Spaelet-based spatio-temporal data co-evolution mode mining method and system

Abstract

The invention provides a spatio-temporal data co-evolution mode mining method and system based on Shaplet, and relates to the technical field of data processing. The method comprises the steps of collecting multivariate spatio-temporal data, and constructing a multivariate spatio-temporal data set; a space-time shaplet with space position information is extracted from the multivariate space-time data set; carrying out clustering and pattern recognition on the space-time shaplet through a clustering algorithm; according to a temporal relationship between every two space-time shapelet modes, dividing the space-time co-evolution modes into three types; and mining the three types of space-time co-evolution modes by using a mode growth mode according to a set vertical support degree threshold value and a set horizontal support degree threshold value to obtain a final space-time co-evolution mode set. The rationality and effectiveness of pattern mining can be improved. The system comprises a data acquisition module, a mode mining algorithm and a monitoring APP which are connected with one another to form a complete system, and beneficial reference is provided for research and development of related fields.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a shapelet-based spatiotemporal data co-evolution pattern mining method and system. Background Art

[0002] In the real world, mining spatiotemporal co-evolution patterns of multivariate spatiotemporal data is of great significance to real-world issues such as urban management and environmental protection. However, the complexity of the patterns of multivariate spatiotemporal data and the complexity and efficiency of computation have brought serious challenges to researchers.

[0003] Inspired by frequent item set mining algorithms and sequential pattern mining algorithms, the extraction of temporal association rules from time series data has developed rapidly. These temporal relationships can be explained using expert knowledge to achieve a better understanding of the current problem in a purely qualitative way. However, different event types have their own life cycles, and there are different temporalities between two events. Many previous algorithms have ignored the problem of multi-temporality.

[0004] Spatiotemporal coevolution can be divided into three modes according to different temporal relationships, including: synchronous spatiotemporal coevolution mode, delayed spatiotemporal coevolution mode and sequential spatiotemporal coevolution mode. In the synchronous spatiotemporal coevolution mode, shapelet pattern 1 and shapelet pattern 2 existing in time series 1 and time series 2 of different variables at different locations evolve synchronously, refer to Figure 1 In the delayed spatiotemporal coevolution model, the time when shapelet pattern 1 in time series 1 starts to change is before the time when shapelet pattern 2 in time series 2 starts to change, and the end time is also before the end time of shapelet pattern 2. Figure 2 In the sequential spatiotemporal coevolution pattern, shapelet pattern 2 of time series 2 changes after shapelet pattern 1 of time series 1 changes for a period of time. This evolution shows a certain regularity, which is more frequent than the number of occurrences of other patterns. The changes of shapelet pattern 1 in time series 1 and shapelet pattern 2 in time series 2 maintain a certain regularity, and the time processes of the two variables have no intersection. Figure 3 shown.

[0005] Existing pattern mining methods mainly focus on the definition of temporal relationship rules and adopt a two-stage mining method. First, all frequent temporal relationship patterns are discovered, and then temporal rules are generated based on the frequent temporal relationship patterns, which leads to the discovery of a large number of temporal patterns that are irrelevant to temporal relationship rules, wasting a lot of time. In addition, these methods are frequency-driven and use the Apriori property to narrow the search space, making the pruning strategy unsuitable for utility-driven mining. Summary of the invention

[0006] In order to solve the above problems existing in the prior art, the embodiment of the present invention provides a method and system for mining spatiotemporal data co-evolution patterns based on shapelets, taking spatiotemporal shapelets with spatial information as the unit of pattern mining, mining spatiotemporal co-evolution patterns of three different temporalities according to the multi-temporal relationship of shapelets in different spatiotemporal shapelet pattern types, and improving the rationality and effectiveness of pattern mining. The technical solution is as follows:

[0007] In one aspect, a shapelet-based spatiotemporal data co-evolution pattern mining method is provided, the method comprising the following steps:

[0008] S1, collect multivariate spatiotemporal data and construct a multivariate spatiotemporal dataset;

[0009] S2, extracting a spatiotemporal shapelet with spatial location information from the multivariate spatiotemporal dataset;

[0010] S3, clustering and pattern recognition of the spatiotemporal shapelet by clustering algorithm;

[0011] S4, based on the temporal relationship between pairwise spatiotemporal shapelet patterns, the spatiotemporal coevolution patterns are divided into three categories;

[0012] S5. According to the set vertical support threshold and horizontal support threshold, the three types of spatiotemporal coevolution patterns are mined using the pattern growth method to obtain the final spatiotemporal coevolution pattern set.

[0013] Optionally, in step S1, the multivariate spatiotemporal data is a spatiotemporal data sequence with spatial information collected by sensors, and the multivariate spatiotemporal data set includes an environmental monitoring data set.

[0014] Optionally, the step S2 specifically includes:

[0015] Extract shapelets with spatial location information as spatiotemporal shapelets and construct a spatiotemporal shapelet set S, which satisfies , the position information of the shapelet in set S is stored in set D; m is the variable number, M is the total number of variables, k is the shapelet number, and K is the total number of shapelets.

[0016] Optionally, the step S3 specifically includes:

[0017] The same spatiotemporal shapelets are clustered by clustering algorithm to obtain the clustering result R; clustering result , where R c It is the c-type spatiotemporal shapelet pattern, which contains multiple spatiotemporal shapelets of the same type. , ,in Represents R c The shapelet structure in represents the spatiotemporal shapelet structure; k∈[1,K] is the shapelet number, c∈[1,C], C is the total number of clustering patterns, i.e., the total number of spatiotemporal shapelet pattern types; m∈[1,M] is the variable number, t s ,t e ∈[1,h] represents the start and end time of shapelet, D n ,n∈[1,N] is The corresponding spatial position is a longitude and latitude coordinate; R c The shapelet structure of each variable in takes the average value as the spatiotemporal shapelet pattern of this class, omitting its start and end time, and using p c,m It is represented by , where c represents the clustering pattern number and m represents the variable number.

