Terminal model identification method and device based on field of Internet of Things, and server
Through the low-dimensional embedding technology of words and the optimized similarity model, the problem of low accuracy of terminal model recognition in the prior art is solved, and more efficient terminal model recognition is achieved.
Patent Information
- Application Number
- CN202411868218.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-17
- Publication Date
- 2025-04-08
AI Technical Summary
The prior art semantic matching method based on large language models is used to identify mobile phone terminal models and Internet of Things terminal models, with poor results and low accuracy.
The word low-dimensional embedding technology is used to embed terminal model data words into low-dimensional space, obtain vectors, and through the optimized terminal model similarity model, the dependencies between data words are captured and the similarity value is output to identify terminal models.
The accuracy of terminal model recognition is improved, and the understanding and recognition ability of terminal model characteristics is enhanced by optimizing the model.
Smart Images
Figure CN119782840A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of Internet of Things, and in particular to a method, device and server for identifying terminal models based on the field of Internet of Things. Background Art
[0002] The application areas of the Internet of Things cover many fields such as industry, agriculture, transportation, energy and smart cities. With the continuous diversification of business scenarios, IoT cards are also widely used in all walks of life. IoT cards are only used by enterprise users for IoT devices and cannot be used on personal terminals such as mobile phones. Therefore, it is necessary to identify IoT cards used on mobile terminals and make rectifications to ensure the correct use of IoT cards.
[0003] Currently, the existing technology performs semantic matching on input data based on a large language model to identify mobile phone terminal models and IoT terminal models.
[0004] However, the effect of identifying mobile phone terminal models and IoT terminal models based on semantic matching is poor and the accuracy is low. Summary of the invention
[0005] The embodiments of the present application provide a method, device and server for identifying terminal models based on the field of Internet of Things, so as to achieve the effect of improving the accuracy of identifying terminal models.
[0006] In a first aspect, an embodiment of the present application provides a terminal model identification method based on the field of the Internet of Things, including: collecting terminal model data; wherein the terminal model data includes terminal models; the terminal models include mobile phone terminal models and Internet of Things terminal models; segmenting the terminal model data to obtain multiple terminal model data words; using a low-dimensional embedding method of words to obtain a vector of each terminal model data word; according to the vector of each terminal model data word, obtaining a vector of any terminal model text data in the terminal model data; adding a relative position relationship to the terminal model data words in the vector of any terminal model text data to obtain a terminal model vector library; inputting the terminal model vector library into an optimized terminal model similarity model, so that the optimized terminal model similarity model outputs a similarity value of each terminal model; obtaining all terminal models whose similarity values exceed a preset similarity threshold; wherein all terminal models whose similarity values exceed the similarity threshold are used to identify mobile phone terminal models and Internet of Things terminal models.
[0007] In a possible implementation, obtaining a vector of any terminal model text data in the terminal model data according to the vector of each terminal model data word includes: obtaining a vector of any terminal model text data in the terminal model data according to the vector corresponding to each terminal model data word in any terminal model text data.
[0008] In a possible implementation manner, the formula for obtaining a vector of any terminal model text data in the terminal model data is:
[0009]
[0010] In the formula, represents a vector of any terminal model text data, n represents the number of terminal model data words in any terminal model text data, The vector representing the i-th terminal model data word.
[0011] In a possible implementation, it also includes: calculating a loss value based on a vector of any terminal model text data and a vector of each terminal model data word in any terminal model text data; and continuously adjusting the vector of each terminal model data word so that the loss value gradually decreases until the loss value no longer decreases.
[0012] In one possible implementation, the formula for calculating the loss value is:
[0013]
[0014] In the formula, o represents the loss value, n represents the number of terminal model data words in any terminal model text data, Represents the vector of the i-th terminal model data word, A vector representing text data for any terminal model.
[0015] In a possible implementation, a relative position relationship is added to the terminal model data words in the vector of any terminal model text data, and the formula is:
[0016]
[0017] In the formula, Indicates the relative position relationship between the terminal model data word corresponding to position i and the terminal model data word corresponding to position j, represents the query vector, represents the key vector, and represents the position offset vector, Represents the relative position encoding between positions i and j.
