Intelligent decision-making method and device for industrial chain data
Through the variational modal decomposition algorithm, the feature decomposition and merging of industrial chain data is solved, and the prediction accuracy of small sample data in intelligent decision-making is established. A reinforcement learning and deep learning combination model is established, which improves the decision-making accuracy of small sample data. and model applicability.
Patent Information
- Application Number
- CN202411344436.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-25
- Publication Date
- 2024-12-24
AI Technical Summary
Existing intelligent decision-making methods have poor prediction accuracy when processing small sample data, and cannot effectively use small sample data for deep learning models, resulting in inaccurate decision results.
The variational modal decomposition algorithm is used to decompose and merge the characteristic information of the industrial chain data, generate a collection of target feature information, and train the deep learning model based on this to establish a small sample industrial chain data intelligent decision-making model.
By decomposing and merging feature information, a reinforcement learning and deep learning combination model is generated, which improves the prediction accuracy of small sample data and the applicability of the model, providing more accurate support for data decisions in small sample industry chain.
Smart Images
Figure CN119180524A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of artificial intelligence technology, and in particular to an intelligent decision-making method and device for industrial chain data. Background Art
[0002] Intelligent decision-making is an important research direction in the field of artificial intelligence, which aims to help people make better decisions in complex environments. The core of intelligent decision-making is to automatically learn rules and patterns by analyzing large amounts of data and information, thereby providing support and suggestions for decision-making.
[0003] At present, a large number of intelligent decision-making studies have attached great importance to mining the internal logical relationship of big data, but have ignored small sample data with clear value demands. In order to achieve accurate prediction and intelligent decision-making, large sample data is needed to capture rules and small sample data to match scenarios. However, the lack of data volume cannot provide data space for deep learning models to learn, resulting in unsatisfactory prediction accuracy. Therefore, it is urgent to seek an intelligent decision-making model for small sample data so as to enable data intelligent decision-making. Summary of the invention
[0004] The main purpose of the present invention is to provide an intelligent decision-making method and device for industrial chain data to solve the technical problems existing in related technologies.
[0005] In order to achieve the above-mentioned purpose, according to the first aspect of the present invention, there is provided an intelligent decision-making method for industrial chain data, including obtaining industrial chain data to be decided; inputting the industrial chain data to be decided into a trained intelligent decision-making model for industrial chain data, and outputting a decision result; wherein, when training the intelligent decision-making model for industrial chain data, a small sample amount of industrial chain data is collected and normalized, and the normalized industrial chain data is used as a training set; specified feature information in the training set is extracted to obtain a feature information set; the feature information set is decomposed based on a variational mode decomposition algorithm, and a target feature information set is obtained after merging the decomposed feature information; a deep learning model is trained based on the target feature information set to obtain a small sample amount of industrial chain data intelligent decision-making model.
[0006] Optionally, decomposing the feature information set based on the variational modal decomposition algorithm includes: setting the number of modes K, performing Hilbert transform on the features in the feature information set, and obtaining analytical features of K modal components; adjusting the analytical features of each modal component to a unified vector space; determining the central features of each analytical feature and the distribution of each analytical feature in the text in the same vector space, and constructing a variational objective function based on this; and solving the objective function based on a preset quadratic penalty factor and a Lagrange multiplication operator.
[0007] Optionally, each analytical feature and the corresponding central feature are continuously updated using an alternating direction multiplier method before solving.
[0008] Optionally, the target feature information set includes category distribution, node distribution, time distribution, space distribution, scenario distribution, event distribution, as well as enterprise distribution, regulatory enforcement distribution, talent distribution, investment and financing distribution, merger and reorganization distribution, technology distribution, patent distribution, policy distribution, product distribution and other feature information in the industrial chain data.
[0009] According to a second aspect of the present invention, there is provided an intelligent decision-making device for industrial chain data, comprising a data acquisition unit, configured to acquire industrial chain data to be decided; a decision-making unit, configured to input the industrial chain data to be decided into a trained intelligent decision-making model for industrial chain data, and output a decision result; wherein, when training the intelligent decision-making model for industrial chain data, a small sample amount of industrial chain data is collected and normalized, and the normalized industrial chain data is used as a training set; specified feature information in the training set is extracted to obtain a feature information set; the feature information set is decomposed based on a variational mode decomposition algorithm, and a target feature information set is obtained after merging the decomposed feature information; a deep learning model is trained based on the target feature information set to obtain an intelligent decision-making model for small sample industrial chain data.
