CN116384410B - Visual processing method and system for digital factory - Google Patents

Abstract

The visual processing method and the visual processing system of the digital factory relate to the field of artificial intelligence, the text vector field corresponding to the debugged text knowledge mining network is extracted and opened up through optimizing and updating the artificial intelligent model, the obtained text vector field is transferred to the basic text knowledge mining network, the debugged first target text knowledge mining network is obtained, independent text training data and corresponding annotation information are not needed for optimizing the network, the network is enabled to be faster in debugging, the follow-up visual application of the visual interaction device of the digital factory is facilitated, and the accuracy and timeliness of visual management of the factory are improved.

Description

Visual processing method and system for digital factory

Technical Field

The application relates to the field of artificial intelligence, in particular to a visual processing method and a visual processing system for a digital factory.

Background

With the continuous development of information technology, three-dimensional digital factories are becoming popular, and three-dimensional digital factories are gradually adopted by more enterprises with the advantages of enterprise management innovation and data integration. As a carrier of company production operation management and control information, the three-dimensional digital factory can realize integrated operation and maintenance management and control of company production, energy, security and equipment, thereby ensuring scientific management decision. One of the functions of the digital factory is to realize visual management of the equipment, and the visual function can be in data docking with the enterprise equipment management system to provide data integration for equipment operation and maintenance management. And enterprises can also utilize the visual function to provide intelligent equipment guidance, intelligent inspection, intelligent monitoring and rapid fault handling for the user. In addition, text analysis processing by visualization in combination with artificial intelligence is also one of the demands for visual management of devices.

Disclosure of Invention

The application provides a visual processing method and a visual processing system for a digital factory.

According to an aspect of the present application, there is provided a visual processing method of a digital factory, applied to an AI processing system, the method including: acquiring a current debugging text sequence, wherein the current debugging text sequence is acquired from a debugging text sequence deployed in advance; loading each debug text in the current debug text sequence into a debugged text knowledge mining network and a basic text knowledge mining network to be debugged respectively for text knowledge vector mining, and obtaining a debugged text knowledge vector corresponding to each debug text and a to-be-debugged text knowledge vector corresponding to each debug text, wherein the basic text knowledge mining network is obtained by pre-configuring network configuration variables of the debugged text knowledge mining network; obtaining common measurement coefficients among the debugged text knowledge vectors corresponding to the debugged texts, obtaining a debugged common measurement coefficient set, and obtaining common measurement coefficients among the to-be-debugged text knowledge vectors corresponding to the debugged texts, thus obtaining a to-be-debugged common measurement coefficient set; obtaining a cost value between the to-be-debugged common measurement coefficient set and the debugged common measurement coefficient set, optimizing the to-be-debugged basic text knowledge mining network through the cost value, and then jumping to a step of obtaining a current debugging text sequence to roll until meeting the debugging cut-off requirement, determining the debugged basic text knowledge mining network as a first target text knowledge mining network, wherein the first target text knowledge mining network is configured to mine text knowledge vectors of visual management texts of target equipment; and visually managing text knowledge vectors of the text through the target equipment to recognize text intention.

As one embodiment, the obtaining the commonality measurement coefficient between the debugged text knowledge vectors corresponding to the respective debug texts, to obtain a debugged commonality measurement coefficient set, includes: obtaining a debugged text knowledge array through the debugged text knowledge vectors corresponding to the debugged texts, and performing standardized operation on the debugged text knowledge array to obtain a debugged standardized array; acquiring a turnover array corresponding to the debugged standardized array, and acquiring a debugged turnover array; obtaining the debugged commonality measurement coefficient set through the debugged flip array and the debugged standardized array; the obtaining the commonality measurement coefficient between the knowledge vectors of the text to be debugged corresponding to each debug text, obtaining the commonality measurement coefficient set to be debugged, includes: obtaining a to-be-debugged text knowledge array through to-be-debugged text knowledge vectors corresponding to the respective debug texts, and performing standardization operation on the to-be-debugged text knowledge array to obtain a to-be-debugged standardization array; acquiring a turnover array corresponding to the to-be-debugged standardized array, and acquiring a to-be-debugged turnover array; and obtaining the to-be-debugged commonality measurement coefficient set through the to-be-debugged overturn array and the to-be-debugged standardization array.

As an implementation manner, the step of obtaining the cost value between the to-be-debugged commonality measurement coefficient set and the debugged commonality measurement coefficient set, optimizing the to-be-debugged basic text knowledge mining network through the cost value, and then jumping to obtain the current debug text sequence for rolling includes: obtaining standard deviation values of the to-be-debugged commonality measurement coefficient set and the debugged commonality measurement coefficient set, obtaining a basic error result, and determining the basic error result as the cost value; the cost value is reversely propagated to optimize network configuration variables in the basic text knowledge mining network, and an optimized text knowledge mining network is obtained; determining the optimized text knowledge mining network as a basic text knowledge mining network, and then jumping to a step of acquiring a current debugging text sequence for rolling; the obtaining the standard deviation value of the to-be-debugged commonality measurement coefficient set and the debugged commonality measurement coefficient set to obtain the cost value includes: obtaining standard deviation values of the to-be-debugged commonality measurement coefficient set and the debugged commonality measurement coefficient set, and obtaining a basic error result; obtaining the number of debugging texts corresponding to the current debugging text sequence, and obtaining the proportional coefficient of the basic error result and the number of the debugging texts to obtain a mean error result; acquiring a preset adjusting variable, and adjusting and calculating the mean error result through the preset adjusting variable to acquire an adjusting error result; and acquiring an intention recognition error result corresponding to the basic text knowledge mining network to be debugged, and acquiring error results and values of the intention recognition error result and the adjustment error result to obtain the cost value.

As an implementation manner, the obtaining the cost value between the set of to-be-debugged commonality measurement coefficients and the set of debugged commonality measurement coefficients, optimizing the to-be-debugged basic text knowledge mining network through the cost value, and then jumping to the step of obtaining the current debug text sequence to perform rolling until meeting the debug cut-off requirement, determining the debugged basic text knowledge mining network as a first target text knowledge mining network, including: loading the to-be-debugged commonality measurement coefficient set into a basic transformation module to perform coefficient transformation to obtain a target transformation commonality measurement coefficient set; obtaining standard deviation values between the target transformation commonality measurement coefficient set and the debugged commonality measurement coefficient set, obtaining a target cost value, and reversely propagating through the target cost value to optimize the basic transformation module and the basic text knowledge mining network, and obtaining an optimized transformation module and an optimized text knowledge mining network; and determining the optimized transformation module as a basic transformation module, determining the optimized text knowledge mining network as a basic text knowledge mining network, and then jumping to the step of acquiring the current debugging text sequence to roll until meeting the debugging cut-off requirement, and obtaining a second target text knowledge mining network through the debugged basic text knowledge mining network and the debugged basic transformation module.

As one embodiment, the basic text knowledge mining network is a basic mining network; the step of obtaining the cost value between the to-be-debugged commonality measurement coefficient set and the debugged commonality measurement coefficient set, optimizing the to-be-debugged basic text knowledge mining network through the cost value, and then jumping to the step of obtaining the current debugging text sequence to roll until meeting the debugging cut-off requirement, determining the debugged basic text knowledge mining network as a first target text knowledge mining network, wherein the step of obtaining the current debugging text sequence comprises the following steps of: loading the to-be-debugged commonality measurement coefficient set and the debugged commonality measurement coefficient set into a basic identification module for identification to obtain a commonality measurement coefficient identification result; and optimizing the basic recognition module and the basic text knowledge mining network through the commonality measurement coefficient recognition result, and then jumping to the step of acquiring the current debugging text sequence to roll until the current debugging text sequence meets the debugging cut-off requirement, and determining the debugged basic text knowledge mining network as a third target text knowledge mining network.

As one embodiment, the current debugging text sequence comprises multiple text sets, and the multiple text sets comprise combined text groups; the method further comprises the steps of: loading each multi-element text set into a debugged text knowledge mining network and a basic text knowledge mining network to be debugged respectively to carry out text knowledge vector mining, so as to obtain debugged multi-element set knowledge vectors corresponding to each multi-element text set and to-be-debugged multi-element set knowledge vectors corresponding to each multi-element text set; and acquiring a multi-element set cost value through the debugged multi-element set knowledge vector and the multi-element set knowledge vector to be debugged, obtaining a basic multi-element set error result, reversely spreading the basic multi-element set error result to optimize the basic text knowledge mining network, and then jumping to the step of acquiring the current debugging text sequence to roll until the current debugging text sequence meets the debugging cut-off requirement, and determining the debugged basic text knowledge mining network as a fourth target text knowledge mining network.

As an implementation manner, after the step of obtaining the cost value between the set of to-be-debugged commonality measurement coefficients and the set of debugged commonality measurement coefficients, optimizing the to-be-debugged basic text knowledge mining network through the cost value, and then jumping to the step of obtaining the current debugged text sequence to roll until meeting the requirement of the cutoff of debugging, determining the debugged basic text knowledge mining network as the first target text knowledge mining network, further includes: acquiring a target device visual management text, loading the target device visual management text into the first target text knowledge mining network to mine text knowledge vectors, and acquiring a to-be-recognized text knowledge vector; acquiring a historical intent text knowledge vector corresponding to a visual management text set of a historical intent device, and acquiring a commonality measurement coefficient of the text knowledge vector to be identified and the historical intent text knowledge vector; and determining text intention identification information corresponding to the target device visual management text through the commonality measurement coefficient.

