Source code problem classification method and device, equipment, readable storage medium and product

Abstract

The invention relates to a source code problem classification method and device, equipment, a readable storage medium and a product. The method comprises the steps of obtaining at least one source code problem in a preset time period; for any source code problem, determining an initial classification result of the source code problem according to the problem content of the source code problem; and according to the initial classification result and the problem content of the source code problem, classifying the source code problem to obtain a target classification result of the source code problem. By adopting the method, the source code problems can be better classified.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a source code problem classification method, device, equipment, readable storage medium and product. Background Art

[0002] With the rapid development of artificial intelligence technology and the rapid expansion of corporate knowledge bases, traditional question-answering systems have been unable to meet R&D needs due to their difficulty in dynamic adjustment and accurate classification.

[0003] Although there have been breakthroughs in intelligent classification and code analysis technologies based on pre-trained models (such as Bidirectional Encoder Representations from Transformers, BERT) and Robustly Optimized BERT Pretraining Approach, RoBERTa), most of these technologies lack real-time updates and personalized responses.

[0004] Existing question-answering systems mostly rely on static knowledge bases. In the scenario of source code question classification, their classification efficiency and accuracy still need to be improved. Summary of the invention

[0005] Based on this, it is necessary to provide a source code problem classification method, device, equipment, readable storage medium and product for the above technical problems, which can better classify source code problems.

[0006] In a first aspect, the present application provides a source code problem classification method, comprising:

[0007] Obtain at least one source code problem within a preset period of time;

[0008] For any source code problem, determine an initial classification result of the source code problem according to the problem content of the source code problem;

[0009] According to the initial classification result and the problem content of the source code problem, the source code problem is classified to obtain the target classification result of the source code problem.

[0010] In one embodiment, if the initial classification result indicates that the source code question is a text question, the source code question is classified according to the initial classification result and the question content of the source code question to obtain a target classification result of the source code question, including:

[0011] Get text data from source code issues;

[0012] Input the text data into the large text model to obtain the target classification results corresponding to the source code problems.

[0013] In one embodiment, text data is input into a large text model to obtain target classification results corresponding to source code problems, including:

[0014] Input text data into the large text model to obtain the intermediate classification results corresponding to the source code problems;

[0015] According to the preset natural language processing model, the intermediate classification results are adjusted to determine the target classification results corresponding to the source code problem.

[0016] In one embodiment, if the initial classification result indicates that the source code problem is a code type problem, the source code problem is classified according to the initial classification result and the problem content of the source code problem to obtain a target classification result of the source code problem, including:

[0017] Get code data from source code issues;

[0018] Determine the similarity between the code data and different standard code data in the standard code library;

[0019] According to the size relationship between the similarities, the source code problems are classified to obtain the target classification results of the source code problems.

[0020] In one embodiment, source code problems are classified according to the magnitude relationship between the similarities to obtain target classification results of the source code problems, including:

[0021] Determine candidate classification results of the source code problem according to the magnitude relationship between each similarity and a preset similarity threshold;

[0022] Process the code data according to the code macro model to obtain a processing result;

[0023] According to the processing result, a target classification result of the source code problem is selected from the candidate classification results.

[0024] In one embodiment, after obtaining the target classification result of the source code problem, the method further includes:

[0025] From the predetermined code maintenance strategies corresponding to different problem types, a code maintenance strategy matching the target classification result is selected.

[0026] In a second aspect, the present application also provides a source code problem classification device, comprising:

[0027] A problem acquisition module, used to acquire at least one source code problem within a preset period of time;

[0028] A first determination module is used to determine an initial classification result of a source code problem according to the problem content of the source code problem for any source code problem;

[0029] The second determination module is used to classify the source code problem according to the initial classification result and the problem content of the source code problem to obtain a target classification result of the source code problem.

[0030] In a third aspect, the present application further provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0031] Obtain at least one source code problem within a preset period of time;

[0032] For any source code problem, determine an initial classification result of the source code problem according to the problem content of the source code problem;

[0033] According to the initial classification result and the problem content of the source code problem, the source code problem is classified to obtain the target classification result of the source code problem.

