Matching method and matching device based on Linux system and Linux system

Abstract

The invention provides a matching method and a matching device based on a Linux system and the Linux system, which are applied to the field of system interaction, and comprise the following steps: responding to a user input operation, and obtaining text description information; the character description information is input into a neural network model, first information is obtained, and the first information comprises a target intention corresponding to the character description information, entities in character description and the relation between the entities; constructing query information according to the first information; according to the query information and a first mapping relation, instruction information based on the Linux system is confirmed, and the first mapping relation is the mapping relation between the query information and the instruction information. According to the method, after character description information is converted into first information through a neural network model, instruction information is obtained through construction and executed, high matching between the character description information and the instruction information is achieved, single correspondence between the character description information and the instruction information is achieved, and instruction execution errors caused by the one-to-many situation are avoided.

Description

Technical Field

[0001] The present application relates to the field of system interaction, and more specifically, to a matching method, a matching device and a Linux system based on a Linux system. Background Art

[0002] The Linux system has extremely strong stability and reliability in servers and embedded systems. The system can run for a long time without restarting, which is essential for environments that require high availability. In addition, the Linux system can provide high customizability. Users can choose different distributions, desktop environments, applications, etc. according to their needs. During the interaction between the user and the Linux system, the Linux system needs to convert the text description sent by the user into instruction information so that the instruction information can be executed through the shell environment in the Linux system.

[0003] In view of this, how to provide an information matching method with high accuracy during the conversion process is a technical problem that needs to be solved urgently. Summary of the invention

[0004] The present application provides a matching method, a matching device and a Linux system based on a Linux system, which aims to convert text description information into first information through a neural network model, construct instruction information and execute the instruction information, thereby achieving a high matching degree between the text description information and the instruction information, and achieving a single correspondence between the two, avoiding a one-to-many situation that causes instruction execution errors.

[0005] In a first aspect, a matching method based on a Linux system is provided, the matching method comprising: obtaining text description information in response to a user input operation; inputting the text description information into a neural network model to obtain first information, wherein the first information comprises a purpose intent corresponding to the text description information, entities in the text description information and the relationship between the entities; constructing query information based on the first information; confirming instruction information based on the Linux system based on the query information and the first mapping relationship, wherein the first mapping relationship is a mapping relationship between the query information and the instruction information.

[0006] Based on the above scheme, the text description information is used through a neural network model to obtain the purpose, entity and relationship between entities corresponding to the text description, and then the query information is constructed. This can achieve a high match between the text description information and the query information; and the instruction information finally generated has a high match with the text description information entered by the user.

[0007] The first mapping relationship belongs to the database, and there is a relationship table corresponding to the query information and the instruction information in the database, which can be queried. For example, the text description information is: "I want to view all .log files in the home directory of user John", and the query information constructed is: find, -user, -name. By querying the first mapping relationship, the final instruction information is: find / home / John -type f -name "*.log.

[0008] In some possible implementations, constructing query information according to the first information includes: constructing query information according to the first information and a second mapping relationship, wherein the second mapping relationship is a mapping relationship between the first information and the query information.

[0009] In some possible implementations, the neural network model includes a first network, a second network, and a third network, and text description information is input into the neural network model to obtain first information, including: inputting the text description information into the first network to obtain purpose intent; inputting the purpose intent and text description information into the second network to obtain entities; inputting entities and text description information into the third network to obtain relationships between entities.

[0010] In some possible implementations, the first network includes an embedding layer, an LSTM layer, and an output layer, wherein the embedding layer is used to convert text description information into word vectors; the LSTM layer is used to obtain time series and structural information in the word vectors, the time series is used to indicate the order of word vectors, and the structural information is used to characterize the relationship between the text description information and the purpose intent; the output layer is used to determine the purpose intent based on the time series and structural information.

[0011] Based on the above scheme, in the process of converting text description information into purpose intent through the first network, since the LSTM layer in the first network can extract the characteristics of time series, the time series in the text description information can be extracted. The time series is the order in which each word appears. Through the time series, the purpose intent can be obtained more accurately.