[0018] In a possible implementation manner, before collecting the terminal model data, the method further includes: adding a learnable four-dimensional array to the terminal model similarity model to obtain an optimized terminal model similarity model.
[0019] In a second aspect, an embodiment of the present application provides a terminal model identification device based on the field of Internet of Things, including:
[0020] A collection module is used to collect terminal model data; wherein the terminal model data includes terminal models; terminal models include mobile terminal models and Internet of Things terminal models;
[0021] A segmentation module, used for segmenting the terminal model data to obtain a plurality of terminal model data words;
[0022] The first acquisition module is used to obtain the vector of each terminal model data word by using a low-dimensional word embedding method;
[0023] A second acquisition module is used to acquire a vector of any terminal model text data in the terminal model data according to the vector of each terminal model data word;
[0024] A third acquisition module is used to add a relative position relationship to the terminal model data words in the vector of any terminal model text data to obtain a terminal model vector library;
[0025] An output module, used for inputting the terminal model vector library into the optimized terminal model similarity model, so that the optimized terminal model similarity model outputs the similarity value of each terminal model;
[0026] The fourth acquisition module is used to acquire all terminal models whose similarity values exceed a preset similarity threshold; wherein all terminal models whose similarity values exceed the similarity threshold are used to identify mobile phone terminal models and Internet of Things terminal models.
[0027] In a third aspect, an embodiment of the present application provides a server, including: a memory, a processor;
[0028] The memory stores computer-executable instructions;
[0029] The processor executes the computer-executable instructions stored in the memory, so that the processor executes the above first aspect and / or various possible implementations of the first aspect.
[0030] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the first aspect above and / or various possible implementations of the first aspect.
[0031] In a fifth aspect, an embodiment of the present application provides a computer program product, including a computer program, which, when executed by a processor, implements the above first aspect and / or various possible implementation methods of the first aspect.
[0032] The terminal model identification method, device and server based on the field of Internet of Things provided by the embodiment of the present application, through the low-dimensional embedding technology of words, multiple terminal model data words are embedded into a low-dimensional space to obtain the vector of each terminal model data word; according to the vector of each terminal model data word, the vector of any terminal model text data in the terminal model data is obtained. At the same time, a relative position relationship is added to the terminal model data words in the vector of any terminal model text data to obtain a terminal model vector library. The terminal model vector library is input into the optimized terminal model similarity model to output the similarity value of each terminal model. On the basis of obtaining the vector of each terminal model data word, the vector of any terminal model text data is obtained, and the relative position relationship is added to the terminal model data words in the vector of any terminal model text data, which can capture the dependency relationship between the terminal model data words, and then through the optimized terminal model similarity model, the similarity value of each terminal model can be accurately obtained. Therefore, according to the similarity value, the accuracy of identifying the mobile phone terminal model and the Internet of Things terminal model is increased. In addition, by continuously adjusting the vector of each terminal model data word, the loss value is gradually reduced until the loss value no longer decreases. The smaller the loss value, the higher the degree of match between the vector of the terminal model data word and any terminal model text data, and the more accurate the understanding of the terminal model data by the optimized terminal model similarity model; in the process of loss value calculation, by taking the square root based on the Euclidean distance, the vector of the terminal model data word is easier to encode and calculate, the encoding length is shorter, and the accuracy of loss value calculation is improved. BRIEF DESCRIPTION OF THE DRAWINGS
[0033] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.
[0034] Figure 1 A schematic diagram of a scenario of a method for identifying a terminal model based on the Internet of Things field provided in an embodiment of the present application;
[0035] Figure 2 A schematic diagram of a process for identifying a terminal model based on the Internet of Things field provided in an embodiment of the present application;
[0036] Figure 3 A schematic diagram of the structure of a terminal model identification device based on the Internet of Things field provided in an embodiment of the present application;
[0037] Figure 4 A schematic diagram of the structure of a server provided in an embodiment of the present application.
[0038] The above drawings have shown clear embodiments of the present application, which will be described in more detail later. These drawings and text descriptions are not intended to limit the scope of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION
[0039] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.
