The Assessment of Vulnerability of the Global Container Shipping
Network Based on MATLAB Software
Jing Lyu, Qinran Zhang
*
and Wan Su
Dalian Maritime University, Dalian, China
Keywords: Network Disconnection, Vulnerability, Layered Weighted Network Efficiency, Complex Network, Maritime.
Abstract: Extreme Events seriously affect the normal flow of goods on routes and exacerbate the vulnerability of the
global container shipping network. To assess the impact of extreme events on the vulnerability of the global
container shipping network, the global container shipping network was constructed based on the route data of
the top seven liner shipping companies in the world against the backdrop of the COVID-19. Based on the
network disintegration theory and the layered weighted network efficiency metric, routes suffering from
different levels and degrees of disconnection are simulated. The results show that the disconnection of each
segment has a wide impact on the vulnerability of the global container shipping network, the disconnection
between Singapore Port and Santos Port will have the greatest impact on the shipping network, which reduce
the efficiency of the shipping network by 2%. The reduction of layered weighted network efficiency is
consistent with the degree of segments disconnection. The global container shipping network has strong anti-
interference resistance to international segments partial disconnection, and presents great vulnerability to
international segments complete disconnection. While the number of international segments remains at 10%,
the performance of the maritime network can still reach 48.4% of the original network. However, if the
international segments are completely disconnected, the performance of the maritime network will drop to
8.2% of the original network. This paper further improves the understanding of the global container shipping
network under extreme events, and provides a useful reference for studying the vulnerability of the global
shipping network under interference in the future.
1 INTRODUCTION
Ships carry more than 80% of global trade, and
emergencies seriously affect the normal operation of
shipping companies, thus affecting the normal
operation of world trade. Therefore, the simulation of
different levels and degrees of routes disconnection
and the assessment of vulnerability of the global
container shipping network under emergencies can
help the decision-makers of various countries to
formulate reasonable policies, optimize the shipping
network and formulate emergency plans after
accidents.
At present, scholars' research results on the
vulnerability of global container shipping network
have been rich. The research results of maritime
network from the point of port disconnection are
systematic and perfect. LIU used complex network
theory to see the topology characteristics of shipping
network, and studied the vulnerability of global
container shipping network under two attack
methods: random interference and intentional attack
(Liu Chanjuan, 2016). WANG et al. studied the
impact of random interference and intentional attack
of Sino-US container shipping network (Wang Liehui,
2020). HE et al. studied the impact of port
disconnection on the vulnerability of the shipping
network of China's coastal container ports (He Yao,
2022). JIANG et al. evaluated port vulnerability from
the perspective of supply chain (Jiang M, 2021). Xu
et al. proposed a new cascade model to quantify the
impact of port disconnection on network vulnerability
by reducing connectivity and network efficiency (Xu
X, 2022).
In the process of studying the vulnerability of
global container shipping network, scholars have
found that when global emergencies occur, the
probability of route changes is greater, so route
disconnection has gradually become a new research
hotspot. VILJOEN et al. studied the vulnerability of
global container transport network under the
disconnection of target connection, and proposed a
new possibility of network disconnection, namely,
route connection disconnection (Viljoen N M, 2016).
442
Lyu, J., Zhang, Q. and Su, W.
The Assessment of Vulnerability of the Global Container Shipping Network Based on MATLAB Software.
DOI: 10.5220/0012285600003807
Paper published under CC license (CC BY-NC-ND 4.0)
In Proceedings of the 2nd International Seminar on Artificial Intelligence, Networking and Information Technology (ANIT 2023), pages 442-448
ISBN: 978-989-758-677-4
Proceedings Copyright © 2024 by SCITEPRESS – Science and Technology Publications, Lda.
DIRZKA et al. used the ship schedule cancellations
released by liner companies to quantify the initial
degree and dynamics of the COVID-19 pandemic's
disconnection to the shipping network (Dirzka C,
2022).
To sum up, the research on the vulnerability of
global container shipping network is mainly aimed at
the disconnection of port nodes, and few research on
route disconnection, and there is no study on the level
and degree of route disconnection. The layered
weighted network efficiency metric can better
describe these problems. Therefore, this paper
simulates different levels and degrees of route
disconnection and introduces layered weighted
network efficiency metric to calculate the
vulnerability changes of the global container shipping
network, providing reference for relevant maritime
managers.
