Software fault feedback processing method, device and equipment and readable storage medium

Abstract

The application relates to natural language processing technology and discloses a software fault feedback processing method, which comprises the following steps: performing entity extraction on fault feedback text to obtain first fault entities; performing text recognition on fault feedback images to obtain image text; performing entity extraction on the image text to obtain second fault entities; identifying a fault category corresponding to the fault feedback text; screening all processing method knowledge bases by using the fault category to obtain a target processing method knowledge base; screening processing method text in the target processing method knowledge base by using the first fault entities and the second fault entities to obtain target processing method text; and sending the target processing method text to a preset terminal device. The application also relates to a blockchain technology, and the processing method knowledge base can be deployed in the blockchain. The application further provides a software fault feedback processing device, an electronic device and a readable storage medium. The application can improve the accuracy of software fault feedback processing.

Description

Technical Field

[0001] This invention relates to the field of natural language processing, and more particularly to a software fault feedback processing method, apparatus, electronic device, and readable storage medium. Background Technology

[0002] With the development of internet technology, applications are becoming increasingly indispensable in people's lives. However, when users report problems with the software, how to handle these reports, provide users with solutions, and help developers identify the corresponding problem categories have become increasingly important issues.

[0003] Current feedback handling methods can only match user fault feedback text information within a single processing method knowledge base. This single matching dimension results in low accuracy in handling software fault feedback. Summary of the Invention

[0004] This invention provides a software fault feedback processing method, apparatus, electronic device, and computer-readable storage medium, the main purpose of which is to improve the accuracy of software fault feedback processing.

[0005] To achieve the above objectives, the present invention provides a software fault feedback processing method, comprising:

[0006] Obtain the fault feedback text and fault feedback image corresponding to the same fault feedback in the software;

[0007] The fault feedback text is subjected to fault category identification to obtain the target fault category;

[0008] Entity extraction is performed on the fault feedback text to obtain the first fault entity;

[0009] The error reporting area in the fault feedback image is segmented, and the text in the segmented image is identified to obtain the image text;

[0010] Entity extraction is performed on the image text to obtain the second fault entity;

[0011] One or more processing method knowledge bases are obtained, wherein each processing method knowledge base has a corresponding knowledge base tag, and the similarity between the target fault category and the knowledge base tag is used to filter all processing method knowledge bases to obtain the target processing method knowledge base;

[0012] The first fault entity and the second fault entity are used as filtering conditions to filter the processing method text in the target processing method knowledge base to obtain the target processing method text.

[0013] The target processing method text is sent to a preset terminal device.

[0014] Optionally, the step of identifying the target fault category from the fault feedback text includes:

[0015] The trained BERT model is used to extract features from the fault feedback text to obtain a text feature vector;

[0016] The softmax function is used to calculate the recognition probability of different fault categories.

[0017] The fault category corresponding to the highest recognition probability is identified as the target fault category;

[0018] Optionally, the step of segmenting the error reporting area in the fault feedback image and identifying the text in the segmented image to obtain image text includes:

[0019] The error reporting area in the fault feedback image is cropped to obtain the cropped image;

[0020] The cropped image is converted to grayscale to obtain a grayscale image;

[0021] The grayscale image is filtered to obtain a standard image;

[0022] The standard image is processed by a preset text extraction algorithm to obtain the image text.

[0023] Optionally, the step of filtering all the processing method knowledge bases based on the similarity between the target fault category and the knowledge base tags to obtain the target processing method knowledge base includes:

[0024] The target fault categories are converted into vectors to obtain category vectors;

[0025] The knowledge base tags are converted into vectors to obtain tag vectors;

[0026] Calculate the similarity between the category vector and the label vector to obtain a similarity value;

[0027] The target processing method knowledge base is obtained by filtering all the processing method knowledge bases based on the similarity value.

[0028] Optionally, converting the knowledge base tags into vectors to obtain tag vectors includes:

[0029] Each character in the knowledge base tags is converted into a vector to obtain a character vector;

[0030] Combine all the character vectors according to the order of the corresponding characters in the knowledge base tags to obtain the knowledge base tag matrix;

[0031] The knowledge base tag matrix is ​​convolved using a preset convolution kernel to obtain a convolution matrix;

[0032] The dimensionality of the convolution matrix is ​​reduced to obtain the label vector.