[0012] In some possible implementations, the second network includes a feature extraction layer and an output layer, wherein the feature extraction layer is used to extract semantic information from user description information and purpose intent; and the output layer is used to determine entities based on the semantic information.

[0013] In some possible implementations, the purpose includes at least one of the following: viewing Linux system information, managing files or directories, monitoring Linux system resources, managing user permissions, and configuring the Linux system.

[0014] In some possible implementations, the entity includes object information corresponding to the instruction information, wherein the object information includes at least one of the following: a file name, a directory path, and a user name.

[0015] In some possible implementations, the method further includes: executing the instruction information based on a shell environment in the Linux system, and feeding back the execution result.

[0016] In some possible implementations, before inputting the text description information into the neural network model, the method further includes: segmenting the text description information using a word segmentation algorithm to obtain word segmentation data, the word segmentation algorithm including a Vibit algorithm; and marking the word segmentation data to obtain the first information.

[0017] In a second aspect, a matching device based on a Linux system is provided, which includes: a response unit for responding to user input operations; an acquisition unit for acquiring text description information; a processing unit for inputting the text description information into a neural network model to obtain first information, wherein the first information includes the purpose intention corresponding to the text description information, entities in the text description, and the relationship between the entities; based on the first information, query information is constructed; based on the query information and the first mapping relationship, instruction information based on the Linux system is obtained, wherein the first mapping relationship is a mapping relationship between the query information and the instruction information.

[0018] In a third aspect, a matching device based on a Linux system is provided, the matching device comprising: a processor and a memory, the memory being used to store a computer program, the processor being used to call and run the computer program stored in the memory to execute: the matching method in the first aspect.

[0019] In a fourth aspect, a Linux system is provided, comprising: a computing device, and a matching device as in the second aspect or the third aspect, wherein the matching device is used to generate instruction information on the computing device and execute the instruction information. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 A schematic architecture diagram of a user-Linux system interaction process applicable to an embodiment of the present application is shown.

[0021] Figure 2 A schematic flowchart of a matching method based on a Linux system provided in an embodiment of the present application is shown.

[0022] Figure 3 A schematic flowchart of an information matching process provided in an embodiment of the present application is shown.

[0023] Figure 4A schematic flowchart of another information matching process provided in an embodiment of the present application is shown.

[0024] Figure 5 A schematic structural block diagram of a matching device based on a Linux system provided in an embodiment of the present application is shown.

[0025] Figure 6 A schematic structural block diagram of another matching device based on a Linux system provided in an embodiment of the present application is shown.

[0026] Figure 7 A schematic structural block diagram of a Linux system provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0027] The technical solution in this application will be described below in conjunction with the accompanying drawings.

[0028] This application involves the interaction process between users and Linux systems, that is, completing tasks such as file management, software installation and updating, and system configuration through the interaction between users and Linux systems.

[0029] Figure 1 A schematic architecture diagram of a user-Linux system interaction process applicable to an embodiment of the present application is shown.

[0030] like Figure 1 As shown, during the interaction process, the user 110 inputs a text description to the terminal device 120, for example, the user 110 inputs a text description of "please search for file A", and the natural language processing engine 1202 installed on the Linux system 1201 is used to parse the text description input by the user 110. Specifically, the natural language processing engine 1202 converts the text description input by the user 110 into a structured query statement, and searches and matches it in the Linux command database 1203. When the natural language processing engine 1202 finds a command that matches the description of the user 110 in the database 1203, the Linux system 1201 will further parse the grammatical structure and parameter requirements of the command. According to the specific content input by the user 110 (for example, whether it contains additional parameters or options), the Linux system 1201 constructs a complete Linux command string and automatically executes the command through the shell environment in the Linux system 1201. Finally, the execution result will be fed back to the user 110 in real time, and the necessary logging function will be provided for subsequent viewing and analysis.

[0031] Among them, the natural language processing engine 1202 has the capabilities of word segmentation, part-of-speech tagging, semantic analysis, etc. to accurately understand the purpose and intention of the user 110.

[0032] The terminal device 120 may be a computer or a server, etc.