2 CONSTRUCTION OF
VULNERABILITY
ASSESSMENT MODEL OF
GLOBAL CONTAINER
SHIPPING NETWORK UNDER
EMERGENCIES
2.1 Problem Description
Maritime network vulnerability refers to the degree to
which the connectivity and efficiency of the maritime
network are disturbed when external factors affect the
network (Jin L, 2022). Based on the process, nature
and mechanism, emergencies are divided into four
types: accident disasters, natural disasters, public
health incidents and social security incidents.
According to the third type of specific scenario, this
paper selects the COVID-19 as research background.
The COVID-19 has posed enormous challenges to
countries and societies and disrupted the global
economy (Tuti R W, McKibbin W). The Clarkson
report pointed out that the COVID-19 has had a
significant impact on the container shipping industry,
and the impact has become a major determinant of
market trends, leading to major operational
disconnections across the industry, in the first half of
2020, the global container trade volume dropped
sharply, only 93.2% in the same period last year.
As for route disconnection level, this paper studies
two levels: segment disconnection and international
segment disconnection. Segment disconnection refers
to the disconnection of the segment between any two
directly connected ports in the shipping network,
simulating the decline of connectivity between ports
due to the impact of the pandemic. International
segment disconnection refers to the disconnection of
a segment between any two directly connected ports
in the shipping network that are not in the same
country. It simulates the decline in port connectivity
between countries due to certain isolation policies
adopted by all countries under the pandemic. The
degree of route disconnection refers to the degree of
reduction in the number of segments between two
directly connected ports.
The layered weighted network efficiency metric
can measure the vulnerability change of shipping
network when the route is interrupted. The layered
weighted network efficiency metric can better
measure the vulnerability change of shipping network
when the route is interrupted. Firstly, the global
container shipping network is divided into several
sub-networks, which will change due to route
interruption. Then calculate the network efficiency of
all sub-networks, get the layered weighted network
efficiency, and then quantitatively measure the
vulnerability of the whole network.
2.2 MATLAB Software Introduction
MATLAB is a comprehensive mathematical and
technical calculation software. It uses a high-level
programming language containing hundreds of built-
in instructions to carry out mathematical calculations.
It includes a complete library of built-in algorithms to
meet the needs of more advanced users (Bonakdari H,
2022). MATLAB software is widely used in
transportation research and can solve modeling
problems well.
2.3 Construction of Global Container
Shipping Network Model
The global container shipping network consists of
canals, ports and other nodes and the sides (segments)
between nodes. is the set of nodes in the network,
is the set of edges, then the global container
shipping network is . Among them,


; The sequence 
indicates
that node and node are connected, node and node
are the two ports of call adjacent to any route. The
connection relationship between any two nodes in the
network can be represented by the adjacency matrix
of 


of order .

 
 
(1)
The Assessment of Vulnerability of the Global Container Shipping Network Based on MATLAB Software
443
2.4 Construction of Vulnerability
Measurement Model of Global
Container Shipping Network
Network efficiency measurement is one of network
structure measurement methods, which is derived
from the concept of average shortest path length. The
measurement of network efficiency is not limited by
the weight, connectivity and sparsity of the system,
and can better measure the efficiency of network
information transmission and achieve accurate
quantitative analysis of weighted and unweighted
networks (Latora V, 2001). Therefore, in the relevant
research of shipping network, network efficiency
measurement is widely used to describe the change of
the transportation efficiency of the entire network
when the network is attacked, and measure the
connectivity of network. The network efficiency
measure can be expressed as



(2)
In formula (2),
is the network efficiency of
the shipping network; is the total number of nodes
in the shipping network ;

is the shortest path
length between node and node .
The global container shipping network is a
weighted network, and weight is an important part of
it. This paper takes the number of segments between
two ports as weight. During the COVID-19
pandemic, faced with uncertain cargo demand and
continuous market fluctuations, liner companies
rescheduled or canceled route services, resulting in
disconnection of container shipping network links
and decreased connectivity (Pan J J, 2022).
Therefore, this paper aims to study how the
vulnerability of the maritime network is affected by
the removal of a large number of segments. The
layered weighted network efficiency can better
describe this process. It can accurately identify the
impact of addition or deletion of segments and
increase or decrease of the number of segments on the
efficiency of the maritime network, which is helpful
to study the impact of route disconnection on the
overall network efficiency. Using the idea of network
disintegration, the shipping network is regarded as a
multi-layer network, and the number of layers on each
side is determined by the number of segments. The
weighted network is decomposed into several
unweighted subnetworks by deleting all the directly
connected edge segments from the remaining
network and forming a subnet with these deleted
segments in each step (Zhou Y, 2021). Then, the
network efficiency of each sub-network is calculated
using formula (2). The layered weighted network
efficiency of the original network is obtained by
summing the network efficiency of all subnetworks.