[0033] Optionally, the step of reducing the dimensionality of the convolution matrix to obtain the label vector includes:

[0034] Calculate the average value of each column element in the convolution matrix to obtain the column feature value;

[0035] The label vector is obtained by replacing each column in the convolution matrix with the corresponding column feature value.

[0036] To address the above problems, the present invention also provides a software fault feedback processing device, the device comprising:

[0037] The text processing module is used to acquire the fault feedback text and fault feedback image corresponding to the same fault feedback in the software; to perform fault category identification on the fault feedback text to obtain the target fault category; and to perform entity extraction on the fault feedback text to obtain the first fault entity.

[0038] The image processing module is used to segment the error area in the fault feedback image, identify the text in the segmented image to obtain image text, and perform entity extraction on the image text to obtain a second fault entity.

[0039] The processing method filtering module is used to acquire one or more processing method knowledge bases, wherein each processing method knowledge base has a corresponding knowledge base tag, and to filter all processing method knowledge bases by using the similarity between the target fault category and the knowledge base tag to obtain a target processing method knowledge base; to use the first fault entity and the second fault entity as filtering conditions to filter the processing method text in the target processing method knowledge base to obtain the target processing method text; and to send the target processing method text to a preset terminal device.

[0040] Optionally, the step of reducing the dimensionality of the convolution matrix to obtain the label vector includes:

[0041] Calculate the average value of each column element in the convolution matrix to obtain the column feature value;

[0042] The label vector is obtained by replacing each column in the convolution matrix with the corresponding column feature value.

[0043] To address the above problems, the present invention also provides an electronic device, the electronic device comprising:

[0044] Memory, storing at least one computer program; and

[0045] The processor executes the computer program stored in the memory to implement the software fault feedback processing method described above.

[0046] To address the aforementioned problems, the present invention also provides a computer-readable storage medium storing at least one computer program, which is executed by a processor in an electronic device to implement the software fault feedback processing method described above.

[0047] This invention utilizes the similarity between the target fault category and the knowledge base tags to filter all processing method knowledge bases, obtaining a target processing method knowledge base. The similarity between the target fault category and the knowledge base tags narrows the filtering range of the processing method knowledge base. Furthermore, the first fault entity and the second fault entity are used as filtering conditions to filter the processing method text in the target processing method knowledge base. By utilizing both text and image dimensions, the corresponding processing methods for matching fault entities are obtained, resulting in more diverse matching dimensions and thus more accurate filtering of processing methods, thereby improving the accuracy of software fault feedback processing. Therefore, the software fault feedback processing method, apparatus, electronic device, and computer-readable storage medium proposed in this invention improve the accuracy of software fault feedback processing. Attached Figure Description

[0048] Figure 1 A flowchart illustrating a software fault feedback processing method provided in an embodiment of the present invention;

[0049] Figure 2 This is a schematic diagram of a software fault feedback processing device provided in an embodiment of the present invention;

[0050] Figure 3 This is a schematic diagram of the internal structure of an electronic device that implements a software fault feedback processing method according to an embodiment of the present invention.

[0051] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0052] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0053] This invention provides a software fault feedback processing method. The executing entity of the software fault feedback processing method includes, but is not limited to, at least one of the following: a server, a terminal, or other electronic devices that can be configured to execute the method provided in this application embodiment. In other words, the software fault feedback processing method can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster.

[0054] Reference Figure 1 The flowchart shown is a schematic diagram of a software fault feedback processing method provided in an embodiment of the present invention. In this embodiment, the software fault feedback processing method includes:

[0055] S1. Obtain the fault feedback text and fault feedback image corresponding to the same fault feedback in the software.

[0056] In this embodiment of the invention, the fault feedback text is a textual description of the software problem reported by the user, and the fault feedback image is a screenshot of the problem corresponding to the software problem reported by the user; further, in this embodiment of the invention, the fault feedback text and the corresponding fault feedback image can be obtained through the fault feedback text and fault feedback image input module of the software or webpage.