[0010] Optionally, decomposing the feature information set based on the variational modal decomposition algorithm includes: setting the number of modes K, performing Hilbert transform on the features in the feature information set, and obtaining analytical features of K modal components; adjusting the analytical features of each modal component to a unified vector space; determining the central features of each analytical feature and the distribution of each analytical feature in the text in the same vector space, and constructing a variational objective function based on this; and solving the objective function based on a preset quadratic penalty factor and a Lagrange multiplication operator.
[0011] Optionally, each analytical feature and the corresponding central feature are continuously updated using an alternating direction multiplier method before solving.
[0012] According to a third aspect of the present invention, there is provided a computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause the computer to execute the method described in any one of the implementations of the first aspect.
[0013] According to a fourth aspect of the present invention, there is provided an electronic device comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor executes the method described in any one implementation of the first aspect.
[0014] According to a fifth aspect of the present invention, a computer program product is provided, which, when executed by a processor, implements the method described in any one of the implementation modes of the first aspect.
[0015] This embodiment is directed to an intelligent decision-making method and device for industrial chain data, wherein the method includes obtaining industrial chain data to be decided; inputting the industrial chain data to be decided into a trained industrial chain data intelligent decision-making model, and outputting a decision result; wherein, when training the industrial chain data intelligent decision-making model, a small sample of industrial chain data is collected and normalized, and the normalized industrial chain data is used as a training set; the specified feature information in the training set is extracted to obtain a feature information set; the feature information set is decomposed based on a variational mode decomposition algorithm, and a target feature information set is obtained after merging the decomposed feature information; a deep learning model is trained based on the target feature information set to obtain a small sample industrial chain data intelligent decision model. By decomposing a small sample industrial chain data set, multiple feature information in the small sample industrial chain data is obtained, thereby establishing a small sample industrial chain data reinforcement learning and deep learning combined model, and generating a small sample industrial chain data intelligent decision model, thereby enabling industrial data intelligent decision-making.
[0016] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0017] Figure 1 is a flow chart of an intelligent decision-making method for industrial chain data according to an embodiment of the present invention;
[0018] Figure 2 is a schematic diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION
[0019] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.
[0020] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so as to describe the embodiments of the present invention described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0021] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.
[0022] According to an embodiment of the present invention, an intelligent decision-making method for industrial chain data is provided. Figure 1 As shown, it includes the following steps 101 to 102:
[0023] Step 101: Obtain data of the industry chain to be decided.
[0024] Step 102: Input the industrial chain data to be decided into the trained industrial chain data intelligent decision-making model, and output the decision result, wherein, when training the industrial chain data intelligent decision-making model, collect a small sample of industrial chain data and then normalize it, and use the normalized industrial chain data as a training set; extract the specified feature information in the training set to obtain a feature information set; decompose the feature information set based on the variational mode decomposition algorithm, and obtain the target feature information set after merging the decomposed feature information; train the deep learning model based on the target feature information set to obtain a small sample industrial chain data intelligent decision-making model.
[0025] In this embodiment, small sample industrial chain data is collected, normalized, and the original small sample industrial chain data set is divided into a test data set and a training data set; the feature information of the small sample industrial chain data is extracted, and a feature information set of the small sample industrial chain data is established; the variational mode decomposition (VMD) algorithm is used to decompose the above industrial chain feature information data set to obtain the category distribution, node distribution, time distribution, spatial distribution, scene distribution, event distribution, enterprise distribution, supervision and law enforcement distribution, technology distribution, talent distribution, investment and financing distribution, merger and reorganization distribution, patent distribution, policy distribution, product distribution and other feature information of the industry in the small sample industrial chain data; based on the above multiple feature information sets, a small sample industrial chain data reinforcement learning and deep learning combination model is established, and pre-training is performed to generate a small sample industrial chain data intelligent decision model. The test data is input into the above model and the test results are output.