[0034] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the following steps are implemented:

[0035] Obtain at least one source code problem within a preset period of time;

[0036] For any source code problem, determine an initial classification result of the source code problem according to the problem content of the source code problem;

[0037] According to the initial classification result and the problem content of the source code problem, the source code problem is classified to obtain the target classification result of the source code problem.

[0038] In a fifth aspect, the present application further provides a computer program product, including a computer program, which implements the following steps when executed by a processor:

[0039] Obtain at least one source code problem within a preset period of time;

[0040] For any source code problem, determine an initial classification result of the source code problem according to the problem content of the source code problem;

[0041] According to the initial classification result and the problem content of the source code problem, the source code problem is classified to obtain the target classification result of the source code problem.

[0042] The above-mentioned source code problem classification method, device, equipment, readable storage medium and product perform initial classification of source code problems based on the problem content of the source code problems to obtain corresponding initial classification results. Furthermore, in order to make the target classification result more accurate, after determining the initial classification result, the source code problems are further classified according to the initial classification result and the problem content of the source code problems to obtain the final classification result of the source code problems. In other words, in the process of classifying source code problems, the above-mentioned process adopts different methods to perform secondary classification of source code problems, which can make the classification effect better. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0044] Figure 1 A schematic diagram of a flow chart of a source code problem classification method in one embodiment;

[0045] Figure 2 A schematic diagram of a flow chart of a source code problem classification step in one embodiment;

[0046] Figure 3 It is a flowchart of a source code problem classification step in another embodiment;

[0047] Figure 4 A schematic diagram of a process for selecting a maintenance strategy in an embodiment;

[0048] Figure 5 is a flow chart of a source code problem classification method in another embodiment;

[0049] Figure 6 is a structural block diagram of a source code problem classification device in one embodiment;

[0050] Figure 7 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0052] In one embodiment, if Figure 1As shown, a source code problem classification method is provided. This embodiment takes the method applied to a terminal as an example for illustration. It can be understood that the method can also be applied to a server, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server. In this embodiment, the method includes the following steps:

[0053] S110, obtaining at least one source code problem within a preset period of time.

[0054] The preset time period may be a preset historical time period before and including the current time. For example, the preset time period may be one day or one week. The present application does not impose any limitation on the specific duration of the preset time period.

[0055] Among them, the source code problem is used to describe the problems that occur during the source code processing process. Exemplarily, the source code problem can be a problem described in pure text, a source code problem can also be a problem that includes both code and text descriptions, or a source code problem can also be a problem described in pure code. This application does not impose any limitation on the problem content of the source code problem.

[0056] Exemplarily, there are many ways to obtain at least one source code problem within a preset period. For example, source code problems within a preset historical period may be pre-stored in a preset database, and when there is a need to classify source code problems, at least one source code problem is obtained from the corresponding preset database. For source code problems at the current moment, source code problems may be obtained based on a source code upload portal.

[0057] S120, for any source code problem, determine an initial classification result of the source code problem according to the problem content of the source code problem.

[0058] Among them, the initial classification result can represent the classification result of the source code problem in a large classification direction. It can be understood that the initial classification result cannot represent the final classification result of the source code problem.

[0059] In an optional embodiment, the problem content of the source code problem may be input into an initial classification model to obtain an initial classification result of the corresponding source code problem.

[0060] In another optional embodiment, the problem content of the source code problem can be detected based on a preset classification strategy and an initial classification result of the source code problem can be determined. Exemplarily, the initial classification result of the source code problem can be determined based on the data type in the problem content of the source code problem. For example, when there is code in the problem content, the source code problem is determined to be a code-type problem; when there is no code in the problem content, the source code problem is determined to be a text-type problem.

[0061] It should be noted that, in this embodiment, the preset classification strategy can be determined based on manual experience, and this application does not impose any limitation on the specific content of the preset classification strategy.