As an embodiment, the method further comprises: acquiring a text to be analyzed and an analyzed text sequence; loading the text to be analyzed and the text sequence to be analyzed into a debugged text knowledge mining network to perform text knowledge vector mining, obtaining a text knowledge vector to be analyzed corresponding to the text to be analyzed and a text knowledge vector to be analyzed corresponding to the text sequence to be analyzed, and obtaining a common measurement coefficient of the text knowledge vector to be analyzed and the text knowledge vector to be analyzed, so as to obtain a first common measurement coefficient set; loading the to-be-analyzed text and the analyzed text sequence into a target text knowledge mining network to perform text knowledge vector mining, obtaining a to-be-analyzed target text knowledge vector corresponding to the to-be-analyzed text and an analyzed text knowledge vector set corresponding to the analyzed text sequence, and obtaining a commonality measurement coefficient of the to-be-analyzed target text knowledge vector and the analyzed text knowledge vector set to obtain a second commonality measurement coefficient set, wherein the target text knowledge mining network is obtained by performing migration learning and debugging based on a debugged text knowledge mining network; analyzing through the first commonality measurement coefficient set and the second commonality measurement coefficient set to obtain an analysis result corresponding to the to-be-analyzed text, and determining the commonality measurement analysis result corresponding to the to-be-analyzed text through the analysis result corresponding to the to-be-analyzed text; the obtaining the commonality measurement coefficient of the text knowledge vector to be analyzed and the text knowledge vector to be analyzed, obtaining a first commonality measurement coefficient set, including: normalizing the text knowledge vector to be analyzed to obtain a normalized text knowledge vector to be analyzed, and normalizing the text knowledge vector to be analyzed to obtain a normalized text knowledge vector to be analyzed; and performing inversion operation on the standardized analyzed text knowledge vector to obtain an analyzed inversion array, and obtaining the first common measurement coefficient set through the product of the standardized analyzed text knowledge vector and the analyzed inversion array.

As one embodiment, the obtaining the commonality metric coefficient of the target text knowledge vector to be analyzed and the analyzed text knowledge vector set, to obtain a second commonality metric coefficient set, includes: normalizing the target text knowledge vector to be analyzed to obtain a normalized target text knowledge vector to be analyzed, and normalizing the text knowledge vector set to obtain a normalized text knowledge vector set to be analyzed; performing overturn operation on the standardized analyzed text knowledge vector set to obtain an analyzed object overturn array, and obtaining the second commonality measurement coefficient set through the standardized analyzed text knowledge vector set and the analyzed object overturn array; analyzing through the first commonality measurement coefficient set and the second commonality measurement coefficient set to obtain an analysis result corresponding to the to-be-analyzed text, and determining the commonality measurement analysis result corresponding to the to-be-analyzed text through the analysis result corresponding to the to-be-analyzed text, including: obtaining a standard deviation value between the first commonality measurement coefficient set and the second commonality measurement coefficient set to obtain a target error result; determining the number of texts corresponding to the text to be analyzed and the analyzed text sequence, acquiring a proportionality coefficient of the target error result and the number of texts, and determining an analysis result corresponding to the text to be analyzed through the proportionality coefficient; and if the analysis result is larger than a preset value, obtaining a matching analysis indication result corresponding to the text to be analyzed.

According to another aspect of the present application, there is provided an AI processing system including: at least one processor; and a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method described above.

The application at least has the following beneficial effects:

According to the visual processing method and the visual processing system for the digital factory, each debugging text in the current debugging text sequence is respectively loaded into a debugged text knowledge mining network and a basic text knowledge mining network to be debugged to carry out text knowledge vector mining, and a debugged text knowledge vector corresponding to each debugging text and a to-be-debugged text knowledge vector corresponding to each debugging text are obtained. And finally, optimizing a basic text knowledge mining network to be debugged through a cost value based on the cost value, and then jumping to the step of acquiring a current debugging text sequence for rolling until meeting the debugging cut-off requirement, determining the debugged basic text knowledge mining network as a first target text knowledge mining network, transferring the obtained text vector field to the basic text knowledge mining network by extracting text vector fields corresponding to the text knowledge mining network after opening up the debugged text knowledge mining network, and obtaining the debugged first target text knowledge mining network without optimizing the network by using independent text training data and corresponding annotation information.

It should be understood that the description in this section is not intended to identify key or critical features of the embodiments of the application or to delineate the scope of the application. Other features of the present application will become apparent from the description that follows.

Drawings

The accompanying drawings illustrate exemplary embodiments and, together with the description, serve to explain exemplary implementations of the embodiments. The illustrated embodiments are for exemplary purposes only and do not limit the scope of the claims. Throughout the drawings, identical reference numerals designate similar, but not necessarily identical, elements.

Fig. 1 shows an application scenario diagram of a visualization processing method of a digital factory according to an embodiment of the present application.

Fig. 2 shows a flow chart of a method of visualization processing for a digital factory according to an embodiment of the application.

Fig. 3 shows a functional block diagram of a processing device according to an embodiment of the application.

Fig. 4 shows a composition schematic diagram of an AI processing system according to an embodiment of the application.

Detailed Description

Exemplary embodiments of the present application will now be described with reference to the accompanying drawings, in which various details of the embodiments of the present application are included to facilitate understanding, and are to be considered merely exemplary. Accordingly, one of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope of the application. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.

In the present application, the use of the terms "first," "second," etc. to describe various elements is not intended to limit the positional relationship, timing relationship, or importance relationship of these elements unless otherwise indicated, and such terms are merely used to distinguish one element from another element. In some examples, a first element and a second element may refer to the same instance of the element, and in some cases, they may also refer to different instances based on the description of the context.

The terminology used in the description of the various illustrated examples in this disclosure is for the purpose of describing particular examples only and is not intended to be limiting. Unless the context clearly indicates otherwise, the elements may be one or more if the number of the elements is not specifically limited. Furthermore, the term "and/or" as used in the present application encompasses any and all possible combinations of the listed items.

Fig. 1 shows a schematic diagram of a scenario provided according to an embodiment of the present application. Including one or more terminal devices 101, an AI processing system 120, and one or more communication networks 110 coupling the one or more terminal devices 101 to the AI processing system 120. Terminal device 101 may be configured to execute one or more applications.

In an embodiment of the present application, the AI processing system 120 can run one or more services or software applications that enable execution of the visualization processing method of the digital plant. In some embodiments, the AI processing system 120 can also provide other services or software applications, which can include non-virtual environments and virtual environments. In some embodiments, these services may be provided as web-based services or cloud services, for example, provided to the user of terminal device 101 under a software as a service (SaaS) model.

In the configuration shown in fig. 1, the AI processing system 120 can include one or more components that implement the functionality performed by the AI processing system 120. These components may include software components, hardware components, or a combination thereof that are executable by one or more processors. The user operating the terminal device 101 can, in turn, utilize one or more applications to interact with the AI processing system 120 to utilize the services provided by these components. It should be appreciated that a variety of different system configurations are possible, which may differ from system 100. Accordingly, FIG. 1 is one example of a system for implementing the various methods described herein and is not intended to be limiting. The terminal device 101 may send network debug commands to the AI processing system over the network 110, and the terminal device 101 may include various types of computer devices, such as portable handheld devices, general purpose computers (such as personal computers and laptop computers), workstation computers, wearable devices, smart screen devices, self-service terminal devices, service robots, gaming systems, thin clients, various messaging devices, sensors or other sensing devices, and the like. These computer devices may run various types and versions of software applications and operating systems, such as MICROSOFT Windows, APPLE iOS, UNIX-like operating systems, linux, or Linux-like operating systems (e.g., GOOGLE Chrome OS); or include various mobile operating systems such as MICROSOFT Windows Mobile OS, iOS, windows Phone, android. Portable handheld devices may include cellular telephones, smart phones, tablet computers, personal Digital Assistants (PDAs), and the like. Wearable devices may include head mounted displays (such as smart glasses) and other devices. The gaming system may include various handheld gaming devices, internet-enabled gaming devices, and the like. The terminal device 101 is capable of executing various different applications, such as various Internet-related applications, communication applications (e.g., email applications), short Message Service (SMS) applications, and may use various communication protocols. Network 110 may be any type of network known to those skilled in the art that may support data communications using any of a number of available protocols, including but not limited to TCP/IP, SNA, IPX, etc. For example only, the one or more networks 110 may be a Local Area Network (LAN), an ethernet-based network, a token ring, a Wide Area Network (WAN), the internet, a virtual network, a Virtual Private Network (VPN), an intranet, an extranet, a blockchain network, a Public Switched Telephone Network (PSTN), an infrared network, a wireless network (e.g., bluetooth, WIFI), and/or any combination of these and/or other networks.

The AI processing system 120 can include one or more general purpose computers, special purpose server computers (e.g., PC (personal computer) servers, UNIX servers, mid-end servers), blade servers, mainframe computers, server clusters, or any other suitable arrangement and/or combination. The AI processing system 120 can include one or more virtual machines running a virtual operating system, or other computing architecture that involves virtualization (e.g., one or more flexible pools of logical storage devices that can be virtualized to maintain virtual storage devices of a server). In various embodiments, the AI processing system 120 can run one or more services or software applications that provide the functionality described below.

The computing units in the AI processing system 120 can run one or more operating systems including any of the operating systems described above as well as any commercially available server operating systems. The AI processing system 120 can also run any of a variety of additional server applications and/or middle tier applications, including HTTP servers, FTP servers, CGI servers, JAVA servers, database servers, and the like. In some implementations, the AI processing system 120 can be a server of a distributed system or a server that incorporates a blockchain. The AI processing system 120 can also be a cloud server, or an intelligent cloud computing server or intelligent cloud host with artificial intelligence techniques. The cloud server is a host product in a cloud computing service system, so as to solve the defects of large management difficulty and weak service expansibility in the traditional physical host and Virtual special server (VPS PRIVATE SERVER) service.

