Data type identification model training method, data processing method, and electronic device

Abstract

The application discloses a data type identification model training method, a data processing method and an electronic device. The scheme provided by the application comprises the following steps: acquiring various Internet of Things data collected by different types of Internet of Things devices, wherein the Internet of Things data comprises first Internet of Things data with a data length less than a preset length threshold and second Internet of Things data with a data length not less than the preset length threshold; adding corresponding feature data to the first Internet of Things data to obtain third Internet of Things data with a data length not less than the preset length threshold, wherein the feature data is used to indicate the source information of the corresponding first Internet of Things data; constructing a sample data set based on a target Internet of Things data set, wherein the target Internet of Things data set comprises the second Internet of Things data and the third Internet of Things data; and training a data type identification model by taking sample data in the sample data set as input and taking the type of the Internet of Things data corresponding to the sample data as a label.

Description

Technical Field

[0001] This application relates to the field of Internet of Things (IoT) technology, and in particular to a training method for a data type identification model, a data processing method, an electronic device, and a computer-readable storage medium. Background Technology

[0002] There are many types of IoT devices, such as IoT sensors. Different types of IoT sensors collect IoT data with different content and formats, and the data is mainly transmitted and processed via wireless communication. Current methods for collecting IoT data from IoT sensors involve using corresponding communication or collection devices for different types of IoT sensors to classify the data.

[0003] In existing technologies, IoT data collection devices are highly specialized and lack versatility, making it impossible to collect and identify diverse IoT data from various business scenarios. This necessitates significant investment in infrastructure and resources. Furthermore, with the continuous development of large-scale, multi-business IoT sensors, the current methods for identifying and classifying IoT data are becoming increasingly difficult to implement. Summary of the Invention

[0004] The purpose of this application is to provide a training method, a data processing method, an electronic device, and a computer-readable storage medium for a data type identification model, in order to solve the problem of high cost in collecting different types of IoT data.

[0005] To solve the above-mentioned technical problems, this specification is implemented as follows:

[0006] Firstly, a training method for a data type recognition model is provided, including:

[0007] Acquire various IoT data collected by different types of IoT devices, wherein the IoT data includes first IoT data with a data length less than a preset length threshold and second IoT data with a data length not less than the preset length threshold;

[0008] Add corresponding feature data to the first IoT data to obtain third IoT data with a data length not less than the preset length threshold. The feature data is used to indicate the source information of the corresponding first IoT data.

[0009] A sample dataset is constructed based on a target IoT dataset, wherein the target IoT dataset includes the second IoT data and the third IoT data.

[0010] Using sample data from the aforementioned sample dataset as input and the type of IoT data corresponding to the sample data as a label, a data type recognition model is trained.

[0011] Optionally, the IoT data is hexadecimal code stream data. Corresponding feature data is added to each first IoT data set to obtain third IoT data with a data length not less than the preset length threshold, including:

[0012] Determine the data length of the hexadecimal code stream corresponding to each IoT data;

[0013] Based on the data length, determine the first IoT data in each IoT data set;

[0014] Based on the communication device information corresponding to each first IoT data, the characteristic data corresponding to each first IoT data is determined. The communication device information includes the location information of the communication device that collects the IoT data and the information of the IoT device corresponding to the communication device.

[0015] The feature data of a preset length is added to the corresponding first IoT data to obtain the third IoT data.

[0016] Optionally, a sample dataset is constructed based on the target IoT dataset, including:

[0017] The hexadecimal code stream data corresponding to the second IoT data and the third IoT data included in the target IoT data set are respectively converted into binary code stream data according to the ASCII code table;

[0018] The set of binary code stream data corresponding to the target IoT data set is converted into a two-dimensional matrix;

[0019] The elements of the two-dimensional matrix are normalized, and the normalized two-dimensional matrix is ​​then dimensionality-reduced to obtain the sample dataset.

[0020] Optionally, after normalizing the elements of the two-dimensional matrix and reducing the dimensionality of the normalized two-dimensional matrix, the method further includes:

[0021] In the one-dimensional matrix obtained by the dimensionality reduction process, determine each first element with added feature data and each second element without added feature data;

[0022] The binary code stream data corresponding to each first element is compressed and encoded to obtain the binary code stream data with reduced length.

