Question matching method and device, equipment and storage medium

Abstract

The invention relates to the technical field of data processing, and discloses a problem matching method, device and equipment and a storage medium, and the method comprises the steps: determining a to-be-solved problem based on notification information, and carrying out the vectorization and label extraction of the to-be-solved problem, and obtaining a to-be-solved problem vector and a to-be-solved problem label; respectively performing label matching and vector retrieval based on the to-be-solved problem labels and the to-be-solved problem vectors to obtain a plurality of historical problems; performing semantic analysis based on the plurality of historical problems to obtain a target historical problem; and determining a solution of the to-be-solved problem based on the target historical problem. According to the method, the retrieval efficiency and accuracy can be effectively improved, then the problem solution can be quickly and accurately determined, and the overall efficiency of project management is improved.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a question matching method, apparatus, device and storage medium. Background Art

[0002] As the complexity of software project management continues to increase, project management tools such as JIRA are becoming more and more important. JIRA is not only used to track issues and tasks, but also to manage team collaboration and project progress. However, there are still many problems with JIRA in use: 1. As the scale of the project expands, the number of issues also increases rapidly, making manual matching and finding related issues very time-consuming and error-prone; 2. Traditional keyword retrieval methods are inefficient when dealing with a large number of issues and cannot accurately capture the implicit relationship between issues; 3. Sorting out the logical relationship between similar issues requires experts in the field to spend a lot of human resources.

[0003] Therefore, how to effectively improve retrieval efficiency and accuracy, and then quickly and accurately determine problem solutions and improve the overall efficiency of project management is a problem that needs to be solved urgently.

[0004] The above contents are only used to assist in understanding the technical solution of the present application and do not constitute an admission that the above contents are prior art. Summary of the invention

[0005] The main purpose of this application is to provide a problem matching method, device, equipment and storage medium, aiming to solve the technical problem of how to effectively improve retrieval efficiency and accuracy, and then quickly and accurately determine problem solutions and improve the overall efficiency of project management.

[0006] To achieve the above purpose, the present application proposes a question matching method, which includes:

[0007] Determine the problem to be solved based on the notification information, and vectorize and extract labels for the problem to be solved to obtain a problem vector and a problem label;

[0008] Based on the labels of the unsolved problems and the vectors of the unsolved problems, label matching and vector retrieval are performed to obtain a number of historical problems;

[0009] Performing semantic analysis based on the several historical issues to obtain target historical issues;

[0010] A solution to the problem to be solved is determined based on the target historical problem.

[0011] In one embodiment, determining the problem to be solved based on the notification information, and vectorizing and extracting labels for the problem to be solved to obtain the problem vector and the problem label include:

[0012] Obtaining notification information, and determining whether the notification information is information of interest;

[0013] When the notification information is concerned information, extracting the data of the unresolved issues in the notification information;

[0014] Clean and structure the data of the unsolved problems to generate unsolved problems;

[0015] Vectorize the problem to be solved by using an embedded model to obtain a vector of the problem to be solved;

[0016] The label of the problem to be solved is extracted by a large language model to obtain a label of the problem to be solved.

[0017] In one embodiment, label matching and vector retrieval are performed based on the problem label and the problem vector to be solved, respectively, to obtain several historical problems, including:

[0018] Perform label matching on the historical question vectors in the target question library based on the labels of the questions to be solved, to obtain a historical question vector set;

[0019] A vector search is performed on the historical problem vector set according to the unsolved problem vector to obtain a plurality of historical problems.

[0020] In one embodiment, the vector search of the historical problem vector set is performed according to the problem vector to be solved to obtain a number of historical problems, including:

[0021] Obtaining vector similarities between the to-be-solved problem vector and each historical problem vector in the historical problem vector set;

[0022] If there is a vector similarity greater than a preset vector similarity threshold, determining a number of target historical question vectors in the historical question vector set based on the vector similarity;

[0023] A plurality of historical issues are determined according to the plurality of target historical issue vectors and the issue identifier-vector index table.