[0018] Optionally, in step S4, the spatiotemporal shapelet patterns include three temporal relationships: synchronous, delayed, and ordered. The three temporal relationships can occur between any two spatiotemporal shapelet patterns of different types. c,m and p c’,m’ and exist separately and Two shapelet structures, r represents three different temporal relations, r∈{Syn,Del,Ord};

[0019] Using triples Represents a spatiotemporal co-evolutionary pattern consisting of a shapelet structure of two different types of spatiotemporal shapelet patterns;

[0020] The step S4 specifically includes:

[0021] Initialize three types of spatiotemporal co-evolution pattern subsets P Syn , P Del , P Ord ;

[0022] The spatiotemporal shapelet patterns in the spatiotemporal shapelet set S are combined into different types of spatiotemporal co-evolution patterns, and their temporal relations r are queried;

[0023] According to different temporal relations r, the spatiotemporal shapelet patterns are divided into corresponding spatiotemporal co-evolution pattern subsets;

[0024] The three types of spatiotemporal coevolution pattern subsets are combined to obtain the initial spatiotemporal coevolution pattern set.

[0025] Optionally, the step S5 specifically includes:

[0026] Initialize the cardinality of the spatiotemporal coevolution pattern;

[0027] Scanning the spatiotemporal coevolution patterns in the initial spatiotemporal coevolution pattern set, and calculating the vertical support and horizontal support of the current spatiotemporal coevolution pattern;

[0028] Performing pruning operations by using set vertical support thresholds and horizontal support thresholds to filter the spatiotemporal coevolution patterns in the initial spatiotemporal coevolution pattern set, and adding new patterns to achieve pattern growth;

[0029] Update the cardinality value and return the final set of spatiotemporal coevolution patterns.

[0030] Optionally, the vertical support refers to: the number of spatiotemporal shapelet patterns in which the spatiotemporal coevolution pattern P appears in the spatiotemporal shapelet set divided by the total number of spatiotemporal shapelet patterns in the spatiotemporal shapelet set, wherein the spatiotemporal coevolution pattern P is one of the three types of spatiotemporal coevolution patterns;

[0031] The horizontal support refers to: assuming that the spatiotemporal coevolution pattern P is in the time series T n appears, then the spatiotemporal coevolution pattern P in the time series T n The horizontal support on the time series T n The number of occurrences of .

[0032] On the other hand, a shapelet-based spatiotemporal data co-evolution pattern mining system is provided, which is used to implement any of the above methods, and the system includes:

[0033] Data collection module, used to collect multivariate spatiotemporal data and construct multivariate spatiotemporal data sets;

[0034] A shapelet extraction module, for extracting a spatiotemporal shapelet with spatial location information from the multivariate spatiotemporal dataset;

[0035] A shapelet clustering module, used for clustering and pattern recognition of the spatiotemporal shapelet by using a clustering algorithm;

[0036] The pattern classification module is used to classify the spatiotemporal coevolution patterns into three categories based on the temporal relationship between pairwise spatiotemporal shapelet patterns;

[0037] The pattern mining module is used to mine three types of spatiotemporal coevolution patterns using the pattern growth method according to the set vertical support threshold and horizontal support threshold, and obtain the final spatiotemporal coevolution pattern set.

[0038] Optionally, the data acquisition module includes a sensor component, a serial port server, a management workstation, an interface server, an information center storage server, and a platform core switch;

[0039] Among them, the sensor component is connected to the management workstation through the serial port server, the management workstation is connected to the interface server, the interface server is connected to the information center storage server through a unidirectional isolation gateway, the information center storage server is connected to the platform core switch through optical fiber, and the platform core switch is connected to multiple cloud computing nodes and web servers.

[0040] Optionally, the system further includes a monitoring APP, and the monitoring APP includes: a client, a server, and a system management background;

[0041] The client is used for user registration and login, online query, modification and logout; the server is used for registration and login verification as well as data transmission, addition, modification and deletion; the system management background is used to manage the database.

[0042] In another aspect, an electronic device is provided, the electronic device comprising:

[0043] processor;

[0044] A memory having computer-readable instructions stored thereon, wherein when the computer-readable instructions are loaded and executed by the processor, the steps of the shapelet-based spatiotemporal data co-evolution pattern mining method are implemented.

[0045] On the other hand, a computer-readable storage medium is provided, in which program code is stored. The program code can be called by a processor to execute the steps of the shapelet-based spatiotemporal data co-evolution pattern mining method as described above.

[0046] The beneficial effects brought about by the technical solution provided by the present invention include at least:

[0047] The present invention provides a shapelet-based spatiotemporal data co-evolution pattern mining method, which uses spatiotemporal shapelet as a pattern mining unit, uses shapelet extraction technology to extract spatiotemporal shapelet with spatial location information from a multivariate spatiotemporal data set, performs pattern recognition through a clustering algorithm, and mines spatiotemporal co-evolution patterns of three different temporal states according to the multi-temporal relationship of shapelet, thereby improving the rationality and effectiveness of pattern mining. The present invention also constructs a shapelet-based spatiotemporal data co-evolution pattern mining system, which includes a data acquisition module, a pattern mining algorithm, and a monitoring APP. The three parts are interconnected to form a complete system, which provides a useful reference for research and development in related fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0049] Figure 1 It is a schematic diagram of the synchronous spatiotemporal co-evolution pattern;