[0023] Based on the reduced-length binary bitstream data corresponding to each first element and the binary bitstream data corresponding to each second element, the sample dataset is obtained, and one sample data in the sample dataset corresponds to one element in the one-dimensional matrix.

[0024] Optionally, using sample data from the sample dataset as input and the type of IoT data corresponding to the sample data as a label, a data type recognition model is trained, including:

[0025] An initial one-dimensional convolutional neural network (CNN) model is established, which includes an input layer, a convolutional layer, a pooling layer, a global pooling layer, a fully connected layer, and an output layer.

[0026] Using the array data as samples and the type of IoT data corresponding to each array data as labels, the parameters of the initial one-dimensional CNN model are trained until the parameters converge, thus obtaining the data type recognition model.

[0027] Secondly, an IoT data processing method is provided, including:

[0028] Acquire IoT data collected by the target IoT device;

[0029] Determine the data length of the IoT data;

[0030] If the data length is less than a preset length threshold, feature data indicating the source information of the IoT data is added to the IoT data.

[0031] The IoT data with added feature data is input into a preset data type recognition model to identify the data type of the IoT data;

[0032] The IoT data is processed based on the data type described above.

[0033] Optionally, if the data length is less than a preset length threshold, feature data indicating the source information of the IoT data is added to the IoT data, including:

[0034] Obtain communication device information corresponding to the IoT data, wherein the communication device information includes the location information of the communication device that collects the IoT data and the information of the IoT device corresponding to the communication device;

[0035] Based on the communication device information, determine the feature data corresponding to the IoT data;

[0036] Add the feature data to the IoT data.

[0037] Optionally, processing the IoT data based on the data type includes:

[0038] Based on the preset mapping relationship between data type and priority, the priority corresponding to the IoT data of the specified data type is determined;

[0039] IoT data of the specified data type is transmitted according to the specified priority.

[0040] Thirdly, an electronic device is provided, including a memory and a processor electrically connected to the memory, the memory storing a computer program executable by the processor, the computer program, when executed by the processor, implementing the steps of the method as described in the first or second aspect.

[0041] Fourthly, a computer-readable storage medium is provided that stores a computer program thereon, which, when executed by a processor, implements the steps of the method described in the first or second aspect.

[0042] In this embodiment, IoT data collected by a target IoT device is acquired; the data length of the IoT data is determined; if the data length is less than a preset length threshold, feature data indicating the source information of the IoT data is added to the IoT data; the IoT data with added feature data is input into a preset data type recognition model to identify the data type of the IoT data; and the IoT data is processed based on the data type. Thus, for different types of IoT data collected by the same information collection device, the data type can be identified through the preset data type recognition model. This supports the unified collection and classification of different types of IoT data, reduces the cost of IoT data collection and processing, improves the universality of IoT data processing, and has a high accuracy rate in automatically identifying and classifying different types of IoT data. Attached Figure Description

[0043] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0044] Figure 1 This is a flowchart illustrating the training method of the data type recognition model according to an embodiment of this application.

[0045] Figure 2 This is a schematic diagram illustrating an application scenario of the communication device according to an embodiment of this application.

[0046] Figure 3 This is a schematic diagram of the structure of a one-dimensional CNN model according to an embodiment of this application.

[0047] Figure 4 This is a flowchart illustrating the IoT data processing method according to an embodiment of this application.

[0048] Figure 5 This is a structural block diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0049] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application. The drawing numbers in this application are only used to distinguish the various steps in the solution and are not used to limit the execution order of the various steps. The specific execution order is subject to the description in the specification.

[0050] To address the problems existing in the prior art, embodiments of this application provide a training method for a data type recognition model, such as... Figure 1 As shown, it includes the following steps:

[0051] Step 102: Obtain IoT data collected by different types of IoT devices. The IoT data includes first IoT data with a data length less than a preset length threshold and second IoT data with a data length not less than the preset length threshold.

[0052] In this step, the IoT device includes different types of IoT sensors, such as temperature sensors, liquid level sensors, gas sensors, humidity sensors, etc., to collect corresponding IoT data.

