces the procedures of the hybrid framework based on TNNs for determining the industries most affected by the COVID-19 pandemic. Section 5 discusses the assessment procedures. Section 6 discusses the sensitivity analysis, and Section 7 presents the conclusions and future research directions.

Literature review

Researchers have been actively proposing solutions to problems arising from the COVID-19 pandemic. Belhadi et al. (2021) investigated supply chain resiliences in the automobile and airline industries during the COVID-19 pandemic. Singh et al. (2021) developed a simulation model of the public distribution system network with three scenarios to show the impact of COVID-19 on logistics systems and disruptions in the food supply chain. Shahed et al. (2021) introduced a mathematical model to reduce disruptions in the supply chain network subject to a natural disaster like the COVID-19 pandemic. Qin et al. (2021) investigated the effects of the COVID-19 pandemic and public health expenditure on supply chain procedures. Their study applied several models, such as the generalized method of moments (GMM), fixed effect, and random effect, in evaluating processes. Alkahtani et al. (2021) developed a novel COVID-19 supply chain management approach grounded on variable production under uncertain environment states to benefit the industrialization corporation. Butt (2021) introduced strategies to reveal the effect of the COVID-19 pandemic on the supply chain disturbances taken by distributing and buying corporations. They investigated countermeasures to supply chain disturbances. Moreover, Moosavi and Hosseini (2021) presented a discrete event simulation model to cope with supply chain disturbances. They investigated supply chain disturbances in a real case study during the COVID-19 pandemic.

As some information about the pandemic is tainted by ambiguity, inaccuracy, and indeterminacy, scholars have been using neutrosophic theory to deal with such ambiguity. Neutrosophic theory is well known in dealing with data ambiguity, as it considers the principle of indeterminacy. Samad et al. (2021) proposed a methodology for determining the best hand sanitizer to decrease COVID-19 impact. Their methodology is based on the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) method under neutrosophic hypersoft set. Karuppiah et al. (2021) investigated supply chains' challenges during the COVID-19 pandemic. They determined 20 challenges and three strategies through literature reviews. They applied the MCDM methodology that consisted of the analytical hierarchy process technique for assessing challenges and TODIM, an acronym in Portuguese for interactive multicriteria decision-making technique, to determine the preferred strategy under a neutrosophic environment. Sherwani et al. (2021) introduced a test approach for analyzing COVID-19 data by conducting neutrosophic Kruskal–Wallis H Test. They conducted the daily intensive care unit residency test by COVID-19 patients. Çakır et al. (2021) proposed an MCDM methodology for determining vaccination clinics under a neutrosophic environment.

Paul et al. (2021) developed a methodology to anticipate difficulties in supply chains during recovery, especially in the Bangladeshi garment industry. Their methodology applied the Decision Making Trial and Evaluation Laboratory (DEMATEL) based on the Delphi method to determine supply chain recovery difficulties. Nguyen et al. (2021) presented an MCDM methodology to determine governmental intrusion policies for dealing with the COVID-19 pandemic in Vietnam. Tarei & Kumar (2021) developed a decision-making approach for evaluating different dimensions and barriers that have affected the admission process in management educational institutions during the COVID-19 pandemic in India. They applied two MCDM methods; ANP was applied to determine the weights of the dimensions and barriers, and the DEMATEL method is applied to consider the interdependency among the dimensions and highlights the significance of each barrier. Gupta and Soni (2021) developed a decision-making strategy for dealing with supply chain finance after the COVID-19 crisis. They applied their approach to a case study in the food industry to determine the critical factors that are significant to acclimating to the sustainable supply chain. Salehi et al. (2021) introduced an MCDM hybrid approach consisting of ANP and DEMATEL techniques for analyzing and determining information technology barriers in the supply chain of sugarcane in Khuzestan province. Magableh & Mistarihi (2022) proposed an MCDM methodology based on the ANP-TOPSIS approach for helping the organizations to analyze the impact of COVID-19 on supply chains prioritize solutions based on their relative significance. Shanker et al. (2021) presented an approach to analyze the factors influencing perishable food supply chains during the COVID-19 pandemic and improve their resiliency by applying the grey-ANP method. Ahmad et al. (2021) proposed a hybrid MCDM framework for determining strategies to tackle the COVID-19 pandemic.

Many studies related to MCDM have used the MARCOS method despite its recentness. The MARCOS method maintains stability even if the number of criteria and alternatives is large. In addition, it suggests a potential compromise solution, which is closest to the best (Stević & Brković, 2020). Recently, Ecer and Pamucar (2021) presented a novel methodology for identifying the performance of insurance firms during the COVID-19 pandemic. Their approach applied the MARCOS technique for ranking insurance firms in the healthcare services field.

In summary, our study proposes an MCDM approach to the problem of assessing the industries most affected by the COVID-19 pandemic. The proposed approach used two methods of decision-making. First, the ANP method is used to evaluate and determine the weights of the barriers that affect the industry’s performance during a crisis. Second, the MARCOS method is used to evaluate industries' performance and rank them according to the weights of the barriers. The study is conducted under a neutrosophic environment to deal with ambiguity, indeterminacy, and inaccuracy in the data. Fig. 1 . presents the structure of the research methodology.

Fig. 1

General structure of the research methodology.

Industry barriers during the COVID-19 pandemic

This section will explain the barriers to industries that affect supply chains during the COVID-19 pandemic. We marked these barriers with the letter “B”. These barriers are listed as follows:

Lack of inventory ()

Market fluctuations and uncertainty during the COVID-19 pandemic led to a lack of inventory in some industries, which leads to disruption of production and a decline in exports. Therefore, the lack of inventory is a barrier to industries during the crisis.

Lack of transportation ()

During the COVID-19 pandemic, transportation problems are major barriers to industrialization, as international and local travel has been suspended to prevent the spread of the virus (Cayvaz et al., 2022). In this regard, supply chain bottlenecks, factory disruptions, and lack of other inputs have slowed down manufacturing and recovery efforts in demand.

Local law enforcement ()

Since the emergence of COVID-19, economic and social repercussions would be significant and influential at the global and national levels (Biswas & Das, 2020). In this regard, the outbreak forced governments to take a series of internal and external precautionary measures, such as isolation and quarantine, social distancing, travel bans, and the complete closure of all state institutions (Singh et al., 2021). These measures reflected negatively on economies worldwide and have brought the global system into a state of stagnation, which impacted the economic system and caused the decline of industries.

Scarcity of raw materials ()

The global trade movement, especially at the level of providing manufacturing materials, has been disrupted due to the COVID-19 pandemic. This disruption of trade movement and the accumulation of containers (without shipment) in ports reduced the stock of raw materials in factories. Raw materials are essential in establishing industries; thus, their lack is a barrier (Cayvaz et al., 2022).

Fluctuation of demand ()

The COVID-19 pandemic has led to disruption and a downturn in the economic system due to the precautionary measures to contain the virus. This disruption led to an increase in volatility in global financial markets, a fall in the price of assets, and a significant decline in market liquidity. The volatility in demand affects industries; thus, such instability is a barrier.

Deficiency in cash flow in the market ()

Cash flow is a key factor in setting up the economy and industries. The COVID-19 pandemic has significantly affected cash flow in organizations and banks due to lockdowns (Biswas & Das, 2020).

Lack of manpower ()

The workforce is an essential part of any industry. Labour shortages were widespread during the COVID-19 pandemic, and workers suffered abuse to sustain supply chains (Biswas & Das, 2020). Companies are dealing with a massive wave of resignations among employees. This variable, which affects productivity, is a barrier to industries during the COVID-19 pandemic.

Research methodology

This section provides a detailed explanation of the proposed model to determine the industries most affected by the COVID-19 pandemic. The proposed model consists of the ANP method and the MARCOS method. The former is used to calculate the weights of the criteria that affect the identification of alternatives, and the latter is used to determine the ranking of alternatives affected by the COVID-19 pandemic under a neutrosophic environment using TNNs. Fig. 2 . shows the steps of the proposed model.

Fig. 2

Steps of the proposed model.

Step 1: Three experts were identified to collect the necessary information on the problem. Some were interviewed in open interviews, and some were contacted online. The selected experts have sufficient experience in supply chains and decision-making. Some of them are academics, and others have been working in the business for more than 15 years. They shared their knowledge, skills, experiences, and judgment until they reached a consensus (Gumus, 2009). In this regard, the experts collaborated with the authors during the research period and obtained actual results.

Step 2: Determine a set of criteria on the basis of the information gathered from experts and a literature review to determine the alternative most affected by the COVID-19 pandemic. Finalizing the number of alternatives to be evaluated according to the specified criteria is another task in this step.

Step 3: Determine the linguistic terms and their equivalent TNNs to evaluate the criteria and determine their weights, as in Table 1 . Select a set of linguistic terms to evaluate alternatives according to the criteria and to identify the alternative most affected by the COVID-19 pandemic, as shown in Table 2 .

Table 1

Evaluation scale for weighing barriers.