[0050] Figure 2 It is a schematic diagram of the delayed spatiotemporal coevolution pattern;

[0051] Figure 3 It is a schematic diagram of the sequential spatiotemporal coevolution pattern;

[0052] Figure 4 is a flow chart of a shapelet-based spatiotemporal data co-evolution pattern mining method provided by an embodiment of the present invention;

[0053] Figure 5 is a schematic diagram of a temporal relationship provided by an embodiment of the present invention;

[0054] Figure 6 is a support diagram provided by an embodiment of the present invention;

[0055] Figure 7 It is a schematic diagram of the STCEMinerTree() algorithm provided in an embodiment of the present invention;

[0056] Figure 8 is a schematic diagram of experimental results of four data sets provided by an embodiment of the present invention;

[0057] Fig. 9 It is a schematic diagram of comparison between STCEMiner provided by an embodiment of the present invention and three comparison algorithms in four data sets;

[0058] Fig.10 It is a schematic diagram of the comparison of the running time of STCEMiner on four data sets provided by an embodiment of the present invention;

[0059] Fig.11 It is a schematic diagram of a synchronous spatiotemporal co-evolution mode in different regions provided by an embodiment of the present invention;

[0060] Fig.12 It is a schematic diagram of prediction and verification of synchronous spatiotemporal co-evolution patterns in different regions provided by an embodiment of the present invention;

[0061] Fig.13 It is a schematic diagram of the delayed spatiotemporal co-evolution mode of different regions provided by an embodiment of the present invention;

[0062] Fig.14 It is a schematic diagram of prediction and verification of delayed spatiotemporal co-evolution patterns in different regions provided by an embodiment of the present invention;

[0063] Fig.15 It is a schematic diagram of the sequential spatiotemporal co-evolution pattern of different regions provided by an embodiment of the present invention;

[0064] Fig.16 It is a schematic diagram of prediction and verification of sequential spatiotemporal co-evolution patterns in different regions provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0065] In order to make the purpose, technical solution and advantages of the embodiment of the present invention clearer, the technical solution of the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings of the embodiment of the present invention. Obviously, the described embodiment is a part of the embodiment of the present invention, not all of the embodiments. Based on the described embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0066] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "exemplary" in the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the word "exemplary" is used to present concepts in a specific way.

[0067] The embodiment of the present invention provides a method for mining spatiotemporal data co-evolution patterns based on shapelet, which can be implemented by an electronic device, which can be a terminal or a server. The method of the present invention is based on a newly designed spatiotemporal co-evolution pattern mining model (Spatiotemporal co-evolution pattern mining, STCEMiner), which takes spatiotemporal shapelet with spatial information as a unit, and aims to mine spatiotemporal co-evolution patterns with multi-temporality in spatiotemporal data. The input of the model is a set of multivariate spatiotemporal data, and through the preset vertical support threshold and horizontal support threshold, a set of spatiotemporal co-evolution pattern sets that meet the threshold requirements are finally output.

[0068] Specifically, Figure 4 As shown, the processing flow of the method may include the following steps:

[0069] S1. Collect multivariate spatiotemporal data and construct a multivariate spatiotemporal dataset.

[0070] In this step, the multivariate spatiotemporal data is a spatiotemporal data sequence with spatial information collected by sensors. In an embodiment of the present invention, the multivariate spatiotemporal data set mainly includes an environmental monitoring data set, a smart city monitoring data set, etc.

[0071] S2. Extract spatiotemporal shapelets with spatial location information from multivariate spatiotemporal datasets.

[0072] In this step, shapelets with spatial location information are extracted as spatiotemporal shapelets, and a spatiotemporal shapelet set S is constructed. The set S satisfies , the position information of the shapelet in set S is stored in set D; m is the variable number, M is the total number of variables, k is the shapelet number, and K is the total number of shapelets.

[0073] S3. Clustering and pattern recognition of spatiotemporal shapelets are performed through clustering algorithms.

[0074] In this step, the spatiotemporal shapelets of the same type are clustered by a clustering algorithm to obtain the clustering result R; the clustering result , where R c It is the c-type spatiotemporal shapelet pattern, which contains multiple spatiotemporal shapelets of the same type. , ,in Represents R c The shapelet structure in represents the spatiotemporal shapelet structure; k∈[1,K] is the shapelet number, c∈[1,C], C is the total number of clustering patterns, i.e., the total number of spatiotemporal shapelet pattern types; m∈[1,M] is the variable number, t s ,t e ∈[1,h] represents the start and end time of shapelet, D n ,n∈[1,N] is The corresponding spatial position is a longitude and latitude coordinate; R c The shapelet structure of each variable in takes the average value as the spatiotemporal shapelet pattern of this class, omitting its start and end time, and using p c,m It is represented by , where c represents the clustering pattern number and m represents the variable number.

[0075] S4. Based on the temporal relationship between pairwise spatiotemporal shapelet patterns, the spatiotemporal coevolution patterns are divided into three categories.

[0076] Any combination of two different types of spatiotemporal shapelet patterns has a temporal relationship. Figure 5 As shown in Figure 2, there are three temporal relationships between different types of spatiotemporal shapelet patterns: Synchronous, Delayed, and Order. The three temporal relationships can occur between any two different types of spatiotemporal shapelet patterns. c,m and p c’,m’ and exist separately Two shapelet structures, r represents three different temporal relations, r∈{Syn,Del,Ord}.

[0077] In this step, the triplet is used represents a spatiotemporal coevolution pattern composed of shapelet structures of two different types of spatiotemporal shapelet patterns. Moreover, Pr can be used to represent the temporal relationship between two different types of spatiotemporal shapelet patterns within any spatiotemporal coevolution pattern.