[0026] Small sample industrial chain data is obtained and normalized, and the original small sample industrial chain data set is divided into a test data set and a training data set. The small sample industrial chain data includes case information and related data of several industrial chains and several enterprises in the industrial chain nodes; and the above case information and data are normalized.
[0027] Furthermore, a feature information set of small sample industrial chain data is established, which includes industry classification, upstream and downstream nodes of the industrial chain, enterprises corresponding to upstream and downstream nodes of the industry, and distribution of time, space, environment, etc.
[0028] Furthermore, variational mode decomposition (VMD) is a nonlinear and adaptive process that utilizes an iterative process to achieve high-quality reconstruction of features by decomposing the data into simpler, fixed features that are then merged and filtered.
[0029] The variational mode decomposition (VMD) algorithm is used to decompose the above-mentioned industrial chain characteristic information data set to obtain multiple characteristic information such as industry category distribution, node distribution, time distribution, spatial distribution, scenario distribution, event distribution, as well as enterprise distribution, supervision and law enforcement distribution, technology distribution, talent distribution, investment and financing distribution, merger and acquisition distribution, patent distribution, policy distribution, and product distribution in the small sample industrial chain data.
[0030] Furthermore, prediction is performed based on a hidden Markov model.
[0031] As an optional implementation method of this embodiment, decomposing the feature information set based on the variational modal decomposition algorithm includes: setting the number of modes K, performing Hilbert transform on the features in the feature information set, and obtaining analytical features of K modal components; adjusting the analytical features of each modal component to a unified vector space; determining the central features of each analytical feature and the distribution of each analytical feature in the text in the same vector space, and constructing a variational objective function based on this; solving the objective function based on a preset quadratic penalty factor and a Lagrange multiplication operator.
[0032] In this optional implementation, when constructing the variational problem, the features are first subjected to Hilbert transformation to obtain the analytical features of K modal components and obtain the spectrum. In this embodiment, each modal spectrum is the distribution characteristics of the text in different feature spaces or representation levels. The modes here can be different themes, semantic categories, etc. of the text, and the spectrum is the distribution or intensity of these characteristics in the text. The obtained unilateral spectrum is:
[0033]
[0034] Furthermore, the spectrum of each mode is modulated to the baseband to obtain:
[0035] That is, the text features are transformed into a unified representation space or framework, and necessary standardization is performed for subsequent comparison and analysis. For example, when using different text representation methods (such as bag-of-words model, TF-IDF, word embedding, etc.), these representation methods are converted to the same dimensional space and the feature vectors are normalized or standardized to ensure that they are comparable.
[0036] Furthermore, the square norm of the gradient of the above formula is calculated, and the bandwidth of each modal feature is estimated. The bandwidth of each modality is the distribution breadth or variation range of the text feature in a certain modality. For example, when using word embedding to represent text, the distribution range of different words in the embedding space may represent the semantic variation range of the word, which can be analogized to "bandwidth". The constructed variational problem is as follows
[0037]
[0038] The goal of the above formula is to find the decomposed IMF modal function ik and the center frequency of each component when the sum of the bandwidths of the center frequencies of each modal component is minimized.
[0039] Furthermore, when solving the problem, the quadratic penalty factor α and the Lagrangian multiplication operator λ are introduced to transform the above constraints into an unconstrained variational problem, so the extended Lagrangian expression is obtained as follows:
[0040]
[0041] As an optional implementation of this embodiment, an alternating direction multiplier method is used to continuously update each analytical feature and the corresponding central feature before solving.
