Analysis of influence factors in communication and collaboration between controllers and commanders

Abstract

To cope with the smooth implementation of apron control transfer at Chinese airports, two new departments were established, namely apron tower and airport operation command center. Therefore, based on interview texts of controllers and commanders of these two departments, this paper uses text mining and Decision-making Trial and Evaluation Laboratory-Interpretative Structural Modeling Method methods to determine key influence factors and factor hierarchy that affect communication and collaboration in their daily work. The results show that for controllers, key influence factors are mainly personnel development and professional abilities. These factors are located at the bottom of the factor hierarchy and are the basis for ensuring smooth communication and collaboration. For commanders, key influence factors are mainly personnel professional abilities and flight status. These factors are at the top of the factor hierarchy and are focus points that affect communication and collaboration. Hence, the case analysis results show the application potential of these three methods in the field of civil aviation. The combined use of these three methods can enable airport managers to clearly understand the degree of influence between factors.

Keywords

Decision-making Trial and Evaluation Laboratory interpretative structural modeling method text mining communication and collaboration

1 Introduction

With the increasing scale of air transportation, large ground traffic flow and high density have become normalization. Contradictions and problems of airport control tower acting on behalf of aircraft apron operation management are constantly emerging, which can not meet the needs of transportation airport operation safety and efficiency. Therefore, the Civil Aviation Administration of China (CAAC) puts forward the plan of aircraft apron control transfer operation. Meanwhile, to ensure the smooth progress of the plan, two new departments were established, namely the apron tower (AT) and the airport operation command center (AOC). AT belongs to a sub-department of the air traffic control tower (ATC), and AOC belongs to a department established by the airport. Therefore, in daily work of communication and collaboration, air traffic controllers (Hereinafter referred to as controllers) and airport operation commanders (Hereinafter referred to as commanders) in AT and AOC pay attention to different focuses. Based on this, we interviewed controllers and commanders from three airports in South China, Central China, and North China that have annual passenger flow exceeding 10 million. The purpose is to find out the factors that affect communication and collaboration between the controllers and commanders by analyzing the interview content, to provide managers with a reference for improving personnel management, to continuously improve the effect of apron control operation.

This paper is divided into five sections. In Section 1, problem statement of this study. In Section 2, review associated literature and introduction to methods of this study. In Section 3, case study analysis. In Section 4, discusses the results obtained. In Section 5, concludes this study and recommendations for future research.

2 Materials and methods

The literature [1] studied the controller’s work status after the Southeast Asian air traffic control system changed from paper flight strips to a stripless flight management system. The results of the questionnaire were analyzed by Lexical analysis to extract 10 high-frequency words with most times are mentioned by these controllers of their daily work. Then these high-frequency words as hierarchy evaluation indices, and calculate index weights and sort through the AHP. Sorting results showed that the most important issue for controllers is training. The literature [2] conducted an empirical analysis of work pressure and workload of controllers of the Mactan Civil Aviation Authority of the Philippines. Through the comprehensive use of DEMATEL-ANP and PROMETHEE II methods. The results showed that the controller’s operational responsibilities have the greatest impact on workload, and partitioning can effectively reduce workload. The literature [3] comprehensively used text mining and fuzzy comprehensive evaluation methods to evaluate the current situation of controllers and commanders of A airport in domestic. The evaluation results show that the factors of personnel and management are the main reasons affecting their work efficiency. The literature [4] interviewed psychologists and project managers of military oil and gas projects to determine the factors that affect face-to-face communication in different situations. The results showed that the degree of influence of these effective factors in face-to-face communication will change with the situation, and there is a mutual influence between the factors. To make accurate decisions on personnel selection under uncertain circumstances, literature [5] used the DEMATEL and the ELECTRE methods in an intuitionistic fuzzy situation to conduct an empirical analysis on the selection of personnel for an air filter manufacturing company. The literature [6] proposed a fault analysis method for CNC equipment based on the DEMATEL-ISM and ANP to accurately describe the CNC equipment failure. The results showed that the method can quantify the failure factors of CNC equipment, and learning the relationship between these factors and the weak links of equipment reliability. The literature [7] used human dynamics and complex system methods to conduct an empirical analysis on the activities of controllers to detect the interaction between air traffic activities and the controller’s communication activities. The results indicated that the basic mechanism of the controller’s activity is different from the general mechanism proposed by Barabasi’s, and the Lévy process with positive drift may be able to explain the controller’s adaptive behavior. The literature [8] used “ conversation analysis” technology to study the voice recordings of aircraft approaching at Bangkok International Airport, and explored the potential impact of “ non-English” on information transmission between pilots and controllers. The results showed that in the case of complex information and involving digital information, the probability of communication errors caused by pilots not understanding the event is greater. The literature [9] proposed a research plan based on laboratory and field evaluation to study the impact of datalink and freeflight on communication and collaboration between pilots, controllers, and other personnel. The literature [10] based on ergonomics, and focused on analyzing the communication between controllers teams. The analysis results showed how operators can use the natural redundancy and diversity of human communication spontaneously to enable team members to better understand each other, and it is also the key to cooperation between members.