[0057] S2. Perform fault category identification on the fault feedback text to obtain the target fault category;

[0058] In this embodiment of the invention, the fault feedback text is a question that the user wants to inquire about or provide feedback on, such as "Why can't the software play in landscape mode?" or "Why can't the software connect to the internet?" The fault feedback text can be obtained through the text input module of a webpage.

[0059] Furthermore, in this embodiment of the invention, in order to better match the corresponding processing method to the fault feedback, the fault feedback text is subjected to fault category identification to obtain the target fault category.

[0060] For example, the target's fault category is "network connection failure" or "playback stuttering failure," etc. This embodiment of the invention does not limit the fault category.

[0061] In detail, in this embodiment of the invention, the fault feedback text is subjected to fault category identification to obtain the target fault category, including:

[0062] The trained BERT model is used to extract features from the fault feedback text to obtain a text feature vector;

[0063] Calculate the text feature vector using the softmax function to obtain the recognition probabilities of preset different fault categories;

[0064] Identify the fault category corresponding to the maximum recognition probability as the target fault category;

[0065] Further, before the method for identifying the fault category of the fault feedback text in the embodiments of the present invention, the method further includes:

[0066] Step A: Obtain a historical fault text set, mark the fault categories of the historical fault text set to obtain a training set;

[0067] In the embodiments of the present invention, the historical fault text set is a set of different historical fault texts, and the historical fault texts can be text data with different contents but of the same type as the fault text. Further, in the embodiments of the present invention, the fault categories of each historical fault text in the historical fault text set are marked to obtain the training set.

[0068] Step B: Use the training set to iteratively train a pre-constructed BERT model until the BERT model converges to obtain a trained BERT model.

[0069] By identifying the fault category of the fault text, the screening range of subsequent processing methods is narrowed, and the screening efficiency is improved.

[0070] S3. Perform entity extraction on the fault feedback text to obtain a first fault entity;

[0071] In the embodiments of the present invention, in order to locate the fault problem in the fault feedback text, entity extraction is performed on the fault feedback text to obtain a first fault entity.

[0072] Optionally, in the embodiments of the present invention, the entity extraction process includes: using named entity recognition technology (Named Entity Recognition, abbreviated as NER) to recognize the segmented fault feedback text to obtain a first fault entity. Among them, using named entity recognition technology can identify entities with specific meanings in the fault feedback text (such as, person names, place names, organization names, proper nouns, etc.) to represent corresponding faults. For example: the fault feedback text is "The network connection is interrupted and cannot be connected", and the segmented fault feedback text is "The network connection is interrupted", "le", "cannot", "be connected". Using the named entity recognition technology to recognize the fault feedback text can obtain the first fault entity as "The network connection is interrupted".

[0073] S4. Segment the error reporting area in the fault feedback image and recognize the text in the segmented image to obtain image text;

[0074] In this embodiment of the invention, the fault feedback image is a screenshot of a software problem reported by a user. The information in the fault feedback image is more helpful in identifying the problem reported by the user. Therefore, the fault feedback image is preprocessed and text is recognized to obtain the image text.

[0075] In detail, in order to remove the influence of irrelevant areas in the fault feedback image, the error reporting area in the fault feedback image is cropped to obtain a cropped image. Preferably, in this embodiment of the invention, the error reporting area is the error text display area of ​​the fault feedback image.

[0076] In detail, since the key regions may have different colors, in order to reduce the amount of data, storage space, and image processing time, the key regions are grayscale processed. Furthermore, since there is image noise in the key regions, in order to reduce the impact of image noise on subsequent processing, the embodiments of the present invention perform filtering processing on the key regions. Preferably, the embodiments of the present invention use a median filtering algorithm to filter the cropped image.

[0077] Furthermore, in order to obtain the text information in the cropped image, this embodiment of the invention employs a text extraction algorithm to perform text extraction processing on the cropped image, so as to extract the text from the cropped image. In one embodiment of the invention, the text extraction algorithm can be a known OCR (Optical Character Recognition) algorithm.