[0024] In one embodiment, before performing label matching on the historical question vectors in the target question library based on the labels of the unsolved questions to obtain the historical question set, the method further includes:

[0025] Acquire historical question data, and perform data cleaning and structuring on the historical question data to obtain historical questions in a target format;

[0026] Based on the historical question database in the target format and database vectorization, an initial question database is obtained;

[0027] Perform semantic analysis based on the historical issues to generate hot spots and graphic elements of the historical issues;

[0028] A label of each historical question is generated based on the heat map and the graphic element, and the corresponding historical questions in the initial question library are marked according to the label of each historical question to obtain a target question library.

[0029] In one embodiment, the semantic analysis is performed based on the plurality of historical questions to obtain a target historical question, including:

[0030] Performing semantic analysis on the several historical issues to obtain semantic similarities between the problem to be solved and each historical issue;

[0031] If there is a semantic similarity greater than a preset semantic similarity threshold, a target historical question among the plurality of historical questions is determined according to the semantic similarity.

[0032] In one embodiment, determining a solution to the problem to be solved based on the target historical problem includes:

[0033] Obtaining a solution to the target historical problem;

[0034] Determine whether the solution to the target historical problem can be transferred to the problem to be solved by using a large language model according to preset prompt words;

[0035] When the solution to the target historical problem can be transferred to the unresolved problem, generating a solution to the unresolved problem according to the solution to the target historical problem;

[0036] When the solution to the target historical problem cannot be transferred to the unsolved problem, the information of the unsolved problem is obtained, and a solution to the unsolved problem is generated according to the unsolved problem and the information of the unsolved problem through a large language model.

[0037] In addition, to achieve the above purpose, the present application also proposes a question matching device, which includes:

[0038] An extraction module, used to determine the problem to be solved based on the notification information, and vectorize and extract labels for the problem to be solved to obtain a vector of the problem to be solved and a label of the problem to be solved;

[0039] A retrieval module, used to perform label matching and vector retrieval based on the labels of the unsolved problems and the vectors of the unsolved problems, respectively, to obtain a number of historical problems;

[0040] An analysis module, used for performing semantic analysis based on the plurality of historical issues to obtain target historical issues;

[0041] A determination module is used to determine a solution to the problem to be solved based on the target historical problem. In addition, to achieve the above purpose, the present application also proposes a problem matching device, the device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the problem matching method as described above.

[0042] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, which is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the question matching method described above are implemented.

[0043] In addition, to achieve the above-mentioned purpose, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, the steps of the question matching method described above are implemented.

[0044] The present application provides a problem matching method, which first determines the problem to be solved based on notification information, and vectorizes and extracts labels for the problem to be solved to obtain a problem vector and a problem label; performs label matching and vector retrieval based on the problem label and the problem vector to obtain several historical problems; performs semantic analysis based on the several historical problems to obtain target historical problems; and determines a solution to the problem based on the target historical problem, which can effectively improve retrieval efficiency and accuracy, thereby quickly and accurately determining a solution to the problem and improving the overall efficiency of project management.

[0045] In summary, the present application obtains several historical problems by quickly and accurately performing label matching and vector retrieval based on the problem vectors obtained by vectorization and label extraction of the problem to be solved and the problem labels to be solved, and then quickly and accurately determines the target historical problems through semantic analysis and generates solutions to the problems to be solved, overcoming the technical defects of low retrieval efficiency and poor accuracy, and can effectively improve retrieval efficiency and accuracy, thereby quickly and accurately determining problem solutions and improving the overall efficiency of project management. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0047] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0048] Figure 1 A flowchart of the first embodiment of the problem matching method of the present application is provided;

[0049] Figure 2 A flowchart of the second embodiment of the problem matching method of the present application is provided;

[0050] Figure 3 This is a schematic diagram of the module structure of the problem matching device in the embodiment of the present application;

[0051] Figure 4 A schematic diagram of the device structure of the hardware operating environment involved in the problem matching method of the embodiment of the present application.