From the above literature review, the research content mainly focuses on workload, personnel development, and the ability of controllers or commanders. The methods used in these researches mainly include Analytic Hierarchy Process (AHP), fuzzy analytic hierarchy process (FAHP), text mining, and DEMATEL-Based ANP (D-ANP), etc. These methods are only a macroscopic assessment of the current situation and cannot further explore the interaction and hierarchical structure of factors within the micro-level system. Therefore, this study uses the decision making trial and evaluation laboratory (DEMATEL) to analyze the importance of the key factors that affect the communication between controllers and commanders and then uses the interpretive structural modeling (ISM) method to determine the factor hierarchical structure. The purpose is to deeply analyze the hierarchical relationship between the factors that affect the communication and collaboration between controllers and commanders so that managers can clearly understand the degree of influence of each factor. Moreover, it is found that there is not much literature comprehensively use text mining and DEMATEL-ISM methods for research. So this study uses the text mining method to extract high-frequency words or phrases of interview texts as influence factors and based on the result of word frequency comparison, a direct influence matrix is established. Compared with the traditional method of determining the direct influence matrix through expert scoring, it not only reduces the subjectivity of the results but also saves the time and labor cost of consulting experts.

Innovation of this paper to further study are listed as follows:

There is much literature based on text mining and DEMATEL methods respectively, but there are few studies that combine the two methods to analyze the factors that affect communication and collaboration between controllers and commanders. This paper uses text mining methods to extract high-frequency words or phrases from interview texts of controllers and commanders. Compare the frequency of high-frequency words or phrases in pairs, instead of the expert scoring to construct a direct influence matrix between factors. This can make the results closer to the actual situation, and reduce the impact of subjective factors while saving time and labor costs.

Based on the DEMATEL-ISM method, key influence factors and factor hierarchy that affect communication and collaboration between controllers and commanders are respectively determined. The case analysis results not only provide managers concerning improving personnel management but also verifies that the method can be widely and deeply applied in the civil aviation field.

2.1 Text mining

Text mining is to convert a large amount of text data into previously unknown and usable knowledge, and use the knowledge to reintegrate information more systematically and comprehensively for more in-deep data analysis [11]. It can find the most critical criteria based on the entire text data and then quantify the criteria based on the analysis results of most existing reports and literature [12]. In the process of text data processing, unstructured text data needs to be filtered and summarized in the early stage. After collecting the source text data that needs to be analyzed, you can select a suitable text mining tool to do an in-deep analysis of source text data [13]. The criteria extracted by this method have a high reference value and objectivity. Through the two literature [3] and [1], this study determines that the interview content of controllers and commanders is shown in Table 1.

Table 1
Interview content

NO. Interview content

1 What are the main content of your communication with controllers (or commanders) in your daily work?

2 What factors do you think will affect your communication with controllers (or commanders)?

3 How do you deal with a conflict between you and controllers (or commanders) due to work?

4 When you encounter difficulties at work, how do you collaborate with controllers (or commanders) to solve them?

5 What do you think controllers (or commanders) should improve their qualifications?

NO.	Interview content
1	What are the main content of your communication with controllers (or commanders) in your daily work?
2	What factors do you think will affect your communication with controllers (or commanders)?
3	How do you deal with a conflict between you and controllers (or commanders) due to work?
4	When you encounter difficulties at work, how do you collaborate with controllers (or commanders) to solve them?
5	What do you think controllers (or commanders) should improve their qualifications?