[0078] Therefore, in summary, in this embodiment of the invention, the preprocessing and text recognition of the fault feedback image includes: cropping the error reporting area in the fault feedback image to obtain a cropped image; performing grayscale processing on the cropped image to obtain a grayscale image; performing filtering processing on the grayscale image to obtain the standard image; and performing text recognition processing on the standard image using a preset text extraction algorithm to obtain the image text.

[0079] S5. Extract entities from the image text to obtain the second fault entity;

[0080] In this embodiment of the invention, in order to locate the fault in the image text, entity extraction is performed on the image text to obtain a second fault entity.

[0081] Furthermore, in this embodiment of the invention, the entity extraction method in this step is the same as that in S3, and will not be described again here.

[0082] S6. Obtain one or more processing method knowledge bases, wherein each processing method knowledge base has a corresponding knowledge base tag, and filter all processing method knowledge bases by using the similarity between the target fault category and the knowledge base tag to obtain the target processing method knowledge base.

[0083] In this embodiment of the invention, the processing method knowledge base is a database composed of pre-constructed software knowledge with defined functional categories. For example, the software covers multiple functional categories, including video playback, network connection, and video download. In this embodiment, knowledge related to video playback is summarized and stored in the database to obtain a video playback function processing method knowledge base; knowledge related to network connection is summarized and stored in the database to obtain a network connection function processing method knowledge base; and knowledge related to video download is summarized and stored in the database to obtain a video download function processing method knowledge base. The knowledge base tags are short texts used to identify the knowledge categories in the knowledge base.

[0084] Furthermore, in this embodiment of the invention, there are one or more processing method knowledge bases. In order to accurately locate the database where the fault processing method corresponding to the fault feedback is located, the similarity between the target fault category and the knowledge base tags is used to filter all the processing method knowledge bases to obtain the target processing method knowledge base.

[0085] Specifically, in this embodiment of the invention, the similarity between the target fault category and the knowledge base tags is used to filter all the processing method knowledge bases to obtain a target processing method knowledge base, including:

[0086] The target fault categories are converted into vectors to obtain category vectors;

[0087] The knowledge base tags are converted into vectors to obtain tag vectors;

[0088] Calculate the similarity between the category vector and the label vector to obtain a similarity value;

[0089] Specifically, in this embodiment of the invention, algorithms such as Euclidean distance, Manhattan distance, and cosine similarity can be used to calculate the similarity between the category vector and the label vector.

[0090] The target processing method knowledge base is obtained by filtering all the processing method knowledge bases based on the similarity value.

[0091] Specifically, in this embodiment of the invention, the tag vector corresponding to the maximum similarity value is determined as the target tag vector; the knowledge base tag corresponding to the target tag vector is determined as the target knowledge base tag; and the processing method knowledge base corresponding to the target knowledge base tag is determined as the target processing method knowledge base.

[0092] Furthermore, in this embodiment of the invention, the knowledge base tags are converted into vectors to obtain tag vectors, including:

[0093] Each character in the knowledge base tags is converted into a vector to obtain a character vector;

[0094] Combine all the character vectors according to the order of the corresponding characters in the knowledge base tags to obtain the knowledge base tag matrix;

[0095] The knowledge base tag matrix is ​​convolved using a preset convolution kernel to obtain a convolution matrix;

[0096] The dimensionality of the convolution matrix is ​​reduced to obtain the label vector.

[0097] In detail, in this embodiment of the invention, the dimensionality reduction of the convolution matrix to obtain the label vector includes:

[0098] Calculate the average value of each column element in the convolution matrix to obtain the column feature value;

[0099] The label vector is obtained by replacing each column in the convolution matrix with the corresponding column feature value.

[0100] In another embodiment of the present invention, the knowledge base of the processing method is composed of domain-specific knowledge and can be deployed on a blockchain.

[0101] In this embodiment of the invention, both the target fault category and the knowledge base label are text. Furthermore, this embodiment of the invention does not limit the method of converting text into vectors.

[0102] S7. Use the first fault entity and the second fault entity as filtering conditions to filter the processing method text in the target processing method knowledge base to obtain the target processing method text.