[0078] Figure 5 Shows the structure of two shapelets The synchronization relationship between two spatiotemporal shapelet patterns composed of hour . Figure 5 The delay relationship also shows that the two shapelet structures The temporal relationship between two spatiotemporal shapelet patterns is determined if and only if the start and end times of the two spatiotemporal shapelet patterns satisfy hour, . Figure 5 The order relationship shows that there are three shapelet structures The relationship between the three spatiotemporal shapelet patterns is if and only if hour .

[0079] Assuming different mode types The spatiotemporal shapelet , the spatial position is , the two spatiotemporal shapelet structures are represented as and , the spatiotemporal co-evolution pattern P composed of them is expressed as , where r∈{Syn,Del,Ord}, c1,c2∈[0,C] are the category numbers, m1,m2∈[0,M] are the variable numbers, n1,n2∈[0,N] are the position numbers, k1,k2∈[0,K] are the shapelet numbers, sup V is the vertical support of the pattern, conf represents its confidence, P∈P, P is the set of spatiotemporal coevolution patterns. Since there are three different temporal relations r∈{Syn,Del,Ord}, there are three different types of spatiotemporal coevolution patterns P Syn , P Del , P Ord .

[0080] Based on the above definition, step S4 specifically includes:

[0081] Initialize three types of spatiotemporal co-evolution pattern subsets P Syn , P Del , P Ord ;

[0082] The spatiotemporal shapelet patterns in the spatiotemporal shapelet set S are combined into different types of spatiotemporal co-evolution patterns, and their temporal relations r are queried;

[0083] According to different temporal relations r, the spatiotemporal shapelet patterns are divided into corresponding spatiotemporal co-evolution pattern subsets;

[0084] The three types of spatiotemporal coevolution pattern subsets are combined to obtain the initial spatiotemporal coevolution pattern set.

[0085] S5. According to the set vertical support threshold and horizontal support threshold, the three types of spatiotemporal coevolution patterns are mined using the pattern growth method to obtain the final spatiotemporal coevolution pattern set.

[0086] The vertical support refers to: the number of spatiotemporal shapelet patterns in which the spatiotemporal coevolution pattern P appears in the spatiotemporal shapelet set divided by the total number of spatiotemporal shapelet patterns in the spatiotemporal shapelet set, where the spatiotemporal coevolution pattern P is one of the three types of spatiotemporal coevolution patterns. The vertical support of the spatiotemporal coevolution pattern P is calculated as follows:

[0087]

[0088] in is the number of occurrences of the spatiotemporal coevolution pattern P, is the total number of occurrences of three different types of spatiotemporal coevolution patterns.

[0089] The horizontal support refers to: assuming that the spatiotemporal coevolution pattern P is in the time series T n appears, then the spatiotemporal coevolution pattern P in the time series T n The horizontal support on the time series T n The number of occurrences of spatiotemporal coevolution pattern P in the time series T n The horizontal support on is calculated as follows:

[0090]

[0091] like Figure 6 As shown, two time series (a) and (b) appear and Two spatiotemporal coevolution patterns, P1 appears twice in (a) and once in (b). Figure 6 In the data set shown, the vertical support sup V (P1) = 1. In addition, the horizontal support sup H (P1;(a))=2, and the horizontal support sup H (P1;(b))=1.

[0092] In addition, the confidence level is defined as follows: Assume is a spatiotemporal co-evolution model. is the antecedent of the spatiotemporal coevolution pattern P, and the remaining elements constitute the consequent of the spatiotemporal coevolution pattern P. The confidence of the spatiotemporal coevolution pattern P is the probability that the consequent will occur after the antecedent occurs in the spatiotemporal coevolution pattern, which is calculated as follows:

[0093]

[0094] Among them, conf(P) is the confidence of the spatiotemporal coevolution pattern P.

[0095] The step S5 specifically includes:

[0096] Initialize the cardinality of the spatiotemporal coevolution pattern;

[0097] Scanning the spatiotemporal coevolution patterns in the initial spatiotemporal coevolution pattern set, and calculating the vertical support and horizontal support of the current spatiotemporal coevolution pattern;

[0098] Performing pruning operations by using set vertical support thresholds and horizontal support thresholds to filter the spatiotemporal coevolution patterns in the initial spatiotemporal coevolution pattern set, and adding new patterns to achieve pattern growth;

[0099] Update the cardinality value and return the final set of spatiotemporal coevolution patterns.

[0100] In an embodiment of the present invention, spatiotemporal shapelets are used as the unit of pattern mining, and shapelet extraction technology is used to extract spatiotemporal shapelets with spatial location information from a multivariate spatiotemporal data set. Pattern recognition is performed through a clustering algorithm, and spatiotemporal co-evolution patterns of three different temporal states are mined based on the multi-temporal relationship of shapelets, thereby improving the rationality and effectiveness of pattern mining.

[0101] Furthermore, the specific implementation process of the method is as follows:

[0102] For different shapelets, it is necessary to determine the temporal relationship between different shapelets before conducting spatiotemporal co-evolution pattern mining. This part of the work is completed by Algorithm 1 and Algorithm 2. Algorithm 1 describes the framework of the spatiotemporal co-evolution pattern mining process based on pattern growth. First, in Step 1, shapelets with spatial location information are extracted from the spatiotemporal dataset through the shapelet extraction algorithm, and a shapelet set S is constructed. The location information of the shapelets in the set is stored in the set D; secondly, in Step 2, several shapelets are spatiotemporally clustered, and pattern recognition is performed on the clustering result R; Step 3 calls Algorithm 2 GetClass(), and divides the spatiotemporal shapelet patterns into three corresponding categories according to the temporal relationship between the two spatiotemporal shapelet patterns; Steps 4-7 call Algorithm 3 STCEMinerTree() to generate spatiotemporal co-evolution patterns; Step 8 outputs the spatiotemporal co-evolution pattern set P final .