To avoid losing important interview content, we have recorded each interview with the consent of the interviewees. After the interview, the recording is processed uniformly, and the voice content is converted into texts. These texts are the interview texts that need to be processed by text mining. The text mining process is shown in Fig. 1.

Fig. 1

The text mining process.

In Fig. 1, high-frequency words or phrases with a frequency of more than 11 are extracted from interview texts as influence factors that affect communication and collaboration between controllers and commanders [1, 14]. These influence factors and frequency are shown in Table 2.

Table 2

High-frequency words or phrases and frequency

NO.	Controllers		Commanders
	Influence factors	Frequency	Influence factors	Frequency
1	Career development	39	Command skills	79
2	Command skills	37	ATC	61
3	Length of learning	33	Flight flow	51
4	Department collaboration	33	Apron operation management	40
5	Pilot	25	Airport control center (ACC)	37
6	Regulations	25	Department collaboration	37
7	Work time	24	Work time	27
8	Flight flow	23	Service guarantee	26
9	Personnel ability	20	Information transfer	26
10	Work pressure	19	Pronunciation clear	26
11	Team communication	19	Gate position	26
12	Training	18	Request identification	22
13	Night shift	17	Runway allocation	21
14	Service guarantee	16	Skilled workflow	17
15	AOC	15	Seat coordination	16
16	Group discussion	15	Emergency management	15
17	Runway allocation	13	Application for release	14
18	Simulator operation	13	Pilot	13
19	Wages and benefits	13	Aircraft handover	12
20	Job responsibilities	12	Order issuance	12
21	Conflict	12	Training	11
22	Case analysis	12	Flight delay	11

In Table 2, some of the influence factors are the same. This is because controllers and commanders have the same work content to a certain extent. For example, coordinating runway allocation and gate position for flights are a part of their same work content. This study constructs the direct influence matrix A required for DEMATEL analysis based on these frequencies are listed in Table 2. The a_ij in matrix A are obtained by comparing the frequency of these influence factors, for example, compare the frequency of influence factor 1 and 2 indicates the influence relationship of 1 to 2 [15].

2.2 The DEMATEL-ISM method

2.2.1 The DEMATEL method

The decision making trial and evaluation laboratory (DEMATEL) is a method procedure that originated from the Geneva Research Centre of the Battelle Memorial Institute [16, 17]. This method is based on graph theory and matrix, through qualitative judgments on the mutual influence relationship between factors in complex systems, and then constructs a direct influence matrix. The DEMATEL method can reveal the internal cause-effect relationship of factors in the system and identify key influence factors by calculating the influence degree, centrality degree, and cause degree of each factor [18].

The specific steps of the DEMATEL method are as follows:

Determine influence factors: a₁,a₂,..., a_n.

Determine the influence relationship between different factors and establish a direct influence matrix A , as follows: $A = {[a_{ij}]}_{n \times n} = [\begin{matrix} 0 & a_{12} & \dots & a_{1 n} \\ a_{21} & 0 & \dots & a_{2 n} \\ ⋮ & ⋮ & ⋱ & ⋮ \\ a_{n 1} & a_{n 2} & \dots & 0 \end{matrix}]$ (1)

In matrix A , a_ij represents the direct influence of factor a_i on a_j, if i = j, a_ij = 0.

Normalizing direct influence matrix A , referred to as matrix R , is calculated using formula (2).