[0053] The collected IoT data is typically hexadecimal raw code stream data, i.e., the payload in the application layer message, also known as the actual data or data body. The length and corresponding indication characteristics of the IoT data collected by different types of IoT sensors may vary. In the embodiments of this application, IoT data includes IoT data with a data length less than a preset length threshold and IoT data with a data length not less than the preset length threshold.

[0054] IoT data shorter than a preset length threshold contains too little data and therefore lacks clear or sufficient feature dimensions, making it difficult to extract features during subsequent model training. IoT data longer than the preset length threshold contains more feature dimensions and can comprehensively reflect the relevant information of the corresponding IoT device.

[0055] Step 104: Add corresponding feature data to the first IoT data to obtain third IoT data with a data length not less than the preset length threshold. The feature data is used to indicate the source information of the corresponding first IoT data.

[0056] As mentioned above, it is difficult to extract features from IoT data that is too short, thus failing to obtain a high-accuracy data type identification model when used for training. Therefore, in one embodiment, relevant features can be added to short IoT data.

[0057] Optionally, the IoT data is hexadecimal code stream data. Adding corresponding feature data to each first IoT data to obtain third IoT data with a data length not less than the preset length threshold includes: determining the data length of the hexadecimal code stream data corresponding to each IoT data; determining the feature data corresponding to each first IoT data based on the communication device information corresponding to each first IoT data, wherein the communication device information includes the location information of the communication device collecting the IoT data and the information of the IoT device corresponding to the communication device; and adding the feature data of the preset length to the corresponding first IoT data to obtain the third IoT data.

[0058] For IoT data collected by IoT devices in hexadecimal raw bitstream format, the first step is to determine the data length of the hexadecimal raw bitstream data corresponding to each IoT data point. If the data length is less than a preset length threshold, it is determined that feature data needs to be added to that IoT data point, and thus it is designated as the first IoT data point.

[0059] The added feature data is used to indicate the source information of the target first IoT data, that is, the relevant information of the communication device that collects the target first IoT data. The communication device information includes the location information of the communication device that collects the target IoT data and the information of the corresponding IoT device acquired by the communication device. For example, if the communication device is a drone responsible for collecting IoT data information collected by various IoT devices, then the location of the drone and the information of the corresponding acquired IoT devices, including the appearance pictures or videos of the IoT devices taken by the drone, can be used to extract the type of the IoT device. The relevant communication device information can be added as relevant feature data of the corresponding collected target IoT devices to the shorter IoT data. By adding feature data, the information content of the IoT data collected by the target IoT device can be enriched, and the feature dimensions of the IoT data can be increased, thereby obtaining third IoT data with more dimensional features.

[0060] like Figure 2 As shown, communication device 1 can collect IoT data from different types of IoT sensors, including sensor 1, sensor 2, sensor 3, sensor 4, sensor 5 and sensor 6. One communication device can collect different IoT data for multiple services, which can reduce infrastructure investment costs, reduce resource consumption, and has good versatility.

[0061] Step 106: Construct a sample dataset based on the target IoT data set, which includes the second IoT data and the third IoT data.

[0062] The IoT data set is obtained by combining long-term IoT data collected by IoT devices with IoT data after adding feature data, and sample data is constructed based on each IoT data in the set.

[0063] Optionally, constructing a sample dataset based on the target IoT data set includes: converting the hexadecimal code stream data corresponding to the second IoT data and the third IoT data included in the target IoT data set into binary code stream data according to the ASCII code table; converting the set of converted binary code stream data corresponding to the target IoT data set into a two-dimensional matrix; normalizing the elements of the two-dimensional matrix and reducing the dimensionality of the normalized two-dimensional matrix to obtain the sample dataset.

[0064] For the second and third IoT data included in the target IoT dataset, the original hexadecimal bitstream data corresponding to each IoT data is translated into corresponding binary integer bitstream data according to the ASCII code table. After all the IoT data included in the target IoT dataset has been converted into binary bitstream data, a set of multiple binary bitstream data is obtained. Then, the binary bitstream data in this set is converted into a two-dimensional matrix. For example, each binary bitstream data in the set is sequentially input into a buffer, and then adjacent binary codes are paired to form two-dimensional data pairs and output from the buffer. Thus, the binary bitstream data in the set can be presented in the form of a two-dimensional matrix.