[0052] The purpose, features and advantages of this application will be further described in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0053] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0054] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0055] The main solution of the embodiment of the present application is: determine the unresolved problem based on the notification information, and vectorize and extract labels for the unresolved problem to obtain an unresolved problem vector and an unresolved problem label; perform label matching and vector retrieval based on the unresolved problem label and the unresolved problem vector respectively to obtain several historical problems; perform semantic analysis based on the several historical problems to obtain target historical problems; determine the solution to the unresolved problem based on the target historical problem.

[0056] As the complexity of software project management continues to increase, project management tools such as JIRA are becoming more and more important. JIRA is not only used to track issues and tasks, but also to manage team collaboration and project progress. However, there are still many problems with JIRA in use: 1. As the scale of the project expands, the number of issues also increases rapidly, making manual matching and finding related issues very time-consuming and error-prone; 2. Traditional keyword retrieval methods are inefficient when dealing with a large number of issues, and cannot accurately capture the implicit relationship between issues; 3. Sorting out the logical relationship between similar issues requires a lot of human resources from experts in the field. Therefore, how to effectively improve retrieval efficiency and accuracy, and then quickly and accurately determine problem solutions and improve the overall efficiency of project management is a problem that needs to be solved urgently.

[0057] The present application obtains several historical problems by quickly and accurately performing label matching and vector retrieval based on the vectors of unresolved problems obtained by vectorization and label extraction of the unresolved problems and the labels of the unresolved problems, and then quickly and accurately determines the target historical problems through semantic analysis and generates solutions to the unresolved problems, thereby overcoming the technical defects of low retrieval efficiency and poor accuracy, and can effectively improve retrieval efficiency and accuracy, thereby quickly and accurately determining problem solutions and improving the overall efficiency of project management.

[0058] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device capable of realizing the above functions, a question matching device, etc. The following takes the question matching device as an example to illustrate this embodiment and the following embodiments.

[0059] Based on this, the embodiment of the present application provides a question matching method, referring to Figure 1 , Figure 1 This is a flowchart of the first embodiment of the question matching method of the present application.

[0060] In this embodiment, the question matching method includes steps S10 to S40:

[0061] Step S10: determining the problem to be solved based on the notification information, and vectorizing and extracting labels for the problem to be solved to obtain a vector of the problem to be solved and a label of the problem to be solved.

[0062] It should be noted that the problem to be solved refers to a work item that needs to be solved in JIRA or a component of the project progress. JIRA is a widely used project management tool. Vectorization refers to converting the problem to be solved into a numerical form that can be processed by a computer for subsequent processing and analysis. Tag extraction is to extract key information from the problem to form tags for subsequent matching and retrieval.

[0063] In a feasible implementation, step S10 may include: obtaining notification information and determining whether the notification information is information of concern; when the notification information is information of concern, extracting data of unresolved issues in the notification information; performing data cleaning and structuring on the data of unresolved issues to generate unresolved issues; vectorizing the unresolved issues through an embedded model to obtain an unresolved issue vector; and extracting labels for the unresolved issues through a large language model to obtain labels for the unresolved issues.

[0064] It should be noted that, in this embodiment, WebHook technology is used to establish a monitoring of JIRA on the JiraBot side. When a new issue is created, the status is updated, or a new comment is created in JIRA, a notification message is sent to the JiraBot monitoring port through WebHook. If the notification message is in the watch list, that is, the notification message is a watch message, the subsequent process is triggered, otherwise the process is interrupted.

[0065] It can be understood that data cleaning refers to removing and organizing irrelevant information and duplicate information in the problem data to ensure the accuracy and availability of the data. Structuring is to organize the cleaned data according to a certain format and structure. For example, all problem data is formatted using JSON. JSON (JavaScript Object Notation) format is a lightweight data exchange format that is easy for people to read and write, and is also easy for machines to parse and generate. By structuring the problem data into JSON format, data storage, transmission and processing can be conveniently performed.

[0066] It is worth noting that an embedded model is a model that can convert text data into numerical vectors, usually based on deep learning technology. Through the embedded model, the text information of the problem to be solved can be converted into a numerical vector form that can be processed by the computer, which is convenient for subsequent calculation and analysis. The large language model is a model that can understand and generate natural language text. By training a large amount of language data, it can perform semantic analysis and label extraction on the text.