[0103] Specifically, in this embodiment of the invention, the first fault entity and the second fault entity are used as query conditions to query the processing methods in the target processing method knowledge base to obtain the target processing method, including:

[0104] The first faulty entity is converted into a vector to obtain the first entity vector;

[0105] The second fault entity is transformed into a vector to obtain the second entity vector;

[0106] Each processing method text in the target processing method knowledge base is converted into a vector to obtain a processing method vector;

[0107] Calculate the correlation between the first entity vector and the processing method vector to obtain the first correlation.

[0108] Optionally, in embodiments of the present invention, algorithms such as Euclidean distance, Manhattan distance, and cosine similarity can be used to calculate the correlation between the first entity vector and the processing method vector.

[0109] Calculate the correlation between the second entity vector and the processing method vector to obtain the second correlation.

[0110] Based on preset weighting coefficients, the matching coefficient corresponding to each processing method vector is obtained by weighting the first relevance and the second relevance.

[0111] The matching coefficient is used to filter the processing method text in the target processing method knowledge base to obtain the target processing method text.

[0112] In detail, in this embodiment of the invention, based on a preset weighting coefficient, a weighted calculation is performed according to the first relevance and the second relevance corresponding to each processing method vector to obtain the matching coefficient corresponding to the processing method vector, including:

[0113] Extract the first similarity weighting coefficient from the weighting coefficients, calculate the product of the first similarity coefficient and the first similarity, and obtain the first similarity weighting value;

[0114] Extract the second similarity weighting coefficient from the weighting coefficients, calculate the product of the second similarity coefficient and the second similarity, and obtain the second similarity weighted value;

[0115] The first similarity weighted value and the second similarity weighted value corresponding to each processing method vector are added together to obtain the matching coefficient corresponding to the processing method vector.

[0116] Specifically, in this embodiment of the invention, the matching coefficient and a preset matching coefficient threshold are used to filter the processing method text in the target processing method knowledge base to obtain the target processing method text, including:

[0117] The processing method vector corresponding to the maximum matching coefficient is determined as the target processing method vector;

[0118] The processing method text corresponding to the target processing method vector is determined as the target processing method text.

[0119] S8. Send the target processing method text to a preset terminal device.

[0120] In this embodiment of the invention, the target processing method text is sent to a preset terminal device. The terminal device can be the terminal device that issues the fault feedback, including: a mobile phone, computer, tablet and other smart terminals.

[0121] Furthermore, in this embodiment of the invention, the target processing method text is pushed to the user, and feedback information on whether the user approves of the target processing method text is obtained. Based on the feedback information, the data in the processing method knowledge base is updated accordingly to achieve error correction processing of the processing method knowledge base.

[0122] like Figure 2 The diagram shown is a functional block diagram of the software fault feedback processing device of the present invention.

[0123] The software fault feedback processing device 100 of the present invention can be installed in an electronic device. Depending on the functions implemented, the software fault feedback processing device may include a text processing module 101, an image processing module 102, and a processing method filtering module 103. The module mentioned in the present invention may also be referred to as a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, and are stored in the memory of the electronic device.

[0124] In this embodiment, the functions of each module / unit are as follows:

[0125] The text processing module 101 is used to acquire fault feedback text and fault feedback image corresponding to the same fault feedback in the software; to perform fault category identification on the fault feedback text to obtain the target fault category; and to perform entity extraction on the fault feedback text to obtain the first fault entity.

[0126] The image processing module 102 is used to segment the error area in the fault feedback image, identify the text in the segmented image, and obtain image text; and to extract entities from the image text to obtain a second fault entity.

[0127] The processing method filtering module 103 is used to acquire one or more processing method knowledge bases, wherein each processing method knowledge base has a corresponding knowledge base tag, and to filter all processing method knowledge bases by using the similarity between the target fault category and the knowledge base tag to obtain a target processing method knowledge base; to use the first fault entity and the second fault entity as filtering conditions to filter the processing method text in the target processing method knowledge base to obtain the target processing method text; and to send the target processing method text to a preset terminal device.

[0128] In detail, each module in the software fault feedback processing device 100 described in this embodiment of the invention adopts the same characteristics as described above during use. Figure 1 The same technical means are used in the software fault feedback handling method described in the article, and can produce the same technical effect, so they will not be repeated here.