[0065] Next, the elements in the two-dimensional matrix are normalized, and the normalized two-dimensional matrix is ​​then reduced in dimensionality to obtain a one-dimensional matrix. Each element in the one-dimensional matrix contains one piece of one-dimensional binary code stream data, corresponding to a single IoT data point. Each element in the one-dimensional matrix corresponds to one sample data point, and the collection of one-dimensional binary code stream data points corresponding to multiple elements in the one-dimensional matrix constitutes the corresponding sample dataset.

[0066] Optionally, after normalizing the elements of the two-dimensional matrix and reducing the dimensionality of the normalized two-dimensional matrix, the method further includes: determining each first element with added feature data and each second element without added feature data in the one-dimensional matrix obtained by the dimensionality reduction process; compressing and encoding the binary code stream data corresponding to each first element to obtain a reduced-length binary code stream data; and obtaining the sample dataset based on the reduced-length binary code stream data corresponding to each first element and the binary code stream data corresponding to each second element, wherein one sample data in the sample dataset corresponds to one element in the one-dimensional matrix.

[0067] In this embodiment, after obtaining the one-dimensional matrix after dimensionality reduction, the elements in the one-dimensional matrix are first classified. Each element corresponds to a set of binary code stream data. The binary code stream data carrying the binary code corresponding to the feature data added in step 104 can be identified by the value of each set of binary code stream data, that is, the first elements with added feature data and the second elements without added feature data are identified.

[0068] The number of bits in the binary representation of added target feature data can be further compressed to a shorter number of bits. For example, if the added binary bit indicating the location of the communication device corresponds to binary code 010, it can be further compressed into a single binary code 1 through classification encoding. Simplifying the number of bits in the binary code of added feature data allows for faster indexing of the corresponding original IoT data from the dataset storing IoT data.

[0069] Thus, based on the set of binary code stream data corresponding to the first and second elements, the sample dataset is determined, and one sample data in the sample dataset corresponds to one first or second element.

[0070] Step 108: Using the sample data in the sample dataset as input and the type of IoT data corresponding to the sample data as the label, train the data type recognition model.

[0071] The sample data corresponds to IoT data. The type of IoT sensor that collects IoT data can be determined in advance, and the corresponding IoT data type can also be known in advance, which can be used as a tag.

[0072] Optionally, using sample data from the sample dataset as input and the type of IoT data corresponding to the sample data as a label, a data type recognition model is trained, including: establishing an initial one-dimensional convolutional neural network (CNN) model, wherein the initial one-dimensional CNN model includes an input layer, a convolutional layer, a pooling layer, a global pooling layer, a fully connected layer, and an output layer; using the array data as samples and the type of IoT data corresponding to each array data as a label, the parameters of the initial one-dimensional CNN model are trained until the parameters converge, thereby obtaining the data type recognition model.

[0073] In one embodiment, such as Figure 3 As shown, the convolutional neural network corresponding to the one-dimensional CNN model includes an input layer 10, two convolutional layers 20, two pooling layers 30, a global pooling layer 40, a fully connected layer 50, and an output layer 60. The convolutional layers 20 have, for example, 32 kernels, a kernel size of, for example, 25*25, and a stride of, for example, 1. The pooling layer 30 can use max pooling, with a size of, for example, 3 and a stride of, for example, 3. The fully connected layer 50 has the same number of neurons as the input data, and uses a predefined regularization algorithm to prevent overfitting, and uses a softmax logistic regression function to output the classification result.

[0074] In this embodiment, IoT data collected by different types of IoT devices are acquired. The IoT data includes first IoT data with a length less than a preset length threshold and second IoT data with a length not less than the preset length threshold. Corresponding feature data is added to the first IoT data to obtain third IoT data with a length not less than the preset length threshold. The feature data indicates the source information of the corresponding first IoT data. A sample dataset is constructed based on a target IoT data set, which includes the second and third IoT data. Using sample data from the sample dataset as input and the type of IoT data corresponding to the sample data as a label, a data type identification model is trained. The resulting data type identification model can support unified collection and classification of different types of IoT data, reducing the cost of IoT data collection and processing, improving the universality of IoT data processing, and achieving high accuracy in automatically identifying and classifying different types of IoT data.