[0062] S130, classifying the source code problem according to the initial classification result and the problem content of the source code problem to obtain a target classification result of the source code problem.

[0063] Among them, the target classification result is the final classification result of the source code problem.

[0064] In an optional embodiment, the initial classification result may be directly used as the target classification result of the source code problem.

[0065] In another optional embodiment, the initial classification result and the problem content of the source code problem may be input into a pre-trained target classification model to obtain a target classification result of the source code problem.

[0066] In another optional embodiment, different methods may be used to process the problem content of the corresponding source code problem according to different initial classification results to obtain the target classification result of the corresponding source code problem.

[0067] In the above source code problem classification method, the source code problem is initially classified based on the problem content of the source code problem to obtain the corresponding initial classification result. Furthermore, in order to make the target classification result more accurate, after determining the initial classification result, the source code problem is further classified according to the initial classification result and the problem content of the source code problem to obtain the final classification result of the source code problem. In other words, in the process of classifying the source code problem, the above process adopts different methods to perform secondary classification on the source code problem, which can achieve better classification effect.

[0068] Based on the technical solutions of the above embodiments, the present application also provides an optional embodiment. In this optional embodiment, a specific process is provided for classifying the source code problem according to the initial classification result and the problem content of the source code problem to obtain the target classification result of the source code problem when the initial classification result indicates that the source code problem is a text problem.

[0069] See also Figure 2 The steps for classifying source code issues shown include:

[0070] S210, obtaining text data in the source code problem.

[0071] Specifically, in this embodiment, the text content in the source code question is extracted to obtain text data.

[0072] S220. Input the text data into the text large model to obtain the target classification result corresponding to the source code problem.

[0073] Among them, the text large model can be the BERT large model.

[0074] In an alternative implementation, the text data can be directly input into the text large model to obtain the target classification result corresponding to the source code problem.

[0075] In another alternative implementation, in order to make the target classification result more accurate, the text data can also be processed before being input into the text large model to obtain the input data, and then the input data is input into the text large model to obtain the target classification result corresponding to the corresponding source code. Exemplarily, the data processing method can be as follows: extract the keywords in the text data to obtain a word cloud diagram, and based on the mining algorithm, perform data mining on the text data carrying the word cloud diagram annotation information to obtain the pattern association data. Among them, the pattern association data can characterize the correlation between each keyword. Further, perform data preprocessing on the pattern association data to obtain the input data.

[0076] Optionally, the data preprocessing process can include at least one of denoising processing, stop word filtering, stemming or lemmatization, semantic information extraction, and data clustering.

[0077] Among them, the denoising processing can be completed based on regular expressions. For example, clean the irrelevant content such as special characters, redundant spaces, and Hyper Text Markup Language (HTML) tags in the text data to improve the purity of the data. Stop word filtering can use a stop word list to remove the common but meaningless words in the text data (such as "de", "he"), which can help improve the accuracy of subsequent semantic extraction. Stemming or lemmatization: For English data, stemming algorithms (such as Porter Stemmer or Snowball Stemmer) or lemmatization (such as WordNet Lemmatizer) can reduce the vocabulary to its basic form, which helps to reduce the interference of word deformation on semantics. Semantic information extraction can be completed based on common neural networks, converting words into semantic vectors and encoding the semantic relationships of the vocabulary into the vector space. Data clustering can be completed based on common clustering algorithms, clustering the technical Q&A data to discover the potential semantic relationships between different questions and help achieve efficient classification.

[0078] It should be noted that data mining of the text data can discover the potential associations between different questions and establish a multi-dimensional and multi-layer strategy framework for more accurate classification and clustering.

[0079] In order to improve the accuracy of the target classification result, in another optional implementation, the text data is input into a large text model to obtain an intermediate classification result corresponding to the source code problem; the intermediate classification result is adjusted according to a preset natural language processing model to determine the target classification result corresponding to the source code problem.

[0080] Exemplarily, a classifier can be trained based on a preset natural language algorithm, and the intermediate classification results of the large text model can be input into the classifier. The intermediate classification results can be fine-tuned based on the classifier to obtain more accurate target classification results.

