Event coincidence analysis based quantification of the synchronicity of specific events for directed weighted network construction

Abstract

The application discloses a kind of directed weighted network building based on event coincidence analysis quantification specific event synchronism, and threshold method is used to identify specific event in original space-time sequence, and event sequence is extracted from space-time sequence according to the occurrence time of specific event;Synchronous event on different geographical positions is identified using static coincidence interval ΔT and time delay τ;Geographical position is defined as the node of network, and the direction and weight of edge in network are determined according to the order of event occurrence time and the number of synchronization in synchronous event between different geographical positions, so as to map space-time sequence into directed weighted network with the geographical position of specific event occurrence as base element.The application establishes directed weighted network from the geographical position of specific event occurrence in space-time sequence and the synchronism of specific event, can identify the correlation of specific event at different geographical positions, and analyzes the correlation degree and propagation direction of specific event.

Description

TECHNICAL FIELD

[0001] The present application relates to a network construction method. In particular, it relates to a directed weighted network construction method based on event coincidence analysis to quantify the synchronization of specific events. BACKGROUND

[0002] Complex network is a method of connecting components in a complex system meaningfully, and it reveals the mutual relationship and dynamic changes between components through network topology characteristics. Complex network has been successfully applied to many real-world systems, and it shows great potential as a general tool to explore the dynamics and structural characteristics of complex systems in the fields of physics, biology, social science, economics, etc.

[0003] In different fields such as climate science, social science, neuroscience, epidemiology, transportation and earth science, more and more spatiotemporal data are collected and studied. Using complex network modeling method to mine the relationship between spatiotemporal sequences has been used in various fields. Constructing network according to the similarity between time series is a common method for network construction of spatiotemporal sequences. Taking geographical location as node, the mutual relationship between time series is quantified based on different similarity measurement methods, which is used as the basis for determining the edges between nodes. Commonly used similarity measurement methods include pearson correlation coefficient, spearman correlation coefficient and mutual information, etc. This kind of network construction method faces difficulties in analyzing specific information in spatiotemporal data.

[0004] In recent years, specific events in nature and society, especially extreme events, have attracted more and more attention from academia and the public. In order to understand the driving factors and interactions between specific events, network can be constructed according to the synchronization degree between specific events in spatiotemporal sequence data points. The quantification of event synchronization is the focus of a large number of different studies and methods, and event synchronization method and event coincidence analysis method are two widely used nonlinear measures in the development of methodology aiming to achieve this goal. Event synchronization method has defects in quantifying event synchronization when there are event cluster occurrences. In addition, although the undirected network constructed based on event coincidence analysis method can measure the correlation between specific events, it ignores the influence of the time sequence of the occurrence of synchronous events at different geographical locations, and lacks the exploration of the directionality of the edges in the network. SUMMARY

[0005] The technical problem to be solved by the present application is to provide a directed weighted network construction method based on event coincidence analysis to quantify the synchronization of specific events for spatiotemporal sequence data with geographical location attribute and occurrence time attribute, which can identify the correlation between specific events in spatiotemporal sequence data points.

[0006] The technical scheme adopted by the present application is: a directed weighted network building method based on event coincidence analysis for quantifying synchronization of specific events, comprising the following steps:

[0007] 1) According to the threshold method, data not less than the threshold in the original space-time sequence at the geographical position is identified as a specific event, and an event sequence is constructed according to the occurrence time of the specific event in the original space-time sequence, that is, the event sequence is an ordered sequence with the occurrence time of the specific event as an element, and the elements are arranged in the order of the occurrence time of the specific event, that is: the original space-time sequence at the geographical position i is X(i)={x i l ,l=1,2,…,T i} wherein T i represents the length of the space-time sequence at the geographical position i, and the event sequence E(i)={t i l ,l=1,2,…,N i} at the geographical position i is extracted by the threshold method, wherein N i represents the total number of specific events occurring at the geographical position i, that is, the length of the event sequence E(i), t i l represents the time of the lth occurrence of the specific event at the geographical position i;

[0008] 2) According to the static coincidence interval ΔT and the time delay τ, it is judged whether the events at different geographical positions occur synchronously, if the interval between the time of the lth occurrence of the specific event at the geographical position i and the time of the mth occurrence of the specific event at the geographical position j is within the static coincidence interval ΔT, it is considered that the events occur synchronously, and the difference between the time t i l of the lth occurrence of the specific event at the geographical position i and the time t j m of the mth occurrence of the specific event at the geographical position j is d lm ij , that is