$\begin{matrix} R = & \frac{1}{max_{1 ⩽ i ⩽ n} \sum_{j}^{n} a_{ij}} A, 0 ⩽ r_{ij} ⩽ 1, and \\ max_{1 ⩽ i ⩽ n} \sum_{j}^{n} r_{ij} = 1 \end{matrix}$ (2)

Therefore, the relationship between the matrix R and r_ij can be obtained: $R = {[r_{ij}]}_{n \times n} = [\begin{matrix} 0 & r_{12} & \dots & r_{1 n} \\ r_{21} & 0 & \dots & r_{2 n} \\ ⋮ & ⋮ & ⋱ & ⋮ \\ r_{n 1} & r_{n 2} & \dots & 0 \end{matrix}]$

Using formula (3) to calculate the total-influential matrix T [19]. $T = {[t_{ij}]}_{n \times n} = R {(I - R)}^{- 1}$ (3)

The matrix T represents the combined effect of the indirect influences between factors in the matrix R . Therefore, formula (3) can be used to determine the comprehensive impact of the influence factor a_i on the whole system. I is the identity matrix.

Calculating sum of rows and sum of columns within the total-influential matrix T through formula (4): $\begin{matrix} f_{i} = \sum_{j = 1}^{n} t_{ij}, (i = 1, \dots, n) \\ e_{i} = \sum_{j = 1}^{n} t_{ji}, (i = 1, \dots, n) \end{matrix}$ (4)

Where f_i and e_i denote sum of a row and sum of a column, respectively.

Calculating the degree of centrality and cause of each influence factor to use formula (5) [20]. $\begin{matrix} m_{i} = f_{i} + e_{i}, (i = 1, \dots, n) \\ n_{i} = f_{i} - e_{i} \end{matrix}$ (5)

m_I represents the importance of factors, and when n_i > 0, it indicates the factor belongs to the cause result; otherwise, the factor belongs to the effect result.

2.2.2 The ISM method

Interpretive structural modeling (ISM) was first proposed by Warfield in 1973 [21, 22]. The ISM method was originally used to analyze complex socioeconomic systems. Now, this method has been widely used in various fields, such as supply chains, product design, and risk analysis. The ISM method can transform vague thoughts and concepts into a structural relationship model to intuitively understand the relationship between variables [23].

The various steps of the ISM method are as follows:

Due to the total-influential matrix T and the reachability matrix K have something in common. Non-zero of t_ij and k_ij indicates corresponding factors are dependent, while zero of t_ij and k_ij indicates factors are independent [24].

Step 1: Calculating the matrix H based on total-influential matrix T . Since the matrix T only shows the degree of mutual influence between factors a_i and a_j, it does not consider the factor a_i on itself. To make up for the deficiencies of matrix T , the identity matrix I is introduced, because I represents the one-to-one correspondence between factors and themselves. So add the matrix T and I to get the matrix H . $H = {[h_{ij}]}_{n \times n} = T + I, (i, j = 1, \dots, n)$ (6)

The matrix H can not only reflect the mutual influence between factors but also influence of a_i on itself. For example, if h_ij= 0 indicates that there is no mutual influence between factors a_i and a_j, otherwise it indicates that there is a mutual influence between a_i and a_j [25].

Step 2: Determining the reachability matrix K according to formula (7). $k_{ij} = {\begin{matrix} 1, h_{ij} ⩾ λ \\ 0, h_{ij} < λ \end{matrix}, (i, j = 1, \dots, n)$ (7)

If the system includes many factors, the threshold λ can be set according to the actual situation of study problems. The purpose of setting λ is to eliminate influence relationships with small influence degree to simplify the system structure [25].

Step 3: According to formula (8), the reachability set B _i and antecedent set C _i of each factor are obtained [26]. $\begin{matrix} B_{i} = {a_{i} | a_{i} \in A, k_{ij} \neq 0} \\ C_{i} = {a_{j} | a_{j} \in A, k_{ji} \neq 0} \end{matrix}$ (8) $, (i = 1, \dots, n)$ (9)

Step 4: Determining influence factors of each hierarchy by formula (9) and gradually delete rows and columns corresponding to k_ij in reachability matrix K . $B_{i} = B_{i} \cap C_{i}, (i = 1, \dots, n)$ (10)

Formula (9) represents that in a factor hierarchy structure, the influence factor a_i can reach outside itself, but cannot reach the set of other factors. Such as, if a_i satisfies the condition, it means that a_i is the bottom factor, and the row and column corresponding to a_i in the matrix K are also eliminated.

Step 5: Repeating steps 3 and 4 until all k_ij have been deleted.