[0040] Figure 1 A schematic diagram of a scenario of a method for identifying a terminal model based on the Internet of Things field provided in an embodiment of the present application, such as Figure 1 As shown, the specific application scenario of the present application includes: a receiving device 101, a processor 102 and a display device 103.
[0041] It is understandable that the structure illustrated in the embodiment of the present application does not constitute a specific limitation on the terminal model identification method based on the field of the Internet of Things. In other feasible implementations of the present application, the above architecture may include more or fewer components than shown in the figure, or combine certain components, or split certain components, or arrange the components differently, which can be determined according to the actual application scenario and is not limited here. Figure 1 The components shown may be implemented in hardware, software, or a combination of software and hardware.
[0042] In a specific implementation process, the receiving device 101 may be an input / output interface or a communication interface for collecting terminal model data.
[0043] The processor 102 can process the terminal model data to identify the mobile phone terminal model and the Internet of Things terminal model.
[0044] The display device 103 can be used to display the identified mobile phone terminal model and Internet of Things terminal model.
[0045] It should be understood that the above-mentioned processor can be implemented by the processor reading instructions in the memory and executing the instructions, or it can be implemented by a chip circuit.
[0046] In addition, the network architecture and business scenarios described in the embodiments of the present application are intended to more clearly illustrate the technical solutions of the embodiments of the present application, and do not constitute a limitation on the technical solutions provided in the embodiments of the present application. Ordinary technicians in this field can know that with the evolution of network architecture and the emergence of new business scenarios, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.
[0047] The application areas of the Internet of Things cover many fields such as industry, agriculture, transportation, energy and smart cities. With the continuous diversification of business scenarios, IoT cards are also widely used in all walks of life. IoT cards are only used by enterprise users for IoT devices and cannot be used on personal terminals such as mobile phones. Therefore, it is necessary to identify IoT cards used on mobile terminals and make rectifications to ensure the correct use of IoT cards. At present, the existing technology is based on a large language model to perform semantic matching on the input data to identify the mobile terminal model and the IoT terminal model. However, the effect of identifying the mobile terminal model and the IoT terminal model based on semantic matching is poor and the accuracy is low.
[0048] In order to solve the above technical problems, the embodiment of the present application proposes the following technical concept: through the low-dimensional embedding technology of words, multiple terminal model data words are embedded into a low-dimensional space, the vector of each terminal model data word is obtained, and then the vector of any terminal model text data in the terminal model data is obtained. At the same time, the relative position relationship of the terminal model data words in the vector of any terminal model text data is added, which can capture the dependency relationship between the terminal model data words. Through the optimized terminal model similarity model, the similarity value of each terminal model can be accurately obtained, thereby increasing the accuracy of identifying the mobile phone terminal model and the Internet of Things terminal model according to the similarity value.
[0049] The technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems are described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.
[0050] Figure 2 A schematic diagram of a process flow of a terminal model identification method based on the Internet of Things field provided in an embodiment of the present application is shown in FIG. Figure 2 As shown, the method includes:
[0051] S201: Collect terminal model data; the terminal model data includes terminal models; terminal models include mobile terminal models and Internet of Things terminal models.
[0052] The terminal model data includes the current mobile phone terminal model data and the terminal model data obtained from the IoT data center. The terminal model data obtained from the IoT data center includes the IoT terminal model and the mobile phone terminal model used by the IoT card in violation of regulations.
[0053] In this embodiment, the existing network mobile phone terminal model data includes all mobile phone terminal models released in the past 15 years.
[0054] Optionally, the terminal model data is saved. Optionally, the terminal model data can be saved in a distributed file system (Hadoop Distributed File System, HDFS).
[0055] S202: Segment the terminal model data to obtain a plurality of terminal model data words.
[0056] Optionally, segment the terminal model data, eliminate unnecessary symbols, emojis, hashtags, and Unicode characters, and convert the text to lower case for consistency.
[0057] S203: Obtain a vector for each terminal model data word using a low-dimensional word embedding method.