In addition, one or more databases 130 may be included in the scene. In some embodiments, these databases may be used to store data and other information. For example, one or more of databases 130 may be used to store text. Database 130 may reside in various locations. For example, the database used by the AI processing system 120 can be local to the AI processing system 120 or can be remote from the AI processing system 120 and can communicate with the AI processing system 120 via a network-based or dedicated connection. Database 130 may be of different types. In some embodiments, the database used by the AI processing system 120 can be, for example, a relational database. One or more of these databases may store, update, and retrieve the databases and data from the databases in response to the commands. In some embodiments, one or more of databases 130 may also be used by applications to store application data. The databases used by the application may be different types of databases, such as key value stores, object stores, or conventional stores supported by the file system.

The system 100 of fig. 1 may be configured and operated in various ways to enable application of the various methods and apparatus described in accordance with the present application.

Referring to fig. 2, the method for visualizing a digital factory according to an embodiment of the present application includes the following steps:

s101, acquiring a current debugging text sequence, wherein the current debugging text sequence is acquired from a preset debugging text sequence.

The current debug text sequence includes a plurality of current debug texts, the current debug texts representing texts taken when the neural network is currently debugged. In the embodiment of the application, the debug text sequence is used as a training sample and can be generated by a historical device visual management text selected from a historical user dialogue, such as Chatbots application interfaces. The pre-deployed debug text sequence is a text sequence of debug text used in a pre-set debug process, and the current debug text sequence is a part of the pre-deployed debug text sequence. The debug text in the pre-deployed debug text sequence is, for example, text obtained after the debugged text knowledge mining network is debugged. For example, the current debug text sequence may be obtained directly in the local storage space, where the current debug text sequence is obtained by obtaining in a debug text sequence deployed in advance. For example, after a debug text sequence deployed in advance is obtained, the debug text in the debug text sequence deployed in advance is divided according to the generation time or a preset start-stop symbol, so that each debug text is obtained, and a current debug text sequence is obtained.

S102, loading each debug text in the current debug text sequence into a debugged text knowledge mining network and a basic text knowledge mining network to be debugged respectively for text knowledge vector mining, and obtaining the debugged text knowledge vector corresponding to each debug text and the to-be-debugged text knowledge vector corresponding to each debug text, wherein the basic text knowledge mining network is obtained by pre-configuring network configuration variables of the debugged text knowledge mining network.

The debugged text knowledge mining network is a network for mining text knowledge vectors, which is obtained after a past debugging text is debugged through a neural network. Network optimization is needed to be carried out on the text knowledge mining network after debugging is completed, and updating is completed. The basic text knowledge mining network to be debugged is a text knowledge mining network in which network configuration variables to be debugged are subjected to pre-configuration (namely initialization), or the network configuration variables of the debugged text knowledge mining network are directly subjected to pre-configuration. The pre-configuration of the network configuration variables is, for example, random. The basic text knowledge mining network to be debugged can be any feasible neural network, and the embodiment of the application can pre-configure the network configuration variables of the basic text knowledge mining network to be debugged based on the network configuration variables of the debugged text knowledge mining network, and determine the network configuration variables of the debugged text knowledge mining network as the pre-configured network configuration variables of the basic text knowledge mining network to be debugged.

The text knowledge vector is a text knowledge vector of a debug text obtained by mining based on a text knowledge mining network which is debugged, and the text knowledge vector to be debugged is a text knowledge vector of a debug text obtained by mining the text knowledge vector based on a basic text knowledge mining network which is to be debugged. The text knowledge vector characterizes characteristic information of the text.

For example, loading each debug text in the current debug text sequence into a debugged text knowledge mining network to perform text knowledge vector mining, obtaining a debugged text knowledge vector corresponding to each debug text in the current debug text sequence, loading each debug text in the current debug text sequence into a basic text knowledge mining network to perform text knowledge vector mining, and obtaining a to-be-debugged text knowledge vector corresponding to each debug text in the current debug text sequence. As one embodiment, the debugged text knowledge mining network is Teacher model in transfer learning (abbreviated as TN), and the basic text knowledge mining network to be debugged is Student model in transfer learning (abbreviated as SN).

S103, obtaining common measurement coefficients among the debugged text knowledge vectors corresponding to each debug text, obtaining a debugged common measurement coefficient set, and obtaining common measurement coefficients among the to-be-debugged text knowledge vectors corresponding to each debug text, thus obtaining the to-be-debugged common measurement coefficient set.

For example, the debugged commonality measurement coefficient set includes each debugged commonality measurement coefficient, the debugged commonality measurement coefficient indicates the commonality degree between the debugged text knowledge vectors of two different debug texts, and the commonality measurement coefficient between the debugged text knowledge vectors corresponding to every two debug texts in the current debug text sequence is obtained. The set of to-be-debugged commonality measurement coefficients comprises each to-be-debugged commonality measurement coefficient, and the to-be-debugged commonality measurement coefficient indicates the commonality degree between to-be-debugged text knowledge vectors corresponding to two different debug texts. And (3) browsing each debugging text in the current debugging text sequence, and acquiring the degree of commonality between the current debugging text and each debugging text in the current debugging text sequence. Obtaining a debugged commonality measurement coefficient based on the debugged text knowledge vector, obtaining a debugged commonality measurement coefficient set, and characterizing a text knowledge vector field of a current debugging text sequence obtained by text knowledge vector mining of a debugged text knowledge mining network by the debugged commonality measurement coefficient set. Obtaining a to-be-debugged commonality measurement coefficient based on the to-be-debugged text knowledge vector, obtaining a to-be-debugged commonality measurement coefficient set, and representing a text knowledge vector field of a current debugging text sequence obtained by text knowledge vector mining of a to-be-debugged basic text knowledge mining network by the to-be-debugged commonality measurement coefficient set. Alternatively, the set of debugged commonality metric coefficients may be expressed by an array, such as a two-dimensional matrix, or a multi-dimensional tensor, to debug the set of commonality metric coefficients.

S104, obtaining a cost value between the to-be-debugged commonality measurement coefficient set and the debugged commonality measurement coefficient set, optimizing a to-be-debugged basic text knowledge mining network through the cost value, and then jumping to the step of obtaining a current debugging text sequence to roll until meeting the debugging cut-off requirement, determining the debugged basic text knowledge mining network as a first target text knowledge mining network, wherein the first target text knowledge mining network is configured to mine text knowledge vectors of target equipment visual management texts, and performing text intention recognition through the text knowledge vectors of the target equipment visual management texts.

The cost value indicates a difference between the set of co-metric coefficients to be debugged and the set of debugged co-metric coefficients. For example, the difference between each to-be-debugged common metric coefficient in the to-be-debugged common metric coefficient set and the corresponding debugged common metric coefficient in the debugged common metric coefficient set may be obtained, and then the sum of all the differences may be obtained to obtain the cost value. And then, carrying out reverse propagation through GRADIENT DESCENT based on the cost value to optimize the network configuration variables in the basic text knowledge mining network to be debugged, so as to obtain the optimized text knowledge mining network to be debugged. And then determining the optimized to-be-debugged text knowledge mining network as a to-be-debugged basic text knowledge mining network, and simultaneously, rolling the step of acquiring the next debugging text sequence of the next Batch until meeting the debugging cut-off requirement, determining the debugged basic text knowledge mining network as a first target text knowledge mining network, and traversing all the debugging texts in the previously deployed debugging text sequence to finish the generation. The debugging cut-off requirement, namely network convergence, specifically may be that the number of times of debugging reaches a preset number of times, the cost value obtained by debugging is smaller than a threshold value, the network configuration variable is not changed any more, and the like, and the method is not limited in particular. The first target text knowledge mining network is obtained after the basic text knowledge mining network to be debugged is debugged, and is configured to mine text knowledge vectors of the target device visual management text, and text intention recognition is carried out through the text knowledge vectors of the target device visual management text.

For example, recognition analysis of security management demand intention or line quality detection intention can be performed by a text knowledge vector of a target device visual management text. It should be noted that the specific details intended to be identified are not of great importance to the embodiments of the present application. Optionally, the first target text knowledge mining network may perform text intent recognition by visually managing text knowledge vectors of text by the target device.

According to the visual processing method for the digital factory, provided by the embodiment of the application, the text knowledge vectors are mined by loading each debug text in the current debug text sequence into the debugged text knowledge mining network and the basic text knowledge mining network to be debugged, so that the debugged text knowledge vectors corresponding to each debug text and the to-be-debugged text knowledge vectors corresponding to each debug text are obtained. And then obtaining a common measurement coefficient between debugged text knowledge vectors corresponding to each debugged text, obtaining a debugged common measurement coefficient set, obtaining a common measurement coefficient between to-be-debugged text knowledge vectors corresponding to each debugged text, obtaining a to-be-debugged common measurement coefficient set, finally obtaining a cost value between the to-be-debugged common measurement coefficient set and the debugged common measurement coefficient set, optimizing a to-be-debugged basic text knowledge mining network through the cost value, and then jumping to a step of obtaining a current debugging text sequence to roll until meeting the debugging cut-off requirement, determining the debugged basic text knowledge mining network as a first target text knowledge mining network, opening up text intention recognition through the first target text knowledge mining network, transferring the obtained text vector field to the basic text knowledge mining network, obtaining the debugged first target text knowledge mining network, and optimizing the network without independent text training data and corresponding annotation information, so that the network has higher speed in time, is convenient for subsequent efficient text intention recognition and timeliness recognition.