[0081] In the above embodiment, a method for determining the target classification result corresponding to the text type question is provided. Specifically, the intermediate classification result is determined by using a large text model, and the intermediate classification result is adjusted based on a natural language processing model to obtain a more accurate target classification result, so that the target classification result is more accurate.

[0082] Based on the technical solutions of the above embodiments, the present application also provides an optional embodiment. In this optional embodiment, a specific process is provided for classifying the source code problem according to the initial classification result and the problem content of the source code problem to obtain the target classification result of the source code problem when the initial classification result indicates that the source code problem is a code type problem.

[0083] See also Figure 3 The steps for classifying source code issues shown include:

[0084] S310, obtaining code data in the source code problem.

[0085] Specifically, in this embodiment, the code data content in the source code question is extracted to obtain the code data.

[0086] S320, determining the similarity between the code data and different standard code data in the standard code library.

[0087] Among them, multiple historical standard codes are stored in the standard code library.

[0088] Exemplarily, in this embodiment, the code data and the standard code data may be directly input into a pre-trained similarity determination model to obtain the similarity between the code data and each standard code data.

[0089] Furthermore, in order to make the similarity more accurate, in this embodiment, the code data may be processed, and the processed code data may be input into a similarity determination model to obtain the similarity between the code data and each standard code data.

[0090] Exemplarily, the process of processing the code data may include at least one of code snippet cleaning, code tokenization, and code vectorization.

[0091] Among them, code snippet cleaning is used to convert code snippets into an input format suitable for model processing, including removing noise such as comments and extra blank lines. Code tokenization can use the tokenizer that comes with CodeBERT to tokenize the code and decompose the code into tokens that the model can understand. For example, convert for (int i = 0; i < 10; i++) to the token required by the model. Code vectorization can convert the tokenized code into semantic vectors (Embedding) through the CodeBERT model. These vectors are semantic representations of code snippets, which facilitate subsequent classification and similarity analysis.

[0092] S330, classifying the source code problems according to the magnitude relationship between the similarities, and obtaining a target classification result of the source code problems.

[0093] Exemplarily, in this embodiment, the similarities may be sorted, and the type of standard code data corresponding to the highest similarity may be selected as the target classification result of the source code problem.

[0094] In order to further improve the accuracy of the target classification result, in another optional implementation, the candidate classification results of the source code problem can be determined according to the size relationship between each similarity and a preset similarity threshold; the code data is processed according to the code macro model to obtain a processing result; and according to the processing result, the target classification result of the source code problem is selected from the candidate classification results.

[0095] Exemplarily, the types of standard code data corresponding to each similarity greater than a preset similarity threshold are used as candidate classification results of the source code problem. Further, the code data is input into the code big model to obtain a processing result. And among the candidate classification results, the one that matches the processing result is used as the target classification result of the source code problem.

[0096] In the above embodiment, a method for determining the target classification result corresponding to the code type problem is provided. Specifically, based on the similarity between the code data and each standard code data, the candidate classification result corresponding to the code data is determined, and the code data is processed by the code big model, so that the target classification result is selected from the candidate classification results according to the processing result, so that the target classification result is more accurate.

[0097] Based on the technical solutions of the above embodiments, the present application also provides an optional embodiment. In the optional embodiment, the steps after obtaining the target classification result of the source code problem are limited.

[0098] See also Figure 4 The maintenance strategy selection steps shown include:

[0099] S410, selecting a code maintenance strategy that matches the target classification result from the predetermined code maintenance strategies corresponding to different problem types.

[0100] For example, in this embodiment, each code maintenance strategy corresponding to the problem type to which the source code problem belongs can be used as a candidate strategy corresponding to the source code problem. Further, based on manual experience or a large model, a code maintenance strategy matching the target classification result is selected from the candidate strategies.

[0101] In the above embodiment, a specific method for determining the code maintenance strategy corresponding to the source code problem is provided, which is convenient for the staff to solve the source code problem.