Any change in the route will change the structure
of the shipping network, and then change the network
efficiency of several subnets, which will then be
reflected in the efficiency calculation results of the
original network. Therefore, the layered weighted
network efficiency metric can well measure the
disconnection of the container shipping network
caused by the COVID-19 pandemic. Therefore, this
study uses this index to measure the importance of
segment and the vulnerability of the maritime
network. In this paper, MATLAB software is used to
calculate the efficiency index of the layered weighted
network through the above steps, so as to measure the
vulnerability of the global container shipping network
under global interference.
3 EMPIRICAL ANALYSIS
3.1 Data Selection and Procession
The data of the global container shipping network
constructed in this paper comes from the websites of
the top seven liner companies in terms of global
shipping capacity collected by Alphaliner. The total
capacity of these seven shipping companies is 76.4%
worldwide, this is highly representative. As shown in
Table 1. The total service network of these seven liner
companies covers six continents in the world,
involving the world's most basic maritime ports.
Based on the shipping schedule on the website of each
shipping company, all the ports where liners call on
each route are tracked and obtained respectively, and
the tracking time is from April 2022 to June 2022.
Repeated routes were processed, resulting in 544
ports and 2,464 routes involving 138 countries. The
global container shipping network is abstracted as a
complex network
, consisting of 544
nodes and 2464 edges, as shown in Figure 1.
Table 1. Capacity and market share of the top 7 liner
shipping companies.
Rank
Company
Capacity/teu
Share
1
MSC
4949720
18.5%
2
Maersk
4132026
15.5%
3
GMA-CGM
3449314
12.9%
4
COSCO
2890349
10.8%
5
Hapag-Lloyd
1795909
6.7%
6
Evergreen
1664330
6.2%
7
ONE
1553956
5.8%
Total
20435604
76.40%
ANIT 2023 - The International Seminar on Artificial Intelligence, Networking and Information Technology
444
Figure 1. Global container shipping network.
Among them, the direct link between Shanghai
port and Ningbo Port has the greatest weight, there
are 339 segments between the two ports, 10% of the
port pairs are connected through one segment. The
layered weighted network efficiency of the global
container shipping network is 0.8641, while the
unweighted efficiency is only 0.332. To sum up, the
weight distribution of the global container shipping
network is not uniform.
3.2 Results and Results Analysis
1) The Impact of Segment Disconnection on
Maritime Network Vulnerability
The vulnerability of the global container shipping
network to segment disconnection can be evaluated
by calculating the change of the layered weighted
network efficiency during segment disconnection.
The results of the impact of disconnection of each
segment on the vulnerability of the global container
shipping network were calculated. The top 20
segments ranked by the percentage of decreased
network efficiency by stratified weighting are shown
in Table 2. Columns 2 and 3 represent the two ports
at either end of each segment. In the global container
shipping network constructed in this paper, the two
ports with the greatest weight are Shanghai Port and
Ningbo Port, and the number of segments is 339.
However, it is calculated that the impact of
disconnection of this segment on the vulnerability of
the shipping network ranks 29th. The reason may be
because that the factors affecting the vulnerability of
the maritime network caused by the disconnection of
flight segments do not only depend on the number of
flight segments, other factors may be the
geographical location of the two ports, the country to
which they belong and the connection with other
ports.
It can be seen from Table 2 that the ports at both
ends of the segment that most affect the vulnerability
of the shipping network are Singapore Port and
Santos Port, and the number of segments between the
two ports is 16. If this segment is disconnected, the
layered weighted network efficiency will decrease by
2.01928%. Singapore Port is adjacent to the southeast
side of Malacca Strait in the west and the north side
of Singapore Strait in the south. It is the main port
node connecting the Pacific Ocean and the Indian
Ocean. There are 124 ports directly connected to port
of Singapore. Santos Port, more than 60 kilometers
northwest of Sao Paulo and 210 kilometers northeast
of Rio de Janeiro, is a free port in Brazil and the
largest port in Latin America. Ports directly
connected to port of Santos include port of Singapore,
port of Rotterdam, port of Antwerp, and 15 small
Latin American ports. These 15 small ports are
relatively isolated from the rest of the world, and the
disconnection of the segment between port of Santos
and port of Singapore virtually cuts them off from the
global container shipping network. As a result, this
segment is the most vulnerable in the global container
shipping network.