Optionally, obtaining the commonality metric coefficient between the debugged text knowledge vectors corresponding to each debug text in S103 may specifically include:

S1031, obtaining a debugged text knowledge array through the debugged text knowledge vectors corresponding to the debugged texts, and performing standardization operation on the debugged text knowledge array to obtain a debugged standardization array.

In the embodiment of the application, the debugged text knowledge array may be a two-dimensional array (i.e., a matrix) containing text knowledge vectors of each debugged text knowledge vector. Generating a debugged text knowledge array based on the debugged text knowledge vectors corresponding to the debugged texts, and performing normalization operation on the debugged text knowledge array, for example, completing normalization based on a normalization operator, for example, dividing each component in the text knowledge vector by the Euclidean norm of the text knowledge vector to obtain a debugged normalized array.

S1032, obtaining a turnover array corresponding to the debugged standardized array, obtaining a debugged turnover array, and obtaining a debugged commonality measurement coefficient set through the debugged turnover array and the debugged standardized array.

In the embodiment of the application, taking the debugged text knowledge array as a two-dimensional array as an example, the debugged flip array is a table obtained by transposition. For example, the debugged standardized array is subjected to overturn operation, a debugged overturn array is obtained after transposition, matrix multiplication is performed on the debugged overturn array and the debugged standardized array, matrix multiplication is performed on the debugged text knowledge vector and other debugged text knowledge vectors, so that a common measurement coefficient of the debugged text knowledge vector and other debugged text knowledge vectors is obtained, and a debugged common measurement coefficient set is obtained after the above process of each debugged text knowledge vector is completed.

Optionally, obtaining the set of debugged commonality metric coefficients corresponding to the debugged flip array and the debugged standardized array may be obtaining a vector distance (such as a euclidean distance, a minth distance, a cosine distance, etc.) corresponding to the debugged flip array and the debugged standardized array, to obtain the set of debugged commonality metric coefficients.

Optionally, obtaining the commonality measurement coefficient between the knowledge vectors of the to-be-debugged text corresponding to each debug text in S103 may specifically include:

S1033, obtaining a to-be-debugged text knowledge array through to-be-debugged text knowledge vectors corresponding to the respective debug texts, and performing standardization operation on the to-be-debugged text knowledge array to obtain a to-be-debugged standardization array.

S1034, obtaining a turnover array corresponding to the quasi-debugging standardized array, obtaining the quasi-debugging turnover array, and obtaining a quasi-debugging commonality measurement coefficient set through the quasi-debugging turnover array and the quasi-debugging standardized array.

The process of S1033 and S1034 can refer to the principle of S1031 and S1032.

And obtaining a turnover array through turnover operation, and then obtaining a standardized array and a matrix of the turnover array to multiply to obtain a commonality measurement coefficient, so that the efficiency of obtaining the commonality measurement coefficient is ensured.

Optionally, in S104, obtaining a cost value between the set of to-be-debugged commonality measurement coefficients and the set of debugged commonality measurement coefficients, optimizing the to-be-debugged basic text knowledge mining network through the cost value, and then jumping to the step of obtaining the current debug text sequence for rolling, which specifically includes:

S1041, obtaining standard deviation values of the to-be-debugged commonality measurement coefficient set and the debugged commonality measurement coefficient set, obtaining a basic error result, and determining the basic error result as a cost value.

S1042, the network configuration variables in the basic text knowledge mining network are optimized through the counter-propagation of the cost values, and the optimized text knowledge mining network is obtained.

According to the embodiment of the application, the gradient is acquired through the cost value, and then the gradient is reversely transmitted to the basic text knowledge mining network, so that the optimization of the network configuration variables in the basic text knowledge mining network is completed, and the optimized text knowledge mining network is obtained.

S1043, determining the optimized text knowledge mining network as a basic text knowledge mining network, and then jumping to a step of acquiring a current debugging text sequence for rolling.

Optionally, in S1041, obtaining a standard deviation value of the set of to-be-debugged commonality measurement coefficients and the debugged commonality measurement coefficient set, so as to obtain a cost value, which may specifically include:

s10411, obtaining standard deviation values of the to-be-debugged commonality measurement coefficient set and the debugged commonality measurement coefficient set, and obtaining a basic error result.

For example, euclidean distance between the coefficient of the common metric to be debugged in the coefficient set of the common metric to be debugged and the coefficient of the common metric to be debugged in the coefficient set of the common metric to be debugged is obtained to obtain a basic error result

S10412, obtaining the number of the debugging texts corresponding to the current debugging text sequence, and obtaining the proportionality coefficient of the basic error result and the number of the debugging texts to obtain a mean error result.

S10412, obtaining a preset adjusting variable, and adjusting and calculating a mean error result through the preset adjusting variable to obtain an adjusting error result.

The preset adjustment variable is a network configuration variable of an eccentric coefficient (weight) which is deployed in advance and used for measuring the migration learning cost and the intention recognition cost. For example, the mean error result is weighted using the adjustment variable to obtain an adjustment error result, i.e. a weighted error result.

S10413, acquiring an intention recognition error result corresponding to the basic text knowledge mining network to be debugged, and acquiring an error result sum value of the intention recognition error result and the adjustment error result to obtain the cost value.

The intention recognition error result is the cost of the basic text knowledge mining network to be debugged when text knowledge vector mining is performed and then text intention recognition tasks are performed. For example, an intention recognition error result corresponding to the basic text knowledge mining network to be debugged is obtained, wherein text intention indication information and the basic text knowledge mining network to be debugged can be obtained to carry out text intention recognition according to the text knowledge vector to be debugged corresponding to the debug text, a basic recognition result is obtained, and then cost between the basic recognition result and the text intention indication information is obtained to obtain an intention recognition error result. The intention recognition error result can be obtained by obtaining a probability distribution distance, and then obtaining an error result sum value of the intention recognition error result and the adjustment error result so as to obtain the cost value.

Based on the method, a basic error result is obtained by obtaining a standard deviation value, and after adjustment is carried out according to an adjustment variable, the obtained sum value of the error result is identified with the intention to obtain a cost value, so that the obtained cost value is more accurate.

Optionally, in S104, obtaining a cost value between the set of to-be-debugged commonality measurement coefficients and the set of debugged commonality measurement coefficients, optimizing a to-be-debugged basic text knowledge mining network through the cost value, and then jumping to a step of obtaining a current debug text sequence to perform rolling until meeting a debug stop requirement, determining the debugged basic text knowledge mining network as a first target text knowledge mining network, including:

(1) And loading the to-be-debugged commonality measurement coefficient set into a basic transformation module to perform coefficient transformation, and obtaining a target transformation commonality measurement coefficient set.

The basic transformation module is a transformation module for pre-configuring network configuration variables, the transformation module is used for performing transformation on the to-be-debugged commonality measurement coefficient set to reduce semantic gaps, the transformation module is a neural network, and the target transformation commonality measurement coefficient set is a set obtained after transformation. Each to-be-debugged commonality measurement coefficient in the to-be-debugged commonality measurement coefficient set can be loaded into the basic transformation module to carry out coefficient transformation, and the target transformation commonality measurement coefficient set output by the basic transformation module is obtained.

(2) The standard deviation value between the target transformation commonality measurement coefficient set and the debugged commonality measurement coefficient set is obtained, the target cost value is obtained, the basic transformation module and the basic text knowledge mining network are optimized through the inverse propagation of the target cost value, and the optimized transformation module and the optimized text knowledge mining network are obtained.

(3) And determining the optimized transformation module as a basic transformation module, determining the optimized text knowledge mining network as a basic text knowledge mining network, and then jumping to the step of acquiring the current debugging text sequence to roll until meeting the debugging cut-off requirement, and obtaining a second target text knowledge mining network through the debugged basic text knowledge mining network and the debugged basic transformation module.

For example, average dividing errors are obtained according to a target transformation commonality measurement coefficient set and a debugged commonality measurement coefficient to obtain a target cost value, then the average dividing errors are reversely propagated based on the target cost value to optimize network configuration variables of a basic transformation module and network configuration variables in a basic text knowledge mining network, an optimized transformation module and an optimized text knowledge mining network are obtained, then the optimized transformation module is determined as the basic transformation module, the optimized text knowledge mining network is determined as the basic text knowledge mining network, and then the step of obtaining a current debugging text sequence is skipped until the current debugging text sequence is obtained, and a second target text knowledge mining network is obtained through the debugged basic text knowledge mining network and the debugged basic transformation module until the debugging cut-off requirement is met. In other words, the second target text knowledge mining network includes a debugged underlying text knowledge mining network and a debugged transformation module. According to the embodiment of the application, the basic text knowledge mining network and the basic transformation module are debugged together by arranging the basic transformation module in the basic text knowledge mining network, so that the second target text knowledge mining network is obtained, and the text knowledge vector mining precision can be improved based on the second target text knowledge mining network.

Optionally, the basic text knowledge mining network is a basic mining network, and in S104, obtaining a cost value between the to-be-debugged commonality measurement coefficient set and the debugged commonality measurement coefficient set, optimizing the to-be-debugged basic text knowledge mining network through the cost value, and then jumping to a step of obtaining a current debug text sequence to perform rolling until meeting a debug cut-off requirement, determining the debugged basic text knowledge mining network as a first target text knowledge mining network, which specifically includes: loading the to-be-debugged commonality measurement coefficient set and the debugged commonality measurement coefficient set into a basic recognition module for recognition, obtaining a commonality measurement coefficient recognition result, optimizing the basic recognition module and the basic text knowledge mining network through the commonality measurement coefficient recognition result, and then jumping to the step of obtaining the current debugging text sequence for rolling until the current debugging text sequence meets the debugging cut-off requirement, and determining the debugged basic text knowledge mining network as a third target text knowledge mining network.