[0058] Optionally, low-dimensional embedding of words converts each terminal model data word into a vector in a low-dimensional vector space.
[0059] In this embodiment, the co-occurrence information in the terminal model data is encoded in order to maintain the second-order similarity, that is, not only the characteristics of the terminal model data words themselves, but also the association relationship between the terminal model data words is considered.
[0060] S204: Obtain a vector of any terminal model text data in the terminal model data according to the vector of each terminal model data word.
[0061] Specifically, according to the vector corresponding to each terminal model data word in any terminal model text data, the vector of any terminal model text data in the terminal model data is obtained.
[0062] In this embodiment, all terminal model data words in any terminal model text data are obtained, and the vector mean of all terminal model data words is used to represent the vector of any terminal model text data.
[0063] Specifically, the formula for obtaining the vector of any terminal model text data in the terminal model data is:
[0064]
[0065] In the formula, represents a vector of any terminal model text data, n represents the number of terminal model data words in any terminal model text data, The vector representing the i-th terminal model data word.
[0066] Exemplarily, any terminal model text data is , w represents the terminal model data word in any terminal model text data, n represents the number of terminal model data words in any terminal model text data, The vector is .
[0067] In this embodiment, the loss value is calculated based on the vector of any terminal model text data and the vector of each terminal model data word in any terminal model text data; the vector of each terminal model data word is continuously adjusted so that the loss value gradually decreases until the loss value no longer decreases.
[0068] Specifically, the formula for calculating the loss value is:
[0069]
[0070] In the formula, o represents the loss value, n represents the number of terminal model data words in any terminal model text data, Represents the vector of the i-th terminal model data word, A vector representing text data for any terminal model.
[0071] In this embodiment, the vector of any terminal model text data is determined by minimizing the loss value. The loss value is gradually reduced by continuously adjusting the vector of each terminal model data word to achieve minimization.
[0072] In this embodiment, It is the Euclidean distance, which is used to measure the difference between the vector of the terminal model data word and the vector of any terminal model text data. Taking the square root based on the Euclidean distance makes the vector easier to encode and calculate, the encoding length is shorter, and the accuracy is improved.
[0073] S205: adding relative position relationships to terminal model data words in a vector of any terminal model text data to obtain a terminal model vector library.
[0074] Specifically, the formula for adding a relative position relationship to the terminal model data words in the vector of any terminal model text data is:
[0075]
[0076] In the formula, Indicates the relative position relationship between the terminal model data word corresponding to position i and the terminal model data word corresponding to position j, represents the query vector, represents the key vector, and represents the position offset vector, Represents the relative position encoding between positions i and j.
[0077] In this embodiment, an offset term and a relative position code are introduced into the relative position relationship, which can capture the dependency relationship between terminal model data words in any terminal model text data.
[0078] S206: Inputting the terminal model vector library into the optimized terminal model similarity model, so that the optimized terminal model similarity model outputs a similarity value of each terminal model.
[0079] In this embodiment, the optimized terminal model similarity model establishes a pattern for distinguishing mobile phone model terminals and IoT terminal models by learning the differences and similarity patterns between different models in the terminal model vector library. The optimized terminal model similarity model can learn the similarity between the mobile phone terminal model data word vectors and the similarity between the IoT terminal model data word vectors to output the similarity value of each terminal model.
[0080] S207: Acquire all terminal models whose similarity values exceed a preset similarity threshold; wherein all terminal models whose similarity values exceed the similarity threshold are used to identify mobile phone terminal models and Internet of Things terminal models.
[0081] In this embodiment, if the terminal model with a similarity value exceeding a preset similarity threshold has a high similarity in a feature dimension related to a mobile phone terminal model, but a low similarity in a feature dimension related to an Internet of Things terminal, then it is identified as a mobile phone terminal model; otherwise, it is identified as an Internet of Things terminal model.
[0082] Optionally, the Internet of Things card used in the identified mobile phone terminal model is functionally restricted and shut down.