[0103]

[0104] The process of classifying spatiotemporal shapelet patterns into three categories according to the temporal relationship between them is shown in Algorithm 2. Step 1, initialize the three types of spatiotemporal co-evolution pattern subsets; Step 2-7, combine the spatiotemporal shapelet patterns in the shapelet set S into different types of spatiotemporal co-evolution patterns, and query their temporal relationship r, Step 8-11 divides them into corresponding spatiotemporal co-evolution pattern subsets according to different temporal relationships, and Step 12 combines three different spatiotemporal co-evolution pattern subsets to obtain the initial spatiotemporal co-evolution pattern set; Step 17 returns the initial spatiotemporal co-evolution pattern set, and the initial spatiotemporal co-evolution pattern set at this time still needs to be processed by the subsequent Algorithm 3.

[0105]

[0106] After completing the determination of the temporal relationship between shapelets, Algorithm 3 performs spatiotemporal co-evolution pattern mining. Through the horizontal support threshold parameter and the vertical support threshold parameter, Algorithm 3 can use the pattern growth method to mine spatiotemporal co-evolution patterns for spatiotemporal data. First, in Step 1, the initialization work is performed, the cardinality of the spatiotemporal co-evolution pattern is initially set to 2, and the root node is cleared; Step 2-3 scans the database and obtains the vertical support and horizontal support of the current spatiotemporal co-evolution pattern; Step 4-5 performs pruning operations through two thresholds, filters the original pattern set, and adds new patterns to achieve pattern growth; Step 6 updates the cardinality Level value; Finally, in Step 8, the final spatiotemporal co-evolution pattern set P is returned. final .

[0107]

[0108] Figure 7 The mining process of a STCEMinerTree() is demonstrated. The start and end time information of the spatiotemporal shapelet pattern in the example data set is not displayed. This is because in the actual spatiotemporal coevolution pattern mining process, the type of spatiotemporal shapelet pattern is mainly considered, and the same type of spatiotemporal shapelets have different start and end times. When there is no need to specifically analyze the temporal relationship, the start and end times of the spatiotemporal shapelet pattern can be omitted. In the example, the spatiotemporal coevolution patterns are mapped to a path of the STCEMinerTree, and each node represents the corresponding shapelet pattern type. The support is calculated for pruning operations. No pruning operation occurs in the example data set, and the corresponding spatiotemporal coevolution patterns can be obtained by retrieving all paths.

[0109] The complexity analysis of the STCEMiner algorithm is as follows, assuming is the number of spatiotemporal shapelets in the spatiotemporal dataset T. The time complexity of scanning the database is Assume that z is the average number of spatiotemporal coevolution patterns in cardinality k-1. Then the complexity of mining spatiotemporal coevolution patterns with cardinality k is Assuming h is the maximum cardinality of spatiotemporal coevolution patterns, the spatiotemporal complexity of the spatiotemporal coevolution pattern mining algorithm depends on the size of the search space. and the complexity of the mining process itself .

[0110] Accordingly, an embodiment of the present invention further provides a shapelet-based spatiotemporal data co-evolution pattern mining system, the system comprising:

[0111] Data collection module, used to collect multivariate spatiotemporal data and construct multivariate spatiotemporal data sets;

[0112] A shapelet extraction module, for extracting a spatiotemporal shapelet with spatial location information from the multivariate spatiotemporal dataset;

[0113] A shapelet clustering module, used for clustering and pattern recognition of the spatiotemporal shapelet by using a clustering algorithm;

[0114] The pattern classification module is used to classify the spatiotemporal coevolution patterns into three categories based on the temporal relationship between pairwise spatiotemporal shapelet patterns;

[0115] The pattern mining module is used to mine three types of spatiotemporal coevolution patterns using the pattern growth method according to the set vertical support threshold and horizontal support threshold, and obtain the final spatiotemporal coevolution pattern set.

[0116] The system of this embodiment can be used to execute Figure 4 The technical solution of the method embodiment shown has similar implementation principles and technical effects, which will not be repeated here.

[0117] Furthermore, the data acquisition module includes a sensor component, a serial port server, a management workstation, an interface server, an information center storage server, and a platform core switch.

[0118] Among them, the sensor component is connected to the management workstation through the serial port server, the management workstation is connected to the interface server, the interface server is connected to the information center storage server through a unidirectional isolation gateway, the information center storage server is connected to the platform core switch through optical fiber, and the platform core switch is connected to multiple cloud computing nodes and web servers.

[0119] The data acquisition module is a supporting system responsible for sensor monitoring data acquisition. It uses C++ as the development language and is embedded with multiple data communication protocols such as 101, 102, 103, 104, Modbus, CDT, DISA, etc. of IEC 60870-5. The modeling complies with the requirements of the interface reference model, common information model (CIM) and component interface specification (CIS) in IEC 61970, complies with international standards, and can be seamlessly integrated with various systems as middleware; it realizes the access of system data such as monitoring system, integrated energy management and control system, metering, fault analysis and alarm push. The system supports the access of multiple devices and has the ability to parse multiple protocols.

[0120] The data acquisition module adopts a two-level architecture, namely a single integrated management and control system and a cloud platform centralized monitoring system. The integrated management and control system is used to collect real-time operation monitoring data, realize local data monitoring, historical data sampling and storage, and upload key real-time data to the cloud platform centralized monitoring system. The cloud platform centralized monitoring system obtains real-time monitoring data. The communication protocol between the two-level systems can adopt the power standard IEC104 protocol or other protocols. The real-time data acquisition frequency supports seconds according to the protocol requirements, and can support change quantity upload, cyclic upload and call modes.