[0009]

[0010] If d lm ij satisfies: 0≤|d lm ij |≤ΔT, it is considered that the lth occurrence of the specific event at the geographical position i and the mth occurrence of the specific event at the geographical position j are synchronous events, if d lm ij satisfies: 0≤|d lm ijIf |τ|≤ΔT, the specific event occurring for the l-th time at the geographic location i and the specific event occurring for the m-th time at the geographic location j are considered to be a lag-synchronous event, where the time delay τ represents the length of the lag time between the spatiotemporal sequences;

[0011] 3) Different geographic locations are taken as nodes of the network, the edges between the nodes are determined according to the synchronization relationship between the specific events at different geographic locations, the direction of the edges is determined according to the chronological order of the occurrence of the synchronous events at different geographic locations, and the weight of the edges is determined according to the number of the synchronous events, so that the spatiotemporal sequence is mapped into a directed and weighted network with the geographic locations where the specific events occur as the basic elements;

[0012] 4) The topological features of the directed and weighted network are calculated, including the node degree k i , the network cumulative degree distribution P(k), the node strength s i , the network cumulative strength distribution P(s), and the node directed clustering coefficient C i In the directed and weighted network, the degree of a node includes the out-degree k i out and the in-degree k i in ,

[0013] k i = k i out + k i in (2)

[0014] In the above formula, N represents the total number of nodes in the network, a ij and a ji represent the directed connection relationship between the node i and the node j, if the node i has an edge pointing to the node j, then a ij = 1, otherwise a ij = 0, if the node j has an edge pointing to the node i, then a ji = 1, otherwise a ji = 0,

[0015]

[0016] In the above formula, N k' represents the number of nodes with the degree k', and N represents the total number of nodes in the network, the strength of a node includes the out-strength s i out and the in-strength s i in ,

[0017] s i = s i out + si in (4)

[0018] In the above formula, N represents the total number of nodes in the network, w ij represents the weight of the edge from node i to node j, w ji represents the weight of the edge from node j to node i,

[0019]

[0020] In the above formula, N s' represents the number of nodes with strength s', N represents the total number of nodes in the network,

[0021]

[0022] In the above formula, b ij represents the connection relationship between node i and node j, if there is an edge between node i and node j, then b ij = 1, otherwise b ij = 0, a jk and a kj represent the directed connection relationship between node j and node k, if node j has an edge pointing to node k, then a jk = 1, otherwise a jk = 0, if node k has an edge pointing to node j, then a kj = 1, otherwise a kj = 0.

[0023] 5) Realize the visualization of the directed and weighted network, the topological characteristics of the network and the geographical distribution of the topological characteristics by means of Matlab software.

[0024] In step 2), in order to make the static consistent with the interval ΔT statistically significant, ΔT needs to satisfy

[0025] ΔT < T / N max (7)

[0026] In the above formula, T is the span of the time of the original spatiotemporal sequence record at all geographical positions, N max is the maximum value of the length of all event sequences.

[0027] In step 3), the specific method for determining the edges between nodes and the direction and weight of the edges is as follows: if there is a synchronous event between geographical position i and geographical position j, then there is an edge between the corresponding nodes, in all synchronous events, if the time of the occurrence of a specific event at geographical position i is earlier than the time of the occurrence of a specific event at geographical position j, then the direction of the edge is from the node corresponding to geographical position i to the node corresponding to geographical position j,

[0028] The weight of the edge depends on the number of synchronous events between different geographical locations, and the method for measuring the number of synchronous events includes precursor event coincidence and trigger event coincidence. In order to explicitly express the synchronous relationship between the mth occurrence of a specific event at geographical location j and the specific event at geographical location i, a coefficient h ij lm For

[0029]

[0030] In the precursor event coincidence, when the mth occurrence of a specific event at geographical location j and the specific events at geographical location i are all synchronous events, the synchronous events are counted only once. At this time, the number of synchronous events between the mth occurrence of a specific event at geographical location j and the specific event at geographical location i is ep ij ●m For

[0031]

[0032] Then, among all the specific events at geographical location j, the number of synchronous events of the specific event at geographical location i occurring before the specific event at geographical location j is r p ij It can be obtained by the following formula:

[0033]

[0034] At this time, the direction of the edge is from the node corresponding to geographical location i to the node corresponding to geographical location j, and the weight of the edge is r p ij Similarly, among all the specific events at geographical location i, the number of synchronous events of the specific event at geographical location j occurring before the specific event at geographical location i is r p ji At this time, the direction of the edge is from the node corresponding to geographical location j to the node corresponding to geographical location i, and the weight of the edge is r p ji ,

[0035] In the trigger event coincidence, when the lth occurrence of a specific event at geographical location i and the specific events at geographical location j are all synchronous events, the synchronous events are counted only once. At this time, the number of synchronous events between the lth occurrence of a specific event at geographical location i and the specific event at geographical location j is et ij l● For

[0036]

[0037] The number r of synchronous events in which the specific event at geographical position i occurs after the specific event at geographical position j among all specific events at geographical position i t ij The number r of synchronous events in which the specific event at geographical position i occurs after the specific event at geographical position j among all specific events at geographical position i

[0038]

[0039] At this time, the direction of the edge is from the node corresponding to geographical position i to the node corresponding to geographical position j, and the weight of the edge is r t ij Similarly, the number r of synchronous events in which the specific event at geographical position i occurs after the specific event at geographical position j among all events at geographical position j At this time, the direction of the edge is from the node corresponding to geographical position j to the node corresponding to geographical position i, and the weight of the edge is r t ij ,

[0040] Then, the element w in the weighted adjacency matrix W of the directed weighted network ij can be obtained by selecting r p ij or r in the trigger event match t ij According to the precursor event match and the trigger event match, two directed weighted networks can be constructed respectively.

[0041] The directed weighted network building method based on event match analysis and quantifying the synchronization of specific events of the present application identifies specific events from spatiotemporal sequence data with geographical position attributes and occurrence time attributes, and constructs event sequences. Based on event match analysis, the number of synchronous events and the order of occurrence time between event sequences are measured, and a directed weighted network is established. The topology of the established network can accurately reflect the correlation between specific events in the spatiotemporal sequence. The features of the present application are reflected in the following aspects:

[0042] 1. The directed weighted network proposed by the present application can accurately reflect the correlation between specific events in the spatiotemporal sequence, and the network topology features can analyze the correlation, correlation degree and propagation direction of specific events at different geographical positions.

[0043] 2. The nodes in the directed weighted network proposed by the present application have geographical position attributes. Therefore, the network not only reflects the correlation between specific events in the spatiotemporal sequence but also reflects the spatial structure features of specific events. At the same time, the construction of the directed weighted network based on event match analysis makes the network topology not affected by the occurrence of event aggregation in the event sequence.

[0044] 3、The directed weighted network provided by the application is for spatiotemporal sequence data with geographical position attribute and occurrence time attribute, and is easy to realize by software programming. It can be widely applied to spatiotemporal sequence data with geographical position attribute and occurrence time attribute, such as meteorological data, traffic data, communication data, internet throughput data, electroencephalogram data, etc. BRIEF DESCRIPTION OF DRAWINGS

[0045] Figure 1 is a flow chart of the directed weighted network building method for quantifying synchronization of specific events based on event coincidence analysis of the application;

[0046] Figure 2 is a directed weighted network generated by triggering event coincidence in event coincidence analysis of base station time flow spatiotemporal sequence;

[0047] Figure 3 is a directed weighted network generated by precursor event coincidence in event coincidence analysis of base station time flow spatiotemporal sequence;

[0048] Figure 4 is a directed weighted network corresponding to Figure 2 and Figure 3 is a comparison chart of node in-degree, out-degree and degree of two directed weighted networks corresponding to

[0049] Figure 5 is a comparison chart of node in-strength, out-strength and strength of two directed weighted networks corresponding to Figure 2 and Figure 3 is a comparison chart of node in-strength, out-strength and strength of two directed weighted networks corresponding to

[0050] Figure 6 is a topological feature of a directed weighted network corresponding to Figure 2 is a geographical distribution chart of node in-degree, node out-degree and node degree of a directed weighted network corresponding to

[0051] Figure 7 is a geographical distribution chart of node in-strength, node out-strength and node strength of a directed weighted network corresponding to Figure 2 is a geographical distribution chart of node in-strength, node out-strength and node strength of a directed weighted network corresponding to