[0067] Step S20, performing label matching and vector retrieval based on the labels of the unsolved problems and the vectors of the unsolved problems, respectively, to obtain a number of historical problems.

[0068] It should be noted that label matching refers to matching the labels of the problem to be solved and the labels of the historical problems based on the similarity, so as to find the historical problems related to the problem to be solved. Vector retrieval uses the vector of the problem to be solved to search in the historical problem vector library, so as to find the historical problems that are closest to the problem to be solved in terms of vector similarity. By combining label matching and vector retrieval, the accuracy and efficiency of problem matching can be improved.

[0069] It can be understood that historical issues refer to issues that have been resolved during the project management process. These historical issues are stored in the issue library for future reference and reuse. Historical issues usually contain detailed information such as solutions, related discussions, and modification records. This information is of great value for quickly locating and solving new problems.

[0070] Step S30, performing semantic analysis based on the plurality of historical questions to obtain a target historical question.

[0071] It should be noted that semantic analysis refers to the in-depth understanding of historical issues through natural language processing technology, and the identification of the target historical issues that are semantically closest to the problem to be solved based on semantic similarity. Semantic analysis can include lexical, syntactic and semantic analysis of the question text to extract key information and understand the true intention of the question. Through semantic analysis, historical issues that are related to the problem to be solved and have reference value can be effectively screened out to provide support for subsequent solutions.

[0072] In a feasible implementation, step S30 includes: performing semantic analysis on the several historical issues to obtain the semantic similarity between the problem to be solved and each historical issue; if there is a semantic similarity greater than a preset semantic similarity threshold, determining the target historical issue among the several historical issues based on the semantic similarity.

[0073] It should be noted that semantic similarity refers to the degree of similarity between the problem to be solved and the historical problem at the semantic level. It can be calculated using a variety of natural language processing techniques, such as cosine similarity, Jaccard similarity, or a semantic embedding model based on deep learning. Cosine similarity measures the similarity between two question text vectors by calculating the cosine value of the two question text vectors, while Jaccard similarity focuses on the proportion of words that appear in common in the two question texts. The deep learning model obtains an embedding vector that can capture complex semantic relationships through training, thereby more accurately measuring the semantic similarity between questions.

[0074] It is understood that the preset semantic similarity threshold refers to a standard value set during semantic analysis to determine the semantic relevance between the problem to be solved and the historical problem. Only when the semantic similarity between the historical problem and the problem to be solved exceeds this threshold, the historical problem is considered to be a target historical problem with reference value. The setting of the threshold can be adjusted according to the actual application scenario and needs to ensure the accuracy and effectiveness of the screening results.

[0075] In the specific implementation, after the vector-based search, the semantic similarity between the retrieved N historical questions and the problem to be solved is determined based on semantics. If there is a semantic similarity greater than the preset semantic similarity threshold, the top K questions with semantic similarity greater than the set threshold are taken as the target historical questions.

[0076] It is worth noting that if all semantic similarities are less than the preset semantic similarity threshold, that is, no similar historical problems are found, it is necessary to generate possible solutions based on the domain-wide model and database, where the database is a professional domain knowledge base built using the knowledge graph. Generating possible solutions based on the domain-wide model and database specifically includes: retrieving domain-wide data based on the knowledge graph to obtain data on the problem to be solved, submitting the data and the problem as input to the domain-wide model, and the domain-wide model generates a solution based on the database.

[0077] Step S40: determining a solution to the problem to be solved based on the target historical problem.

[0078] It should be noted that the solution to the problem to be solved can be obtained by adjusting and optimizing the solution to the target historical problem to adapt to the specific circumstances of the current problem. When determining the solution, a variety of strategies can be adopted, such as directly borrowing the solution to the target historical problem, or combining the solutions to multiple historical problems to form a more complete solution through comprehensive analysis and innovation, and this embodiment does not impose specific restrictions on this.