[0121] Furthermore, the system also includes a monitoring APP, and the monitoring APP includes: a client, a server and a system management background.

[0122] The client is used for user registration and login, online query, modification and logout; the server is used for registration and login verification as well as data transmission, addition, modification and deletion; the system management background is used to manage the database.

[0123] The client is developed and designed using the MUI front-end framework, and HTML5, CSS, and JavaScript are used for front-end development, which is used for user registration and login, online query, modification, and logout. The server is developed using the ThinkJS server-side framework, combined with the MySQL database, which is used for registration and login verification as well as data transmission, addition, modification, and deletion. The system management backend is developed using HTML5, CSS, and JavaScript to manage the database.

[0124] The monitoring APP is easy to operate and has a simple and beautiful interface. It is real-time, and registered users can log in to the system through the mobile APP no matter where they are. The system provides automatic query and display functions, as well as user registration information management functions.

[0125] The shapelet-based spatiotemporal data collaborative evolution pattern mining system constructed by the present invention consists of three main parts: data collection, pattern mining, and monitoring APP, which are interconnected to form a complete system, providing a useful reference for research and development in related fields.

[0126] In the experimental evaluation part, the evaluation focus of this paper is the performance comparison of STCEMiner algorithm with other advanced temporal pattern mining algorithms, as well as its own performance under different parameters. All experiments are run in Python 3.9, i7-1165G7 @ 2.80GHz processor, 16GB memory environment.

[0127] In order to evaluate the performance of the proposed algorithm, the present invention will compare some well-known indicators in experiments: runtime consumption, the ability of pruning strategies and its own performance under different parameters.

[0128] The algorithm proposed in the present invention is called STCEMiner. In the experiment, it is compared with three comparison algorithms, namely Z-Miner, TPMiner and IEMiner. The introduction of the three algorithms is shown below.

[0129] 1) Z-Miner: Z-Miner is an effective algorithm for mining frequent permutations of event intervals, which uses Z-Table and Z permutations to efficiently generate candidates and support calculations.

[0130] 2) TPMiner: TPMiner mines sequential patterns from time interval-based data. The algorithm uses endpoint representation and end time to simplify the complex relationship between event intervals.

[0131] 3) IEMiner: The IEMiner algorithm is an Apriori-based pattern mining algorithm for mining frequent temporal patterns from time interval-based events. The algorithm designs an effective support counting procedure and adopts a window blacklist strategy and a prefix counting strategy to further improve the algorithm performance.

[0132] To evaluate the STCEMiner algorithm, several real-world datasets from different fields, namely the Motes dataset, the KitchenCmu1 dataset, the KitchenCmu2 dataset, and the Xi’an environmental monitoring dataset, are used. The detailed information of the datasets is shown in Table 1.

[0133] Table 1 Dataset information

[0134]

[0135] 1) Motes: The Motes dataset is a collection of reading logs from 54 sensors deployed in a research lab. Each sensor collects 4 types of data, namely temperature, humidity, light, and voltage. All methods are evaluated on one day’s log, with a total of 2880 time steps.

[0136] 2) KitchenCmu1: The KitchenCmu2 dataset is a dataset collected by 39 subjects while cooking five different recipes using 4 wireless IMU sensors and 5 wired IMU sensors.

[0137] 3) KitchenCmu2: The KitchenCmu2 dataset is a dataset of abnormal situations of three subjects when cooking five different recipes. Four wireless IMU sensors and four wired IMU sensors are placed on the subjects as spatial position information.

[0138] 4) Xi'an environmental monitoring data set: Environmental monitoring data of 13 districts and cities in Xi'an, including AQI, PM2.5, PM10, S02, N02, O3, CO and other monitoring items. All data are collected every hour and processed into daily detection values. This example uses a time step of 320 days from July 5, 2021 to May 30, 2022.

[0139] First, in order to verify the integrity of pattern mining and the necessity of horizontal support, the present invention designed a quantitative experiment and conducted experiments on the Motes dataset and the KitchenCmu1 dataset. In the experiment, the present invention set the parameters, set the vertical support threshold between 20% and 70%, and conducted algorithm integrity experiments with and without setting the horizontal support threshold. Figure 8 (a), (b), (c), and (d) show the number of patterns found on the four datasets respectively. From the experimental results of the Motes dataset, we can see that the number of spatiotemporal coevolution patterns found by the algorithm with horizontal support threshold setting is much higher than that of the algorithm without setting horizontal support threshold. In the KitchenCmu1 dataset, the difference is even greater, which shows that horizontal support can discover some overlooked spatiotemporal coevolution patterns.

[0140] In addition, Fig. 9 (a), (b), (c), and (d) show the number of patterns found by STCEMiner and the three comparison algorithms in the four data sets. In the four data sets, due to the horizontal support, the algorithm of the present invention found the largest number of patterns, and in some individual data, it led by 10%. This is because in the process of pattern mining, the other algorithms have pruned too many possible related patterns.

[0141] In the comparative experiment of running duration, the present invention uses STCEMiner algorithm and IEMiner algorithm to conduct comparative analysis on four data sets, referring to Fig.10 As shown in (a), (b), (c), and (d), the running time results of the four data sets are shown respectively. In this figure, it is clearly shown that STCEMiner is much faster than IEMiner in all data sets. This is because the algorithm of the present invention uses the pattern growth method to perform pattern mining, while the comparison algorithm uses the Apriori method to perform pattern mining.