[0075] In another embodiment of this application, an Internet of Things (IoT) data processing method is also proposed, such as... Figure 4 As shown, it includes the following steps:

[0076] Step 202: Obtain IoT data collected by the target IoT device;

[0077] Step 204: Determine the data length of the IoT data;

[0078] Step 206: If the data length is less than a preset length threshold, add feature data to the IoT data to indicate the source information of the IoT data;

[0079] Step 208: Input the IoT data with added feature data into a preset data type recognition model to identify the data type of the IoT data;

[0080] Step 210: Process the IoT data based on the data type.

[0081] The acquisition of IoT data collected by the target IoT device can be data of a specific type collected by a particular IoT device. After acquiring the collected IoT data, the length of the data is used to determine whether it is long or short. If the data length is less than a preset length threshold, the IoT data is determined to be short. Then, feature data indicating the source information of the IoT data is added to this IoT data.

[0082] Based on the solution provided in the above embodiments, optionally, in step 206, when the data length is less than a preset length threshold, feature data for indicating the source information of the IoT data is added to the IoT data, including: obtaining communication device information corresponding to the IoT data, wherein the communication device information includes the location information of the communication device that collects the IoT data and the information of the IoT device obtained by the communication device; determining the feature data corresponding to the IoT data based on the communication device information; and adding the feature data to the IoT data.

[0083] In step 208, the shorter IoT data is preprocessed and then input into a preset data type identification model. This preset data type identification model can be based on the above... Figures 1 to 3 The data type recognition model obtained by training the data type recognition model in this embodiment can thus identify the data type of the corresponding Internet of Things (IoT) data.

[0084] Of course, if the length of the target IoT data collected is not less than the preset length threshold, the target IoT data can be directly input into the preset data type recognition model to identify the data type of the IoT data.

[0085] Based on the solution provided in the above embodiments, optionally, in step 210 above, processing the IoT data based on the data type includes: determining the priority corresponding to the IoT data of the data type based on a preset mapping relationship between data type and priority; and transmitting the IoT data of the data type according to the priority.

[0086] The preset mapping relationship between data type and priority is established in advance according to specific rules. For example, if the priority of data type A is higher than that of data type B, then the collected IoT data of data type A can be transmitted with priority.

[0087] After identifying the data type of IoT data collected by the target IoT device through a preset data type identification model, the priority of that data type is determined. If it is the highest priority, the transmission of other types of IoT data that need to be transmitted can be postponed until the highest priority IoT data is transmitted.

[0088] In this embodiment, IoT data collected by a target IoT device is acquired; the data length of the IoT data is determined; if the data length is less than a preset length threshold, feature data indicating the source information of the IoT data is added to the IoT data; the IoT data with added feature data is input into a preset data type recognition model to identify the data type of the IoT data; and the IoT data is processed based on the data type. Thus, for different types of IoT data collected by the same information collection device, the data type can be identified through the preset data type recognition model. This supports the unified collection and classification of different types of IoT data, reduces the cost of IoT data collection and processing, improves the universality of IoT data processing, and has a high accuracy rate in automatically identifying and classifying different types of IoT data.

[0089] Optionally, embodiments of this application also provide an electronic device. Figure 5 This is a structural block diagram of an electronic device according to an embodiment of this application.

[0090] like Figure 5 As shown, the electronic device 2000 includes a memory 2200 and a processor 2400 electrically connected to the memory 2200. The memory 2200 stores a computer program that can be run on the processor 2400. When the computer program is executed by the processor, it implements the various processes of any of the above-described data type recognition model training method embodiments or any of the Internet of Things data processing method embodiments, and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0091] This application also provides a computer-readable storage medium storing a computer program. When executed by a processor, this computer program implements the various processes of any of the above-described data type recognition model training method embodiments or any of the IoT data processing method embodiments, and achieves the same technical effect. To avoid repetition, it will not be described again here. The computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, etc.