Step 6: Establishing a factor hierarchy in the order in which factors are deleted.

3 Case study

Interview texts of this study come from three large international airports with a passenger flow of tens millions in China and these three airports are Z , N , and S from North China, Central China, and South China. The three airports are chosen mainly due to their busy daily flight tasks. Controllers need to frequency communicate with commanders to collaborate aircraft demands, such as takeoffs and landings, runways allocation, and gate position,.etc.. Therefore, the analysis of key factors affecting communication and collaboration between controllers and commanders of these three busy airports is representative and systemic.

3.1 DEMATEL-ISM analysis

3.1.1 DEMATEL analysis

According to the comparison of the frequency in Table 2 and the calculation steps of the DEMATEL method in Section 2, influence factors of communication and collaboration are gradually determined. Due to space limitations, only the direct influence matrix A _Controller _s of controllers is listed, as shown in Fig. 2.

Fig. 2

The direct influence matrix A_Controllers.

To show conciseness, use serial numbers in Table 2 to indicate influence factors. According to formulas (2)–(5), the m_i and n_i are calculated, as shown in Table 3.

Table 3

The m_i and n_i of controllers and commanders

NO.	Controllers		Commanders
	m_i	n_i	m_i	n_i
1	2.070417259	1.338746367	2.847797861	2.397905399
2	2.002854486	1.228639011	2.403563922	1.837124824
3	1.882172703	1.018540308	2.318156215	1.721529628
4	1.873314096	0.984020945	2.046288609	1.323387558
5	1.668952746	0.496519977	2.020490868	1.282184332
6	1.668952746	0.496519977	1.723009393	0.708653227
7	1.650115144	0.426522515	1.735901889	0.726603945
8	1.633501884	0.35430071	1.584621101	0.177031843
9	1.600344667	0.120952734	1.579100664	0.115551312
10	1.596316498	0.036145838	1.593146237	0.151495464
11	1.596316498	0.036145838	1.605971686	0.177965615
12	1.596776037	–0.053148765	1.582824753	–0.158154663
13	1.602515231	–0.147723024	1.596586693	–0.250750962
14	1.614524016	–0.248566873	1.650310373	–0.497587739
15	1.634056309	–0.356934231	1.681820247	–0.596098727
16	1.634056309	–0.356934231	1.721522367	–0.702801996
17	1.702636201	–0.603184252	1.771071076	–0.819351903
18	1.702636201	–0.603184252	1.836351497	–0.953186712
19	1.702636201	–0.603184252	1.909646992	–1.091504015
20	1.756603016	–0.745986131	1.995230464	–1.240570281
21	1.756603016	–0.745986131	2.00448192	–1.253340146
22	1.756603016	–0.745986131	2.104016521	–1.414344026

The greater the m_i, the more important the influence factor, so it can be considered a key influence factor. Therefore, in Table 3, for controllers, the m_i of influence factors Career development, Command skills, Length of learning, and Department coordination are 2.070417259, 2.002854486, 1.882172703, and 1.873314096, respectively. These four factors can be considered as key influence factors that have an important impact on the controller’s communication and collaboration. Meanwhile, for commanders, the m_i of influence factors Command skills, ATC, Flight flow, Apron operation Management, ACC, Training, and Flight delay are 2.847797861, 2.403563922, 2.318156215, 2.046288609, 2.020490868, 2.00448192, 2.104016521, respectively. These seven factors can be considered as the key influence factors.

If the n_i is positive, it means that influence factors are cause factors, otherwise, the factors are outcome factors. In Table 3 that cause factors affecting the communication and collaboration of controllers are as follows: Career development, Command skills, Length of learning, Department coordination, Pilot, Regulations, Work time, Flight flow, Personnel ability, Work pressure, and Team communication. Similarly, cause factors of commanders are as follows: Command skills, ATC, Flight flow, Apron operation management, ACC, Department coordination, Work time, Service guarantee, Information transfer, Pronunciation clear, and Gate position.

3.1.2 ISM analysis

Calculating factor hierarchy distribution according to steps of the ISM method in Section 2. Due to many influence factors in Table 2, this study compares and verifies threshold λ multiple times. The purpose is to eliminate factors with weak influence and simplify the hierarchy. The value of λ can directly affect the matrix K and the factor hierarchy. Due to space limitations, Table 4 only lists comparison results of controllers’.