Linguistic terms	Definitions	TNN 〈lo,mo_1,mo_2,up;α,β,θ〉
Equal significance	〈ESB〉	〈0.20,0.15,0.25,0.20;0.50,0.15,0.25〉
Moderate significance	〈MSB〉	〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉
Strong significance	〈SSB〉	〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉
Very strong significance	〈VSB〉	〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉
Extreme significance	〈EMB〉	〈1.00,0.90,0.90,1.00;0.90,0.05,0.05〉

Table 2

Evaluation scale for alternatives.

Linguistic terms	Definitions	TNN 〈lo,mo_1,mo_2,up;α,β,θ〉
Equally important	〈EIB〉	〈0.10,0.20,0.20,0.30;0.20,0.10,0.15〉
Equally to moderately important	〈EPB〉	〈0.20,0.35,0.35,0.40;0.40,0.40,0.35〉
Moderately important	〈MIB〉	〈0.30,0.45,0.50,0.55;0.50,0.25,0.35〉
Moderately to robustly important	〈MRB〉	〈0.40,0.50,0.55,0.65;0.60,0.25,0.25〉
Strongly important	〈SIB〉	〈0.50,0.65,0.65,0.70;0.70,0.15,0.15〉
Strongly to very robustly important	〈SVB〉	〈0.60,0.70,0.75,0.75;0.80,0.15,0.15〉
Very robustly important	〈VRB〉	〈0.70,0.75,0.80,0.85;0.90,0.10,0.10〉
Very robustly to extremely important	〈VEB〉	〈0.80,0.85,0.85,0.90;0.95,0.10,0.10〉
Extremely important	〈EXB〉	〈1.00,0.90,0.90,1.00;1.00,0.0,00.00〉

Step 4: Create a network model concerning the objective, criteria, and alternatives as shown in Fig. 3 . The MCDM problem constructs as a network and includes defining the elements, criteria, and alternatives. It involves the aggregation of an expert panel to show the interdependency between criteria and feedback between network elements.

Fig. 3

Network model of barriers according to six industries.

Step 5: Create a decision matrix (D) by experts between the criteria to detect the impact of each criterion on another and obtain the relative significance of criteria by using the linguistic terms (shown in Eq. [1]) and neutrosophic scales (shown in Eq. [2]). The number of comparisons N can be computed as n(n-1)/2. The eigenvector of the pairwise comparison matrix is utilized in the supermatrix. where [ ], r, t = 1, 2… n, and is the performance valuation of the criterion of the criterion , , …, concerning the criterion , , …, . The diagonal criteria of the matrix are equal to 0.5, such that  = 0.5.

where is a single-valued neutrosophic set in which lo, , , up to represent to the lower,, , and upper of neutrosophic number, with truth, indeterminate, and falsity membership , and respectively (Abdel-Basset et al., 2021b).

Step 6: Conduct de-neutrosophication of the TNNs to crisp values by applying score function s(x) by using Eq. (3) to build the evaluation matrix among criteria using crisp values.

Step 7: Compute the weights matrix. Normalize the elements of each column by dividing each element by the sum of the column. Then, the elements in each resulting row are added, and this sum is divided by the number of elements in the row as shown in Eq. (4).

Step 8: Calculate the consistency ratio (CR) on the basis of the decision matrix by applying Eq. (5) to assess the consistency of comparison.

where , is the mean of the weighted sum vector divided by the corresponding criterion, and n is the number of criteria. RI is a random index assigned with the number of selected criteria considered, as shown in Table 3 . If CR is less than or equal to 0.1, then the level of consistency is adequate; otherwise, the comparison is inconsistent and the expert is recommended to change the comparison components to identify more suitable consistency (Asadabadi et al., 2019).

Table 3

Randomness index (Saaty, 1977).

N	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15
RI	0	0	0.58	0.9	1.12	1.24	1.32	1.41	1.45	1.49	1.51	1.48	1.56	1.57	1.59

Step 9: Determine the terminal weights of the criteria by using the ordered weighted aggregation (OWA) operator (Kacprzyk et al., 2019) of the terminal weights of experts.

Step 10: Obtain the sorted vector according to the obtained weights in Eq. (4). If the obtained weights above pass the consistency test, it will be formed as a matrix, local weight matrix can be obtained as in Eq. (6).

Step 11: Obtain the internal interdependency of the criterion and other criteria sets as shown in Eq. (7), and then get the unweighted supermatrix which composed of the sorting vectors affected by each criterion (Yuan et al., 2019). The unweighted supermatrix involves only the indirect influence, but not the intermediate elements, with the effect among a pair of elements. The resulting matrices are applied to identify the influence of the relations between interdependent criteria.

Step 12: Obtain the final weights of the criteria by multiplying the local weight obtained from experts' comparison matrices of criteria by the weight of interdependence of criteria as shown in Eq. (8).

Step 10: Create a decision matrix by experts between criteria and alternatives to determine the alternative most affected by the COVID-19 pandemic by employing the linguistic terms according to Eq. (9) then by neutrosophic scales as presented in Eq. (10).

where i is the number of industries (alternatives) and j is the number of barriers (criteria). If is the performance valuation of the alternative , , …, concerning the criteria , , …, .

Step 11: Conduct de-neutrosophication of the TNNs to crisp values by applying score function s(x) by using Eq. (3) to build the evaluation matrix among criteria and alternatives using crisp values.

Step 12: Create an extended decision matrix as shown in Eq. (11) by determining the ideal (AI) and unideal (AAI) solutions, which are the worst alternative and the best alternative, respectively, by applying Eqs. (12), (13).

where B denotes benefit criteria and C denotes cost criteria.

Step 13: Obtain the normalized decision matrix as shown in Eq. (14) by using Eqs. (15), (16).

where elements and represent the elements of the matrix P.

Step 14: Calculate the weighted matrix V =  by multiplying the normalized matrix Q with the weight coefficients of the barrier as applied in Eq. (17).

Step 15: Compute the utility degree of alternatives by using Eqs. (18), (19).

where (t = 1, 2… j) denotes the sum of the elements of the weighted matrix V according to Eq. (20).

Step 16: Define the utility function of alternatives ƒ () according to Eq. (21). The utility function consists of the observed alternative concerning the ideal and unideal solutions.

where f denotes the utility function in relation to the unideal solution as shown in Eq. (22), whereas f denotes the utility function in relation to the ideal solution as shown in Eq. (23).

Step 17: Rank the alternatives according to the highest possible value of the utility function.

Application of the methodology

This section applies the proposed neutrosophic–ANP–MARCOS methodology to identify the most affected industry during the COVID-19 pandemic. The industries evaluated are: pharmaceutical, medical devices and supplies, tourism, textile, food, electronics, and heavy equipment, which are labeled , , , , , and , respectively. We will list the steps to implement the proposed methodology, starting from collecting data and identifying the barriers to arranging industries according to the extent of their impact from these barriers.

Step 1: Three experts (i.e., academics and economists) in the field of the supply chain, industry, and decision-making were identified. The experts cooperated with the authors until actual results were obtained during the research term.

Step 2: A set of barriers affecting industries under the COVID-19 pandemic has been identified according to a review of literature and the experts. The barriers identified are lack of inventory (), lack of transportation (), local law enforcement (), scarcity of raw materials (), fluctuation of demand (), deficiency in cash flow in the market (), and lack of manpower (). Six industries would be evaluated according to the identified barriers.

Step 3: Linguistic scales and their equivalent TNNs are defined in two parts. First, a set of linguistic scales and their equivalent TNNs were specified to evaluate the barriers and determine the relative importance of each barrier, as shown in Table 1. Second, a set of linguistic scales and their equivalent TNNs were specified to determine the industry most affected during the COVID-19 pandemic, as presented in Table 2.

Step 4: A network model concerning the objective, criteria, and alternatives has been created, as shown in Fig. 3. The set of assessment criteria used in the problem was defined, and the main objective of the problem was described, which is to identify the industry most affected by the COVID-19 pandemic, and the network structure was designed based on expert opinions and literature searches. Also, the complicated problem has been broken down into a logical system as a network to analyze the interdependencies between criteria.

Step 5: A decision matrix of the barriers was created separately for each expert to reveal the effect of each barrier on the other by using the linguistic scales in Table 1 according to Eq. (1) as presented in Table A.1 (See Appendix A). Then, a decision matrix was created for each expert separately between the barriers to reveal the effect of each barrier on the other using TNNs in Table 1 according to Eq. (2), as presented in Table A.2 (See Appendix A).

Table A.1

Evaluation of barriers using linguistic terms.