[0083] In summary, through the low-dimensional embedding technology of words, multiple terminal model data words are embedded into a low-dimensional space to obtain the vector of each terminal model data word; according to the vector of each terminal model data word, the vector of any terminal model text data in the terminal model data is obtained. At the same time, a relative position relationship is added to the terminal model data words in the vector of any terminal model text data to obtain a terminal model vector library. The terminal model vector library is input into the optimized terminal model similarity model to output the similarity value of each terminal model. On the basis of obtaining the vector of each terminal model data word, the vector of any terminal model text data is obtained, and the relative position relationship is added to the terminal model data words in the vector of any terminal model text data, which can capture the dependency relationship between the terminal model data words, and then through the optimized terminal model similarity model, the similarity value of each terminal model can be accurately obtained. Therefore, according to the similarity value, the accuracy of identifying mobile phone terminal models and Internet of Things terminal models is increased. In addition, by continuously adjusting the vector of each terminal model data word, the loss value is gradually reduced until the loss value no longer decreases. The smaller the loss value, the higher the degree of match between the vector of the terminal model data word and any terminal model text data, and the more accurate the understanding of the terminal model data by the optimized terminal model similarity model; in the process of loss value calculation, by taking the square root based on the Euclidean distance, the vector of the terminal model data word is easier to encode and calculate, the encoding length is shorter, and the accuracy of loss value calculation is improved.
[0084] Based on the above embodiment, in this embodiment, the optimized terminal model similarity model is introduced, which is described in detail as follows:
[0085] A learnable four-dimensional array is added to the terminal model similarity model to obtain an optimized terminal model similarity model.
[0086] In this embodiment, a variable low-rank matrix method is proposed to solve the problems that may arise when fine-tuning the model in the traditional LoRA method. In the process of optimizing the terminal model similarity model, LoRA usually converts the four-dimensional tensor of the Conv2D convolution layer parameters into a two-dimensional tensor, but this easily leads to the occurrence of rank reduction. If rank reduction occurs, some important information may be lost, affecting the learning and recognition of terminal model features by the terminal model similarity model.
[0087] Optionally, the variable low-rank matrix method uses only 20% of the original LoRA parameters for the optimization process by adding a small learnable four-dimensional array as a matrix multiplication prefix in the terminal model similarity model.
[0088] In this embodiment, by adding a learnable four-dimensional array, LoRA is prevented from converting the four-dimensional tensor of the Conv2D convolution layer parameters into a two-dimensional tensor, resulting in a reduced rank. In addition, the number of parameter adjustments is reduced, and reducing parameter adjustments means faster computing speed and lower resource consumption.
[0089] Figure 3 A schematic diagram of the structure of a terminal model identification device based on the Internet of Things field provided in an embodiment of the present application, such as Figure 3 As shown, the terminal model identification device based on the Internet of Things field provided in this embodiment includes: a collection module 301, a segmentation module 302, a first acquisition module 303, a second acquisition module 304, a third acquisition module 305, an output module 306 and a fourth acquisition module 307.
[0090] The collection module 301 is used to collect terminal model data; wherein the terminal model data includes terminal models; terminal models include mobile terminal models and Internet of Things terminal models;
[0091] A segmentation module 302 is used to segment the terminal model data to obtain a plurality of terminal model data words;
[0092] A first acquisition module 303 is used to acquire a vector of each terminal model data word by using a low-dimensional word embedding method;
[0093] A second acquisition module 304 is used to acquire a vector of any terminal model text data in the terminal model data according to the vector of each terminal model data word;
[0094] The third acquisition module 305 is used to add a relative position relationship to the terminal model data words in the vector of any terminal model text data to obtain a terminal model vector library;
[0095] An output module 306, used to input the terminal model vector library into the optimized terminal model similarity model, so that the optimized terminal model similarity model outputs the similarity value of each terminal model;
[0096] The fourth acquisition module 307 is used to acquire all terminal models whose similarity values exceed a preset similarity threshold; wherein all terminal models whose similarity values exceed the similarity threshold are used to identify mobile phone terminal models and Internet of Things terminal models.
[0097] In a possible implementation manner, the second acquisition module 304 is specifically configured to acquire a vector of any terminal model text data in the terminal model data according to a vector corresponding to each terminal model data word in any terminal model text data.