[0052] Figure 8 is a geographical distribution chart of node in-strength, node out-strength and node strength of a directed weighted network corresponding to Figure 2 is a geographical distribution chart of node in-strength, node out-strength and node strength of a directed weighted network corresponding to

[0053] Figure 9 is a geographical distribution chart of node in-strength, node out-strength and node strength of a directed weighted network corresponding to Figure 2 is a geographical distribution chart of node in-strength, node out-strength and node strength of a directed weighted network corresponding to DETAILED DESCRIPTION

[0054] The directed weighted network building method for quantifying synchronization of specific events based on event coincidence analysis of the application will be described in detail below in combination with examples and drawings

[0055] The technical scheme adopted by the present application is: a directed weighted network building method based on event coincidence analysis for quantifying the synchronization of specific events, like Figure 1 , comprising the following steps:

[0056] Step 1: According to the threshold method, the data not less than the threshold in the original space-time sequence at the geographical position is identified as a specific event, and an event sequence is constructed according to the occurrence time of the specific event in the original space-time sequence, that is, the event sequence is an ordered sequence with the occurrence time of the specific event as the element, and the elements are arranged in the order of the occurrence time of the specific event. That is: the original space-time sequence at geographical position i is X(i)={x i l ,l=1,2,…,T i}, wherein T i represents the length of the space-time sequence at geographical position i, and the event sequence E(i)={t i l ,l=1,2,…,N i} is extracted at geographical position i by the threshold method, wherein N i represents the total number of specific events occurring at geographical position i, that is, the length of the event sequence E(i), and t i l represents the time of the lth occurrence of the specific event at geographical position i.

[0057] Step 2: According to the static coincidence interval ΔT and the time delay τ, it is judged whether the events at different geographical positions occur synchronously. In order to make the static coincidence interval ΔT have statistical significance, ΔT needs to satisfy

[0058] ΔT<T / N ma x (1)

[0059] In the above formula, T is the span of the time recorded by the original space-time sequence at all geographical positions, and N max is the maximum value of the lengths of all event sequences.

[0060] If the interval between the time t i l of the lth occurrence of the specific event at geographical position i and the time t j m of the mth occurrence of the specific event at geographical position j is within the static coincidence interval ΔT, it is considered that the events occur synchronously. Let the difference between the time t lm ij of the lth occurrence of the specific event at geographical position i and the time t lm ij of the mth occurrence of the specific event at geographical position j be d lm ij , that is

[0061]

[0062] If dlm ij satisfies: 0≤|d lm ij satisfies: 0≤|d lm ij satisfies: 0≤|d lm ij satisfies: 0≤|d

[0063] Step 3: Different geographical locations are taken as nodes of the network, the edges between nodes are determined according to the synchronization relationship between specific events at different geographical locations, the direction of the edges is determined according to the chronological order of the occurrence of the synchronous events at different geographical locations, and the weight of the edges is determined according to the number of synchronous events. Thus, the spatiotemporal sequence is mapped into a directed and weighted network with geographical locations where specific events occur as the basic elements. The specific method for determining the edges between nodes and the direction and weight of the edges is as follows:

[0064] If there are synchronous events between geographical location i and geographical location j, there is an edge between the nodes corresponding to the two geographical locations. Among all the synchronous events, if the time of the occurrence of the specific event at geographical location i is earlier than the time of the occurrence of the specific event at geographical location j, the direction of the edge is from the node corresponding to geographical location i to the node corresponding to geographical location j.

[0065] The weight of the edge depends on the number of synchronous events between different geographical locations. The method for measuring the number of synchronous events includes precursor event coincidence and trigger event coincidence. In order to clearly define the synchronization relationship between the specific event occurring at geographical location j for the mth time and the specific event occurring at geographical location i, a coefficient h ij lm is defined as

[0066]

[0067] In the precursor event coincidence, when the specific event occurring at geographical location j for the mth time is a synchronous event with multiple specific events occurring at geographical location i, the synchronous event is counted only once. At this time, the number of synchronous events between the specific event occurring at geographical location j for the mth time and the specific event occurring at geographical location i is ep ij ●m is defined as

[0068]

[0069] So, the number of synchronous events in which the specific event at geographical location j occurs before the specific event at geographical location i among all specific events at geographical location i is r p ij The number of synchronous events in which the specific event at geographical location j occurs before the specific event at geographical location i among all specific events at geographical location i can be obtained by the following formula:

[0070]

[0071] At this time, the direction of the edge is from the node corresponding to geographical location i to the node corresponding to geographical location j, and the weight of the edge is r p ij Similarly, the number of synchronous events in which the specific event at geographical location i occurs before the specific event at geographical location j among all specific events at geographical location j is r p ji At this time, the direction of the edge is from the node corresponding to geographical location j to the node corresponding to geographical location i, and the weight of the edge is r p ji .