Expert_1	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP
B_LIY	–	〈EMB〉	〈VSB〉	〈SSB〉	〈VSB〉	〈EMB〉	〈ESB〉
B_LTN	1/〈EMB〉	–	〈MSB〉	〈VSB〉	〈MSB〉	〈SSB〉	〈SSB〉
B_LLE	1/〈VSB〉	1/〈MSB〉	–	〈EMB〉	〈ESB〉	〈VSB〉	〈MSB〉
B_SRM	1/〈SSB〉	1/〈VSB〉	1/〈EMB〉	–	〈SSB〉	〈ESB〉	〈SSB〉
B_FDD	1/〈VSB〉	1/〈MSB〉	1/〈ESB〉	1/〈SSB〉	–	〈SSB〉	〈EMB〉
B_DCF	1/〈EMB〉	1/〈SSB〉	1/〈VSB〉	1/〈ESB〉	1/〈SSB〉	–	〈MSB〉
B_LMP	1/〈ESB〉	1/〈SSB〉	1/〈MSB〉	1/〈SSB〉	1/〈EMB〉	1/〈MSB〉	–
Expert_2	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP
B_LIY	–	〈EMB〉	〈VSB〉	〈MSB〉	〈VSB〉	〈VSB〉	〈ESB〉
B_LTN	1/〈EMB〉	–	〈MSB〉	〈VSB〉	〈MSB〉	〈SSB〉	〈SSB〉
B_LLE	1/〈VSB〉	1/〈MSB〉	–	〈EMB〉	〈ESB〉	〈VSB〉	〈MSB〉
B_SRM	1/〈MSB〉	1/〈VSB〉	1/〈EMB〉	–	〈SSB〉	〈ESB〉	〈VSB〉
B_FDD	1/〈VSB〉	1/〈MSB〉	1/〈ESB〉	1/〈SSB〉	–	〈SSB〉	〈EMB〉
B_DCF	1/〈VSB〉	1/〈SSB〉	1/〈VSB〉	1/〈ESB〉	1/〈SSB〉	–	〈MSB〉
B_LMP	1/〈ESB〉	1/〈SSB〉	1/〈MSB〉	1/〈VSB〉	1/〈EMB〉	1/〈MSB〉	–
Expert_3	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP
B_LIY	–	〈VSB〉	〈VSB〉	〈SSB〉	〈VSB〉	〈EMB〉	〈MSB〉
B_LTN	1/〈VSB〉	–	〈MSB〉	〈VSB〉	〈MSB〉	〈SSB〉	〈SSB〉
B_LLE	1/〈VSB〉	1/〈MSB〉	–	〈EMB〉	〈ESB〉	〈EMB〉	〈MSB〉
B_SRM	1/〈SSB〉	1/〈VSB〉	1/〈EMB〉	–	〈SSB〉	〈ESB〉	〈SSB〉
B_FDD	1/〈VSB〉	1/〈MSB〉	1/〈ESB〉	1/〈SSB〉	–	〈SSB〉	〈EMB〉
B_DCF	1/〈EMB〉	1/〈SSB〉	1/〈EMB〉	1/〈ESB〉	1/〈SSB〉	–	〈MSB〉
B_LMP	1/〈MSB〉	1/〈SSB〉	1/〈MSB〉	1/〈SSB〉	1/〈EMB〉	1/〈MSB〉	–

Table A.2

Evaluation of barriers using TNNs

Expert_1	B_LIY	B_LTN	B_LLE	B_SRM
B_LIY	–	〈1.00,0.90,0.90,1.00;0.90,0.05,0.05〉	〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉	〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉
B_LTN	1/〈1.00,0.90,0.90,1.00;0.90,0.05,0.05〉	–	〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉	〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉
B_LLE	1/〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉	1/〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉	–	〈1.00,0.90,0.90,1.00;0.90,0.05,0.05〉
B_SRM	1/〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉	1/〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉	1/〈1.00,0.90,0.90,1.00;0.90,0.05,0.05〉	–
B_FDD	1/〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉	1/〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉	1/〈0.20,0.15,0.25,0.20;0.50,0.15,0.25〉	1/〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉
B_DCF	1/〈1.00,0.90,0.90,1.00;0.90,0.05,0.05〉	1/〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉	1/〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉	1/〈0.20,0.15,0.25,0.20;0.50,0.15,0.25〉
B_LMP	1/〈0.20,0.15,0.25,0.20;0.50,0.15,0.25〉	1/〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉	1/〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉	1/〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉
Expert_1	B_FDD	B_DCF	B_LMP
B_LMP	〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉	〈1.00,0.90,0.90,1.00;0.90,0.05,0.05〉	〈0.20,0.15,0.25,0.20;0.50,0.15,0.25〉
B_LTN	〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉	〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉	〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉
B_LLE	〈0.20,0.15,0.25,0.20;0.50,0.15,0.25〉	〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉	〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉
B_SRM	〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉	〈0.20,0.15,0.25,0.20;0.50,0.15,0.25〉	〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉
B_FDD	–	〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉	〈1.00,0.90,0.90,1.00;0.90,0.05,0.05〉
B_DCF	1/〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉	–	〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉
B_LIY	1/〈1.00,0.90,0.90,1.00;0.90,0.05,0.05〉	1/〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉	–
Expert_2	B_LIY	B_LTN	B_LLE	B_SRM
B_LIY	–	〈1.00,0.90,0.90,1.00;0.90,0.05,0.05〉	〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉	〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉
B_LTN	1/〈1.00,0.90,0.90,1.00;0.90,0.05,0.05〉	–	〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉	〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉
B_LLE	1/〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉	1/〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉	–	〈1.00,0.90,0.90,1.00;0.90,0.05,0.05〉
B_SRM	1/〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉	1/〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉	1/〈1.00,0.90,0.90,1.00;0.90,0.05,0.05〉
B_FDD	1/〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉	1/〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉	1/〈0.20,0.15,0.25,0.20;0.50,0.15,0.25〉	1/〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉
B_DCF	1/〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉	1/〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉	1/〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉	1/〈0.20,0.15,0.25,0.20;0.50,0.15,0.25〉
B_LMP	1/〈0.20,0.15,0.25,0.20;0.50,0.15,0.25〉	1/〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉	1/〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉	1/〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉
Expert_2	B_FDD	B_DCF	B_LMP
B_LIY	〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉	〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉	〈0.20,0.15,0.25,0.20;0.50,0.15,0.25〉
B_LTN	〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉	〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉	〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉
B_LLE	〈0.20,0.15,0.25,0.20;0.50,0.15,0.25〉	〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉	〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉
B_SRM	〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉	〈0.20,0.15,0.25,0.20;0.50,0.15,0.25〉	〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉
B_FDD	–	〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉	〈1.00,0.90,0.90,1.00;0.90,0.05,0.05〉
B_DCF	1/〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉	–	〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉
B_LMP	1/〈1.00,0.90,0.90,1.00;0.90,0.05,0.05〉	1/〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉	–
Expert_3	B_LIY	B_LTN	B_LLE	B_SRM
B_LIY	–	〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉	〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉	〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉
B_LTN	1/〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉	–	〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉	〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉
B_LLE	1/〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉	1/〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉	–	〈1.00,0.90,0.90,1.00;0.90,0.05,0.05〉
B_SRM	1/〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉	1/〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉	1/〈1.00,0.90,0.90,1.00;0.90,0.05,0.05〉	–
B_FDD	1/〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉	1/〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉	1/〈0.20,0.15,0.25,0.20;0.50,0.15,0.25〉	1/〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉
B_DCF	1/〈1.00,0.90,0.90,1.00;0.90,0.05,0.05〉	1/〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉	1/〈1.00,0.90,0.90,1.00;0.90,0.05,0.05〉	1/〈0.20,0.15,0.25,0.20;0.50,0.15,0.25〉
B_LMP	1/〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉	1/〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉	1/〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉	1/〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉
Expert_3	B_FDD	B_DCF	B_LMP
B_LIY	〈0.80,0.85,0.85,0.90;0.90,0.10,0.10〉	〈1.00,0.90,0.90,1.00;0.90,0.05,0.05〉	〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉
B_LTN	〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉	〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉	〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉
B_LLE	〈0.20,0.15,0.25,0.20;0.50,0.15,0.25〉	〈1.00,0.90,0.90,1.00;0.90,0.05,0.05〉	〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉
B_SRM	〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉	〈0.20,0.15,0.25,0.20;0.50,0.15,0.25〉	〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉
B_FDD	–	〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉	〈1.00,0.90,0.90,1.00;0.90,0.05,0.05〉
B_DCF	1/〈0.60,0.55,0.65,0.80;0.90,0.25,0.50〉	–	〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉
B_LMP	1/〈1.00,0.90,0.90,1.00;0.90,0.05,0.05〉	1/〈0.40,0.25,0.45,0.50;0.70,0.25,0.35〉	–

Step 6: TNNs were converted to real values by using the score function using Eq. (3) as presented in Table A.3 (See Appendix A).Step 7: The weights of the barriers were calculated by each expert separately using Eq. (4) as shown in Table A.3 (See Appendix A). The consistency level of each matrix was verified by each expert using Eq. (5) and the RI in Table 3. Afterward, the consistency level was confirmed to be less than 0.1 for all the matrices in Table A.3 (See Appendix A).

Table A.3

Normalized matrix for barriers.