Table 4
Comparison of the threshold λ

λ=0.01 λ=0.02 λ=0.03 λ=0.04 λ=0.05

Factor hierarchy 1 3 10 3 2

	λ=0.01	λ=0.02	λ=0.03	λ=0.04	λ=0.05
Factor hierarchy	1	3	10	3	2

In Table 4, different threshold λ corresponds to different factor hierarchy structures. The less the factor hierarchy, the more influence factors included in each hierarchy, which makes it impossible to clearly express the relationship between factors. Therefore, according to the comparison results, the λ as 0.03 is selected. After the same calculation process, the commander’s threshold λ is also 0.03. Based on formulas (6) and (7), the matrix K is calculated. Due to space limitations, only the matrix K of controllers is listed, as shown in Fig. 3.

Fig. 3

The reachability matrix K_controllers.

According to formulas (8) and (9), the factor hierarchy of controllers and commanders is determined, as shown in Table 5.

Table 5

Factor hierarchy

Hierarchy	Controllers	Commanders
H₁	Job responsibilities, Conflict, Case analysis	Aircraft handover, Order issuance, Training, Flight delay
H₂	Runway allocation, Simulator operation, Wages and benefits	Seat coordination, Emergency management, Application for release, Pilot
H₃	Service guarantee, AOC, Group discussion	Request identification, Runway allocation, Skilled workflow,
H₄	Training, Night shift	Service guarantee, Information transfer, Pronunciation clear, Gate position
H₅	Personnel ability, Work pressure, Team communication	Department coordination, Work time
H₆	Work time, Flight flow	Flight flow, Apron operation management, ACC
H₇	Pilot, Regulations	ATC
H₈	Department coordination	Command skills
H₉	Length of learning
H₁₀	Career development, Command skills

In Table 5, some of the same influence factors are at the different hierarchy, respectively. This is because commanders are facing a new working environment and partners. However, for controllers, their working environment has remained unchanged, so factor hierarchical distribution is different from commanders.

4 Discussion

According to the factor hierarchy in Table 5, it can be roughly divided into three layers: the top hierarchy H1, the middle hierarchy, and the bottom hierarchy. In the top hierarchy H1, for controllers, the greatest influence factor of communication and collaboration are Job responsibilities, Conflict, and Case analysis. And the greatest influence factors of commanders are Aircraft handover, Order issuance, Training, and Flight delay. For controllers, clear job responsibilities and case analysis can avoid conflicts when communicating with commanders, pilots, and other department personnel. While, because commanders are responsible for all aircraft activities on the apron. Therefore, personnel skills training and clear and accurate order issuance are factors that have the greatest impact on communication and collaboration, especially when aircraft are handed over and flights are delayed. In the middle hierarchy, the factor hierarchies of controllers and commanders include H2, H3, H4, H5, H6, H7 and H2, H3, H4, H5, H6, respectively. For controllers, it can be roughly divided into two aspects: personal needs and daily work. Personal needs include influence factors, such as wages and benefits, personnel ability training, work time, and work pressure,.etc.. If these needs are not met, it will affect their daily work effectiveness and efficiency such as aircraft runway allocation, case discussions, and communication within groups. For commanders, the middle hierarchy is mainly based on personal skills, such as emergency response ability, information transmission ability, pronunciation clear, runway, and gate allocation, etc.. These influence factors have a more direct impact on communication and collaboration. In the bottom hierarchy, factor hierarchies of controllers and commanders are H8, H9, H10, and H7, H8, respectively. The indirect influence factors of controllers are mainly personal development, such as length of study, career development, and command skills. The direct influencing factor of commanders is command skills. Table 6 lists key influence factors and corresponding hierarchy.