When the basic text knowledge mining network is debugged, the text knowledge mining network is obtained according to countermeasure training debugging, and the basic text knowledge mining network is used as the basic mining network for generating text knowledge vectors of the debug text. The basic recognition module is used for recognizing whether the loaded text knowledge vector is the text knowledge vector mined by the debugged text knowledge mining network or the text knowledge vector mined by the basic text knowledge mining network. Determining the text knowledge vector mined by the debugged text knowledge mining network as positive, determining the text knowledge vector mined by the basic text knowledge mining network to be debugged as negative, debugging the basic text knowledge mining network to enable the text knowledge vector mined by the debugged text knowledge mining network to be closer to the text knowledge vector mined by the basic text knowledge mining network so as to interfere with the judgment of the recognition module, and performing countermeasure learning on the positive and negative text knowledge vectors by the debugging recognition module. After the debugging is finished, the text knowledge vector field obtained by the basic text knowledge mining network mining obtained by the debugging is closer to the text knowledge vector field obtained by the text knowledge mining network mining already debugged, and the identification module can better identify.

For example, the to-be-debugged commonality measurement coefficient set and the debugged commonality measurement coefficient set are loaded into a basic recognition module to be recognized, so that a commonality measurement coefficient recognition result is obtained, and the commonality measurement coefficient recognition result is taken as the basis recognition module to recognize whether the obtained input commonality measurement coefficient belongs to the commonality measurement coefficient corresponding to the positive text knowledge vector or the commonality measurement coefficient corresponding to the negative text knowledge vector. And (3) carrying out cost acquisition on the actual text knowledge vector based on the input commonality measurement coefficient and the identification result of the commonality measurement coefficient, adopting GRADIENT DESCENT counter-propagation to optimize a basic identification module and a basic text knowledge mining network based on the cost, and then jumping to the step of acquiring the current debugging text sequence to roll until the current debugging text sequence meets the debugging cut-off requirement, and determining the debugged basic text knowledge mining network as a third target text knowledge mining network.

Based on the above, the countermeasure learning is executed by adding the basic recognition module, and the debugged basic text knowledge mining network is determined to be a third target text knowledge mining network, so that the precision of the second target text knowledge mining network obtained by debugging is higher.

Alternatively, multiple sets of multiple text, e.g., triples, may be included in the current debug text sequence, with combined text sets included in the multiple sets of text; based on this, the visualization processing method for the digital factory provided by the embodiment of the application further comprises the following steps:

s100, loading each multi-element text set into a debugged text knowledge mining network and a basic text knowledge mining network to be debugged respectively to carry out text knowledge vector mining, and obtaining debugged multi-element set knowledge vectors corresponding to each multi-element text set and to-be-debugged multi-element set knowledge vectors corresponding to each multi-element text set.

In the embodiment of the application, the current debugging text sequence comprises multiple text sets, wherein the multiple text sets comprise combined text groups, namely, two debugging texts in the multiple text sets, namely, the triples, are the same type of texts, the other debugging text is different from the two text of the same type, the same type of texts are positive text groups, and one text in the same type of texts and the text of the different type form negative text groups. And the multi-element set knowledge vector is obtained by constructing each knowledge vector after text knowledge vector mining is carried out on each debugging text in the multi-element text set. That is, the text knowledge vectors of each debug text are combined to obtain a multi-element set knowledge vector, and the debugged multi-element set knowledge vector is a knowledge vector obtained by carrying out text knowledge vector mining on the multi-element text set by the debugged text knowledge mining network. The knowledge vector of the multi-element set to be debugged is a knowledge vector obtained by carrying out text knowledge vector mining on the multi-element text set by a basic text knowledge mining network to be debugged. For example, each multi-element text set is loaded into a debugged text knowledge mining network and a basic text knowledge mining network to be debugged to carry out text knowledge vector mining, a debugged multi-element set knowledge vector corresponding to each multi-element text set is obtained, each multi-element text set is loaded into the basic text knowledge mining network to be debugged to carry out text knowledge vector mining, and a to-be-debugged multi-element set knowledge vector corresponding to each multi-element text set is obtained.

S200, acquiring a multi-element set cost value through the debugged multi-element set knowledge vector and the multi-element set knowledge vector to be debugged, obtaining a basic multi-element set error result, reversely spreading the basic multi-element set error result to optimize a basic text knowledge mining network, and then jumping to a step of acquiring a current debugging text sequence to roll until the current debugging text sequence meets the debugging cut-off requirement, and determining the debugged basic text knowledge mining network as a fourth target text knowledge mining network.

The base multi-element set error result indicates errors corresponding to the debugged multi-element set knowledge vector and the multi-element set knowledge vector to be debugged. For example, the multi-element set cost value is obtained based on the debugged multi-element set knowledge vector, the debugged multi-element set cost value is obtained, meanwhile, the multi-element set cost value is obtained based on the multi-element text set to be debugged, the multi-element set cost value to be debugged is obtained, and then the difference value between the debugged multi-element set cost value and the multi-element set cost value to be debugged is obtained, so that a basic multi-element set error result is obtained. And then, reversely propagating through a basic multi-element set error result by using GRADIENT DESCENT to optimize a basic text knowledge mining network, and then jumping to the step of acquiring a current debugging text sequence to roll until the current debugging text sequence meets the debugging cut-off requirement, and determining the debugged basic text knowledge mining network as a fourth target text knowledge mining network. According to the method, text knowledge vector mining is carried out by loading each multi-element text set into a debugged text knowledge mining network and a basic text knowledge mining network to be debugged respectively to obtain a debugged multi-element set knowledge vector corresponding to each multi-element text set and a to-be-debugged multi-element set knowledge vector corresponding to each multi-element text set, multi-element set cost acquisition is carried out by the debugged multi-element set knowledge vector and the to-be-debugged multi-element set knowledge vector to obtain a basic multi-element set error result, the basic text knowledge mining network is optimized by reversely spreading the basic multi-element set error result, and then the step of obtaining a current debugging text sequence is skipped until the debugged basic text knowledge mining network is determined to be a fourth target text knowledge mining network until the debugging cut-off requirement is met, so that the precision of the obtained text knowledge mining network is improved.

Optionally, after obtaining the cost value between the set of to-be-debugged commonality measurement coefficients and the set of debugged commonality measurement coefficients, optimizing the to-be-debugged basic text knowledge mining network through the cost value, and then jumping to the step of obtaining the current debug text sequence to roll until meeting the debug cut-off requirement, determining the debugged basic text knowledge mining network as the first target text knowledge mining network, the method may further include:

s110, acquiring a target device visual management text, loading the target device visual management text into a first target text knowledge mining network to perform text knowledge vector mining, and acquiring a text knowledge vector to be identified.

The target device visual management text is a text to be subjected to intention recognition, and the text knowledge vector to be recognized is a text knowledge vector corresponding to the target device visual management text. For example, a target device visual management text is obtained, the target device visual management text is a dialogue record in an intercepted Chatbots session, and the target device visual management text is loaded into a first target text knowledge mining network to perform text knowledge vector mining, so that a text knowledge vector to be identified is obtained.

S120, obtaining a historical intention text knowledge vector corresponding to the visual management text set of the historical intention equipment, and obtaining a commonality measurement coefficient of the text knowledge vector to be identified and the historical intention text knowledge vector.

S130, determining text intention identification information corresponding to the visual management text of the target equipment through the commonality measurement coefficient.

The historical intent device visualization management text stores the identified text and corresponding historical intent text knowledge vectors in a set. The identified text is text whose text content contains user intent, and the historical intent text knowledge vector is a text knowledge vector corresponding to the identified text. For example, a historical intent device visualization management text set is deployed in advance, the historical intent device visualization management text set storing a historical intent text knowledge vector corresponding to the identified text. And acquiring a historical intent text knowledge vector in the visual management text set of the historical intent equipment, acquiring a commonality measurement coefficient of the text knowledge vector to be identified and the historical intent text knowledge vector, and if the commonality measurement coefficient is larger than a preset commonality measurement coefficient, representing that the visual management text of the target equipment contains corresponding user intentions, such as information of service requirements, emotion tendencies and the like. At this time, the target device visually manages text intention recognition information corresponding to the text as recognizing the target intention. If the commonality measurement coefficient of each historical intent text knowledge vector in the visual management text set of the text knowledge vector to be identified and the historical intent device is smaller than the preset commonality measurement coefficient, representing that no intent information exists in the visual management text of the target device, and the text intent identification information corresponding to the visual management text of the target device is that no target intent is identified. And carrying out text knowledge vector mining on the visual management text of the target equipment through the first target text knowledge mining network to obtain a to-be-recognized text knowledge vector, then carrying out commonality measurement coefficient acquisition on the to-be-recognized text knowledge vector and each recognized text in the visual management text set of the historical intention equipment, and determining text intention recognition information corresponding to the visual management text of the target equipment through the commonality measurement coefficient, so that the efficiency of obtaining the text intention recognition information is improved.

As an embodiment, the debugging of the corresponding text knowledge mining network may be performed, knowledge transfer learning may be performed, a debugging text sequence is loaded into TN and SN, both TN and SN are text knowledge vector mining networks, TN outputs text knowledge vectors of each debugging text, each TN knowledge vector is obtained, then a common metric coefficient between TN knowledge vectors is obtained, a TN common metric coefficient array is obtained, SN outputs text knowledge vectors of each debugging text, each SN knowledge vector is obtained, then a common metric coefficient between SN knowledge vectors is obtained, an SN common metric coefficient array is obtained, cost of the TN common metric coefficient array and the SN common metric coefficient array is obtained based on a cost algorithm, back propagation is performed based on the cost to optimize SN, rolling is performed, and after the debugging is completed, the debugged SN is determined as the obtained text knowledge mining network.