[0033] The natural language processing engine 1202 can be installed as a local application on the Linux system 1201 (eg Figure 1 As shown in the figure, it will run directly on the terminal device 120. The natural language processing engine 1202 can also be located on a cloud server instead of the local terminal device 120, that is, the natural language processing engine 1202 is a cloud NLP service, such as Google Cloud Natural Language API, Amazon Comprehend, IBM Watson Natural Language Understanding, etc. The user 110 interacts with the cloud natural language processing engine 1202 by calling the API interface.

[0034] The database 1203 can be installed on the terminal device 120 of the local Linux system 1201 (eg Figure 1 As shown). The database system includes MySQL, PostgreSQL, MongoDB, etc. The database 1203 can also be installed on a cloud server, and users can access it through the network.

[0035] In addition, the user 110 can input through the user interaction interface GUI, wherein the GUI supports the user 110 to input text descriptions, view matching results and execution results, etc. The GUI can provide intelligent prompts and error handling mechanisms to help the user 110 input descriptions more accurately and handle possible abnormal situations.

[0036] During the interaction between user 110 and Linux system 1201, the text description input by user 110 is converted by natural language processing engine 1202. During the conversion process, since there are a large number of instructions with similar functions but different specific usages in database 1203 of Linux system 1201, the text description may not be able to accurately distinguish these differences, thereby causing ambiguity and resulting in the execution of unexpected operations.

[0037] In view of this, an embodiment of the present application provides an information matching method based on a Linux system, which aims to learn the purpose, entities and relationships between entities of the user's text description by passing the text description through a network model, and construct a query statement through the purpose, entities, relationships between entities and a database, and finally find the instruction information corresponding to the query data in the database and execute it. The process of constructing a query statement in this way can accurately match the text description with the instruction information to obtain the final instruction.

[0038] Figure 2A schematic flowchart of a matching method based on a Linux system provided in an embodiment of the present application is shown.

[0039] like Figure 2 As shown, the matching method based on the Linux system includes the following steps.

[0040] S210: Responding to a user input operation, obtaining text description information.

[0041] S220: Input the text description information into the neural network model to obtain first information, wherein the first information includes the purpose intention corresponding to the text description information, entities in the text description information, and the relationship between the entities.

[0042] S230: Construct query information according to the first information.

[0043] S240: Confirming Linux system-based instruction information according to the query information and the first mapping relationship, wherein the first mapping relationship is a mapping relationship between the query information and the instruction information.

[0044] Specifically, the execution subject of the embodiment of the method is a computing device. The computing device may include a general computing device, an embedded computing device, a mobile computing device, a network computing device, and the like. Among them, the general computing device may be a personal computer, a server, a supercomputer, and the like. The embedded computing device may be an embedded system, a smart TV, a smart home device, a car entertainment system, and the like. The mobile computing device may be a smart phone, a tablet computer, a wearable device, and the like. The network computing device may be a router, a network terminal, a cloud infrastructure, and the like.

[0045] In the above step S210, the computing device obtains text description information, which may be the editing, querying, and modifying of the file, etc. For example, the text description information may be user input "Please open file B in folder A", or it may be "Find file A and copy it to folder B".

[0046] Among them, the user can enter text description information through the command line interface (CLI) or the graphical user interface (GUI). In this application, an intuitive and easy-to-use GUI interface can be designed to support functions such as user input of text description information, viewing matching results and execution results, and the interface can provide intelligent prompts and error handling mechanisms to help users enter descriptions more accurately and handle possible abnormal situations.

[0047] Specifically, in this application, the GUI interface can support functions such as user input of text descriptions, viewing of matching results and execution results, and can also have good user experience and operability, reduce the user's learning cost, and provide intelligent prompts during the user input process to help users enter command descriptions more accurately. Errors or abnormal situations input by users are processed and fed back in a timely manner to improve the robustness and reliability of the system. In addition, the GUI interface can be implemented using front-end technologies (such as HTML, CSS, JavaScript, etc.). A development model with front-end and back-end separation can also be adopted to improve the maintainability and scalability of the system.