[0042] In this optional implementation, before formally performing VMD decomposition, some parameters need to be initialized, including the number of intrinsic mode functions to be decomposed K (the number of features or feature dimensions obtained in text feature decomposition), noise capacity τ (in text processing, noise usually refers to those useless and interfering information for text analysis, such as spelling errors, punctuation marks, stop words, etc. Noise capacity can be understood as the robustness or stability of the text feature decomposition method in processing these noise information. A good text feature decomposition method should be able to tolerate the existence of noise to a certain extent, while extracting features that are useful for text analysis), control error ε (the features obtained by feature decomposition can accurately reflect the content and information of the original text. Controlling error refers to how to ensure that the extracted features can minimize information loss and errors during the feature decomposition process, so that subsequent text analysis tasks can obtain more accurate results), penalty parameter α, etc. Then, the alternating direction multiplier method (ADMM) is used to continuously update each feature and its central frequency, and a set of intrinsic mode functions and their corresponding weights are solved so that their linear combination best approximates the original features. This process is usually solved using the variational method. In each update, a constrained optimization problem needs to be solved to minimize the residual between features. Finally, the saddle point of the unconstrained model is obtained, which is the optimal solution to the original problem. All features can be obtained according to the frequency domain space by the following formula:
[0043] In the above formula, ω represents the frequency. Represent the Fourier transform of S(t) and λ(t), respectively. yes After the Wiener filter, the algorithm re-estimates the centroid frequency based on the centroid of the power spectrum of each component. The specific process is as follows:
[0044]
[0045] renew
[0046] τ is the noise capacity. When the feature contains strong noise, τ=0 can be set to achieve better denoising effect;
[0047] Repeat the update until the stop condition is met
[0048] This model not only has obvious advantages in the accuracy of predicted values, high efficiency of the model and operational efficiency, but is also applicable to different types of small sample industrial chain data. It can provide guiding suggestions and theoretical basis for intelligent decision-making of small sample industrial chain data.
[0049] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.
[0050] According to an embodiment of the present invention, there is also provided an intelligent decision-making device for industrial chain data, comprising a data acquisition unit, configured to acquire industrial chain data to be decided; a decision-making unit, configured to input the industrial chain data to be decided into a trained intelligent decision-making model for industrial chain data, and output a decision result; wherein, when training the intelligent decision-making model for industrial chain data, a small sample amount of industrial chain data is collected and normalized, and the normalized industrial chain data is used as a training set; specified feature information in the training set is extracted to obtain a feature information set; the feature information set is decomposed based on a variational mode decomposition algorithm, and a target feature information set is obtained after merging the decomposed feature information; a deep learning model is trained based on the target feature information set to obtain a small sample amount of industrial chain data intelligent decision-making model.
[0051] As an optional implementation method of this embodiment, decomposing the feature information set based on the variational modal decomposition algorithm includes: setting the number of modes K, performing Hilbert transform on the features in the feature information set, and obtaining analytical features of K modal components; adjusting the analytical features of each modal component to a unified vector space; determining the central features of each analytical feature and the distribution of each analytical feature in the text in the same vector space, and constructing a variational objective function based on this; solving the objective function based on a preset quadratic penalty factor and a Lagrange multiplication operator.
[0052] As an optional implementation of this embodiment, an alternating direction multiplier method is used to continuously update each analytical feature and the corresponding central feature before solving.
[0053] As an optional implementation method of this embodiment, the target feature information set includes category distribution, node distribution, time distribution, spatial distribution, scene distribution, event distribution, as well as enterprise distribution, supervision and law enforcement distribution, talent distribution, investment and financing distribution, merger and reorganization distribution, technology distribution, patent distribution, policy distribution, product distribution and other feature information in the industrial chain data.
[0054] According to an embodiment of the present invention, the present invention also provides an electronic device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can implement the method described in any of the above embodiments when executing.
[0055] According to an embodiment of the present invention, the present invention further provides a readable storage medium, which stores computer instructions, and the computer instructions are used to enable a computer to implement the method described in any of the above embodiments when executed.
[0056] According to an embodiment of the present invention, the present invention further provides a computer program product, which can implement the method described in any of the above embodiments when executed by a processor.
[0057] Figure 2 A schematic block diagram of an example electronic device 300 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices.
[0058] like Figure 2 As shown, the electronic device 300 includes a computing unit 301, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 302 or a computer program loaded from a storage unit 308 into a random access memory (RAM) 303. In the RAM 303, various programs and data required for the operation of the electronic device 300 can also be stored. The computing unit 301, the ROM 302, and the RAM 303 are connected to each other via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.
[0059] A number of components in the electronic device 300 are connected to the I / O interface 305, including: an input unit 306, such as a keyboard, a mouse, etc.; an output unit 307, such as various types of displays, speakers, etc.; a storage unit 308, such as a disk, an optical disk, etc.; and a communication unit 309, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 309 allows the electronic device 300 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.