[0142] Finally, a case analysis is conducted on the Xi’an environmental monitoring dataset. Fig.11 , Fig.13 and Fig.15 The spatiotemporal coevolution patterns obtained for three different temporalities are visualized.

[0143] Fig.11 (a) shows the spatiotemporal shapelet pattern The spatiotemporal shapelet structure , Fig.11 (b) shows the spatiotemporal shapelet pattern The spatiotemporal shapelet structure , the two appear in a synchronous spatiotemporal co-evolution mode, with a vertical support of 0.537 and a confidence of 0.614. The shapelet superscript 1 represents the first variable PM2.5, and the subscript 6 represents the sixth shapelet. By querying D6, we can find that its location is in Chang'an District. Similarly, The third shapelet representing variable O3 is located in the Economic Development Zone after querying D3. It can be expressed as: .

[0144] In order to verify the synchronous spatiotemporal coevolution model In this embodiment, the data of the first 320 days of the pattern are used for training, and the pattern is found to reappear from the 388th day to the 408th day, such as Fig.12 As shown in (a) and (b), when the O3 variable in Chang'an District experienced a sharp change, first increasing and then decreasing, the PM2.5 variable in the Economic Development Zone also experienced a similar sharp change. This is consistent with the air survey report at the time.

[0145] Fig.13 (a) shows the spatiotemporal shapelet pattern The spatiotemporal shapelet structure , Fig.13 (b) shows the spatiotemporal shapelet pattern The spatiotemporal shapelet structure , the two appear in a delayed spatiotemporal co-evolution mode, with a vertical support of 0.723 and a confidence of 0.749. The shapelet superscript 2 indicates the second variable PM10, and the subscript 1 indicates the first shapelet. The query of D5 shows that its location is at the Municipal People's Stadium. The fifth shapelet representing variable NO2 is located in Yanliang District after querying D1. The gray shaded area in the figure indicates the lag time between the two shapelet patterns. The spatiotemporal co-evolution pattern is expressed as: .

[0146] In order to verify the delayed spatiotemporal coevolution model In the embodiment of the present invention, the data of the first 320 days of the model is used for training, and a verification is performed after 100 days of use, such as Fig.14 As shown in (a) and (b), from the 377th to the 395th day, the PM10 variable concentration at the People's Stadium of the city first increased and then decreased, and for a period of time, the NO2 concentration in Yanliang District first increased and then decreased. The model accurately predicted this behavior, which is consistent with the air survey report of that year.

[0147] Fig.15 (a) shows the spatiotemporal shapelet pattern The spatiotemporal shapelet structure , Fig.15 (b) shows the spatiotemporal shapelet pattern The spatiotemporal shapelet structure , Fig.15 (c) shows the spatiotemporal shapelet pattern The spatiotemporal shapelet structure and , the three appear in a sequential spatiotemporal co-evolution pattern, located in Lintong District, Qujiang and Caotan, with a vertical support of 0.584 and a confidence of 0.675. The sequential spatiotemporal co-evolution pattern can be expressed similarly using the formula: .

[0148] Similarly, the sequential spatiotemporal co-evolution pattern was predicted and verified by using the first 320 days of data for training and the last 100 days for verification, as shown in Fig.15 As shown in (a), (b), and (c), from the 324th to the 394th day, after the PM2.5 variable concentration in Lintong District increased, the PM10 variable concentration in Qujiang first increased and then decreased, and the O3 concentration also decreased. This phenomenon verifies the credibility of the model of the present invention.

[0149] In summary, the spatiotemporal co-evolution pattern mined by the present invention can predict the changes in air indicators in a certain area, which helps managers to make predictions in advance.

[0150] In an exemplary embodiment, the present invention further provides an electronic device, the electronic device comprising:

[0151] processor;

[0152] A memory having computer-readable instructions stored thereon, wherein when the computer-readable instructions are loaded and executed by the processor, the steps of the shapelet-based spatiotemporal data co-evolution pattern mining method are implemented.

[0153] In an exemplary embodiment, the present invention further provides a computer-readable storage medium, wherein at least one instruction is stored in the computer-readable storage medium, and the at least one instruction is loaded and executed by a processor to implement the steps of the shapelet-based spatiotemporal data co-evolution pattern mining method as described above. For example, the computer-readable storage medium may be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, and the like.

[0154] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.

[0155] References in the specification to "one embodiment", "an embodiment", "an exemplary embodiment", "some embodiments", etc. indicate that the described embodiments may include a particular feature, structure, or characteristic, but not every embodiment may include the particular feature, structure, or characteristic. In addition, when a particular feature, structure, or characteristic is described in conjunction with an embodiment, it should be within the knowledge of a person skilled in the relevant art to implement such feature, structure, or characteristic in conjunction with other embodiments (whether or not explicitly described).

[0156] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0157] In the present invention, "at least one" means one or more, and "plurality" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b, or c can mean: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple.

[0158] It should be understood that in various embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0159] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0160] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0161] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0162] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.

[0163] The present invention covers any substitution, modification, equivalent method and scheme made on the essence and scope of the present invention. In order to make the public have a thorough understanding of the present invention, specific details are described in detail in the following preferred embodiments of the present invention, but those skilled in the art can fully understand the present invention without the description of these details. In addition, in order to avoid unnecessary confusion about the essence of the present invention, well-known methods, processes, procedures, components and circuits are not described in detail.