[0048] Optionally, before inputting the text description information into the neural network model, the text description information can be input into a word segmentation model to perform word segmentation and labeling operations on the text description information, so that it is easier to obtain the user's purpose after it is subsequently sent into the neural network model.

[0049] Specifically, the text description information is preprocessed, that is, text cleaning, normalization, and sentence segmentation are performed. Among them, text cleaning is used to remove irrelevant characters in the text description information, such as special symbols and extra spaces. Normalization is used to convert the text description information into a unified format, such as full-width to half-width, traditional Chinese to simplified Chinese, etc. Sentence segmentation can be used to divide the text description information into sentences, providing a basis for subsequent word segmentation and annotation.

[0050] One possible implementation method is to input the preprocessed text into a word segmentation model and parse the text according to pre-trained parameters, where the word segmentation model can be a rule-based word segmentation model or a statistical-based word segmentation model, or you can directly use word segmentation tools such as Jieba word segmentation or NLTK to segment the text description information.

[0051] Specifically, the word segmentation model extracts the features of each possible segmentation point, such as the character context, word frequency, language model probability, etc. Based on the extracted features, a word segmentation algorithm, such as the Viterbi algorithm, is used to determine the best word segmentation path. The optimal word sequence is finally output. For example, the text description information is "find file A and copy it to folder B", and the final word segmentation is "search file A", "copy file A", and "copy file A to file B".

[0052] One possible implementation method is to mark the part of speech of each word, such as noun, verb, adjective, etc., based on word segmentation. In addition, identify specific entities in the text, such as names of people, places, and organizations, and mark their types. Or mark the dependency relationship between words, such as subject-predicate relationship, verb-object relationship, etc.

[0053] In the above step S220, use Figure 3 To explain, Figure 3 In the information matching process, it can be seen that the computing device inputs the text description information into the neural network model to obtain the purpose intent, entities and relationships between entities corresponding to the text description information, and the purpose intent, entities and relationships between entities are constructed by the construction unit to obtain the query statement. It should be understood that the construction unit is a unit in the computing device, which is used to construct the query statement.

[0054] The purpose intent may be what type of Linux command the user wants to execute, including: viewing system information, managing files or directories, monitoring system resources, managing users and permissions, and network configuration, etc. The entity includes object information corresponding to the instruction information, wherein the object information includes at least one of the following: file name, directory path, and user name, etc. The relationship between entities is described in the form of examples. For example, if the user enters "copy file A to directory B", the relationship between entity "file A" and entity "directory B" is a copy relationship.

[0055] Specifically, the processing of more specific neural network models is as follows Figure 4 As shown, in Figure 4 In the information matching process, it can be seen that the neural network model includes the first network, the second network and the third network. The text description information is input into the first network to obtain the purpose intent; the purpose intent and text description information are input into the second network to obtain the entity; the entity and text description information are input into the third network to obtain the relationship between the entities.

[0056] The first network includes an input layer, an embedding layer, an LSTM layer, a fully connected layer, and an output layer.

[0057] Among them, the embedding layer converts the input text description information into dense word vectors, which can capture the semantic information of the vocabulary. The LSTM layer can capture long-distance dependencies in sequence data (i.e., word vectors). Among them, the forget gate in the LSTM layer determines which information is discarded from the cell state. The input gate determines which new information is stored in the cell state. The "memory" in the cell state network stores long-term information in the sequence. The output gate determines which part of the cell state will be output to the next hidden state. In this application, the LSTM layer is used to obtain the time series and structural information in the word vector, the time series is used to indicate the order of the word vector, and the structural information is used to characterize the relationship between the text description information and the purpose intent. The fully connected layer further nonlinearly transforms the output of the LSTM layer, i.e., the time series and structural information, and converts the output vector of the LSTM layer into a vector with more abstract features. Among them, more abstract features can include high-level semantics and structural information that can capture text description information. The output layer uses a softmax activation function to output the probability distribution of each intent category.

[0058] Since it is necessary to pay attention to the order, format and type of words in the process of constructing query information, the LSTM layer in the second network can be used to obtain the time series in the text description information, thereby avoiding this problem.