Expert_1	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP	Sum of rows	Weights	Consistency ratio = 0.098
B_LIY	0.034083	0.062743	0.052059	0.033148	0.140858	0.130556	0.046373	0.499821	0.071
B_LTN	0.077028	0.035368	0.019054	0.054417	0.051556	0.06859	0.154356	0.460369	0.066
B_LLE	0.089298	0.252529	0.034025	0.063096	0.025778	0.112599	0.092746	0.670071	0.096
B_SRM	0.146558	0.092665	0.076897	0.035567	0.085804	0.020606	0.154356	0.612452	0.087
B_FDD	0.089298	0.252529	0.48588	0.152938	0.092064	0.06859	0.293806	1.435104	0.205
B_DCF	0.077028	0.152083	0.089146	0.507896	0.395876	0.073594	0.092746	1.388369	0.198
B_LMP	0.486708	0.152083	0.24294	0.152938	0.208065	0.525464	0.165618	1.933815	0.276
Expert_2	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP	Sum of rows	Weights	Consistency ratio = 0.100
B_LIY	0.030731	0.062743	0.052059	0.021486	0.140858	0.114658	0.042194	0.464729	0.066
B_LTN	0.069453	0.035368	0.019054	0.058702	0.051556	0.069844	0.140446	0.444423	0.063
B_LLE	0.080516	0.252529	0.034025	0.068063	0.025778	0.114658	0.084388	0.659958	0.094
B_SRM	0.219422	0.092665	0.076897	0.038367	0.085804	0.020983	0.230561	0.764698	0.109
B_FDD	0.080516	0.252529	0.48588	0.164979	0.092064	0.069844	0.26733	1.413141	0.202
B_DCF	0.080516	0.152083	0.089146	0.547882	0.395876	0.07494	0.084388	1.424831	0.204
B_LMP	0.438844	0.152083	0.24294	0.100522	0.208065	0.535072	0.150693	1.828219	0.261
Expert_3	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP	Sum of rows	Weights	Consistency ratio = 0.099
B_LIY	0.044326	0.054584	0.052704	0.033148	0.140858	0.1239	0.088636	0.538157	0.077
B_LTN	0.116135	0.035676	0.01929	0.054417	0.051556	0.065093	0.147515	0.489682	0.070
B_LLE	0.116135	0.254727	0.034447	0.063096	0.025778	0.157843	0.088636	0.740662	0.106
B_SRM	0.190603	0.093471	0.07785	0.035567	0.085804	0.019556	0.147515	0.650366	0.093
B_FDD	0.116135	0.254727	0.491905	0.152938	0.092064	0.065093	0.280785	1.453647	0.208
B_DCF	0.100177	0.153407	0.07785	0.507896	0.395876	0.069842	0.088636	1.393684	0.199
B_LMP	0.316489	0.153407	0.245952	0.152938	0.208065	0.498673	0.158278	1.733802	0.248

Step 8: The final weights of the main barriers were determined by using the OWA operator as presented in Table 4 . Also, the obtained local weights calculated in Table 4 pass the consistency test, they have been formed as a matrix by applying the Eq. (6).

Table 4

The local weights of the applied barriers.

Barriers	Weights by expert_1	Weights by expert_2	Weights by expert_3	W_local
B_LIY	0.071	0.066	0.077	0.071
B_LTN	0.066	0.063	0.070	0.065
B_LLE	0.096	0.094	0.106	0.099
B_SRM	0.087	0.109	0.093	0.096
B_FDD	0.205	0.202	0.208	0.206
B_DCF	0.198	0.204	0.199	0.200
B_LMP	0.276	0.261	0.248	0.263

Step 9: According to the previous steps for calculating the local weights of the barriers (steps 5–9), pairwise comparison matrices were constructed to show the internal correlation of each criterion and its effect on the other set of criteria by all experts as presented in Tables (A.4-A.17) (See Appendix A). Also, the final weights of the internal interdependency of the criterion and other criteria are obtained in Table 5 according to Eq. (7).

Table 5

The comparative influence of the seven decision barriers.

	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP	Local weight	Final weight
B_LIY	0.095	0.111	0.114	0.268	0.217	0.296	0.074	0.071	0.170
B_LTN	0.065	0.199	0.201	0.104	0.152	0.104	0.059	0.065	0.115
B_LLE	0.103	0.150	0.160	0.189	0.165	0.139	0.140	0.099	0.149
B_SRM	0.074	0.072	0.078	0.077	0.079	0.079	0.169	0.096	0.102
B_FDD	0.192	0.078	0.084	0.099	0.145	0.118	0.213	0.206	0.145
B_DCF	0.202	0.174	0.140	0.137	0.130	0.151	0.180	0.200	0.156
B_LMP	0.270	0.216	0.223	0.126	0.112	0.112	0.166	0.263	0.163

Step 10: The final weights of the criteria have been obtained by multiplying the local weight obtained from experts' comparison matrices of criteria in Table 4 by the weight of interdependence of criteria as exhibited in Table 5 and Fig. 4 . by applying Eq. (8).

Fig. 4

Final weights of barriers obtained by using ANP.

Step 11: After determining the weights of the barriers, a decision matrix was created for each expert separately between the barriers and the industries used by using the linguistic scales in Table 2 according to Eq. (9) as presented in Table B.1 (See Appendix B). Each expert's decision matrix was created separately between the barriers and the industries used using the TNNs in Table 2 according to Eq. (10) as exhibited in Table B.2 (See Appendix B).

Table B.1

Decision matrices of six industries according to barriers using linguistic terms

Expert_1	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP
M_1	〈EPB〉	〈SVB〉	〈EXB〉	〈VRB〉	〈VEB〉	〈EPB〉	〈MRB〉
M_2	〈MIB〉	〈EPB〉	〈MIB〉	〈EIB〉	〈SIB〉	〈MRB〉	〈VEB〉
M_3	〈VRB〉	〈VEB〉	〈MRB〉	〈SVB〉	〈VRB〉	〈SVB〉	〈SVB〉
M_4	〈VRB〉	〈EPB〉	〈SVB〉	〈EPB〉	〈EXB〉	〈VEB〉	〈EIB〉
M_5	〈EXB〉	〈SVB〉	〈EIB〉	〈MIB〉	〈MIB〉	〈EXB〉	〈MIB〉
M_6	〈EIB〉	〈VEB〉	〈SIB〉	〈VRB〉	〈EXB〉	〈SIB〉	〈VRB〉
Expert_2	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP
M_1	〈VEB〉	〈SVB〉	〈EXB〉	〈VRB〉	〈VEB〉	〈EPB〉	〈MRB〉
M_2	〈MIB〉	〈EPB〉	〈MIB〉	〈EIB〉	〈SVB〉	〈MRB〉	〈VEB〉
M_3	〈VRB〉	〈VEB〉	〈VEB〉	〈SVB〉	〈VRB〉	〈SVB〉	〈SVB〉
M_4	〈VRB〉	〈EPB〉	〈SVB〉	〈EPB〉	〈EXB〉	〈VEB〉	〈EIB〉
M_5	〈EXB〉	〈SVB〉	〈EIB〉	〈MIB〉	〈MIB〉	〈EXB〉	〈MIB〉
M_6	〈EIB〉	〈VEB〉	〈SIB〉	〈VRB〉	〈EXB〉	〈SIB〉	〈VRB〉
Expert_3	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP
M_1	〈EPB〉	〈SVB〉	〈EXB〉	〈VRB〉	〈VEB〉	〈EPB〉	〈MRB〉
M_2	〈SVB〉	〈EPB〉	〈MIB〉	〈EIB〉	〈SVB〉	〈MRB〉	〈VEB〉
M_3	〈VRB〉	〈VEB〉	〈MRB〉	〈SVB〉	〈VRB〉	〈SVB〉	〈SVB〉
M_4	〈VRB〉	〈EPB〉	〈SVB〉	〈EPB〉	〈EXB〉	〈VEB〉	〈EIB〉
M_5	〈EXB〉	〈SVB〉	〈EIB〉	〈MIB〉	〈MIB〉	〈EXB〉	〈MIB〉
M_6	〈EIB〉	〈VEB〉	〈SIB〉	〈VRB〉	〈EXB〉	〈SIB〉	〈VRB〉