[0098] In a possible implementation, the formula of the second acquisition module 304′ is:
[0099]
[0100] In the formula, represents a vector of any terminal model text data, n represents the number of terminal model data words in any terminal model text data, The vector representing the i-th terminal model data word.
[0101] In a possible implementation, the terminal model identification device based on the Internet of Things field also includes: a calculation module, which is used to calculate the loss value based on the vector of any terminal model text data and the vector of each terminal model data word in any terminal model text data; continuously adjust the vector of each terminal model data word so that the loss value gradually decreases until the loss value no longer decreases.
[0102] In a possible implementation, the calculation module , has a formula of:
[0103]
[0104] In the formula, o represents the loss value, n represents the number of terminal model data words in any terminal model text data, Represents the vector of the i-th terminal model data word, A vector representing text data for any terminal model.
[0105] In a possible implementation, the formula of the third acquisition module 305′ is:
[0106]
[0107] In the formula, Indicates the relative position relationship between the terminal model data word corresponding to position i and the terminal model data word corresponding to position j, represents the query vector, represents the key vector, and represents the position offset vector, Represents the relative position encoding between positions i and j.
[0108] In a possible implementation, the terminal model recognition device based on the Internet of Things field further includes: a fifth acquisition module, which is used to add a learnable four-dimensional array to the terminal model similarity model to obtain an optimized terminal model similarity model.
[0109] The terminal model identification device based on the Internet of Things field provided in this embodiment can execute the method provided in the above method embodiment. Its implementation principle and technical effect are similar, and this embodiment will not be described in detail here.
[0110] Figure 4 This is a schematic diagram of the structure of the server provided in the embodiment of the present application. Figure 4 As shown, the server provided in this embodiment includes: at least one processor 401 and a memory 402. Optionally, the server also includes a communication component 403. The processor 401, the memory 402 and the communication component 403 are connected via a bus 404.
[0111] In a specific implementation process, at least one processor 401 executes the computer-executable instructions stored in the memory 402, so that at least one processor 401 executes the above method.
[0112] The specific implementation process of the processor 401 can be found in the above method embodiment, and its implementation principle and technical effect are similar, so this embodiment will not be repeated here.
[0113] In the above embodiments, it should be understood that the processor can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), etc. A general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in the invention can be directly implemented as a hardware processor, or can be implemented by a combination of hardware and software modules in the processor.
[0114] The memory may include a high-speed memory (Random Access Memory, RAM), and may also include a non-volatile memory (NVM), such as at least one disk storage.
[0115] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, the bus in the drawings of this application is not limited to only one bus or one type of bus.
[0116] The present application also provides a computer program product, including a computer program, which implements the above method when executed by a processor.
[0117] The present application also provides a computer-readable storage medium, in which computer-executable instructions are stored. When a processor executes the computer-executable instructions, the above method is implemented.
[0118] The above-mentioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk. The readable storage medium can be any available medium that can be accessed by a general or special-purpose computer.
[0119] An exemplary readable storage medium is coupled to a processor so that the processor can read information from the readable storage medium and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can be located in an application specific integrated circuit (Application Specific Integrated Circuits, referred to as: ASIC). Of course, the processor and the readable storage medium can also exist in the device as discrete components.
[0120] The division of units is only a logical function division, and there may be other divisions in actual implementation, such as multiple units or components can be combined or integrated into another system, 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 an indirect coupling or communication connection through some interface, device or unit, which can be electrical, mechanical or other forms.
[0121] 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.
[0122] 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.
[0123] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it 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 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 of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.
[0124] Those skilled in the art can understand that all or part of the steps of implementing the above-mentioned method embodiments can be completed by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, the steps of the above-mentioned method embodiments are executed; and the aforementioned storage medium includes: ROM, RAM, disk or optical disk and other media that can store program codes.