Optionally, the method provided by the embodiment of the application further comprises the following steps:

S210, obtaining a text to be analyzed and an analyzed text sequence.

In this embodiment, the sequence of analyzed text includes a plurality of analyzed text that is analyzed based on the debugged text knowledge mining network, and the text to be analyzed is text to be analyzed, such as text in the sequence of analyzed text. And after the text knowledge mining network which is debugged is rolled and optimized, the analyzed text is needed to be analyzed again.

S220, loading the text to be analyzed and the text sequence to be analyzed into a debugged text knowledge mining network to perform text knowledge vector mining, obtaining the text knowledge vector to be analyzed corresponding to the text to be analyzed and the text knowledge vector to be analyzed corresponding to the text sequence to be analyzed, obtaining the commonality measurement coefficient of the text knowledge vector to be analyzed and the text knowledge vector to be analyzed, and obtaining the first commonality measurement coefficient set.

The text knowledge mining network is obtained based on past debugging text debugging, the first common metric coefficient set comprises a plurality of first common metric coefficients, the first common metric coefficients indicate common metric coefficients of text knowledge vectors to be analyzed and analyzed text knowledge vectors in the text knowledge vectors to be analyzed, the text knowledge vectors to be analyzed are text knowledge vectors corresponding to the text to be analyzed, and the text knowledge vectors to be analyzed comprise a plurality of text knowledge vectors corresponding to the analyzed text. For example, a debugged text knowledge mining network is called, a to-be-analyzed text and an analyzed text sequence are respectively loaded into the debugged text knowledge mining network to conduct text knowledge vector mining, to obtain to-be-analyzed text knowledge vectors corresponding to the to-be-analyzed text and analyzed text knowledge vectors corresponding to the analyzed text sequence, and then to obtain the commonality measurement coefficients of each analyzed feature in the to-be-analyzed text knowledge vectors and the analyzed text knowledge vectors, to obtain a first commonality measurement coefficient set.

S230, loading the to-be-analyzed text and the analyzed text sequence into a target text knowledge mining network to perform text knowledge vector mining, obtaining a to-be-analyzed target text knowledge vector corresponding to the to-be-analyzed text and an analyzed text knowledge vector set corresponding to the analyzed text sequence, obtaining a commonality measurement coefficient of the to-be-analyzed target text knowledge vector and the analyzed text knowledge vector set, and obtaining a second commonality measurement coefficient set, wherein the target text knowledge mining network is obtained by performing migration learning debugging based on the debugged text knowledge mining network.

In the embodiment of the application, the target text knowledge mining network is obtained by performing migration learning and debugging based on the debugged text knowledge mining network, the target text knowledge vector to be analyzed is a text knowledge vector corresponding to the text to be analyzed, which is obtained by extracting through the target text knowledge mining network, the text knowledge vector set to be analyzed comprises text knowledge vectors obtained by mining all the analyzed texts through the target text knowledge mining network, the second common measurement coefficient set comprises all the second common measurement coefficients, and the second common measurement coefficients indicate the common measurement coefficients of the target text knowledge vector to be analyzed and the analyzed target text knowledge vector in the text knowledge vector set to be analyzed.

For example, a target text knowledge mining network is called, a to-be-analyzed text and an analyzed text sequence are respectively loaded into the target text knowledge mining network to conduct text knowledge vector mining, a to-be-analyzed target text knowledge vector corresponding to the to-be-analyzed text and an analyzed text knowledge vector set corresponding to the analyzed text sequence are obtained, and then a common measurement coefficient of each analyzed target feature in the to-be-analyzed target text knowledge vector and the analyzed text knowledge vector set is obtained, so that a second common measurement coefficient set is obtained.

S240, analyzing through the first commonality measurement coefficient set and the second commonality measurement coefficient set to obtain an analysis result corresponding to the to-be-analyzed text, and determining the commonality measurement analysis result corresponding to the to-be-analyzed text through the analysis result corresponding to the to-be-analyzed text.

The analysis result indicates a commonality metric coefficient of the text to be analyzed and the analyzed text sequence, and the commonality metric analysis result indicates an analysis result of the text to be analyzed, which may specifically include close and non-close. For example, an error is obtained based on the first common metric coefficient set and the second common metric coefficient set, an analysis coefficient corresponding to the analysis text is obtained based on the error, if the analysis coefficient is larger than a preset analysis coefficient, the analysis result of the to-be-analyzed text is determined to be similar, and the to-be-analyzed text knowledge vector corresponding to the to-be-analyzed text can be used for replacing the analyzed text knowledge vector corresponding to the same text in the analyzed text sequence; otherwise, the auxiliary determination can be performed by other modes, such as manpower.

In summary, the to-be-analyzed text and the analyzed text sequence are obtained and loaded to the debugged text knowledge mining network and the target text knowledge mining network for analysis, and because the target text knowledge mining network is obtained by performing migration learning and debugging based on the debugged text knowledge mining network, the analysis speed of the to-be-analyzed text can be improved, meanwhile, the to-be-analyzed text and the analyzed text sequence are subjected to text knowledge vector mining based on the target text knowledge mining network and the debugged text knowledge mining network, a first common measurement coefficient set and a second common measurement coefficient set are determined, and then the common measurement analysis results corresponding to the to-be-analyzed text are determined according to the first common measurement coefficient set and the second common measurement coefficient set, so that the analysis precision of the common measurement analysis results is improved.

Optionally, obtaining the commonality metric coefficient of the text knowledge vector to be analyzed and the text knowledge vector to be analyzed, and obtaining the first commonality metric coefficient set may specifically include: normalizing the text knowledge vector to be analyzed to obtain a normalized text knowledge vector to be analyzed, normalizing the text knowledge vector to obtain a normalized text knowledge vector to be analyzed, performing a flipping operation on the normalized text knowledge vector to obtain an analyzed flipped array, and obtaining a first set of common metric coefficients by the product of the normalized text knowledge vector and the analyzed flipped array.

The normalized text knowledge vector to be analyzed is a normalized text knowledge vector to be analyzed, and the analyzed flip array is a flip array obtained by flipping the normalized text knowledge vector to be analyzed.

Optionally, obtaining the commonality metric coefficient of the target text knowledge vector to be analyzed and the analyzed text knowledge vector set, and obtaining the second commonality metric coefficient set may specifically include: normalizing the target text knowledge vector to be analyzed to obtain a normalized target text knowledge vector to be analyzed, and normalizing the text knowledge vector set to obtain a normalized text knowledge vector set to be analyzed; and performing overturn operation on the standardized analyzed text knowledge vector set to obtain an analyzed target overturn array, and obtaining a second commonality measurement coefficient set through the standardized analyzed text knowledge vector set and the analyzed target overturn array.

Optionally, the analyzing by the first commonality measurement coefficient set and the second commonality measurement coefficient set obtains an analysis result corresponding to the text to be analyzed, and determining the commonality measurement analysis result corresponding to the text to be analyzed by the analysis result corresponding to the text to be analyzed specifically may include:

(A) And obtaining a standard deviation value between the first commonality measurement coefficient set and the second commonality measurement coefficient set, and obtaining a target error result.

(B) Determining the number of texts corresponding to the text to be analyzed and the text sequence to be analyzed, obtaining a proportionality coefficient of a target error result and the number of texts, and determining an analysis result corresponding to the text to be analyzed through the proportionality coefficient.

For example, a difference between the first and second sets of commonality metric coefficients is obtained based on a standard deviation cost algorithm to obtain a target error result, the target error result being indicative of a commonality metric coefficient error between the first and second sets of commonality metric coefficients. Then, determining the number of texts corresponding to the text to be analyzed and the text sequence to be analyzed, namely determining the number of the text to be analyzed in the text sequence to be analyzed, adding the number of the text to be analyzed to obtain the number of the texts, dividing the target error result and the number of the texts to obtain a proportionality coefficient, and determining the analysis result corresponding to the text to be analyzed based on the proportionality coefficient. Wherein the analysis result indicates a change in the distribution of the text to be analyzed in the debugged text knowledge mining network text knowledge vector field as compared to the target text knowledge mining network text knowledge vector field.

(C) And if the analysis result is larger than the preset value, obtaining a matching analysis indication result corresponding to the text to be analyzed.

The selection value of the preset value is set according to the actual situation, and the text knowledge vector of which the matching analysis indication result represents the text to be analyzed can replace the same text in the analyzed text sequence. In a specific scene, a text to be analyzed and a key text sequence are obtained and used as a comparison of the text to be analyzed, the text to be analyzed is a text analyzed based on text knowledge vector mining TN, in the process of obtaining text knowledge vector mining SN through debugging, the text is required to be analyzed again, the text to be analyzed and the key text sequence are loaded to the text knowledge vector mining TN to carry out text knowledge vector mining, a text knowledge vector array of the text knowledge vector to be analyzed and a text knowledge vector array of the key text sequence are obtained, then a common measurement coefficient of the text knowledge vector and the text knowledge vector array is obtained, a TN common measurement coefficient array is obtained, the text knowledge vector to be analyzed and the key text sequence are loaded to the text knowledge vector mining SN to carry out text knowledge vector mining, and then a common measurement coefficient of the text knowledge vector and the text knowledge vector array is obtained, so that the SN common measurement coefficient array is obtained. And obtaining an analysis result corresponding to the text to be analyzed through the TN commonality measurement coefficient array and the SN commonality measurement coefficient array, wherein the analysis result indicates that the distribution of the text to be analyzed in the SN text knowledge vector domain is changed relative to the TN text knowledge vector domain, if the analysis result is larger than a preset analysis result, the text to be analyzed is analyzed and matched, and the analyzed text set can be updated based on the text to be analyzed.