Table B.2

Decision matrices of six industries according to barriers using TNNs

Expert_1	B_LIY	B_LTN	B_LLE	B_SRM
M_1	〈0.20,0.35,0.35,0.40;0.40,0.40,0.35〉	〈0.60,0.70,0.75,0.75;0.80,0.15,0.15〉	〈1.00,0.90,0.90,1.00;1.00,0.0,00.00〉	〈0.70,0.75,0.80,0.85;0.90,0.10,0.10〉
M_2	〈0.30,0.45,0.50,0.55;0.50,0.25,0.35〉	〈0.20,0.35,0.35,0.40;0.40,0.40,0.35〉	〈0.30,0.45,0.50,0.55;0.50,0.25,0.35〉	〈0.10,0.20,0.20,0.30;0.20,0.10,0.15〉
M_3	〈0.70,0.75,0.80,0.85;0.90,0.10,0.10〉	〈0.80,0.85,0.85,0.90;0.95,0.10,0.10〉	〈0.40,0.50,0.55,0.65;0.60,0.25,0.25〉	〈0.60,0.70,0.75,0.75;0.80,0.15,0.15〉
M_4	〈0.70,0.75,0.80,0.85;0.90,0.10,0.10〉	〈0.20,0.35,0.35,0.40;0.40,0.40,0.35〉	〈0.60,0.70,0.75,0.75;0.80,0.15,0.15〉	〈0.20,0.35,0.35,0.40;0.40,0.40,0.35〉
M_5	〈1.00,0.90,0.90,1.00;1.00,0.0,00.00〉	〈0.60,0.70,0.75,0.75;0.80,0.15,0.15〉	〈0.10,0.20,0.20,0.30;0.20,0.10,0.15〉	〈0.30,0.45,0.50,0.55;0.50,0.25,0.35〉
M_6	〈0.10,0.20,0.20,0.30;0.20,0.10,0.15〉	〈0.80,0.85,0.85,0.90;0.95,0.10,0.10〉	〈0.50,0.65,0.65,0.70;0.70,0.15,0.15〉	〈0.70,0.75,0.80,0.85;0.90,0.10,0.10〉
Expert_1	B_FDD	B_DCF	B_LMP
M_1	〈0.80,0.85,0.85,0.90;0.95,0.10,0.10〉	〈0.20,0.35,0.35,0.40;0.40,0.40,0.35〉	〈0.40,0.50,0.55,0.65;0.60,0.25,0.25〉
M_2	〈0.50,0.65,0.65,0.70;0.70,0.15,0.15〉	〈0.40,0.50,0.55,0.65;0.60,0.25,0.25〉	〈0.80,0.85,0.85,0.90;0.95,0.10,0.10〉
M_3	〈0.70,0.75,0.80,0.85;0.90,0.10,0.10〉	〈0.60,0.70,0.75,0.75;0.80,0.15,0.15〉	〈0.60,0.70,0.75,0.75;0.80,0.15,0.15〉
M_4	〈1.00,0.90,0.90,1.00;1.00,0.0,00.00〉	〈0.80,0.85,0.85,0.90;0.95,0.10,0.10〉	〈0.10,0.20,0.20,0.30;0.20,0.10,0.15〉
M_5	〈0.30,0.45,0.50,0.55;0.50,0.25,0.35〉	〈1.00,0.90,0.90,1.00;1.00,0.0,00.00〉	〈0.30,0.45,0.50,0.55;0.50,0.25,0.35〉
M_6	〈1.00,0.90,0.90,1.00;1.00,0.0,00.00〉	〈0.50,0.65,0.65,0.70;0.70,0.15,0.15〉	〈0.70,0.75,0.80,0.85;0.90,0.10,0.10〉
Expert_2	B_LIY	B_LTN	B_LLE	B_SRM
M_1	〈0.80,0.85,0.85,0.90;0.95,0.10,0.10〉	〈0.60,0.70,0.75,0.75;0.80,0.15,0.15〉	〈1.00,0.90,0.90,1.00;1.00,0.0,00.00〉	〈0.70,0.75,0.80,0.85;0.90,0.10,0.10〉
M_2	〈0.30,0.45,0.50,0.55;0.50,0.25,0.35〉	〈0.20,0.35,0.35,0.40;0.40,0.40,0.35〉	〈0.30,0.45,0.50,0.55;0.50,0.25,0.35〉	〈0.10,0.20,0.20,0.30;0.20,0.10,0.15〉
M_3	〈0.70,0.75,0.80,0.85;0.90,0.10,0.10〉	〈0.80,0.85,0.85,0.90;0.95,0.10,0.10〉	〈0.80,0.85,0.85,0.90;0.95,0.10,0.10〉	〈0.60,0.70,0.75,0.75;0.80,0.15,0.15〉
M_4	〈0.70,0.75,0.80,0.85;0.90,0.10,0.10〉	〈0.20,0.35,0.35,0.40;0.40,0.40,0.35〉	〈0.60,0.70,0.75,0.75;0.80,0.15,0.15〉	〈0.20,0.35,0.35,0.40;0.40,0.40,0.35〉
M_5	〈1.00,0.90,0.90,1.00;1.00,0.0,00.00〉	〈0.60,0.70,0.75,0.75;0.80,0.15,0.15〉	〈0.10,0.20,0.20,0.30;0.20,0.10,0.15〉	〈0.30,0.45,0.50,0.55;0.50,0.25,0.35〉
M_6	〈0.10,0.20,0.20,0.30;0.20,0.10,0.15〉	〈0.80,0.85,0.85,0.90;0.95,0.10,0.10〉	〈0.50,0.65,0.65,0.70;0.70,0.15,0.15〉	〈0.70,0.75,0.80,0.85;0.90,0.10,0.10〉
Expert_2	B_FDD	B_DCF	B_LMP
M_1	〈0.80,0.85,0.85,0.90;0.95,0.10,0.10〉	〈0.20,0.35,0.35,0.40;0.40,0.40,0.35〉	〈0.40,0.50,0.55,0.65;0.60,0.25,0.25〉
M_2	〈0.60,0.70,0.75,0.75;0.80,0.15,0.15〉	〈0.40,0.50,0.55,0.65;0.60,0.25,0.25〉	〈0.80,0.85,0.85,0.90;0.95,0.10,0.10〉
M_3	〈0.70,0.75,0.80,0.85;0.90,0.10,0.10〉	〈0.60,0.70,0.75,0.75;0.80,0.15,0.15〉	〈0.60,0.70,0.75,0.75;0.80,0.15,0.15〉
M_4	〈1.00,0.90,0.90,1.00;1.00,0.0,00.00〉	〈0.80,0.85,0.85,0.90;0.95,0.10,0.10〉	〈0.10,0.20,0.20,0.30;0.20,0.10,0.15〉
M_5	〈0.30,0.45,0.50,0.55;0.50,0.25,0.35〉	〈1.00,0.90,0.90,1.00;1.00,0.0,00.00〉	〈0.30,0.45,0.50,0.55;0.50,0.25,0.35〉
M_6	〈1.00,0.90,0.90,1.00;1.00,0.0,00.00〉	〈0.50,0.65,0.65,0.70;0.70,0.15,0.15〉	〈0.70,0.75,0.80,0.85;0.90,0.10,0.10〉
Expert_3	B_LIY	B_LTN	B_LLE	B_SRM
M_1	〈0.20,0.35,0.35,0.40;0.40,0.40,0.35〉	〈0.60,0.70,0.75,0.75;0.80,0.15,0.15〉	〈1.00,0.90,0.90,1.00;1.00,0.0,00.00〉	〈0.70,0.75,0.80,0.85;0.90,0.10,0.10〉
M_2	〈0.60,0.70,0.75,0.75;0.80,0.15,0.15〉	〈0.20,0.35,0.35,0.40;0.40,0.40,0.35〉	〈0.30,0.45,0.50,0.55;0.50,0.25,0.35〉	〈0.10,0.20,0.20,0.30;0.20,0.10,0.15〉
M_3	〈0.70,0.75,0.80,0.85;0.90,0.10,0.10〉	〈0.80,0.85,0.85,0.90;0.95,0.10,0.10〉	〈0.40,0.50,0.55,0.65;0.60,0.25,0.25〉	〈0.60,0.70,0.75,0.75;0.80,0.15,0.15〉
M_4	〈0.70,0.75,0.80,0.85;0.90,0.10,0.10〉	〈0.20,0.35,0.35,0.40;0.40,0.40,0.35〉	〈0.60,0.70,0.75,0.75;0.80,0.15,0.15〉	〈0.20,0.35,0.35,0.40;0.40,0.40,0.35〉
M_5	〈1.00,0.90,0.90,1.00;1.00,0.0,00.00〉	〈0.60,0.70,0.75,0.75;0.80,0.15,0.15〉	〈0.10,0.20,0.20,0.30;0.20,0.10,0.15〉	〈0.30,0.45,0.50,0.55;0.50,0.25,0.35〉
M_6	〈0.10,0.20,0.20,0.30;0.20,0.10,0.15〉	〈0.80,0.85,0.85,0.90;0.95,0.10,0.10〉	〈0.50,0.65,0.65,0.70;0.70,0.15,0.15〉	〈0.70,0.75,0.80,0.85;0.90,0.10,0.10〉
Expert_3	B_FDD	B_DCF	B_LMP
M_1	〈0.80,0.85,0.85,0.90;0.95,0.10,0.10〉	〈0.20,0.35,0.35,0.40;0.40,0.40,0.35〉	〈0.40,0.50,0.55,0.65;0.60,0.25,0.25〉
M_2	〈0.60,0.70,0.75,0.75;0.80,0.15,0.15〉	〈0.40,0.50,0.55,0.65;0.60,0.25,0.25〉	〈0.80,0.85,0.85,0.90;0.95,0.10,0.10〉
M_3	〈0.70,0.75,0.80,0.85;0.90,0.10,0.10〉	〈0.60,0.70,0.75,0.75;0.80,0.15,0.15〉	〈0.60,0.70,0.75,0.75;0.80,0.15,0.15〉
M_4	〈1.00,0.90,0.90,1.00;1.00,0.0,00.00〉	〈0.80,0.85,0.85,0.90;0.95,0.10,0.10〉	〈0.10,0.20,0.20,0.30;0.20,0.10,0.15〉
M_5	〈0.30,0.45,0.50,0.55;0.50,0.25,0.35〉	〈1.00,0.90,0.90,1.00;1.00,0.0,00.00〉	〈0.30,0.45,0.50,0.55;0.50,0.25,0.35〉
M_6	〈1.00,0.90,0.90,1.00;1.00,0.0,00.00〉	〈0.50,0.65,0.65,0.70;0.70,0.15,0.15〉	〈0.70,0.75,0.80,0.85;0.90,0.10,0.10〉

Step 12: TNNs were converted to real values by using the score function using Eq. (3). The average of the three matrices for each expert was aggregated separately into an extended decision matrix using the OWA operator shown in Table B.3 (See Appendix B).