[0129] like Figure 3 The diagram shown is a structural schematic of an electronic device that implements the software fault feedback processing method of the present invention.

[0130] The electronic device 1 may include a processor 10, a memory 11 and a bus, and may also include a computer program, such as a software fault feedback handler 12, stored in the memory 11 and executable on the processor 10.

[0131] The memory 11 includes at least one type of readable storage medium, such as flash memory, portable hard drive, multimedia card, card-type memory (e.g., SD or DX memory), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 11 can be an internal storage unit of the electronic device 1, such as a portable hard drive. In other embodiments, the memory 11 can be an external storage device of the electronic device 1, such as a plug-in portable hard drive, Smart Media Card (SMC), Secure Digital (SD) card, or Flash Card. Furthermore, the memory 11 can include both internal and external storage units of the electronic device 1. The memory 11 can be used not only to store application software and various types of data installed on the electronic device 1, such as the code of software fault feedback handling programs, but also to temporarily store data that has been output or will be output.

[0132] In some embodiments, the processor 10 may be composed of integrated circuits, such as a single packaged integrated circuit or multiple integrated circuits with the same or different functions, including combinations of one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and various control chips. The processor 10 is the control unit of the electronic device, connecting various components of the entire electronic device through various interfaces and lines. It executes programs or modules (such as software fault feedback processing programs) stored in the memory 11, and calls data stored in the memory 11 to perform various functions of the electronic device 1 and process data.

[0133] The bus can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This bus can be divided into an address bus, a data bus, a control bus, etc. The bus is configured to enable communication between the memory 11 and at least one processor 10, etc.

[0134] Figure 3 Only electronic devices with components are shown; it will be understood by those skilled in the art that... Figure 3 The structure shown does not constitute a limitation on the electronic device 1, and may include fewer or more components than shown, or combine certain components, or have different component arrangements.

[0135] For example, although not shown, the electronic device 1 may also include a power supply (such as a battery) to power various components. Preferably, the power supply can be logically connected to the at least one processor 10 through a power management device, thereby enabling functions such as charging management, discharging management, and power consumption management. The power supply may also include one or more DC or AC power supplies, recharging devices, power fault detection circuits, power converters or inverters, power status indicators, and other arbitrary components. The electronic device 1 may also include various sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be described in detail here.

[0136] Furthermore, the electronic device 1 may also include a network interface. Optionally, the network interface may include a wired interface and / or a wireless interface (such as a Wi-Fi interface, a Bluetooth interface, etc.), which is typically used to establish communication connections between the electronic device 1 and other electronic devices.

[0137] Optionally, the electronic device 1 may further include a user interface, which may be a display, an input unit (such as a keyboard), or a standard wired or wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touchscreen. The display may also be appropriately referred to as a screen or display unit, used to display information processed in the electronic device 1 and to display a visual user interface.

[0138] It should be understood that the embodiments described are for illustrative purposes only and are not limited to this structure in the scope of the patent application.

[0139] The software fault feedback processing program 12 stored in the memory 11 of the electronic device 1 is a combination of multiple computer programs. When run in the processor 10, it can achieve the following:

[0140] Obtain the fault feedback text and fault feedback image corresponding to the same fault feedback in the software;

[0141] The fault feedback text is subjected to fault category identification to obtain the target fault category;

[0142] Entity extraction is performed on the fault feedback text to obtain the first fault entity;

[0143] The error reporting area in the fault feedback image is segmented, and the text in the segmented image is identified to obtain the image text;

[0144] Entity extraction is performed on the image text to obtain the second fault entity;

[0145] One or more processing method knowledge bases are obtained, wherein each processing method knowledge base has a corresponding knowledge base tag, and the similarity between the target fault category and the knowledge base tag is used to filter all processing method knowledge bases to obtain the target processing method knowledge base;

[0146] The first fault entity and the second fault entity are used as filtering conditions to filter the processing method text in the target processing method knowledge base to obtain the target processing method text.

[0147] The target processing method text is sent to a preset terminal device.