In one specific use environment, the goal is to construct a set of user intent texts, and to migrate texts having symbolic meanings and that are not misinterpreted to the set of texts to pair with target device visual management texts to identify user intent. When the text knowledge vector mining TN is applied to a certain period, updating and optimizing are needed, constructing a basic text knowledge vector mining SN, carrying out knowledge migration through the text knowledge vector mining TN and the text knowledge vector mining SN, specifically, acquiring a current debugging text sequence, wherein the current debugging text sequence is acquired in a pre-deployed debugging text sequence, loading each debugging text in the current debugging text sequence into the text knowledge vector mining TN and the basic text knowledge vector mining SN respectively for text knowledge vector mining, acquiring a debugged text knowledge vector corresponding to each debugging text and a to-be-debugged text knowledge vector corresponding to each debugging text, acquiring a common measurement coefficient between the debugged text knowledge vectors corresponding to each debugging text, obtaining a debugged commonality measurement coefficient set, obtaining commonality measurement coefficients among the to-be-debugged text knowledge vectors corresponding to each debug text, obtaining the to-be-debugged commonality measurement coefficient set, obtaining a cost value between the to-be-debugged commonality measurement coefficient set and the debugged commonality measurement coefficient set, optimizing a basic text knowledge vector mining SN through the cost value, then jumping to the step of obtaining a current debug text sequence to roll until meeting the debugging cut-off requirement, determining the debugged basic text knowledge vector mining SN as a target text knowledge vector mining SN, screening texts in the user intention text set based on the target text knowledge vector mining SN, analyzing again, and storing texts with analysis results larger than a preset value.

According to another aspect of the present application, there is also provided a processing apparatus, referring to fig. 3, an apparatus 900 includes:

The text obtaining module 910 is configured to obtain a current debug text sequence, where the current debug text sequence is obtained by obtaining the current debug text sequence from a debug text sequence deployed in advance;

The network calling module 920 is configured to load each debug text in the current debug text sequence into a debugged text knowledge mining network and a basic text knowledge mining network to be debugged respectively to perform text knowledge vector mining, so as to obtain a debugged text knowledge vector corresponding to each debug text and a to-be-debugged text knowledge vector corresponding to each debug text, where the basic text knowledge mining network is obtained by pre-configuring network configuration variables of the debugged text knowledge mining network;

The commonality determining module 930 is configured to obtain commonality measurement coefficients between debugged text knowledge vectors corresponding to each debug text, obtain a debugged commonality measurement coefficient set, and obtain commonality measurement coefficients between to-be-debugged text knowledge vectors corresponding to each debug text, and obtain a to-be-debugged commonality measurement coefficient set;

the network debugging module 940 is configured to obtain a cost value between the set of to-be-debugged commonality measurement coefficients and the set of debugged commonality measurement coefficients, optimize a to-be-debugged basic text knowledge mining network through the cost value, and then jump to a step of obtaining a current debugging text sequence to roll until meeting a debugging cut-off requirement, determine the debugged basic text knowledge mining network as a first target text knowledge mining network, where the first target text knowledge mining network is configured to mine text knowledge vectors of a visual management text of a target device; and carrying out text intention recognition by visually managing text knowledge vectors of the text through the target equipment.

There is also provided, in accordance with an embodiment of the present application, an electronic device (i.e., an AI processing system), a readable storage medium, and a computer program product.

Referring to fig. 4, which is a block diagram of the electronic device 1000 of the AI processing system of the present application, the electronic device 1000 includes a computing unit 1001 that can perform various appropriate actions and processes according to a computer program stored in a Read Only Memory (ROM) 1002 or a computer program loaded from a storage unit 1008 into a Random Access Memory (RAM) 1003. In the RAM 1003, various programs and data required for the operation of the electronic apparatus 1000 can also be stored. The computing unit 1001, the ROM 1002, and the RAM 1003 are connected to each other by a bus 1004. An input/output (I/O) interface 1005 is also connected to bus 1004.

Various components in the electronic device 1000 are connected to the I/O interface 1005, including: an input unit 1006, an output unit 1007, a storage unit 1008, and a communication unit 10010. The input unit 1006 may be any type of device capable of inputting information to the electronic device 1000, the input unit 1006 may receive input numeric or character information and generate key signal inputs related to user settings and/or function control of the electronic device, and may include, but is not limited to, a mouse, a keyboard, a touch screen, a trackpad, a trackball, a joystick, a microphone, and/or a remote control. The output unit 1007 may be any type of device capable of presenting information and may include, but is not limited to, a display, speakers, video/audio output terminals, vibrators, and/or printers. Storage unit 1008 may include, but is not limited to, magnetic disks, optical disks. The communication unit 10010 allows the electronic device 1000 to exchange information/data with other devices through a computer network, such as the internet, and/or various telecommunications networks, and may include, but is not limited to, modems, network cards, infrared communication devices, wireless communication transceivers and/or chipsets, such as bluetooth (TM) devices, 802.11 devices, wiFi devices, wiMax devices, cellular communication devices, and/or the like.

The computing unit 1001 may be a variety of general and/or special purpose processing components having processing and computing capabilities. Some examples of computing unit 1001 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various specialized Artificial Intelligence (AI) computing chips, various computing units running machine learning model algorithms, a Digital Signal Processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 1001 performs the various methods and processes described above, such as method 200. For example, in some embodiments, the method 200 may be implemented as a computer software program tangibly embodied on a machine-readable medium, such as the storage unit 1008. In some embodiments, part or all of the computer program may be loaded and/or installed onto the electronic device 1000 via the ROM 1002 and/or the communication unit 1009. One or more of the steps of the method 200 described above may be performed when the computer program is loaded into RAM 1003 and executed by the computing unit 1001. Alternatively, in other embodiments, the computing unit 1001 may be configured to perform the method 200 in any other suitable way (e.g., by means of firmware).

Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuit systems, field Programmable Gate Arrays (FPGAs), application Specific Integrated Circuits (ASICs), application Specific Standard Products (ASSPs), systems On Chip (SOCs), complex Programmable Logic Devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs, the one or more computer programs may be executed and/or interpreted on a programmable system including at least one programmable processor, which may be a special purpose or general-purpose programmable processor, that may receive data and instructions from, and transmit data and instructions to, a storage system, at least one input device, and at least one output device.

Program code for carrying out methods of the present application may be written in any combination of one or more programming languages. These program code may be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus such that the program code, when executed by the processor or controller, causes the functions/operations specified in the flowchart and/or block diagram to be implemented. The program code may execute entirely on the machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.

In the context of the present application, a machine-readable medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. The machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and pointing device (e.g., a mouse or trackball) by which a user can provide input to the computer. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic input, speech input, or tactile input.

The systems and techniques described here can be implemented in a computing system that includes a background component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such background, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), wide Area Networks (WANs), and the internet.

The computer system may include a client and a server. The client and server are typically remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server may be a cloud server, a server of a distributed system, or a server incorporating a blockchain.

It should be appreciated that various forms of the flows shown above may be used to reorder, add, or delete steps. For example, the steps recited in the present application may be performed in parallel, sequentially or in a different order, provided that the desired results of the disclosed embodiments are achieved, and are not limited herein.

Although embodiments or examples of the present application have been described with reference to the accompanying drawings, it is to be understood that the above-described methods, systems, and apparatus are merely illustrative embodiments or examples, and that the scope of the present application is not limited by these embodiments or examples but only by the claims following the grant and their equivalents. Various elements of the embodiments or examples may be omitted or replaced with equivalent elements thereof. Furthermore, the steps may be performed in a different order than described in the present application. Further, various elements of the embodiments or examples may be combined in various ways. It is important that as technology evolves, many of the elements described herein may be replaced by equivalent elements that appear after the application.

Claims (8)

1. A method of visual processing of a digital plant, applied to an AI processing system, the method comprising:

Acquiring a current debugging text sequence, wherein the current debugging text sequence is acquired from a debugging text sequence deployed in advance;

loading each debug text in the current debug text sequence into a debugged text knowledge mining network and a basic text knowledge mining network to be debugged respectively for text knowledge vector mining, and obtaining a debugged text knowledge vector corresponding to each debug text and a to-be-debugged text knowledge vector corresponding to each debug text, wherein the basic text knowledge mining network is obtained by pre-configuring network configuration variables of the debugged text knowledge mining network;

Obtaining common measurement coefficients among the debugged text knowledge vectors corresponding to the debugged texts, obtaining a debugged common measurement coefficient set, and obtaining common measurement coefficients among the to-be-debugged text knowledge vectors corresponding to the debugged texts, thus obtaining a to-be-debugged common measurement coefficient set;

Obtaining a cost value between the to-be-debugged common measurement coefficient set and the debugged common measurement coefficient set, optimizing the to-be-debugged basic text knowledge mining network through the cost value, and then jumping to a step of obtaining a current debugging text sequence to roll until meeting the debugging cut-off requirement, determining the debugged basic text knowledge mining network as a first target text knowledge mining network, wherein the first target text knowledge mining network is configured to mine text knowledge vectors of visual management texts of target equipment;

performing text intention recognition through the text knowledge vector of the visual management text of the target equipment;

The obtaining the commonality measurement coefficient between the debugged text knowledge vectors corresponding to each debug text, obtaining a debugged commonality measurement coefficient set, includes:

Obtaining a debugged text knowledge array through the debugged text knowledge vectors corresponding to the debugged texts, and performing standardized operation on the debugged text knowledge array to obtain a debugged standardized array;

Acquiring a turnover array corresponding to the debugged standardized array, and acquiring a debugged turnover array;

obtaining the debugged commonality measurement coefficient set through the debugged flip array and the debugged standardized array;

The obtaining the commonality measurement coefficient between the knowledge vectors of the text to be debugged corresponding to each debug text, obtaining the commonality measurement coefficient set to be debugged, includes:

Obtaining a to-be-debugged text knowledge array through to-be-debugged text knowledge vectors corresponding to the respective debug texts, and performing standardization operation on the to-be-debugged text knowledge array to obtain a to-be-debugged standardization array;

Acquiring a turnover array corresponding to the to-be-debugged standardized array, and acquiring a to-be-debugged turnover array;

And obtaining the to-be-debugged commonality measurement coefficient set through the to-be-debugged overturn array and the to-be-debugged standardization array.