Table 6
Key influence factors and hierarchy

Controllers Commanders

Key influence factors Hierarchy Key influence factors Hierarchy

Department collaboration H8 Aircraft handover, Training, Order issuance, Flight delay H1

Length of study H9 Flight flow, ACC, Apron operation management H6

Command skills, Career development H10 ATC H7

Job responsibilities, Conflict, Case analysis H1 Command skills H8

Controllers	Commanders
Department collaboration	H8	Aircraft handover, Training, Order issuance, Flight delay	H1
Length of study	H9	Flight flow, ACC, Apron operation management	H6
Command skills, Career development	H10	ATC	H7
Job responsibilities, Conflict, Case analysis	H1	Command skills	H8

In Table 6, it can be found that key influence factors of communication and collaboration for controllers are mainly at the bottom hierarchy. These are mainly personnel development and professional abilities. The top hierarchy influence factors are Job responsibilities, Conflicts, and Case analysis with centrality degree closer to key influence factors. This may be because although controllers are working in a new working environment, their work content is the same as before, so these factors have become the basis for ensuring smooth progress of communication and collaboration. However, for commanders, they are not only in a new working environment but also facing new working content, and key influence factors can have an important impact on communication and collaboration. Moreover, by referring to the literature [3], these factors of commanders can be summarized into two categories: personnel professional ability and flight status.

5 Conclusion

This study uses text mining and DEMATEL-ISM methods to determine the key factors that affect communication and collaboration of controllers and commanders in their daily work and divide the factor hierarchy. For controllers, key influence factors are mainly personal development and professional ability, and these factors are at the bottom hierarchy. Therefore, these factors are the basis for ensuring smooth communication and collaboration between controllers and others. For commanders, key influence factors are mainly personal professional ability and flight status, and these factors are distributed at the top and bottom of the factor hierarchy. Therefore, these factors are both the focus and the basis for guarantee efficient communication and collaboration between commanders and others.

The limitation of this study is that it may not be comprehensive and accurate to determine the key factors affecting the communication and collaboration between controllers and commanders based on the high-frequency words or phrases in the interview texts. Therefore, in future research, on the one hand, scholars can comprehensively analyze the voice call records and interview texts of controllers and commanders in their daily work, to more accurately dig out the key factors that affect their communication and collaboration. On the other hand, text mining and expert scoring can be further compared and appropriate fuzzy rules can be set, so that the direct influence matrix obtained is more scientific.

Footnotes

Acknowledgments

There is no conflict of interest in this study. This research is supported by the Research on Values Basis and Incentive Restriction Mechanism of Safety Risk Control for Civil Aviation Key Post Personnel (Grants No. ASSA2018/17)

References

Surakitbanharn

C.A.

, Landry

S.J.

and Rukijkanpanich

, Development of training for new air traffic control technology implementation, in 6th International Conference on Applied Human Factors and Ergonomics 2015. p. 2738–2745.

Bongo

M.F.

, et al., An application of DEMATEL-ANP and PROMETHEE II approach for air traffic controllers’ workload stress problem: A case of Mactan Civil Aviation Authority of the Philippines, Journal of Air Transport Management 68 (2018), 198–213.

and Feng

, Fuzzy Comprehensive Evaluation of the Apron Control Transfer Management Evaluation System, Arabian Journal for Science and Engineering 2020.

Shakeri

, et al., What do project managers need to know to succeed in face-to-face communication? Economic Research-Ekonomska Istrazivanja: p. 27.

Kilic

H.S.

, Demirci

A.E.

and Delen

, An integrated decision analysis methodology based on IF-DEMATEL and IF-ELECTRE for personnel selection, Decision Support Systems 137 (2020), 19.

Sun

S.G.

, et al., Fault analysis of CNC equipment based on DEMATEL/ISM/ANP, Journal of Mechanical Science and Technology 34(8) (2020), 3181–3188.

Wang

, et al., Empirical analysis of air traffic controller dynamics, Transportation Research Part C-Emerging Technologies, 33 (2013), 203–213.

Tiewtrakul

and Fletcher

S.J.E.

, The challenge of regional accents for aviation English language proficiency standards: A study of difficulties in understanding in air traffic control–pilot communications, 53(2) (2010), 229–239.

Sharples

, et al., Flightdeck and Air Traffic Control Collaboration Evaluation (FACE): Evaluating aviation communication in the laboratory and field, Applied Ergonomics 38(4) (2007), 399–407.