[0060] The computing unit 301 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 301 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 301 performs the various methods and processes described above, such as the object matching method. For example, in some embodiments, the object matching method may be implemented as a computer software program, which is tangibly included in a machine-readable medium, such as a storage unit 308. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 300 via the ROM 302 and / or the communication unit 309. When the computer program is loaded into the RAM 303 and executed by the computing unit 301, one or more steps of the method described above may be performed.
[0061] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include: being implemented in one or more computer programs, which may be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general programmable processor, which may receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.
[0062] The program code for implementing the method of the present invention can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code can be executed entirely on the machine, partially on the machine, partially on the machine as a stand-alone software package and partially on a remote machine, or entirely on a remote machine or server.
[0063] In the context of the present invention, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
Claims
1. An intelligent decision-making method for industrial chain data, characterized in that: include: Obtain data on the industry chain to be decided; Input the to-be-decided industrial chain data into the trained industrial chain data intelligent decision-making model, and output the decision result; When training the intelligent decision-making model for industrial chain data, a small sample of industrial chain data is collected and normalized, and the normalized industrial chain data is used as a training set; Extracting specified feature information from the training set to obtain a feature information set; Decomposing the feature information set based on a variational mode decomposition algorithm, and obtaining a target feature information set after merging the decomposed feature information; Based on the target feature information set, the deep learning model is trained to obtain a small sample industrial chain data intelligent decision-making model.
2. The intelligent decision-making method for industrial chain data according to claim 1 is characterized in that: Decomposing the feature information set based on the variational mode decomposition algorithm includes: Set the number of modes K, perform Hilbert transform on the features in the feature information set, and obtain the analytical features of K modal components; Adjust the analytical features of each modal component to a unified vector space; Determine the central features of each analytical feature and the distribution of each analytical feature in the text in the same vector space, and construct the variational objective function based on this; The objective function is solved based on a preset quadratic penalty factor and a Lagrange multiplication operator.
3. The intelligent decision-making method for industrial chain data according to claim 2 is characterized in that: Before solving, the alternating direction multiplier method is used to continuously update each analytical feature and the corresponding central feature.
4. The intelligent decision-making method for industrial chain data according to claim 1 is characterized in that: The target feature information set includes category distribution, node distribution, time distribution, space distribution, scene distribution, event distribution, enterprise distribution, regulatory enforcement distribution, talent distribution, investment and financing distribution, merger and reorganization distribution, technology distribution, patent distribution, policy distribution, product distribution and other feature information in the industrial chain data.
5. An intelligent decision-making device for industrial chain data, characterized in that: include: A data acquisition unit, configured to acquire data of the industrial chain to be decided; A decision-making unit is configured to input the to-be-decided industrial chain data into the trained industrial chain data intelligent decision-making model, and output a decision result; When training the intelligent decision-making model for industrial chain data, a small sample of industrial chain data is collected and normalized, and the normalized industrial chain data is used as the training set; Extracting specified feature information from the training set to obtain a feature information set; Decomposing the feature information set based on a variational mode decomposition algorithm, and obtaining a target feature information set after merging the decomposed feature information; Based on the target feature information set, the deep learning model is trained to obtain a small sample industrial chain data intelligent decision-making model.
6. The training device for the industrial chain data intelligent decision model according to claim 5 is characterized in that: Decomposing the feature information set based on the variational mode decomposition algorithm includes: Set the number of modes K, perform Hilbert transform on the features in the feature information set, and obtain the analytical features of K modal components; Adjust the analytical features of each modal component to a unified vector space; Determine the central features of each analytical feature and the distribution of each analytical feature in the text in the same vector space, and construct the variational objective function based on this; The objective function is solved based on a preset quadratic penalty factor and a Lagrange multiplication operator.
7. The training device for the industrial chain data intelligent decision model according to claim 6 is characterized in that: Before solving, the alternating direction multiplier method is used to continuously update each analytical feature and the corresponding central feature.
8. A computer storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the method according to any one of claims 1 to 5.
9. An electronic device, characterized in that: include: at least one processor; And a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor executes any one of claims 1-5.
10. A computer program product, comprising a computer program, which, when executed by a processor, implements the method according to any one of claims 1 to 5.