[0059] The second network is an entity recognition model, wherein the second network includes a feature extraction layer and an output layer.

[0060] Specifically, the feature extraction layer in the second network converts the user's text description information and intention into a fixed-length vector representation, which contains the semantic information of the user's text description information. The output layer is a classification layer that assigns an entity label to each word, for example, "PER" represents a person and "LOC" represents a location. Based on the classification results of the output layer, the entity label of each word is decoded. Words that are consecutively labeled as the same entity are combined into one entity.

[0061] The entity and text description information obtained after passing through the second network is input into the third network.

[0062] The third network extracts the relationship between user entities. The third network includes a feature extraction network and an output layer.

[0063] The feature extraction network in the third network is used to extract features of entity pairs and their contexts. The output layer in the third network assigns a relationship label to each entity pair, such as "WorksFor", "LocatedIn", etc. Based on the classification results of the output layer, the relationship between the entity pairs is determined.

[0064] In the above step S230, the computing device constructs query information according to the purpose, entities and the relationship between entities.

[0065] Optionally, the computing device constructs query information according to the first information and the second mapping relationship, wherein the second mapping relationship is a mapping relationship between the first information and the query information.

[0066] For example, the user inputs the text description information as "I want to view all .log files in John user's home directory".

[0067] Among them, the text description information obtains the user's intention to view the file through the first network; the purpose intention and the text description information are passed through the second network to obtain the entity corresponding to the text description information, that is, the user name: John, the directory path: the home directory / home / John and the file type: .log. The entity and the text description information are sent to the third network to determine the relationship between the entities, that is, the home directory of the John user is the starting position of the file search, and the .log file needs to be found. The query statement is constructed through the purpose intention, the entity and the relationship between the entities, that is, the second mapping relationship is used to construct it, and the corresponding find command is found in the second mapping relationship through "search", the -user option is used to specify the user, and the -name option is used to specify the file name pattern. The query statement finally formed is "find, -name, .log".

[0068] In the above step S240, the computing device obtains instruction information corresponding to the Linux system according to the query information and the first mapping relationship, wherein the first mapping relationship is a mapping relationship between the query information and the instruction information.

[0069] The first mapping relationship belongs to a database. Optionally, the first mapping relationship may also be one or more of the following: The mapping relationship between the query information and the example information in the database, the mapping relationship between the query information and the parameter information in the database, and the mapping relationship between the query information and the options corresponding to the instruction information.

[0070] Optionally, the content of the database includes a command table, a parameter table, an option table, an example table, etc. Among them, the command table is used to store instruction information, the parameter table is used to store the description of the instruction information, the option table is used to store the option information of the instruction information, and the example table is used to store the examples in the instruction information. Necessary indexes and associations are established in the database to improve the query efficiency, that is, indexes are established between the above multiple tables, and other tables can be found through one of the tables, for example, the parameter table can be directly found through the command table.

[0071] The database may be located in a computing device or in a cloud service center, etc. This application does not impose any limitation on the location of the database.

[0072] Optionally, in this application, the parameters contained in the database are: command information in common Linux systems, function descriptions corresponding to the command information, syntax structures, parameter options, and actual usage, etc. Among them, indexes and associations are established in the database to improve query efficiency.

[0073] Optionally, the parameters in the database can be collected and obtained through official documents, online tutorials, community forums, books and other channels.

[0074] Optionally, the shell environment in the Linux system automatically executes the command information that is finally generated and feeds back the execution result. During the execution process, it is necessary to capture and handle possible errors or abnormal situations, and feed back the execution result of the command to the user in real time. Record the execution log of the command, including the execution time, execution result, error information, etc., for subsequent viewing and analysis.

[0075] Based on the above scheme, the text description information is passed through a neural network model to obtain the purpose, entity and relationship between entities corresponding to the text description, and then the query information is constructed. This can achieve a high match between the text description information and the query information, and the instruction information finally generated has a high match with the text description information entered by the user.