Table 2. Top 20 vulnerable segments in the global container
shipping network.
Rank
Port 1
Port 2
Percentage
of
efficiency
decrease
1
Singapore
Santos
2.01928%
2
Busan
Manzanillo
1.54472%
3
Shanghai
Nansha New
0.52626%
4
Manzanillo
Balboa
0.47735%
5
Santos
Rio de Janeiro
0.47124%
6
Dakar
Tangier Mediterranee
0.46858%
7
Jeddah
Tangier Mediterranee
0.44481%
8
Singapore
Piraeus
0.44446%
9
Balboa
Buenaventura
0.43423%
10
Tuticorin
Colombo
0.39210%
11
Bangkok
Laem Chabang
0.38986%
12
Singapore
Tangier Mediterranee
0.38920%
13
Manzanillo
Guaranao
0.38190%
14
Kingston
Au Prince
0.35371%
15
Manzanillo
Cartagena
0.34751%
16
Pasir Gudang
Singapore
0.33221%
17
Pointe Des Galets
Louis
0.32899%
18
Kobe
Osaka
0.32607%
19
Jeddah
Aden
0.32141%
20
San Juan
Caucedo
0.31618%
Figure 2 shows the local network between port of
Singapore and port of Santos and ports directly
connected to them. Among them, the black dot
represents the port node, and the width of the line is
proportional to the number of segments. It can be seen
The Assessment of Vulnerability of the Global Container Shipping Network Based on MATLAB Software
445
that there are five indirectly connected routes between
port of Santos and port of Singapore, but the weight
of these five routes is relatively small, unable to bear
the existing volume of Asia-Pacific-Latin American
cargo transported through port of Santos to port of
Singapore, so the disconnection of this segment will
lead to a certain stagnation of global cargo transport,
seriously affecting the connectivity of the global
container shipping network.
2) The Impact of Segment Disconnection Degree on
the Vulnerability of Maritime Network
The purpose of this section is to study the changes
in the vulnerability of the maritime network when
four levels of disconnection occur. Table 3 shows the
degradation of the layered weighted network
efficiency of the maritime network when four levels
of disconnection occur in the top 20 most vulnerable
segments. In general, the decrease of the efficiency of
the layered weighted network is consistent with the
degree of segment disconnection, that is, the higher
the degree of segment disconnection, the more the
efficiency of the layered weighted network of the
maritime network will decline. However, the impact
of different degree of disconnection on the global
container shipping network is not the same. For
example, when calculating the ports of Singapore and
Piraeus, port of Singapore and port Tangier
Mediterranee, Kobe and Osaka ports, the percentage
decline in the efficiency of the layered weighted
network is approximately linear with the degree of
reduction in the number of segments, and the degree
of decline in the layered weighted network efficiency
approximately increases proportionally as the degree
of disconnection deepens. Busan and Manzanillo
ports, Manzanillo and Balboa ports, Dakar and port
Tangier Mediterranee, Jeddah and port Tangier
Mediterranee are convex. The ports of Shanghai and
Nansha New Port, Tuticorin port and Colombo port,
Bangkok port and Laem Chabang port, Kingston port
and Port Au Prince are concave.
Figure 2. Port of Singapore and Port of Santos.
Table 3. Layered weighted network efficiency to different
degrees of disconnection.