Table B.3

Aggregation decision matrix of six industries according to barriers

Expert_s	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP
AAI	0.950	0.779	0.950	0.698	0.950	0.950	0.779
M_1	0.379	0.583	0.950	0.698	0.779	0.179	0.368
M_2	0.384	0.179	0.285	0.130	0.555	0.368	0.779
M_3	0.698	0.779	0.505	0.583	0.698	0.583	0.583
M_4	0.698	0.179	0.583	0.179	0.950	0.779	0.130
M_5	0.950	0.583	0.130	0.285	0.285	0.950	0.285
M_6	0.130	0.779	0.500	0.698	0.950	0.500	0.698
AI	0.130	0.179	0.130	0.130	0.285	0.179	0.130

Step 13: On the basis of the decision matrix in Table B.3, another decision matrix was created using Eq. (11), including the AI and AAI solutions. The newly created matrix was determined using Eqs. (12), (13) as exhibited in Table B.3 (See Appendix B).

Step 14: In this regard, the elements were determined according to whether they have a benefit (Eq. (15)) or have a cost (Eq. (16)). Then, the normalized decision matrix was obtained using Eq. (14) as presented in Table B.4 (See Appendix B).

Table B.4

Normalized decision matrix of six industries according to barriers.

Expert_s	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP
AAI	0.137	0.230	0.137	0.186	0.300	0.188	0.169
M_1	0.343	0.307	0.137	0.186	0.369	1.000	0.353
M_2	0.339	1.000	0.456	1.000	0.514	0.486	0.169
M_3	0.186	0.230	0.257	0.223	0.408	0.307	0.223
M_4	0.186	1.000	0.223	0.726	0.300	0.230	1.000
M_5	0.137	0.307	1.000	0.456	1.000	0.188	0.456
M_6	1.000	0.230	0.260	0.186	0.300	0.358	0.186
AI	1.000	1.000	1.000	1.000	1.000	1.000	1.000

Step 15: Consequently, the weighted matrix was constructed according to the final weights of the barriers identified in Table 5 and the normalized decision matrix obtained in Table B.4 using Eq. (17) as presented in Table B.5 (See Appendix B).

Table B.5

Weighted normalized decision matrix of six industries according to barriers.

Expert_s	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP
AAI	0.023290	0.026450	0.020413	0.018972	0.043500	0.029328	0.027547
M_1	0.058310	0.035305	0.020413	0.018972	0.053505	0.156000	0.057539
M_2	0.057630	0.115000	0.067944	0.102000	0.074530	0.075816	0.027547
M_3	0.031620	0.026450	0.038293	0.022746	0.059160	0.047892	0.036349
M_4	0.031620	0.115000	0.033227	0.074052	0.043500	0.035880	0.163000
M_5	0.023290	0.035305	0.149000	0.046512	0.145000	0.029328	0.074328
M_6	0.170000	0.026450	0.038740	0.018972	0.043500	0.055848	0.030318
AI	0.170000	0.115000	0.149000	0.102000	0.145000	0.156000	0.163000

Step 16: The sum of the elements of a weighted matrix has been calculated by applying Eq. (20), as shown in Table 6 . The utility degree of six industries has been computed by utilizing Eqs. (18), (19) as presented in Table 6.

Table 6

Ranking of six industries using the MARCOS method.

Industries	St	Kt^-	Kt^+	fK^-	fK^+	fKt	Rank
AAI	0.189500
M_1	0.400044	2.111050	0.400044	0.159311	0.840689	0.3883	4
M_2	0.520467	2.746528	0.520467	0.159311	0.840689	0.5052	1
M_3	0.26251	1.385277	0.26251	0.159311	0.840689	0.2548	6
M_4	0.496279	2.618887	0.496279	0.159311	0.840689	0.4817	3
M_5	0.502763	2.653103	0.502763	0.159311	0.840689	0.4880	2
M_6	0.383828	2.025478	0.383828	0.159311	0.840689	0.3726	5
AI	1.000000

Step 17: Finally, the utility function in relation to the unideal solution f has been computed by utilizing Eq. (22) as presented in Table 6. The utility function in relation to the ideal solution f has been calculated by employing Eq. (23) as presented in Table 6. In this regard, the utility function of six industries has been computed according to Eq. (21), as presented in Table 6. Lastly, the six industries are ranked according to the highest possible value of the utility function, as exhibited in Table 6 and Fig. 5 . Fig. 5 presents the relationship between six industries on the vertical axis and the values of the weights on the horizontal axis.

Fig. 5

Final ranking of six industries by using the MARCOS method.

Discussion

In this subsection, the results of the suggested hybrid methodology are explained to identify the industry most affected by the COVID-19 pandemic. The results are divided into two parts. The first part relates to the results of determining the weights of the barriers evaluated using the ANP method under the neutrosophic environment using TNNS. The second part relates to the results of ranking six industries identified in the evaluation process by using the MARCOS method under the neutrosophic environment using TNNs.

Subsequently, the three experts determined the local weights of the barriers used in the evaluation without showing internal interdependency between the criteria. According to Table 4, the first expert assessed the barriers as follows: takes the top weight of 0.276, followed by with a weight value of 0.205, and takes the lowest weight of 0.066. The second expert assessed the barriers as follows: has the highest weight of 0.261, followed by with a weight value 0.205, and has the lowest weight of 0.063. The third expert evaluated the barriers as follows: has the highest weight of 0.248, followed by with a weight value 0.208, and has the lowest weight of 0.070. After aggregating the final weights of the three experts by using the OWA operator is the most important barrier with a value of 0.263, followed by with a weight value of 0.206, and last barrier is with a weight of 0.065.

Then, the inner interdependency between the criteria was calculated to obtain the final weights for the other criteria as shown in Table 5. According to Table 5, it is clear that the inner interdependencies of criteria influence its weights. It's obvious that the weights of the main barriers changed from (0.071, 0.065, 0.099, 0.096, 0.206, 0.200, and 0.263) to (0.170, 0.115, 0.149, 0.102, 0.145, 0.156, and 0.163). Consequently, takes the top weight of 0.170, followed by with a weight value of 0.163, and takes the lowest weight of 0.102.

In the second part, the results of the evaluation of six industries by the MARCOS method are explained. According to Table 6 and Fig. 5, the MARCOS method identified that the tourism industry is the industry most affected by the COVID-19 pandemic, followed by the electronics industry. The textile industry has been ranked last.

Sensitivity analysis

In this section, sensitivity analysis was performed on the data to evaluate six industries according to the weights of the identified barriers. Usually, the results of the MCDM approach explicitly depend on the proportional significance values ​​for each criterion, which refer to the weights. Sensitivity analysis was investigated to assess changes in the ranking of industries through changing barrier weights as presented in Table 7 and Fig. 6 . All cases showed that the food industry is the alternative most affected by the COVID-19 pandemic in light of weight changes, except for three cases. The third case shows that the electronics industry is the most affected by the crisis, followed by the food industry according to the change in weights. The tenth case, shows that the pharmaceutical industry is the most affected alternative by the COVID-19 pandemic, followed by the food industry. Finally, the eleventh case shows that the most affected tourism industry, followed by the electronics industry.

Table 7

Sensitivity analysis for ANP–MARCOS approach.

Consider number	Clarification for tuning barriers’ weight	Industries
		M_1	M_2	M_3	M_4	M_5	M_6	Rank
Consider 1	Various barriers values	0.430	0.461	0.270	0.523	0.518	0.302	M_4
Consider 2	All barriers equal EIB	0.380	0.421	0.230	0.473	0.468	0.262	M_4
Consider 3	All barriers equal EPB	0.440	0.471	0.290	0.533	0.548	0.322	M_5
Consider 4	All barriers equal MIB	0.399	0.452	0.262	0.495	0.465	0.298	M_4
Consider 5	All barriers equal MRB	0.410	0.491	0.288	0.587	0.478	0.272	M_4
Consider 6	All barriers equal SIB	0.428	0.459	0.268	0.521	0.516	0.300	M_4
Consider 7	All barriers equal SVB	0.445	0.476	0.285	0.538	0.533	0.317	M_4
Consider 8	All barriers equal VRB	0.397	0.428	0.237	0.490	0.485	0.269	M_4
Consider 9	All barriers equal VEB	0.408	0.439	0.248	0.501	0.496	0.280	M_4
Consider 10	All barriers equal EXB	0.558	0.489	0.320	0.548	0.543	0.342	M_1
Consider 11	Half barriers equal EIB and half barriers equal EXB	0.503	0.531	0.298	0.489	0.518	0.367	M_2
Consider 12	Half barriers equal EPB and half barriers equal VEB	0.340	0.371	0.220	0.433	0.418	0.234	M_4
Consider 13	Half barriers equal MIB and half barriers equal VRB	0.350	0.381	0.205	0.443	0.428	0.223	M_4
Consider 14	Half barriers equal MRB and half barriers equal SIB	0.365	0.396	0.208	0.458	0.433	0237	M_4

Fig. 6

Ranking of six industries according to changes in evaluation weights.