[0076] Optionally, automatically executing system instructions may involve sensitive or dangerous operations, such as modifying system configurations, deleting important files, etc. If the system cannot accurately determine the security and legality of the instructions, it may perform unauthorized operations, posing a threat to system security. In this application, the Linux system uses a permission model based on users, groups, and other users. Each file and directory has read, write, and execute permissions for the owner, the group to which it belongs, and other users. Specifically, it can be divided into the following steps: Step 1: User authentication. Before performing any operation, the Linux system needs to verify the user's identity. This is usually achieved through the login process of username and password.

[0077] Step 1: Permission verification: Once the user's identity is confirmed, the Linux system will verify the user's permissions based on the user's identity and the operation to be performed.

[0078] Step 3: Implementation of permission check based on text description. First, the Linux system needs to parse the text description information entered by the user to understand the operation the user wants to perform and the files or directories involved. Based on the parsed text description, the Linux system can determine the type of operation the user wants to perform, such as reading a file, writing a file, executing a file, or changing file permissions. The Linux system needs to query the permission information of the file or directory, which can usually be obtained by calling a Linux system call or executing a related command (such as ls -l). The user's identity information is compared with the permission information of the file or directory to verify whether the user has the right to perform the required operation. This includes checking whether the user is the owner of the file or directory, whether the user belongs to the group to which the file or directory belongs, or whether the user has the corresponding permissions of other users. If the user has the required permissions, the system can perform the corresponding operation. If the user does not have the required permissions, the Linux system should return a corresponding error message and may prompt the user to escalate permissions (such as using the sudo command).

[0079] Based on the above solution, the security and reliability of the Linux system can be improved by optimizing and enhancing the permission checking mechanism.

[0080] Combination of the above Figures 2 to 4 The method embodiment provided by the present application is described below. Figure 5 and Figure 6 The device embodiments provided in the present application are described. It is understandable that the device embodiments below may correspond to the method embodiments above, and the specific solutions may refer to the above description, and the repeated parts will not be described in detail.

[0081] Figure 5 A schematic structural block diagram of a matching device 500 based on a Linux system provided in an embodiment of the present application is shown.

[0082] like Figure 5 As shown, the matching device 500 includes: a responding unit 510 , an acquiring unit 520 and a processing unit 530 .

[0083] The response unit 510 is used to respond to user input operations.

[0084] The acquisition unit 520 is used to acquire text description information.

[0085] The processing unit 530 is used to input the text description information into the neural network model to obtain first information, wherein the first information includes the purpose intention corresponding to the text description information, the entities in the text description, and the relationship between the entities; the processing unit 530 is also used to construct query information based on the first information; the processing unit 530 is also used to obtain instruction information based on the Linux system based on the query information and the first mapping relationship, wherein the first mapping relationship is a mapping relationship between the query information and the instruction information.

[0086] Figure 6 A schematic structural block diagram of another matching device 600 based on a Linux system provided in an embodiment of the present application is shown.

[0087] like Figure 6 As shown, the matching device 600 includes: a processor 610 and a memory 620, the memory 620 is used to store a computer program, and the processor 610 is used to call and run the computer program stored in the memory 620 to execute the above Figures 2 to 4 The matching method based on the Linux system provided in any embodiment.

[0088] The embodiment of the present application also provides a Linux system. Figure 7 A schematic structural block diagram of a Linux system 700 is shown.

[0089] like Figure 7 As shown, the Linux system 700 includes: a computing device 710 and a matching device 720, wherein the matching device 720 is used to generate instruction information and execute the instruction information on the computing device 710. Optionally, the matching device 720 may be as follows: Figure 5 or Figure 6 The matching device in .

[0090] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0091] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0092] It should be understood that the specific examples in this article are only intended to help those skilled in the art better understand the embodiments of the present application, rather than to limit the scope of the embodiments of the present application.

[0093] It should also be understood that in the various embodiments of the present application, the size of the serial number of each process does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0094] It should also be understood that the various implementation methods described in this specification can be implemented individually or in combination, and the embodiments of the present application are not limited to this.