[0079] Understandably, when none of the solutions to the target historical problem are applicable, more in-depth analysis and innovation are needed. At this point, the reasoning ability of the large model in the field can be used to generate a completely new solution in combination with the specific conditions and limitations of the current problem. Large models in the field usually have strong logical reasoning and problem-solving capabilities, and can propose innovative solutions based on a rich knowledge base and historical data.

[0080] In a feasible implementation, step S40 includes: obtaining a solution to the target historical problem; determining whether the solution to the target historical problem can be transferred to the problem to be solved through a large language model according to preset prompt words; when the solution to the target historical problem can be transferred to the problem to be solved, generating a solution to the problem to be solved according to the solution to the target historical problem; when the solution to the target historical problem cannot be transferred to the problem to be solved, obtaining information about the problem to be solved, and generating a solution to the problem to be solved according to the problem to be solved and the information about the problem to be solved through a large language model.

[0081] It should be noted that for existing similar problems and solutions, the problem transfer capability of the large model can be used through prompt word instructions to let it determine whether it can be transferred to the new problem. If so, a new solution is generated based on the new problem. Otherwise, it means that the existing solution is not applicable to the new problem. It is necessary to generate a completely new solution based on the specific situation of the new problem, combining the domain knowledge base and the reasoning ability of the large model.

[0082] It is understandable that the natural language processing capabilities of the big model are used to analyze and understand the solution to the target historical problem. By presetting prompt words, the big model can be guided to focus on the key elements in the solution, such as the core characteristics of the problem, the applicable conditions of the solution, etc. Then, the big model compares these key elements with the problem to be solved to determine whether there are similarities, so as to determine whether the solution can be transferred to the new problem.

[0083] It is worth noting that the preset prompt words can be keywords, domain terms or specific instructions related to the problem. These prompt words help the big model to understand the semantic content of the problem and solution more accurately, thereby improving the accuracy of the migration judgment. For example, if the target historical problem involves a specific technical failure, and the problem to be solved also involves a similar failure, then through the preset prompt words "technical failure" and "solution migration", the big model can determine whether the solution to the target historical problem is applicable to the problem to be solved. If the big model determines that the solution to the target historical problem can be migrated to the problem to be solved, then a preliminary solution can be generated based on the solution. Next, according to the specific situation of the problem to be solved, the preliminary solution can be adjusted and optimized to ensure its adaptability and effectiveness. If the big model determines that the solution to the target historical problem is not applicable to the problem to be solved, the information in the field based on the knowledge graph is retrieved to obtain the information of the problem to be solved, and the information and the problem are submitted as input to the big model in the field, and the big model in the field generates a solution based on the database.

[0084] It is worth mentioning that after obtaining the solution to the problem to be solved, the generated solution can be submitted to the corresponding problem discussion area through JiraAPI and relevant personnel can be reminded.

[0085] The present embodiment provides a problem matching method, which first determines the problem to be solved based on notification information, and vectorizes and extracts labels for the problem to be solved to obtain a problem vector and a problem label; performs label matching and vector retrieval based on the problem label and the problem vector to obtain several historical problems; performs semantic analysis based on the several historical problems to obtain target historical problems; and determines a solution to the problem based on the target historical problem, which can effectively improve retrieval efficiency and accuracy, thereby quickly and accurately determining a solution to the problem and improving the overall efficiency of project management.

[0086] In summary, this embodiment quickly and accurately performs label matching and vector retrieval based on the problem vectors and problem labels obtained by vectorization and label extraction of the problem to be solved, and obtains several historical problems, and then quickly and accurately determines the target historical problems through semantic analysis and generates solutions to the problems to be solved, thereby overcoming the technical defects of low retrieval efficiency and poor accuracy, and can effectively improve retrieval efficiency and accuracy, thereby quickly and accurately determining problem solutions and improving the overall efficiency of project management.

[0087] Based on the first embodiment of the present application, in the second embodiment of the present application, the same or similar contents as those in the above-mentioned embodiment 1 can refer to the above introduction, and will not be repeated later. Figure 2 , the step S20 further includes steps S201-S202:

[0088] Step S201 , performing label matching on historical question vectors in a target question library based on the labels of the questions to be solved, to obtain a set of historical question vectors.