Conclusions and future directions

The COVID-19 pandemic resulted in an unprecedented blow to the global economy, which led to the collapse of supply chains and industries due to the lockdowns. In addition to the mandatory precautionary measures, the loss of many jobs, the reduction in working hours, and other economic challenges are existing problems before the crisis. All of the above are barriers borne by industries and supply chains, which have weakened the global economy. Therefore, resilient supply chains are vital for companies and manufacturers. They are imperative in unforeseen events such as the COVID-19 pandemic. Thus, establishing flexible supply chains and identifying barriers to the operation of industries can help industries recover from crises. In this study, the most important barriers facing industries were identified by reviewing previous studies and the opinions of economic and academic experts. Seven barriers have been identified to determine the extent of their impact on the performance of industries. Policymakers are always interested in knowing the most important industry barriers in order to be able to take appropriate measures. Given the multiple barriers, the problem of identifying the industry most affected by the COVID-19 pandemic has been considered an MCDM problem. Therefore, a hybrid MCDM approach under a neutrosophic environment using TNNs has been applied to deal with the uncertainty of industry barriers. According to the results, the lack of inventory is the most crucial barrier in determining the industries most affected by the COVID-19 pandemic. After ranking industries using the MARCOS method, the tourism industry was the most affected industry during the COVID-19 pandemic.

One of the future directions is to investigate other barriers that industries and supply chains may encounter and increase the number of industries in the evaluation process. Also, the existing study can be extended to show the cause-and-effect relationship among these barriers with a dynamic system approach. Furthermore, conducting the study to achieve the principle of sustainability can be another extension of this work.

Funding

This research has no funding source.

Ethical approval.

This article does not involve studies with human participants or animals.

CRediT authorship contribution statement

Abduallah Gamal: Investigation, Methodology, Resources, Visualization, Software, Writing – original draft, Writing – review & editing. Mohamed Abdel-Basset: Investigation, Methodology, Resources, Visualization, Software, Writing – original draft, Writing – review & editing. Ripon K. Chakrabortty: Conceptualization, Investigation, Validation, Methodology, Writing – review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Table A.4

Pairwise comparison matrix based on lack of inventory barrier by all experts.

Lack of inventory	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP
B_LIY	–	1/〈EMB〉	〈VSB〉	〈SSB〉	〈VSB〉	〈ESB〉	1/〈EMB〉
B_LTN	〈EMB〉	–	〈MSB〉	1/〈VSB〉	〈MSB〉	〈SSB〉	〈SSB〉
B_LLE	1/〈VSB〉	1/〈MSB〉	–	〈EMB〉	〈ESB〉	〈MSB〉	〈VSB〉
B_SRM	1/〈SSB〉	〈VSB〉	1/〈EMB〉	–	〈SSB〉	〈SSB〉	〈ESB〉
B_FDD	1/〈VSB〉	1/〈MSB〉	1/〈ESB〉	1/〈SSB〉	–	〈EMB〉	〈SSB〉
B_DCF	1/〈ESB〉	1/〈SSB〉	1/〈MSB〉	1/〈SSB〉	1/〈EMB〉	–	〈MSB〉
B_LMP	〈EMB〉	1/〈SSB〉	1/〈VSB〉	1/〈ESB〉	1/〈SSB〉	1/〈MSB〉	–

Table A.5

Pairwise comparison matrix based on lack of transportation barrier by all experts.

lack of transportation	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP
B_LIY	–	〈ESB〉	1/〈SSB〉	〈SSB〉	〈VSB〉	1/〈MSB〉	〈ESB〉
B_LTN	1/〈ESB〉	–	〈MSB〉	1/〈VSB〉	1/〈SSB〉	〈SSB〉	1/〈MSB〉
B_LLE	〈SSB〉	1/〈MSB〉	–	〈EMB〉	1/〈MSB〉	〈VSB〉	〈MSB〉
B_SRM	1/〈SSB〉	〈VSB〉	1/〈EMB〉	–	〈SSB〉	〈ESB〉	〈SSB〉
B_FDD	1/〈VSB〉	〈SSB〉	〈MSB〉	1/〈SSB〉	–	〈SSB〉	〈EMB〉
B_DCF	〈MSB〉	1/〈SSB〉	1/〈VSB〉	1/〈ESB〉	1/〈SSB〉	–	〈MSB〉
B_LMP	1/〈ESB〉	〈MSB〉	1/〈MSB〉	1/〈SSB〉	1/〈EMB〉	1/〈MSB〉	–

Table A.6

Pairwise comparison matrix based on local law enforcement barrier by all experts.

Local law enforcement	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP
B_LIY	–	〈ESB〉	1/〈SSB〉	〈SSB〉	〈VSB〉	1/〈MSB〉	〈ESB〉
B_LTN	1/〈ESB〉	–	〈MSB〉	1/〈VSB〉	1/〈SSB〉	〈SSB〉	1/〈MSB〉
B_LLE	〈SSB〉	1/〈MSB〉	–	1/〈EMB〉	1/〈MSB〉	〈VSB〉	〈MSB〉
B_SRM	1/〈SSB〉	〈VSB〉	〈EMB〉	–	〈SSB〉	〈MSB〉	〈VSB〉
B_FDD	1/〈VSB〉	〈SSB〉	〈MSB〉	1/〈SSB〉	–	〈SSB〉	〈EMB〉
B_DCF	〈MSB〉	1/〈SSB〉	1/〈VSB〉	1/〈MSB〉	1/〈VSB〉	–	〈MSB〉
B_LMP	1/〈ESB〉	〈MSB〉	1/〈MSB〉	1/〈VSB〉	1/〈EMB〉	1/〈MSB〉	–

Table A.7

Pairwise comparison matrix based on scarcity of raw materials barrier by all experts.

Scarcity of raw materials	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP
B_LIY	–	1/〈ESB〉	〈SSB〉	1/〈SSB〉	〈VSB〉	1/〈MSB〉	1/〈ESB〉
B_LTN	〈ESB〉	–	〈MSB〉	1/〈VSB〉	1/〈SSB〉	〈SSB〉	1/〈MSB〉
B_LLE	1/〈SSB〉	1/〈MSB〉	–	1/〈EMB〉	1/〈MSB〉	〈VSB〉	〈MSB〉
B_SRM	〈SSB〉	〈VSB〉	〈EMB〉	–	〈SSB〉	〈MSB〉	1/〈SSB〉
B_FDD	1/〈VSB〉	〈SSB〉	〈MSB〉	1/〈SSB〉	–	〈SSB〉	〈EMB〉
B_DCF	〈MSB〉	1/〈SSB〉	1/〈VSB〉	1/〈MSB〉	1/〈VSB〉	–	〈EMB〉
B_LMP	〈ESB〉	〈MSB〉	1/〈MSB〉	〈SSB〉	1/〈EMB〉	1/〈EMB〉	–

Table A.8

Pairwise comparison matrix based on fluctuation of demand barrier by all experts.

Fluctuation of demand	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP
B_LIY	–	1/〈VSB〉	〈SSB〉	1/〈SSB〉	〈VSB〉	1/〈MSB〉	1/〈ESB〉
B_LTN	〈VSB〉	–	〈MSB〉	1/〈VSB〉	1/〈SSB〉	1/〈VSB〉	1/〈MSB〉
B_LLE	1/〈SSB〉	1/〈MSB〉	–	1/〈EMB〉	〈MSB〉	〈VSB〉	〈MSB〉
B_SRM	〈SSB〉	〈VSB〉	〈EMB〉	–	〈SSB〉	〈MSB〉	1/〈SSB〉
B_FDD	1/〈VSB〉	〈SSB〉	1/〈MSB〉	1/〈SSB〉	–	〈VSB〉	〈EMB〉
B_DCF	〈MSB〉	〈VSB〉	1/〈VSB〉	1/〈MSB〉	1/〈VSB〉	–	〈EMB〉
B_LMP	〈ESB〉	〈MSB〉	1/〈MSB〉	〈SSB〉	1/〈EMB〉	1/〈EMB〉	–

Table A.9

Pairwise comparison matrix based on deficiency in cash flow in the market barrier by all experts.