[0072] In the coincidence of triggering events, when the specific event occurring for the lth time at geographical location i and the multiple specific events occurring at geographical location j are all synchronous events, the synchronous events are counted only once. At this time, the number of synchronous events between the specific event occurring for the lth time at geographical location i and the specific event occurring at geographical location j is et ij l●

[0073]

[0074] The number of synchronous events in which the specific event at geographical location j occurs after the specific event at geographical location i among all specific events at geographical location i is r t ij The number of synchronous events in which the specific event at geographical location j occurs after the specific event at geographical location i among all specific events at geographical location i can be obtained by the following formula:

[0075]

[0076] At this time, the direction of the edge is from the node corresponding to geographical location i to the node corresponding to geographical location j, and the weight of the edge is r t ij Similarly, the number of synchronous events in which the specific event at geographical location i occurs after the specific event at geographical location j among all events at geographical location j is At this time, the direction of the edge is from the node corresponding to geographical location j to the node corresponding to geographical location i, and the weight of the edge is r t ij .

[0077] So, the element w in the weighted adjacency matrix W of the directed weighted network is​ij r in the precursor event match can be obtained by selecting r p ij or r in the trigger event match t ij Two directed weighted networks can be constructed according to the precursor event match and the trigger event match respectively.

[0078] Step 4: Calculate the topological features of the directed weighted network, including node degree k i , network cumulative degree distribution P(k), node strength s i , network cumulative strength distribution P(s) and node directed clustering coefficient C i In the directed weighted network, the degree of a node includes out-degree k i out and in-degree k i in ,

[0079] k i = k i out + k i in (8)

[0080] In the above formula, N represents the total number of nodes in the network, a ij and a ji represent the directed connection relationship between node i and node j, if node i has an edge pointing to node j, then a ij = 1, otherwise a ij = 0, if node j has an edge pointing to node i, then a ji = 1, otherwise a ji = 0;

[0081]

[0082] In the above formula, N k' represents the number of nodes with degree k', N represents the total number of nodes in the network, and the strength of a node includes out-strength s i out and in-strength s i in ;

[0083] s i = s i out + s i in (10)

[0084] In the above formula, N represents the total number of nodes in the network, w ij represents the weight of the edge pointing from node i to node j, wji denotes the weight of the edge from node j to node i;

[0085]

[0086] In the above formula, N s' denotes the number of nodes with strength s', and N denotes the total number of nodes in the network;

[0087]

[0088] In the above formula, b ij denotes the connection relationship between node i and node j, if there is an edge between node i and node j, then b ij = 1, otherwise b ij = 0, a jk and a kj denote the directed connection relationship between node j and node k, if node j has an edge pointing to node k, then a jk = 1, otherwise a jk = 0, if node k has an edge pointing to node j, then a kj = 1, otherwise a kj = 0.

[0089] Step 5: Realize the visualization of the directed weighted network, the topological characteristics of the network and the geographical distribution of the topological characteristics by means of Matlab software.