[0089] It should be noted that the target question library can be pre-built and contain a large number of historical question vectors with labels. These historical question vectors are obtained by vectorizing the historical question texts.

[0090] It can be understood that by matching the labels of the questions to be solved with the historical question vectors in the target question library to obtain a set of historical question vectors with the same labels, the search scope can be effectively narrowed and the search efficiency can be improved.

[0091] In a feasible implementation, before step S201, it also includes: acquiring historical question data, and performing data cleaning and structuring on the historical question data to obtain historical questions in a target format; constructing a historical question database based on the target format and vectorizing the database to obtain an initial question library; performing semantic analysis based on the historical questions to generate heat maps and graphic elements of the historical questions; generating labels for each historical question based on the heat map and graphic elements, and marking the corresponding historical questions in the initial question library according to the labels of each historical question to obtain a target question library.

[0092] It should be noted that historical issues refer to issues that have been resolved during the project management process. All issue data can be downloaded through the JIRA API interface, and then historical issue data can be obtained based on the issue data.

[0093] It is understandable that the process of data cleaning and structuring includes steps such as removing irrelevant information, unifying data formats, and filling missing values ​​to ensure the quality and consistency of historical question data. The target format can be JSON format, but is not limited to it. When building the initial question library, natural language processing technology can be used to vectorize the historical question text and convert the text into a numerical vector to facilitate subsequent retrieval and matching operations. Use an embedded model to vectorize the database, that is, vectorize all historical questions in the database, and establish an index table between the question id and the vector.

[0094] It is worth noting that the semantic-based generation of heat maps and graphs of existing questions is to identify the core content and related features of the questions by analyzing the keywords, phrases and sentences in the historical question text. The heat map can intuitively display the hot spots of the question, while the graph can represent the key elements of the question. Labels can be generated for each historical question, and each question can contain multiple labels.

[0095] Step S202: performing vector retrieval on the historical problem vector set according to the unsolved problem vector to obtain a plurality of historical problems.

[0096] It should be noted that vector retrieval is achieved by calculating the similarity between the problem vector to be solved and each historical problem vector in the historical problem vector set. Similarity calculation can use methods such as cosine similarity and Euclidean distance to ensure the relevance of the retrieval results. Through vector retrieval, several historical problems that are most similar to the problem to be solved can be quickly screened out from the historical problem vector set, thereby improving the accuracy of retrieval.

[0097] In a feasible implementation, step S202 may include: obtaining the vector similarity between the problem vector to be solved and each historical problem vector in the historical problem vector set; if there is a vector similarity greater than a preset vector similarity threshold, determining a number of target historical problem vectors in the historical problem vector set based on the vector similarity; and determining a number of historical problems based on the several target historical problem vectors and the problem identifier-vector index table.

[0098] It should be noted that the preset vector similarity threshold is a parameter set according to actual application scenarios and requirements, and is used to determine whether the similarity between the problem vector to be solved and the historical problem vector is high enough to ensure the relevance and accuracy of the retrieval results.

[0099] It can be understood that in the historical question vector set with the same label, the vector similarity between the question vector to be solved and each historical question vector in the historical question vector set is calculated. If there is a vector similarity greater than a preset vector similarity threshold, the nearest N historical question vectors greater than the preset vector similarity threshold are matched.

[0100] It is worth noting that after determining a number of historical question vectors, the specific historical question is determined according to the question identifier-vector index table. The question identifier-vector index table is pre-built and is used to associate the question identifier (such as question ID) with the corresponding vector. Through this index table, the specific historical question corresponding to the target historical question vector can be quickly retrieved from the target question library.

[0101] In this embodiment, by performing label matching and vector retrieval according to the problem labels and problem vectors to be solved respectively, the efficiency and accuracy of problem retrieval can be effectively improved, and historical problems matching the problem to be solved can be found quickly and accurately.