[0092] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0093] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0094] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. A training method for a data type recognition model, characterized in that, include: Acquire various IoT data collected by different types of IoT devices, wherein the IoT data includes first IoT data with a data length less than a preset length threshold and second IoT data with a data length not less than the preset length threshold; Add corresponding feature data to the first IoT data to obtain third IoT data with a data length not less than the preset length threshold. The feature data is used to indicate the source information of the corresponding first IoT data. Specifically, it includes: determining the feature data corresponding to each first IoT data according to the communication device information corresponding to each first IoT data. The communication device information includes the location information of the communication device that collects the IoT data and the information of the IoT device that the communication device acquires. Add the feature data of the preset length to the corresponding first IoT data to obtain the third IoT data. A sample dataset is constructed based on a target IoT dataset, wherein the target IoT dataset includes the second IoT data and the third IoT data. Using sample data from the aforementioned sample dataset as input and the type of IoT data corresponding to the sample data as a label, a data type recognition model is trained.

2. The method of claim 1, wherein, The IoT data is hexadecimal code stream data. Corresponding feature data is added to each first IoT data set to obtain third IoT data with a data length not less than the preset length threshold. The data also includes: Determine the data length of the hexadecimal code stream corresponding to each IoT data; Based on the data length, determine the first IoT data in each IoT data set.

3. The method of claim 2, wherein, A sample dataset was constructed based on the target IoT data set, including: The hexadecimal code stream data corresponding to the second IoT data and the third IoT data included in the target IoT data set are respectively converted into binary code stream data according to the ASCII code table; The set of binary code stream data corresponding to the target IoT data set is converted into a two-dimensional matrix; The elements of the two-dimensional matrix are normalized, and the normalized two-dimensional matrix is ​​then dimensionality-reduced to obtain the sample dataset.

4. The method of claim 3, wherein, After normalizing the elements of the two-dimensional matrix and reducing the dimensionality of the normalized two-dimensional matrix, the process further includes: In the one-dimensional matrix obtained by the dimensionality reduction process, determine each first element with added feature data and each second element without added feature data; The binary code stream data corresponding to each first element is compressed and encoded to obtain the binary code stream data with reduced length. Based on the reduced-length binary bitstream data corresponding to each first element and the binary bitstream data corresponding to each second element, the sample dataset is obtained, and one sample data in the sample dataset corresponds to one element in the one-dimensional matrix.

5. The method of claim 3 or 4, wherein, Using sample data from the aforementioned sample dataset as input and the type of IoT data corresponding to the sample data as a label, a data type recognition model is trained, including: An initial one-dimensional convolutional neural network (CNN) model is established, which includes an input layer, a convolutional layer, a pooling layer, a global pooling layer, a fully connected layer, and an output layer. Using array data as samples and the type of IoT data corresponding to each array data as labels, the parameters of the initial one-dimensional CNN model are trained until the parameters converge, thus obtaining the data type recognition model.

6. An Internet of Things data processing method, characterized by, include: Acquire IoT data collected by the target IoT device; Determine the data length of the IoT data; When the data length is less than a preset length threshold, feature data indicating the source information of the IoT data is added to the IoT data. Specifically, this includes: obtaining communication device information corresponding to the IoT data, wherein the communication device information includes the location information of the communication device that collected the IoT data and the information of the IoT device obtained by the communication device; determining the feature data corresponding to the IoT data based on the communication device information; and adding the feature data to the IoT data. The IoT data with added feature data is input into a preset data type recognition model to identify the data type of the IoT data; The IoT data is processed based on the data type described above.

7. The method as described in claim 6, characterized in that, Processing the IoT data based on the data type includes: Based on the preset mapping relationship between data type and priority, the priority corresponding to the IoT data of the specified data type is determined; IoT data of the specified data type is transmitted according to the specified priority.

8. An electronic device, characterized in that, include: A memory and a processor electrically connected to the memory, the memory storing a computer program executable by the processor, the computer program, when executed by the processor, implementing the steps of the method as claimed in any one of claims 1 to 5 or implementing the steps of the method as claimed in any one of claims 6 to 7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the method as claimed in any one of claims 1 to 5 or the steps of the method as claimed in any one of claims 6 to 7.

Images