2. The method of claim 1, wherein the steps of obtaining a cost value between the set of co-metric coefficients to be debugged and the set of debugged co-metric coefficients, optimizing the underlying text knowledge mining network to be debugged by the cost value, and then jumping to obtain a current debug text sequence for rolling comprise:

Obtaining standard deviation values of the to-be-debugged commonality measurement coefficient set and the debugged commonality measurement coefficient set, obtaining a basic error result, and determining the basic error result as the cost value;

The cost value is reversely propagated to optimize network configuration variables in the basic text knowledge mining network, and an optimized text knowledge mining network is obtained;

determining the optimized text knowledge mining network as a basic text knowledge mining network, and then jumping to a step of acquiring a current debugging text sequence for rolling;

the obtaining the standard deviation value of the to-be-debugged commonality measurement coefficient set and the debugged commonality measurement coefficient set to obtain the cost value includes:

Obtaining standard deviation values of the to-be-debugged commonality measurement coefficient set and the debugged commonality measurement coefficient set, and obtaining a basic error result;

Obtaining the number of debugging texts corresponding to the current debugging text sequence, and obtaining the proportional coefficient of the basic error result and the number of the debugging texts to obtain a mean error result;

Acquiring a preset adjusting variable, and adjusting and calculating the mean error result through the preset adjusting variable to acquire an adjusting error result;

and acquiring an intention recognition error result corresponding to the basic text knowledge mining network to be debugged, and acquiring error results and values of the intention recognition error result and the adjustment error result to obtain the cost value.

3. The method according to claim 1, wherein the step of obtaining the cost value between the set of to-be-debugged commonality measurement coefficients and the set of debugged commonality measurement coefficients, optimizing the to-be-debugged underlying text knowledge mining network by the cost value, and then jumping to the step of obtaining the current debugged text sequence to perform rolling until meeting a debug deadline requirement, determining the debugged underlying text knowledge mining network as a first target text knowledge mining network comprises:

Loading the to-be-debugged commonality measurement coefficient set into a basic transformation module to perform coefficient transformation to obtain a target transformation commonality measurement coefficient set;

Obtaining standard deviation values between the target transformation commonality measurement coefficient set and the debugged commonality measurement coefficient set, obtaining a target cost value, and reversely propagating through the target cost value to optimize the basic transformation module and the basic text knowledge mining network, and obtaining an optimized transformation module and an optimized text knowledge mining network;

And determining the optimized transformation module as a basic transformation module, determining the optimized text knowledge mining network as a basic text knowledge mining network, and then jumping to the step of acquiring the current debugging text sequence to roll until meeting the debugging cut-off requirement, and obtaining a second target text knowledge mining network through the debugged basic text knowledge mining network and the debugged basic transformation module.

4. The method of claim 1, wherein the underlying text knowledge mining network is an underlying mining network;

The step of obtaining the cost value between the to-be-debugged commonality measurement coefficient set and the debugged commonality measurement coefficient set, optimizing the to-be-debugged basic text knowledge mining network through the cost value, and then jumping to the step of obtaining the current debugging text sequence to roll until meeting the debugging cut-off requirement, determining the debugged basic text knowledge mining network as a first target text knowledge mining network, wherein the step of obtaining the current debugging text sequence comprises the following steps of:

Loading the to-be-debugged commonality measurement coefficient set and the debugged commonality measurement coefficient set into a basic identification module for identification to obtain a commonality measurement coefficient identification result;

And optimizing the basic recognition module and the basic text knowledge mining network through the commonality measurement coefficient recognition result, and then jumping to the step of acquiring the current debugging text sequence to roll until the current debugging text sequence meets the debugging cut-off requirement, and determining the debugged basic text knowledge mining network as a third target text knowledge mining network.

5. The method of claim 1, wherein each set of multiple text is included in the current debug text sequence, the sets of multiple text including a set of combined text;

the method further comprises the steps of:

Loading each multi-element text set into a debugged text knowledge mining network and a basic text knowledge mining network to be debugged respectively to carry out text knowledge vector mining, so as to obtain debugged multi-element set knowledge vectors corresponding to each multi-element text set and to-be-debugged multi-element set knowledge vectors corresponding to each multi-element text set;

And acquiring a multi-element set cost value through the debugged multi-element set knowledge vector and the multi-element set knowledge vector to be debugged, obtaining a basic multi-element set error result, reversely spreading the basic multi-element set error result to optimize the basic text knowledge mining network, and then jumping to the step of acquiring the current debugging text sequence to roll until the current debugging text sequence meets the debugging cut-off requirement, and determining the debugged basic text knowledge mining network as a fourth target text knowledge mining network.

6. The method of claim 1, wherein after the step of obtaining the cost value between the set of to-be-debugged commonality metric coefficients and the set of debugged commonality metric coefficients and optimizing the to-be-debugged underlying text knowledge mining network by the cost value, and then jumping to the step of obtaining the current debugged text sequence to roll until meeting a debug deadline requirement, determining the debugged underlying text knowledge mining network as a first target text knowledge mining network, further comprises:

Acquiring a target device visual management text, loading the target device visual management text into the first target text knowledge mining network to mine text knowledge vectors, and acquiring a to-be-recognized text knowledge vector;

acquiring a historical intent text knowledge vector corresponding to a visual management text set of a historical intent device, and acquiring a commonality measurement coefficient of the text knowledge vector to be identified and the historical intent text knowledge vector;

And determining text intention identification information corresponding to the target device visual management text through the commonality measurement coefficient.

7. The method according to any one of claims 1-6, further comprising:

acquiring a text to be analyzed and an analyzed text sequence;

Loading the text to be analyzed and the text sequence to be analyzed into a debugged text knowledge mining network to perform text knowledge vector mining, obtaining a text knowledge vector to be analyzed corresponding to the text to be analyzed and a text knowledge vector to be analyzed corresponding to the text sequence to be analyzed, and obtaining a common measurement coefficient of the text knowledge vector to be analyzed and the text knowledge vector to be analyzed, so as to obtain a first common measurement coefficient set;

Loading the to-be-analyzed text and the analyzed text sequence into a first target text knowledge mining network for text knowledge vector mining, obtaining a to-be-analyzed target text knowledge vector corresponding to the to-be-analyzed text and an analyzed text knowledge vector set corresponding to the analyzed text sequence, and obtaining a commonality measurement coefficient of the to-be-analyzed target text knowledge vector and the analyzed text knowledge vector set to obtain a second commonality measurement coefficient set, wherein the first target text knowledge mining network is obtained by performing migration learning and debugging based on a debugged text knowledge mining network;

Analyzing through the first commonality measurement coefficient set and the second commonality measurement coefficient set to obtain an analysis result corresponding to the to-be-analyzed text, and determining the commonality measurement analysis result corresponding to the to-be-analyzed text through the analysis result corresponding to the to-be-analyzed text;

The obtaining the commonality measurement coefficient of the text knowledge vector to be analyzed and the text knowledge vector to be analyzed, obtaining a first commonality measurement coefficient set, including:

Normalizing the text knowledge vector to be analyzed to obtain a normalized text knowledge vector to be analyzed, and normalizing the text knowledge vector to be analyzed to obtain a normalized text knowledge vector to be analyzed;

Performing inversion operation on the standardized analyzed text knowledge vector to obtain an analyzed inversion array, and obtaining the first common measurement coefficient set through the product of the standardized analyzed text knowledge vector and the analyzed inversion array;

The obtaining the commonality measurement coefficient of the target text knowledge vector to be analyzed and the analyzed text knowledge vector set, obtaining a second commonality measurement coefficient set, includes:

normalizing the target text knowledge vector to be analyzed to obtain a normalized target text knowledge vector to be analyzed, and normalizing the text knowledge vector set to obtain a normalized text knowledge vector set to be analyzed;

Performing overturn operation on the standardized analyzed text knowledge vector set to obtain an analyzed object overturn array, and obtaining the second commonality measurement coefficient set through the standardized analyzed text knowledge vector set and the analyzed object overturn array;

Analyzing through the first commonality measurement coefficient set and the second commonality measurement coefficient set to obtain an analysis result corresponding to the to-be-analyzed text, and determining the commonality measurement analysis result corresponding to the to-be-analyzed text through the analysis result corresponding to the to-be-analyzed text, including:

Obtaining a standard deviation value between the first commonality measurement coefficient set and the second commonality measurement coefficient set to obtain a target error result;

determining the number of texts corresponding to the text to be analyzed and the analyzed text sequence, acquiring a proportionality coefficient of the target error result and the number of texts, and determining an analysis result corresponding to the text to be analyzed through the proportionality coefficient;

And if the analysis result is larger than a preset value, obtaining a matching analysis indication result corresponding to the text to be analyzed.

8. An AI processing system, comprising:

At least one processor;

and a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-7.