[0102] Furthermore, in this embodiment, the step of obtaining at least one source code problem within a preset period can be performed regularly, and each source code problem can be classified, and the classification results can be concentrated in the problem system. In the case of a new source code problem, the code maintenance strategy corresponding to the source code problem can be determined directly based on the problem system.

[0103] In addition, by summarizing and analyzing each source code problem, we can organize the code maintenance strategies corresponding to the corresponding problems, and then provide special training for staff based on the code maintenance strategy experience to improve the staff's response speed to source code problems.

[0104] Based on the technical solutions of the above embodiments, the present application also provides an optional embodiment. In this optional embodiment, the source code problem classification method provided by the present application is introduced in detail.

[0105] See also Figure 5 The source code classification method shown includes:

[0106] S510, obtaining at least one source code problem within a preset period of time;

[0107] S520, for any source code problem, determining an initial classification result of the source code problem according to the problem content of the source code problem;

[0108] S530A, if the initial classification result indicates that the source code question is a text question, obtaining text data in the source code question;

[0109] S540A, inputting the text data into the large text model to obtain an intermediate classification result corresponding to the source code problem;

[0110] Before inputting the text data into the large text model, the text data is processed to obtain input data, and the input data is input into the large text model to obtain the target classification result corresponding to the corresponding source code. Exemplarily, the data processing method can be as follows: extract the keywords in the text data to obtain a word cloud diagram, and based on the mining algorithm, perform data mining on the text data carrying the word cloud diagram annotation information to obtain pattern association data. Among them, the pattern association data can characterize the correlation between each keyword. Further, the pattern association data is preprocessed to obtain input data;

[0111] Optionally, the data preprocessing process may include at least one of denoising, stop word filtering, stemming or word form restoration, semantic information extraction and data clustering;

[0112] S550A, adjusting the intermediate classification result according to a preset natural language processing model to determine a target classification result corresponding to the source code problem;

[0113] S530B, if the initial classification result indicates that the source code problem is a code type problem, obtain code data in the source code problem;

[0114] S540B, determining the similarity between the code data and different standard code data in the standard code library;

[0115] In order to make the similarity more accurate, in this embodiment, the code data may also be processed, and the processed code data may be input into the similarity determination model to obtain the similarity between the code data and each standard code data;

[0116] Exemplarily, the process of processing the code data may include at least one of code snippet cleaning, code tokenization, and code vectorization;

[0117] S550B, determining candidate classification results of the source code problem according to the magnitude relationship between each similarity and a preset similarity threshold;

[0118] S560B, processing the code data according to the code big model to obtain a processing result;

[0119] S570B, selecting a target classification result of the source code problem from the candidate classification results according to the processing result;

[0120] S580, selecting a code maintenance strategy that matches the target classification result from the predetermined code maintenance strategies corresponding to different problem types.

[0121] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0122] Based on the same inventive concept, the embodiment of the present application also provides a source code problem classification device for implementing the source code problem classification method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more source code problem classification device embodiments provided below can refer to the limitations of the source code problem classification method above, and will not be repeated here.

[0123] In an exemplary embodiment, Figure 6 As shown, a source code problem classification device is provided, including: a problem acquisition module 610, a first determination module 620 and a second determination module 630, wherein:

[0124] A problem acquisition module 610 is used to acquire at least one source code problem within a preset period of time;

[0125] A first determination module 620 is used to determine an initial classification result of a source code problem according to the problem content of the source code problem for any source code problem;

[0126] The second determination module 630 is used to classify the source code problem according to the initial classification result and the problem content of the source code problem to obtain a target classification result of the source code problem.

[0127] In one embodiment, if the initial classification result indicates that the source code problem is a text type problem, the second determination module 630 includes a text data acquisition unit for acquiring text data in the source code problem; and a first determination unit for inputting the text data into the text large model to obtain a target classification result corresponding to the source code problem.