[0095] Unless otherwise stated, all technical and scientific terms used in the embodiments of the present application are the same as the meanings generally understood by those skilled in the art of the technical field of the present application. The terms used in the present application are only for the purpose of describing specific embodiments and are not intended to limit the scope of the present application. The term "and / or" used in the present application includes any and all combinations of one or more related listed items. The singular forms of "a kind of", "above" and "the" used in the embodiments of the present application and the appended claims are also intended to include majority forms, unless the context clearly indicates other meanings. In addition, the terms "first", "second", "third" etc. are only used for description purposes and cannot be understood as indicating or suggesting relative importance.

[0096] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another 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 through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0097] 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.

[0098] In addition, each functional unit in each embodiment of the present application 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.

[0099] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes 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 described in each embodiment of the present application. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.

[0100] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A matching method based on Linux system, characterized in that: include: Respond to user input operations and obtain text description information; Inputting the text description information into a neural network model to obtain first information, wherein the first information includes the purpose intent corresponding to the text description information, entities in the text description information, and the relationship between the entities; Constructing query information according to the first information; According to the query information and the first mapping relationship, instruction information based on the Linux system is confirmed, wherein the first mapping relationship is a mapping relationship between query information and instruction information.

2. The matching method according to claim 1, characterized in that: The step of constructing query information according to the first information includes: The query information is constructed according to the first information and the second mapping relationship, wherein the second mapping relationship is a mapping relationship between the first information and the query information.

3. The matching method according to claim 1 or 2, characterized in that: The neural network model includes a first network, a second network and a third network. The step of inputting the text description information into a neural network model to obtain first information includes: Inputting the text description information into the first network to obtain the purpose intention; Inputting the purpose intention and the text description information into the second network to obtain the entity; The entities and the text description information are input into the third network to obtain the relationship between the entities.

4. The matching method according to claim 3, characterized in that: The first network includes an embedding layer, an LSTM layer and an output layer, Wherein, the embedding layer is used to convert the text description information into a word vector; The LSTM layer is used to obtain the time series and structural information in the word vector, the time series is used to indicate the order of the word vector, and the structural information is used to characterize the relationship between the text description information and the purpose intent; The output layer is used to determine the purpose intent according to the time series and the structural information.

5. The matching method according to claim 3, characterized in that: The second network includes a feature extraction layer and an output layer, Wherein, the feature extraction layer is used to extract the semantic information in the user description information and the purpose intention; The output layer is used to determine the entity according to the semantic information.

6. The matching method according to claim 1 or 2, characterized in that: The purpose is intended to include at least one of the following: View the Linux system information, manage files or directories, monitor the Linux system resources, manage user permissions, and configure the Linux system.

7. The matching method according to claim 1 or 2, characterized in that: The entity includes object information corresponding to the instruction information, The object information includes at least one of the following: a file name, a directory path, and a user name.

8. The matching method according to claim 1 or 2, characterized in that: The method further comprises: Based on the shell environment in the Linux system, the instruction information is executed and the execution result is fed back.

9. The matching method according to claim 1 or 2, characterized in that: Before inputting the text description information into the neural network model, the method further includes: Segmenting the text description information using a segmentation algorithm to obtain segmentation data, wherein the segmentation algorithm includes a Vibit algorithm; The word segmentation data is marked to obtain the first information.

10. A matching device based on Linux system, characterized in that: include: A response unit, used for responding to user input operations; An acquisition unit, used for acquiring text description information; A processing unit, configured to input the text description information into a neural network model to obtain first information, wherein the first information includes a purpose intent corresponding to the text description information, entities in the text description, and relationships between the entities; Constructing query information according to the first information; According to the query information and the first mapping relationship, instruction information based on the Linux system is obtained, wherein the first mapping relationship is a mapping relationship between query information and instruction information.

11. A matching device based on Linux system, characterized in that: include: A processor and a memory, wherein the memory is used to store a computer program, and the processor is used to call and run the computer program stored in the memory to execute: the matching method according to any one of claims 1 to 9.

12. A Linux system, characterized in that: include: A computing device, and a matching apparatus as claimed in claim 10 or 11, wherein the matching apparatus is used to generate instruction information on the computing device and execute the instruction information.

Images