Rank
Port
1
Port
2
Percentage decline in efficiency
25%
50%
75%
100%
1
Singapore
Santos
0.16243%
0.62682%
1.70263%
2.01928%
2
Busan
Manzanillo
0.34793%
1.03851%
1.51822%
1.54472%
3
Shanghai
Nansha New
0.05700%
0.13182%
0.23689%
0.52626%
4
Manzanillo
Balboa
0.14018%
0.43339%
0.46974%
0.47735%
5
Santos
Rio de Janeiro
0.01616%
0.09683%
0.34281%
0.47124%
6
Dakar
Tangier Mediterranee
0.24566%
0.45537%
0.45961%
0.46858%
7
Jeddah
Tangier Mediterranee
0.04429%
0.37130%
0.40982%
0.44481%
8
Singapore
Piraeus
0.03942%
0.14619%
0.31405%
0.44446%
9
Balboa
Buenaventura
0.13127%
0.42514%
0.43116%
0.43423%
10
Tuticorin
Colombo
0.04952%
0.10627%
0.17456%
0.39210%
11
Bangkok
Laem Chabang
0.03991%
0.10540%
0.22669%
0.38986%
12
Singapore
Tangier Mediterranee
0.07880%
0.18161%
0.30035%
0.38920%
13
Manzanillo
Guaranao
0.00226%
0.15374%
0.15374%
0.38190%
14
Kingston
Au Prince
0.04346%
0.09191%
0.15308%
0.35371%
15
Manzanillo
Cartagena
0.00928%
0.02269%
0.29218%
0.34751%
16
Pasir Gudang
Singapore
0.05642%
0.17859%
0.30368%
0.33221%
17
Pointe Des Galets
Louis
0.03744%
0.07977%
0.32716%
0.32899%
18
Kobe
Osaka
0.02545%
0.08887%
0.20675%
0.32607%
19
Jeddah
Aden
0.00226%
0.09141%
0.09141%
0.32141%
20
San Juan
Caucedo
0.02867%
0.17076%
0.28686%
0.31618%
By analysing the proportion of the number of
segments to the total number of routes, the
relationship between the reduction of the efficiency
of the layered weighted network and the reduction of
the number of segments is obtained. Taking the
segment of Dakar port and port Tangier Mediterranee
as an example, there are 17 routes passing through
port Tangier Mediterranee, and a total of 8 routes
between Dakar port and port Tangier Mediterranee.
This means that the segment between Dakar port and
port Tangier Mediterranee is essentially to serve port
Tangier Mediterranee. Therefore, when part of the
border is interrupted, the cargo transported between
Dakar port and port Tangier Mediterranee is the most
affected, and the route with the port Tangier
Mediterranee as a transit port is almost unaffected.
But when this border is completely interrupted, cargo
using port Tangier Mediterranee as a transit port will
be diverted, and the shipping network will be severely
disrupted. Another example is the segment of
Shanghai port and Nansha New port. Shanghai port is
the largest international container port, there are 420
routes through Shanghai port, there are 106 routes
through Nansha New port, and there are only 9 routes
between the two ports, which indicates that these two
ANIT 2023 - The International Seminar on Artificial Intelligence, Networking and Information Technology
446
ports are important ports for goods to pass through.
As a result, a partial disconnection of this segment
would seriously affect the normal movement of
global cargo, thereby affecting the connectivity of the
maritime network and increasing its vulnerability.
3) The Impact of International Segment
Disconnection on the Vulnerability of Maritime
Networks
During the COVID-19 pandemic, countries have
formulated various pandemic prevention policies, and
airline companies have adjusted their routes,
seriously affecting international transportation.
Therefore, the purpose of this section is to examine
the impact of varying degrees of disconnection in the
international segment on the global container
shipping network. In the global container shipping
network constructed in this paper, the number of all
international segments is reduced to a certain level at
the same time, and the layered weighted network
efficiency of the remaining network is calculated. In
Figure 3, horizontal coordinate represents the
proportional reduction of the number of international
segments, vertical coordinate represents the layered
weighted network efficiency of the remaining
network as a proportion of the original network
efficiency. It can be seen that the global container
shipping network has a strong impact resistance to the
reduction in the number of international segments.
Even if all countries cancelled 40 percent of their
international segments, the maritime network would
still maintain 82 percent of its performance. However,
if all international segments are eliminated, the
efficiency of the maritime network will drop to 8.2%
of the original network. The reason is that when the
international segment is partially interrupted,
although the transportation efficiency between ports
has declined, the global container shipping network is
still a complete network, and the global cargo can
basically achieve circulation. However, if all
international segments are cancelled, the global
container shipping network will become an isolated
subnet, and international transportation will be
greatly affected. As a result, the global container
shipping network is highly vulnerable to complete
disconnection of international segments.
Figure 3. The impact of international segments
disconnection on the vulnerability of shipping network.