[0148] Specifically, the processor 10's implementation method of the above-mentioned computer program can be found in [reference needed]. Figure 1 The descriptions of the relevant steps in the corresponding embodiments are not repeated here.

[0149] Furthermore, if the modules / units integrated in the electronic device 1 are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. The computer-readable medium can be non-volatile or volatile. The computer-readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, or a read-only memory (ROM).

[0150] Embodiments of the present invention may also provide a computer-readable storage medium storing a computer program, which, when executed by a processor of an electronic device, can perform the following:

[0151] Obtain the fault feedback text and fault feedback image corresponding to the same fault feedback in the software;

[0152] The fault feedback text is subjected to fault category identification to obtain the target fault category;

[0153] Entity extraction is performed on the fault feedback text to obtain the first fault entity;

[0154] The error reporting area in the fault feedback image is segmented, and the text in the segmented image is identified to obtain the image text;

[0155] Entity extraction is performed on the image text to obtain the second fault entity;

[0156] One or more processing method knowledge bases are obtained, wherein each processing method knowledge base has a corresponding knowledge base tag, and the similarity between the target fault category and the knowledge base tag is used to filter all processing method knowledge bases to obtain the target processing method knowledge base;

[0157] The first fault entity and the second fault entity are used as filtering conditions to filter the processing method text in the target processing method knowledge base to obtain the target processing method text.

[0158] The target processing method text is sent to a preset terminal device.

[0159] Furthermore, the computer's usable storage medium may mainly include a program storage area and a data storage area, wherein the program storage area may store the operating system, applications required for at least one function, etc.; and the data storage area may store data created based on the use of blockchain nodes, etc.

[0160] In the several embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0161] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0162] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0163] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0164] Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be embraced within the invention. No appended diagram markings in the claims should be construed as limiting the scope of the claims.

[0165] The blockchain referred to in this invention is a novel application model of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanisms, and encryption algorithms. Essentially, a blockchain is a decentralized database, a chain of data blocks linked together using cryptographic methods. Each data block contains information about a batch of network transactions, used to verify the validity of the information (anti-counterfeiting) and generate the next block. A blockchain can include an underlying blockchain platform, a platform product service layer, and an application service layer.

[0166] Furthermore, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices recited in a system claim may also be implemented by a single unit or device through software or hardware. The term "second class" is used to indicate names and does not indicate any specific order.

[0167] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A software fault feedback processing method, characterized in that, The method includes: Obtain the fault feedback text and fault feedback image corresponding to the same fault feedback in the software; Based on the text feature vector of the fault feedback text, the fault category is identified to obtain the target fault category. Entity extraction is performed on the fault feedback text to obtain the first fault entity; The error reporting area in the fault feedback image is segmented, and the text in the segmented image is identified to obtain the image text; Entity extraction is performed on the image text to obtain the second fault entity; Multiple processing method knowledge bases are acquired. The target fault category is converted into a vector to obtain a category vector. The knowledge base tags of each processing method knowledge base are converted into vectors to obtain a tag vector. The similarity value between the category vector and the tag vector is calculated. The processing method knowledge base corresponding to the tag vector with the largest similarity value is determined as the target processing method knowledge base. The step of converting the knowledge base tags of each processing method knowledge base into vectors to obtain a tag vector includes: converting each character in the knowledge base tag into a vector to obtain a character vector; combining all the character vectors according to the order of the corresponding characters in the knowledge base tag to obtain a knowledge base tag matrix; convolving the knowledge base tag matrix with a preset convolution kernel to obtain a convolution matrix; and reducing the dimensionality of the convolution matrix to obtain the tag vector. Using the first fault entity and the second fault entity as filtering conditions to filter processing method texts in the target processing method knowledge base to obtain target processing method texts includes: converting the first fault entity into a vector to obtain a first entity vector; converting the second fault entity into a vector to obtain a second entity vector; converting each processing method text in the target processing method knowledge base into a vector to obtain a processing method vector; calculating the correlation between the first entity vector and the processing method vector to obtain a first correlation; calculating the correlation between the second entity vector and the processing method vector to obtain a second correlation; based on a preset weighting coefficient, performing a weighted calculation according to the first correlation and second correlation corresponding to each processing method vector to obtain a matching coefficient corresponding to the processing method vector; and using the matching coefficient to filter processing method texts in the target processing method knowledge base to obtain the target processing method texts. The target processing method text is sent to a preset terminal device.