[0128] In one embodiment, the first determination unit includes a first determination subunit, which is used to input text data into a text large model to obtain an intermediate classification result corresponding to the source code problem; and a second determination subunit, which is used to adjust the intermediate classification result according to a preset natural language processing model to determine a target classification result corresponding to the source code problem.

[0129] In one embodiment, if the initial classification result indicates that the source code problem is a code type problem, the second determination module 630 includes a code data acquisition unit, which is used to acquire the code data in the source code problem; a second determination unit, which is used to determine the similarity between the code data and different standard code data in the standard code library; and a third determination unit, which is used to classify the source code problem according to the size relationship between each similarity, to obtain a target classification result of the source code problem.

[0130] In one embodiment, the second determination unit includes a third determination subunit, which is used to determine the candidate classification results of the source code problem according to the size relationship between each similarity and a preset similarity threshold; a fourth determination subunit, which is used to process the code data according to the code macro model to obtain a processing result; and a fifth determination subunit, which is used to select a target classification result of the source code problem from the candidate classification results according to the processing result.

[0131] In one embodiment, the source code problem classification device further includes a strategy determination module, which is used to select a code maintenance strategy that matches the target classification result from the predetermined code maintenance strategies corresponding to different problem types.

[0132] Each module in the above source code problem classification device can be implemented in whole or in part by software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each module.

[0133] In an exemplary embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as shown in FIG. Figure 7As shown. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit and an input device. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface, the display unit and the input device are connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a mobile cellular network, near field communication (NFC) or other technologies. When the computer program is executed by the processor, a source code problem classification method is implemented. The display unit of the computer device is used to form a visually visible picture, which can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device shell, or an external keyboard, touchpad or mouse.

[0134] Those skilled in the art will understand that Figure 7 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0135] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:

[0136] Obtain at least one source code problem within a preset period of time;

[0137] For any source code problem, determine an initial classification result of the source code problem according to the problem content of the source code problem;

[0138] According to the initial classification result and the problem content of the source code problem, the source code problem is classified to obtain the target classification result of the source code problem.

[0139] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0140] Get text data from source code issues;

[0141] Input the text data into the large text model to obtain the target classification results corresponding to the source code problems.

[0142] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0143] Input text data into the large text model to obtain the intermediate classification results corresponding to the source code problems;

[0144] According to the preset natural language processing model, the intermediate classification results are adjusted to determine the target classification results corresponding to the source code problem.

[0145] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0146] Get code data from source code issues;

[0147] Determine the similarity between the code data and different standard code data in the standard code library;

[0148] According to the size relationship between the similarities, the source code problems are classified to obtain the target classification results of the source code problems.

[0149] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0150] Determine candidate classification results of the source code problem according to the magnitude relationship between each similarity and a preset similarity threshold;

[0151] Process the code data according to the code macro model to obtain a processing result;

[0152] According to the processing result, a target classification result of the source code problem is selected from the candidate classification results.

[0153] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0154] From the predetermined code maintenance strategies corresponding to different problem types, a code maintenance strategy matching the target classification result is selected.

[0155] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0156] Obtain at least one source code problem within a preset period of time;

[0157] For any source code problem, determine an initial classification result of the source code problem according to the problem content of the source code problem;

[0158] According to the initial classification result and the problem content of the source code problem, the source code problem is classified to obtain the target classification result of the source code problem.

[0159] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0160] Get text data from source code issues;

[0161] Input the text data into the large text model to obtain the target classification results corresponding to the source code problems.

[0162] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0163] Input text data into the large text model to obtain the intermediate classification results corresponding to the source code problems;

[0164] According to the preset natural language processing model, the intermediate classification results are adjusted to determine the target classification results corresponding to the source code problem.

[0165] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0166] Get code data from source code issues;

[0167] Determine the similarity between the code data and different standard code data in the standard code library;

[0168] According to the size relationship between the similarities, the source code problems are classified to obtain the target classification results of the source code problems.

[0169] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0170] Determine candidate classification results of the source code problem according to the magnitude relationship between each similarity and a preset similarity threshold;

[0171] Process the code data according to the code macro model to obtain a processing result;

[0172] According to the processing result, a target classification result of the source code problem is selected from the candidate classification results.