[0164] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A shapelet-based spatiotemporal data co-evolution pattern mining method, characterized in that: The following steps are involved: S1, collect multivariate spatiotemporal data and construct a multivariate spatiotemporal dataset; S2, extracting a spatiotemporal shapelet with spatial location information from the multivariate spatiotemporal dataset; S3, clustering and pattern recognition of the spatiotemporal shapelet by clustering algorithm; S4, based on the temporal relationship between pairwise spatiotemporal shapelet patterns, the spatiotemporal coevolution patterns are divided into three categories; S5. According to the set vertical support threshold and horizontal support threshold, the three types of spatiotemporal coevolution patterns are mined using the pattern growth method to obtain the final spatiotemporal coevolution pattern set.

2. The method according to claim 1, characterized in that In step S1, the multivariate spatiotemporal data is a spatiotemporal data sequence with spatial information collected by sensors, and the multivariate spatiotemporal data set includes an environmental monitoring data set.

3. The method according to claim 1, characterized in that The step S2 specifically includes: Extract shapelets with spatial location information as spatiotemporal shapelets and construct a spatiotemporal shapelet set S, which satisfies , the position information of the shapelet in set S is stored in set D; m is the variable number, M is the total number of variables, k is the shapelet number, and K is the total number of shapelets.

4. The method according to claim 3, characterized in that The step S3 specifically includes: The same spatiotemporal shapelets are clustered by clustering algorithm to obtain the clustering result R; clustering result , where R c It is the c-type spatiotemporal shapelet pattern, which contains multiple spatiotemporal shapelets of the same type. , ,in Represents R c The shapelet structure in represents the spatiotemporal shapelet structure; k∈[1,K] is the shapelet number, c∈[1,C], C is the total number of clustering patterns, i.e., the total number of spatiotemporal shapelet pattern types; m∈[1,M] is the variable number, t s ,t e ∈[1,h] represents the start and end time of shapelet, D n ,n∈[1,N] is The corresponding spatial position is a longitude and latitude coordinate; R c The shapelet structure of each variable in takes the average value as the spatiotemporal shapelet pattern of this class, omitting its start and end time, and using p c,m It is represented by , where c represents the clustering pattern number and m represents the variable number.

5. The method according to claim 4, characterized in that In step S4, there are three temporal relationships between spatiotemporal shapelet patterns, namely, synchronous, delayed, and ordered. The three temporal relationships can occur between any two spatiotemporal shapelet patterns of different types. c,m and p c’,m’ and exist separately Two shapelet structures, r represents three different temporal relations, r∈{Syn,Del,Ord}; Using triples Represents a spatiotemporal co-evolutionary pattern consisting of a shapelet structure of two different types of spatiotemporal shapelet patterns; The step S4 specifically includes: Initialize three types of spatiotemporal co-evolution pattern subsets P Syn , P Del , P Ord ; The spatiotemporal shapelet patterns in the spatiotemporal shapelet set S are combined into different types of spatiotemporal co-evolution patterns, and their temporal relations r are queried; According to different temporal relations r, the spatiotemporal shapelet patterns are divided into corresponding spatiotemporal co-evolution pattern subsets; The three types of spatiotemporal coevolution pattern subsets are combined to obtain the initial spatiotemporal coevolution pattern set.

6. The method according to claim 5, characterized in that The step S5 specifically includes: Initialize the cardinality of the spatiotemporal coevolution pattern; Scanning the spatiotemporal coevolution patterns in the initial spatiotemporal coevolution pattern set, and calculating the vertical support and horizontal support of the current spatiotemporal coevolution pattern; Performing pruning operations by using set vertical support thresholds and horizontal support thresholds to filter the spatiotemporal coevolution patterns in the initial spatiotemporal coevolution pattern set, and adding new patterns to achieve pattern growth; Update the cardinality value and return the final set of spatiotemporal coevolution patterns.

7. The method according to claim 6, characterized in that The vertical support refers to: the number of spatiotemporal shapelet patterns in which the spatiotemporal coevolution pattern P appears in the spatiotemporal shapelet set divided by the total number of spatiotemporal shapelet patterns in the spatiotemporal shapelet set, where the spatiotemporal coevolution pattern P is one of the three types of spatiotemporal coevolution patterns; The horizontal support refers to: assuming that the spatiotemporal coevolution pattern P is in the time series T n appears, then the spatiotemporal coevolution pattern P in the time series T n The horizontal support on the time series T n The number of occurrences of .

8. A shapelet-based spatiotemporal data co-evolution pattern mining system, the system being used to implement the method according to any one of claims 1 to 7, characterized in that: The system comprises: Data collection module, used to collect multivariate spatiotemporal data and construct multivariate spatiotemporal data sets; A shapelet extraction module, for extracting a spatiotemporal shapelet with spatial location information from the multivariate spatiotemporal dataset; A shapelet clustering module, used for clustering and pattern recognition of the spatiotemporal shapelet by using a clustering algorithm; The pattern classification module is used to classify the spatiotemporal coevolution patterns into three categories based on the temporal relationship between pairwise spatiotemporal shapelet patterns; The pattern mining module is used to mine three types of spatiotemporal coevolution patterns using the pattern growth method according to the set vertical support threshold and horizontal support threshold, and obtain the final spatiotemporal coevolution pattern set.

9. The system according to claim 8, characterized in that The data acquisition module includes a sensor component, a serial port server, a management workstation, an interface server, an information center storage server, and a platform core switch; Among them, the sensor component is connected to the management workstation through the serial port server, the management workstation is connected to the interface server, the interface server is connected to the information center storage server through a unidirectional isolation gateway, the information center storage server is connected to the platform core switch through optical fiber, and the platform core switch is connected to multiple cloud computing nodes and web servers.

10. The system according to claim 8, characterized in that The system also includes a monitoring APP, which includes: a client, a server and a system management background; The client is used for user registration and login, online query, modification and logout; the server is used for registration and login verification as well as data transmission, addition, modification and deletion; the system management background is used to manage the database.

Images