2. The software fault feedback processing method as described in claim 1, characterized in that, The step of identifying the fault category of the fault feedback text based on its text feature vector to obtain the target fault category includes: The trained BERT model is used to extract features from the fault feedback text to obtain a text feature vector; The softmax function is used to calculate the recognition probability of different fault categories. The fault category corresponding to the highest recognition probability is identified as the target fault category.

3. The software fault feedback processing method as described in claim 1, characterized in that, The step of segmenting the error reporting area in the fault feedback image and identifying the text in the segmented image to obtain image text includes: The error reporting area in the fault feedback image is cropped to obtain the cropped image; The cropped image is converted to grayscale to obtain a grayscale image; The grayscale image is filtered to obtain a standard image; The standard image is processed by a preset text extraction algorithm to obtain the image text.

4. The software fault feedback processing method as described in claim 1, characterized in that, The step of reducing the dimensionality of the convolution matrix to obtain the label vector includes: Calculate the average value of each column element in the convolution matrix to obtain the column feature value; The label vector is obtained by replacing each column in the convolution matrix with the corresponding column feature value.

5. A software fault feedback processing device, characterized in that, include: The text processing module is used to obtain the fault feedback text and fault feedback image corresponding to the same fault feedback in the software. Based on the text feature vector of the fault feedback text, the fault category is identified to obtain the target fault category; entity extraction is performed on the fault feedback text to obtain the first fault entity. The image processing module is used to segment the error reporting area in the fault feedback image and identify the text in the segmented image to obtain image text; Entity extraction is performed on the image text to obtain the second fault entity; The processing method filtering module is used to acquire multiple processing method knowledge bases, convert the target fault category into a vector to obtain a category vector, convert the knowledge base tags of each processing method knowledge base into a vector to obtain a tag vector, calculate the similarity value between the category vector and the tag vector, and determine the processing method knowledge base corresponding to the tag vector with the maximum similarity value as the target processing method knowledge base. Specifically, converting the knowledge base tags of each processing method knowledge base into a vector to obtain a tag vector includes: converting each character in the knowledge base tag into a vector to obtain a character vector; combining all the character vectors according to the order of the corresponding characters in the knowledge base tag to obtain a knowledge base tag matrix; convolving the knowledge base tag matrix using a preset convolution kernel to obtain a convolution matrix; and reducing the dimensionality of the convolution matrix to obtain the tag vector. The processing method filtering module is further configured to use the first fault entity and the second fault entity as filtering conditions to filter processing method texts in the target processing method knowledge base to obtain target processing method texts, including: converting the first fault entity into a vector to obtain a first entity vector, converting the second fault entity into a vector to obtain a second entity vector, converting each processing method text in the target processing method knowledge base into a vector to obtain a processing method vector, calculating the correlation between the first entity vector and the processing method vector to obtain a first correlation, calculating the correlation between the second entity vector and the processing method vector to obtain a second correlation, performing weighted calculation based on a preset weighting coefficient according to the first correlation and the second correlation corresponding to each processing method vector to obtain a matching coefficient corresponding to the processing method vector, using the matching coefficient to filter the processing method texts in the target processing method knowledge base to obtain the target processing method texts, and sending the target processing method texts to a preset terminal device.

6. The software fault feedback processing device as described in claim 5, characterized in that, The step of reducing the dimensionality of the convolution matrix to obtain the label vector includes: Calculate the average value of each column element in the convolution matrix to obtain the column feature value; The label vector is obtained by replacing each column in the convolution matrix with the corresponding column feature value.

7. An electronic device, characterized in that, The electronic device includes: At least one processor; and, A memory communicatively connected to the at least one processor; wherein, The memory stores computer program instructions that can be executed by the at least one processor to enable the at least one processor to perform the software fault feedback processing method as described in any one of claims 1 to 4.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the software fault feedback processing method as described in any one of claims 1 to 4.

Images