[0090] Embodiment

[0091] The initial data used in this case is the base station time flow space-time sequence recorded by the base stations of a certain operator in a certain region, the time span is from 0:00 on February 22, 2017 to 23:00 on February 27, and the time flow is recorded once an hour. Identify the data not less than the 95th percentile in all base station time flow data as a specific event, and construct an event sequence according to the occurrence time of the specific event in the base station time flow space-time sequence. Set the static interval ΔT to 1 hour, and do not consider the time delay, i.e. τ is set to 0. Take the base station at the geographical position as the node, and if there is a synchronous event between the base stations, there is an edge between the corresponding nodes. According to the order of the occurrence time of the synchronous event between the base stations, determine the direction of the edge, and select the number of synchronous events that trigger the event coincidence and the precursor event coincidence as the edge weight. Finally, two directed weighted complex networks based on event coincidence analysis are obtained. The association between specific events in the system is studied through the largest connected subgraph of the network. As shown in Figure 2 is the directed weighted network generated by the base station time flow space-time sequence based on the event coincidence analysis of the trigger event coincidence. As shown in Figure 3 is the directed weighted network generated by the base station time flow space-time sequence based on the event coincidence analysis of the precursor event coincidence. As shown in Figure 4 isFigure 2 and Figure 3 A comparison diagram of the in-degree, out-degree, and degree of nodes in two directed weighted networks. (See diagram below.) Figure 5 What is shown is Figure 2 and Figure 3 A comparison of the node input strength, output strength, and intensity of two directed weighted networks is shown. The comparison reveals that while precursor event conformance and trigger event conformance are consistent in identifying synchronicity, they differ in measuring the degree of event synchronicity. A directed weighted network generated based on trigger event conformance is selected for further analysis. Figure 6 What is shown is Figure 2 The topological characteristics of a directed weighted network are as follows: the cumulative in-degree, cumulative out-degree, and cumulative degree distributions all follow exponential distributions, while the cumulative in-strength, cumulative out-strength, and cumulative strength distributions all follow power-law distributions. For example... Figure 7 What is shown is Figure 2 The geographical distribution of node in-degree, node out-degree, and node degree in a directed weighted network. Base stations with high activity during extreme event propagation are located in the southwest of the region. For example... Figure 8 What is shown is Figure 2 The geographical distribution map of node inflow strength, node outflow strength, and node strength in a directed weighted network. Base stations located in the southwest of the region are more susceptible to the propagation of extreme events from other base stations, and extreme events occurring at such base stations are also more likely to propagate to other base stations. For example... Figure 9 What is shown is Figure 2 Geographical distribution map of the directed clustering coefficients of nodes in a directed weighted network.