4 CONCLUSION
In this paper, a layered weighted network efficiency
method was introduced to build a global container
shipping network based on the route data of the top
seven liner companies in the world in terms of
shipping capacity against the background of the
COVID-19 pandemic. The vulnerability of the global
container shipping network was studied by simulating
different levels and degrees of disconnection of
routes. The results show that the disconnection of
different segments has a great difference on the
vulnerability of the global container shipping
network, and the disconnection of Singapore Port and
Santos Port has the greatest impact, which will reduce
the efficiency of the shipping network by 2%. When
the segments are interrupted to different degrees, the
decrease of the efficiency of the layered weighted
network is consistent with the variation trend of the
segments interrupted degree. The global container
shipping network has strong anti-interference to the
partial disconnection of the international segment and
great vulnerability to the complete disconnection of
the international segment. As long as the number of
international segments in the maritime network is
maintained at 10% of the original, the maritime
network can still achieve 48.4% of the pre-
disconnection performance; However, if the
international segment is completely disconnected, the
efficiency of the shipping network will drop to 8.2%
of the pre-disconnection level. This paper deepens the
understanding of the global container shipping
network under extreme events, provides a reference
for the study of the vulnerability of the shipping
network under global disturbance, and provides
management enlightenment for relevant government
departments and private enterprises.
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
0%
20%
40%
60%
80%
100%
Residual Marine network performance
International segment reduction
The Assessment of Vulnerability of the Global Container Shipping Network Based on MATLAB Software
447
REFERENCES
Liu Chanjuan, Hu Zhihua. Robustness research of global
container shipping network based on complex network
(J). Journal of Guangxi University (Natural Science
Edition), 2016, 41(05): 1441-1448.
Wang Liehui, Ye Fei, Zheng Yuanbo. The Assessment of
Sino-US Container Shipping Network Evolution and
vulnerability (J). Economic Geography, 2020, 40(05):
136-144.
He Yao, Yang Yongchun, Guo Jianke. Vulnerability of
shipping networks in China's coastal container ports
under disconnection simulation (J). Resources Science,
2022, 44(02): 414-424.
Jiang M , Lu J , Qu Z ,et al. Port vulnerability assessment
from a supply Chain perspective(J). Ocean & Coastal
Management, 2021, 213(7):105851.
https://doi.org/10.1016/j.ocecoaman.2021.105851.
Xu X, Zhu Y, Xu M, et al. Vulnerability analysis of the
global liner shipping network: from static structure to
cascading failure dynamics (J). Ocean & Coastal
Management, 2022, 229: 106325.
https://doi.org/10.1016/j.ocecoaman.2022.106325.
Viljoen N M, Joubert J W. The vulnerability of the global
container shipping network to targeted link disruption
(J). Physica A: Statistical Mechanics and its
Applications, 2016, 462: 396-409.
https://doi.org/10.1016/j.physa.2016.06.111.
Dirzka C, Acciaro M. Global shipping network dynamics
during the COVID-19 pandemic's initial phases (J).
Journal of Transport Geography, 2022, 99: 103265.
https://doi.org/10.1016/j.jtrangeo.2021.103265.
Jin L, Chen J, Chen Z ,et al. Impact of COVID-19 on
China's international liner shipping network based on
AIS data(J). Transport Policy, 2022, 121:90-99.
https://doi.org/10.1016/j.tranpol.2022.04.006.
Tuti R W, Nurmandi A, Zahra A A. Handling COVID-19
in the capital city of Jakarta with innovation policy: the
scale of social restrictions policy (J). Heliyon, 2022,
8(5): e09467. 2022.
https://doi.org/10.1016/j.heliyon.2022.e09467.
McKibbin W, Fernando R. The Global Macroeconomic
Impacts of COVID-19: Seven Scenarios (J). Asian
Economic Papers, 2021, 20(2):1-30.
https://doi.org/10.1162/asep_a_00796.
Bonakdari H, Zeynoddin M. Stochastic Modeling (M),
2022:321-352. https://doi.org/10.1016/B978-0-323-
91748-3.00007-0.
Latora V, Marchiori M. Efficient behavior of small-world
networks (J). Physical review letters, 2001, 87(19):
198701.
https://doi.org/10.1103/PhysRevLett.87.198701.
Pan J J, Zhang Y F, Fan B. Strengthening container
shipping network connectivity during COVID-19: A
graph theory approach (J). Ocean & Coastal
Management, 2022, 229: 106338.
https://doi.org/10.1016/j.ocecoaman.2022.106338.
Zhou Y, Kundu T, Qin W, et al. Vulnerability of the
worldwide air transportation network to global
catastrophes such as COVID-19(J). Transportation
Research Part E: Logistics and Transportation Review,
2021, 154: 102469.
https://doi.org/10.1016/j.tre.2021.102469.
ANIT 2023 - The International Seminar on Artificial Intelligence, Networking and Information Technology
448