[0173] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0174] From the predetermined code maintenance strategies corresponding to different problem types, a code maintenance strategy matching the target classification result is selected.

[0175] In one embodiment, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the following steps:

[0176] Obtain at least one source code problem within a preset period of time;

[0177] For any source code problem, determine an initial classification result of the source code problem according to the problem content of the source code problem;

[0178] According to the initial classification result and the problem content of the source code problem, the source code problem is classified to obtain the target classification result of the source code problem.

[0179] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0180] Get text data from source code issues;

[0181] Input the text data into the large text model to obtain the target classification results corresponding to the source code problems.

[0182] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0183] Input text data into the large text model to obtain the intermediate classification results corresponding to the source code problems;

[0184] According to the preset natural language processing model, the intermediate classification results are adjusted to determine the target classification results corresponding to the source code problem.

[0185] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0186] Get code data from source code issues;

[0187] Determine the similarity between the code data and different standard code data in the standard code library;

[0188] According to the size relationship between the similarities, the source code problems are classified to obtain the target classification results of the source code problems.

[0189] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0190] Determine candidate classification results of the source code problem according to the magnitude relationship between each similarity and a preset similarity threshold;

[0191] Process the code data according to the code macro model to obtain a processing result;

[0192] According to the processing result, a target classification result of the source code problem is selected from the candidate classification results.

[0193] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0194] From the predetermined code maintenance strategies corresponding to different problem types, a code maintenance strategy matching the target classification result is selected.

[0195] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited to this.

[0196] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0197] The above embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A source code problem classification method, characterized in that: The method comprises: Obtain at least one source code problem within a preset period of time; For any source code problem, determining an initial classification result of the source code problem according to the problem content of the source code problem; The source code problem is classified according to the initial classification result and the problem content of the source code problem to obtain a target classification result of the source code problem.

2. The method according to claim 1, characterized in that If the initial classification result indicates that the source code problem is a text problem, classifying the source code problem according to the initial classification result and the problem content of the source code problem to obtain a target classification result of the source code problem includes: Obtaining text data in the source code problem; The text data is input into the text big model to obtain the target classification result corresponding to the source code problem.

3. The method according to claim 2, characterized in that The step of inputting the text data into the large text model to obtain a target classification result corresponding to the source code problem includes: Inputting the text data into the text big model to obtain an intermediate classification result corresponding to the source code problem; The intermediate classification result is adjusted according to a preset natural language processing model to determine a target classification result corresponding to the source code problem.

4. The method according to claim 1, characterized in that: If the initial classification result indicates that the source code problem is a code type problem, then classifying the source code problem according to the initial classification result and the problem content of the source code problem to obtain a target classification result of the source code problem includes: Obtain code data in the source code problem; Determining the similarity between the code data and different standard code data in a standard code library; The source code problem is classified according to the size relationship between the similarities to obtain a target classification result of the source code problem.

5. The method according to claim 4, characterized in that The step of classifying the source code problem according to the size relationship between the similarities to obtain a target classification result of the source code problem includes: Determining a candidate classification result of the source code problem according to a size relationship between each of the similarities and a preset similarity threshold; Processing the code data according to the code macro model to obtain a processing result; According to the processing result, a target classification result of the source code problem is selected from the candidate classification results.

6. The method according to any one of claims 1 to 5, characterized in that After obtaining the target classification result of the source code problem, the method further includes: From the predetermined code maintenance strategies corresponding to different problem types, a code maintenance strategy matching the target classification result is selected.

7. A source code problem classification device, characterized in that: The device comprises: A problem acquisition module, used to acquire at least one source code problem within a preset period of time; A first determination module is used to determine, for any source code problem, an initial classification result of the source code problem according to the problem content of the source code problem; The second determination module is used to classify the source code problem according to the initial classification result and the problem content of the source code problem to obtain a target classification result of the source code problem.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images