Deficiency in cash flow in the market	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP
B_LIY	–	1/〈ESB〉	1/〈MSB〉	1/〈SSB〉	〈VSB〉	1/〈SSB〉	1/〈ESB〉
B_LTN	〈ESB〉	–	〈MSB〉	1/〈VSB〉	1/〈SSB〉	〈SSB〉	1/〈MSB〉
B_LLE	〈MSB〉	1/〈MSB〉	–	1/〈EMB〉	1/〈MSB〉	〈EMB〉	〈MSB〉
B_SRM	〈SSB〉	〈VSB〉	〈EMB〉	–	〈SSB〉	〈MSB〉	1/〈SSB〉
B_FDD	1/〈VSB〉	〈SSB〉	〈MSB〉	1/〈SSB〉	–	〈SSB〉	〈EMB〉
B_DCF	〈SSB〉	1/〈SSB〉	1/〈EMB〉	1/〈MSB〉	1/〈SSB〉	–	〈EMB〉
B_LMP	〈ESB〉	〈MSB〉	1/〈MSB〉	〈SSB〉	1/〈EMB〉	1/〈EMB〉	–

Table A.10

Pairwise comparison matrix based on lack of manpower in the market barrier by all experts.

Lack of manpower	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP
B_LIY	–	1/〈VSB〉	〈VSB〉	〈SSB〉	〈VSB〉	〈ESB〉	1/〈EMB〉
B_LTN	〈VSB〉	–	〈MSB〉	1/〈EMB〉	〈MSB〉	〈SSB〉	〈SSB〉
B_LLE	1/〈VSB〉	1/〈MSB〉	–	〈EMB〉	〈ESB〉	1/〈MSB〉	〈VSB〉
B_SRM	1/〈SSB〉	〈EMB〉	1/〈EMB〉	–	〈SSB〉	〈SSB〉	1/〈ESB〉
B_FDD	1/〈VSB〉	1/〈MSB〉	1/〈ESB〉	1/〈SSB〉	–	〈EMB〉	〈SSB〉
B_DCF	1/〈ESB〉	1/〈SSB〉	〈MSB〉	1/〈SSB〉	1/〈EMB〉	–	〈MSB〉
B_LMP	〈EMB〉	1/〈SSB〉	1/〈VSB〉	〈ESB〉	1/〈SSB〉	1/〈MSB〉	–

Table A.11

Interdependency matrix of the lack of inventory barrier by all experts.

Lack of inventory	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP	Sum of rows	Weights	Consistency ratio = 0.082
B_LIY	0.035266	0.081574	0.052062	0.031924	0.140988	0.022187	0.301202	0.665202	0.095
B_LTN	0.062562	0.036159	0.019055	0.089539	0.051603	0.073851	0.124433	0.457202	0.065
B_LLE	0.092185	0.258244	0.034027	0.060766	0.025802	0.044374	0.204272	0.719671	0.103
B_SRM	0.151432	0.055323	0.076766	0.034254	0.085883	0.073851	0.037383	0.514891	0.074
B_FDD	0.092185	0.258172	0.486117	0.147085	0.092149	0.140571	0.124433	1.340711	0.192
B_DCF	0.503809	0.155265	0.243024	0.147085	0.207888	0.079239	0.074766	1.411076	0.202
B_LMP	0.062562	0.155265	0.088948	0.489347	0.395687	0.565927	0.133511	1.891247	0.270

Table A.12

Interdependency matrix of the lack of transportation barrier by all experts.

Lack of transportation	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP	Sum of rows	Weights	Consistency ratio = 0.097
B_LIY	0.026335	0.017791	0.23304	0.031924	0.071335	0.376728	0.022861	0.780015	0.111
B_LTN	0.376225	0.06354	0.030392	0.089539	0.200205	0.049161	0.583116	1.392178	0.199
B_LLE	0.024544	0.453806	0.054271	0.060766	0.332991	0.080705	0.045722	1.052805	0.150
B_SRM	0.113083	0.097217	0.122436	0.034254	0.043454	0.014769	0.076094	0.501307	0.072
B_FDD	0.06884	0.05922	0.030392	0.147085	0.046624	0.049161	0.14484	0.546162	0.078
B_DCF	0.014748	0.272843	0.141865	0.489347	0.200205	0.052748	0.045722	1.217477	0.174
B_LMP	0.376225	0.035583	0.387604	0.147085	0.105185	0.376728	0.081646	1.510055	0.216

Table A.13

Interdependency matrix of the local law enforcement barrier by all experts.

Local law enforcement	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP	Sum of rows	Weights	Consistency ratio = 0.090
B_LIY	0.026335	0.017791	0.239300	0.041363	0.077398	0.371244	0.021797	0.795229	0.114
B_LTN	0.376225	0.063540	0.031208	0.116013	0.217220	0.048446	0.555971	1.408622	0.201
B_LLE	0.024544	0.453806	0.055729	0.100124	0.361291	0.079530	0.043593	1.118618	0.160
B_SRM	0.113083	0.097217	0.098863	0.044381	0.047147	0.029109	0.119103	0.548904	0.078
B_FDD	0.068840	0.059220	0.031208	0.190573	0.050587	0.048446	0.138097	0.586972	0.084
B_DCF	0.014748	0.272843	0.145675	0.316971	0.132234	0.051980	0.043593	0.978045	0.140
B_LMP	0.376225	0.035583	0.398016	0.190573	0.114124	0.371244	0.077845	1.563610	0.223

Table A.14

Interdependency matrix of the scarcity of raw materials barrier by all experts.

Scarcity of raw materials	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP	Sum of rows	Weights	Consistency ratio = 0.085
B_LIY	0.100402	0.480299	0.063914	0.190573	0.077398	0.497631	0.46338	1.873597	0.268
B_LTN	0.028112	0.033620	0.038404	0.116013	0.21722	0.064939	0.231657	0.729965	0.104
B_LLE	0.431124	0.240116	0.068578	0.100124	0.361291	0.106605	0.018164	1.326003	0.189
B_SRM	0.093574	0.051439	0.121657	0.044381	0.047147	0.039019	0.139280	0.536497	0.077
B_FDD	0.262450	0.031334	0.038404	0.190573	0.050587	0.064939	0.057541	0.695828	0.099
B_DCF	0.056225	0.144365	0.179262	0.316971	0.132234	0.069677	0.057541	0.956276	0.137
B_LMP	0.028112	0.018827	0.489782	0.041363	0.114124	0.157191	0.032436	0.881836	0.126

Table A.15

Interdependency matrix of the fluctuation of demand barrier by all experts.

Fluctuation of demand	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP	Sum of rows	Weights	Consistency ratio = 0.089
B_LIY	0.089206	0.170760	0.044037	0.190573	0.116032	0.445428	0.463380	1.519417	0.217
B_LTN	0.136485	0.065325	0.026460	0.116013	0.325648	0.163029	0.231657	1.064618	0.152
B_LLE	0.383051	0.466553	0.047250	0.100124	0.042469	0.095422	0.018164	1.153034	0.165
B_SRM	0.083140	0.099948	0.083822	0.044381	0.070681	0.034926	0.139280	0.556177	0.079
B_FDD	0.233185	0.060883	0.337460	0.190573	0.075838	0.058126	0.057541	1.013607	0.145
B_DCF	0.049955	0.099948	0.123512	0.316971	0.198241	0.062367	0.057541	0.908536	0.130
B_LMP	0.024978	0.036582	0.337460	0.041363	0.171091	0.140701	0.032436	0.784611	0.112

Table A.16

Interdependency matrix of the deficiency in cash flow in the market barrier by all experts.

Deficiency in cash flow in the market	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP	Sum of rows	Weights	Consistency ratio = 0.100
B_LIY	0.151561	0.480299	0.349516	0.190573	0.071335	0.365509	0.46338	2.072173	0.296
B_LTN	0.042437	0.033620	0.027405	0.116013	0.200205	0.079333	0.231657	0.730671	0.104
B_LLE	0.084874	0.240116	0.048938	0.100124	0.332991	0.151004	0.018164	0.976212	0.139
B_SRM	0.141255	0.051439	0.086816	0.044381	0.043454	0.047668	0.13928	0.554293	0.079
B_FDD	0.396181	0.031334	0.027405	0.190573	0.046624	0.079333	0.057541	0.828992	0.118
B_DCF	0.141255	0.144365	0.110404	0.316971	0.200205	0.085121	0.057541	1.055863	0.151
B_LMP	0.042437	0.018827	0.349516	0.041363	0.105185	0.192033	0.032436	0.781797	0.112

Table A.17

Interdependency matrix of the lack of manpower barrier by all experts.

Lack of manpower	B_LIY	B_LTN	B_LLE	B_SRM	B_FDD	B_DCF	B_LMP	Sum of rows	Weights	Consistency ratio = 0.87
B_LIY	0.035572	0.092498	0.067088	0.062846	0.140988	0.014582	0.10495	0.518523	0.074
B_LTN	0.054425	0.035386	0.024555	0.152124	0.051603	0.048537	0.043357	0.409987	0.059
B_LLE	0.092985	0.252725	0.043848	0.119622	0.025802	0.37194	0.071176	0.978099	0.140
B_SRM	0.152746	0.062774	0.098921	0.067431	0.085883	0.048537	0.664589	1.180881	0.169
B_FDD	0.092985	0.252725	0.626414	0.289548	0.092149	0.092386	0.043357	1.489564	0.213
B_DCF	0.508182	0.151946	0.024555	0.289548	0.207888	0.052078	0.026051	1.260248	0.180
B_LMP	0.063105	0.151946	0.114619	0.018881	0.395687	0.37194	0.04652	1.162699	0.166