10.

Rognin

and Blanquart

J.P.

, Human communication, mutual awareness and system dependability. Lessons learnt from air-traffic control field studies, Reliability Engineering & System Safety 71(3) (2001), 327–336.

11.

Studer

, et al., Knowledge engineering: principles and methods, 25(1-2) (1998), 161–197.

12.

, Jia

and Li

, Synthetic Evaluation of Natural Gas Pipeline Operation Schemes Based on the Multi-level Grey Cluster Decision Method, Arabian Journal for Science and Engineering, 44(6) (2019), 6045–6056.

13.

Szmelter-Jarosz

and Rzesny-Cieplinska

, Priorities of Urban Transport System Stakeholders According to Crowd Logistics Solutions in City Areas. A Sustainability Perspective, Sustainability 12(1) (2020).

14.

and Feng

, Temporary teams: current research focus and future directions, Quanlity & Quantity 53 (2020), 1–18.

15.

Thiranun

and Xu

H.J.A.i.S.S.R.J.

, Evaluating teachers of international business education in NUAA based on efficacy using AHP, 2(11) (2015).

16.

Gabus

and Fontela

E.J.B.G.R.C.

, Geneva, Switzerland, World problems, an invitation to further thought within the framework of DEMATEL, 1972: p. 1–8.

17.

Gabus

and Fontela

E.J.B.G.R.C.

, Geneva, Switzerland, Perceptions of the world problematique: Communication procedure, communicating with those bearing collective responsibility, 1973.

18.

Hori

and Shimizu

Y.J.C.e.p.

, Designing methods of human interface for supervisory control systems, 7(11) (1999), 1413–1419.

19.

Lin

and Wu

W.-W.J.E.J.o.O.R.

, A fuzzy extension of the DEMATEL method for group decision making, 156(1) (2004), 445–455.

20.

W.-W.

, Choosing knowledge management strategies by using a combined ANP and DEMATEL approach, Expert Systems with Applications 35(3) (2008), 828–835.

21.

Warfield

J.N.

, An assault on complexity. 1973: Battelle, Office of Corporate Communications.

22.

Warfield

J.N.J.P.o.t.I.

, SOCIETAL SYSTEMS Planning, Policy and Complexity 66(3) (1978), 362–363.

23.

Chuang

H.-M.

, et al., Evolving MCDM Applications Using Hybrid Expert-Based ISM and DEMATEL Models: An Example of Sustainable Ecotourism. Scientific World Journal, 2013.

24.

Shankar

and Ravi

, Analysis of interactions among the barriers of reverse logistics, Technological Forecasting and Social Change 72(8) (2005), 1011–1029.

25.

Zhou

D.-Q.

, Zhang

and Li

H.-W.

, A study of the system’s hierarchical structure through integration of Dematel and ISM, in Proceedings of 2006 International Conference on Machine Learning and Cybernetics 1-7 (2006), 1449+.

26.

Warfield

J.N.

, Developing interconnection matrices in structural modeling, IEEE Transactions on Systems, Man and Cybernetics SMC-4(1) (1974), 81–87.

Controllers		Commanders
Key influence factors	Hierarchy	Key influence factors	Hierarchy
Department collaboration	H8	Aircraft handover, Training, Order issuance, Flight delay	H1
Length of study	H9	Flight flow, ACC, Apron operation management	H6
Command skills, Career development	H10	ATC	H7
Job responsibilities, Conflict, Case analysis	H1	Command skills	H8

Analysis of influence factors in communication and collaboration between controllers and commanders

Abstract

Keywords

1 Introduction

2 Materials and methods

2.1 Text mining

2.2.1 The DEMATEL method

3.1 DEMATEL-ISM analysis

3.1.1 DEMATEL analysis

Table 4 Comparison of the threshold λ λ=0.01 λ=0.02 λ=0.03 λ=0.04 λ=0.05 Factor hierarchy 1 3 10 3 2

Footnotes

Acknowledgments

References

Table 4
Comparison of the threshold λ

λ=0.01 λ=0.02 λ=0.03 λ=0.04 λ=0.05

Factor hierarchy 1 3 10 3 2