[0125] Finally, it should be noted that those skilled in the art will readily conceive of other embodiments of the present invention after considering the specification and practicing the invention disclosed herein. The present invention is intended to cover any variations, uses or adaptations of the present invention, which follow the general principles of the present invention and include common knowledge or customary technical means in the art not disclosed by the present invention, are not limited to the precise structure described above and shown in the drawings, and may be modified and changed in various ways without departing from the scope thereof. The scope of the present invention is limited only by the appended claims.
Claims
1. A terminal model identification method based on the field of Internet of Things, characterized in that: Applicable to servers, including: Collect terminal model data; wherein the terminal model data includes terminal models; the terminal models include mobile terminal models and Internet of Things terminal models; Segmenting the terminal model data to obtain a plurality of terminal model data words; Use the low-dimensional word embedding method to obtain the vector of each terminal model data word; According to the vector of each terminal model data word, obtaining a vector of any terminal model text data in the terminal model data; Adding relative position relationships to the terminal model data words in the vector of any terminal model text data to obtain a terminal model vector library; Inputting the terminal model vector library into the optimized terminal model similarity model, so that the optimized terminal model similarity model outputs the similarity value of each terminal model; All terminal models whose similarity values exceed a preset similarity threshold are obtained; wherein all terminal models whose similarity values exceed the similarity threshold are used to identify mobile phone terminal models and Internet of Things terminal models.
2. The method according to claim 1, characterized in that The step of obtaining, according to the vector of each terminal model data word, a vector of any terminal model text data in the terminal model data includes: According to the vector corresponding to each terminal model data word in any terminal model text data, a vector of any terminal model text data in the terminal model data is obtained.
3. The method according to claim 2, characterized in that The formula for obtaining the vector of any terminal model text data in the terminal model data is: In the formula, represents a vector of any terminal model text data, n represents the number of terminal model data words in any terminal model text data, The vector representing the i-th terminal model data word.
4. The method according to claim 3, characterized in that Also includes: Calculate the loss value according to the vector of any terminal model text data and the vector of each terminal model data word in any terminal model text data; The vector of each terminal model data word is continuously adjusted so that the loss value gradually decreases until the loss value no longer decreases.
5. The method according to claim 4, characterized in that The formula for calculating the loss value is: In the formula, o represents the loss value, n represents the number of terminal model data words in any terminal model text data, Represents the vector of the i-th terminal model data word, A vector representing text data for any terminal model.
6. The method according to claim 1, characterized in that The formula for adding a relative position relationship to the terminal model data words in the vector of any terminal model text data is: In the formula, Indicates the relative position relationship between the terminal model data word corresponding to position i and the terminal model data word corresponding to position j, represents the query vector, represents the key vector, and represents the position offset vector, Represents the relative position encoding between positions i and j.
7. The method according to any one of claims 1 to 6, characterized in that: Before collecting the terminal model data, the method further includes: A learnable four-dimensional array is added to the terminal model similarity model to obtain an optimized terminal model similarity model.
8. A terminal model identification device based on the field of Internet of Things, characterized in that: Applicable to servers, including: A collection module, used for collecting terminal model data; wherein the terminal model data includes terminal models; the terminal models include mobile terminal models and Internet of Things terminal models; A segmentation module, used for segmenting the terminal model data to obtain a plurality of terminal model data words; The first acquisition module is used to obtain the vector of each terminal model data word by using a low-dimensional word embedding method; A second acquisition module, configured to acquire a vector of any terminal model text data in the terminal model data according to the vector of each terminal model data word; A third acquisition module is used to add a relative position relationship to the terminal model data words in the vector of any terminal model text data to obtain a terminal model vector library; An output module, used for inputting the terminal model vector library into the optimized terminal model similarity model, so that the optimized terminal model similarity model outputs the similarity value of each terminal model; The fourth acquisition module is used to acquire all terminal models whose similarity values exceed a preset similarity threshold; wherein all terminal models whose similarity values exceed the similarity threshold are used to identify mobile phone terminal models and Internet of Things terminal models.
9. A server, characterized in that: include: Memory, processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory, so that the processor performs the method according to any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 7 when executed by a processor.
11. A computer program product, characterized in that The invention comprises a computer program, which implements the method according to any one of claims 1 to 7 when being executed by a processor.