Claims

1. A method for constructing a directed weighted network based on quantifying the synchrony of specific events using event coincidence analysis, characterized in that, Comprising the following steps: 1) According to the threshold method, data not less than the threshold in the original space-time sequence at the geographical position is identified as a specific event, and an event sequence is constructed according to the occurrence time of the specific event in the original space-time sequence. The original space-time sequence is the base station time-flow space-time sequence recorded by the base station; that is, the event sequence is an ordered sequence with the occurrence time of the specific event as the element, and the elements are arranged in the order of the occurrence time of the specific event, that is: the original space-time sequence at the geographical position i is X(i) = {x i l , l = 1, 2, …, T i}, wherein T i represents the length of the space-time sequence at the geographical position i, and the event sequence E(i) = {t i l , l = 1, 2, …, N i} at the geographical position i is extracted by the threshold method, wherein N i represents the total number of specific events occurring at the geographical position i, that is, the length of the event sequence E(i), t i l represents the time of the lth occurrence of the specific event at the geographical position i; 2) According to the static coincidence interval ΔT and the time delay τ, it is judged whether the events at different geographical positions occur synchronously. If the interval between the time of the lth occurrence of a specific event at geographical position i and the time of the mth occurrence of a specific event at geographical position j is within the static coincidence interval ΔT, it is considered that the events occur synchronously, and the time of the lth occurrence of a specific event at geographical position i is defined as t i l the difference between the time of the mth occurrence of a specific event at geographical position j and the time of the lth occurrence of a specific event at geographical position i is defined as d j m lm ij , that is​ If d lm ij satisfies: 0≤|d lm ij |≤ΔT, the specific event occurring at the geographic location i for the lth time and the specific event occurring at the geographic location j for the mth time are considered to be synchronous events, if d lm ij satisfies: 0≤|d lm ij -τ|≤ΔT, the specific event occurring at the geographic location i for the lth time and the specific event occurring at the geographic location j for the mth time are considered to be lag synchronous events, wherein the time delay τ represents the length of the lag time between the spatiotemporal sequences. 3) taking different geographical locations as nodes of the network, determining the edges between the nodes according to the synchronization relationship between specific events at different geographical locations, determining the direction of the edges according to the chronological order of the occurrence of the synchronization events at different geographical locations, and determining the weight of the edges according to the number of the synchronization events, thus mapping the space-time sequence into a directed and weighted network with the geographical locations where the specific events occur as the basic elements; 4) Calculate topological features of the directed weighted network, including node degree k i , network cumulative degree distribution P(k), node strength s i , network cumulative strength distribution P(s), and node directed clustering coefficient C i In a directed weighted network, the degree of a node includes out-degree k i out and in-degree k i in , In the above formula, N represents the total number of nodes in the network, a ij and a ji represents the directed connection relationship between node i and node j, if node i has an edge pointing to node j, then a ij = 1, otherwise a ij = 0, if node j has an edge pointing to node i, then a ji = 1, otherwise a ji = 0, In the above formula, N k' the number of nodes of degree k', N represents the total number of nodes in the network, and the strength of a node includes the out-strength s i out and the in-strength s i in , In the above equation, N represents the total number of nodes in the network, w ij represents the weight of the edge from node i to node j, w ji represents the weight of the edge from node j to node i, In the above equation, N s' represents the number of nodes with strength s', and N represents the total number of nodes in the network, In the above formula, b ij represents the connection relationship between node i and node j, if there is an edge between node i and node j, then b ij = 1, otherwise b ij = 0, a jk and a kj represent the directed connection relationship between node j and node k, if node j has an edge pointing to node k, then a jk = 1, otherwise a jk = 0, if node k has an edge pointing to node j, then a kj = 1, otherwise a kj = 0. 5) realizing the visualization of the directed and weighted network, the topological features of the network, and the geographical distribution of the topological features by means of Matlab software; In step 3), the specific method for determining the edges between the nodes and the direction and weight of the edges is as follows: if there are synchronization events between geographical location i and geographical location j, then there is an edge between the nodes corresponding to the two geographical locations, and in all the synchronization events, if the specific event at geographical location i occurs earlier than the specific event at geographical location j, then the direction of the edge is from the node corresponding to geographical location i to the node corresponding to geographical location j, The weight of the connection depends on the number of simultaneous occurrences of events between different geographical locations, the method of measuring the number of simultaneous events includes precursor event coincidence and trigger event coincidence, in order to clearly define the synchronization relationship between the mth occurrence of a specific event at geographical location j and the occurrence of a specific event at geographical location i, the coefficient h is defined ij lm For In the coincidence of the precursor events, when the mth occurrence of a specific event at geographical location j is a synchronous event with multiple specific events occurring at geographical location i, the synchronous event is only counted once, and the number of synchronous events between the mth occurrence of the specific event at geographical location j and the specific events occurring at geographical location i is ep ij ●m For So, among all the specific events at geographical location j, the number of synchronous events r in which the specific event at geographical location i occurs before the specific event at geographical location j p ij This can be obtained by the following formula: At this time, the direction of the edge is from the node corresponding to the geographic location i to the node corresponding to the geographic location j, and the weight of the edge is r p ij Similarly, among all the specific events at the geographic location i, the number of synchronous events in which the specific event at the geographic location j occurs before the specific event at the geographic location i is r p ji At this time, the direction of the edge is from the node corresponding to the geographic location j to the node corresponding to the geographic location i, and the weight of the edge is r p ji , In the coincidence of the trigger event, when the specific event occurring at the geographic position i for the lth time and the multiple specific events occurring at the geographic position j are all synchronous events, the synchronous event is only counted once, at this time, the number of synchronous events between the specific event occurring at the geographic position i for the lth time and the specific event occurring at the geographic position j is et ij l● For r = number of synchronous events in which the specific event at geographic location j occurs after the specific event at geographic location i t ij may be obtained by the following equation: At this time, the direction of the edge is from the node corresponding to the geographic location i to the node corresponding to the geographic location j, and the weight of the edge is r t ij Similarly, among all events at the geographic location j, the number of synchronous events in which the specific event at the geographic location i occurs after the specific event at the geographic location j is r t ji At this time, the direction of the edge is from the node corresponding to the geographic location j to the node corresponding to the geographic location i, and the weight of the edge is r t ij , So, the element w ij By selecting r p ij or r t ij Two directed and weighted networks are constructed according to the precursor event coincidence and the trigger event coincidence, respectively.

2. The method of claim 1, wherein the method is a method of constructing a directed weighted network for quantifying the synchrony of a specific event based on event coincidence analysis. In step 2), in order to make the static compliance interval ΔT statistically meaningful, ΔT needs to satisfy ΔT < T / N max (7) T is the span of time of the original spatiotemporal sequence recording at all geographical locations, N max is the maximum value of the length of all event sequences.

Images