[0102] It should be noted that the above examples are only used to understand the present application and do not constitute a limitation on the problem matching method of the present application. More simple transformations based on this technical concept are all within the scope of protection of the present application.

[0103] This application also provides a problem matching device, please refer to Figure 3 , the problem matching device comprises:

[0104] The extraction module 10 is used to determine the problem to be solved based on the notification information, and vectorize and extract labels for the problem to be solved to obtain a vector of the problem to be solved and a label of the problem to be solved.

[0105] The retrieval module 20 is used to perform label matching and vector retrieval based on the labels of the unsolved problems and the vectors of the unsolved problems, respectively, to obtain a number of historical problems.

[0106] The analysis module 30 is used to perform semantic analysis based on the plurality of historical questions to obtain a target historical question.

[0107] The determination module 40 is used to determine a solution to the problem to be solved based on the target historical problem.

[0108] The present embodiment provides a problem matching device. The present embodiment determines the problem to be solved based on the notification information, and vectorizes and extracts labels for the problem to be solved to obtain a problem vector and a problem label; performs label matching and vector retrieval based on the problem label and the problem vector, respectively, to obtain several historical problems; performs semantic analysis based on the several historical problems to obtain target historical problems; determines a solution to the problem to be solved based on the target historical problem, which can effectively improve the retrieval efficiency and accuracy, and then quickly and accurately determine the problem solution, thereby improving the overall efficiency of project management.

[0109] In summary, this embodiment quickly and accurately performs label matching and vector retrieval based on the problem vectors and problem labels obtained by vectorization and label extraction of the problem to be solved, and obtains several historical problems, and then quickly and accurately determines the target historical problems through semantic analysis and generates solutions to the problems to be solved, thereby overcoming the technical defects of low retrieval efficiency and poor accuracy, and can effectively improve retrieval efficiency and accuracy, thereby quickly and accurately determining problem solutions and improving the overall efficiency of project management.

[0110] Optionally, the extraction module 10 is also used to obtain notification information and determine whether the notification information is attention information; when the notification information is attention information, extract the unresolved problem data in the notification information; perform data cleaning and structuring on the unresolved problem data to generate unresolved problems; vectorize the unresolved problems through an embedded model to obtain unresolved problem vectors; perform label extraction on the unresolved problems through a large language model to obtain unresolved problem labels.

[0111] Optionally, the retrieval module 20 is further used to perform label matching on historical question vectors in the target question library based on the labels of the unsolved problems to obtain a historical question vector set; and perform vector retrieval on the historical question vector set according to the vectors of the unsolved problems to obtain a number of historical problems.

[0112] Optionally, the retrieval module 20 is also used to obtain the vector similarity between the problem vector to be solved and each historical problem vector in the historical problem vector set; if there is a vector similarity greater than a preset vector similarity threshold, determining a number of target historical problem vectors in the historical problem vector set based on the vector similarity; and determining a number of historical problems based on the several target historical problem vectors and the problem identifier-vector index table.

[0113] Optionally, the retrieval module 20 is also used to obtain historical question data, and perform data cleaning and structuring on the historical question data to obtain historical questions in a target format; construct a historical question database based on the target format and vectorize the database to obtain an initial question library; perform semantic analysis based on the historical questions to generate heat maps and graphic elements for the historical questions; generate labels for each historical question based on the heat map and graphic elements, and mark the corresponding historical questions in the initial question library according to the labels of each historical question to obtain a target question library.

[0114] Optionally, the analysis module 30 is also used to perform semantic analysis on the several historical issues to obtain the semantic similarity between the problem to be solved and each historical issue; if there is a semantic similarity greater than a preset semantic similarity threshold, the target historical issue among the several historical issues is determined according to the semantic similarity.

[0115] Optionally, the determination module 40 is also used to obtain a solution to the target historical problem; determine whether the solution to the target historical problem can be transferred to the problem to be solved according to preset prompt words through a large language model; when the solution to the target historical problem can be transferred to the problem to be solved, generate a solution to the problem to be solved according to the solution to the target historical problem; when the solution to the target historical problem cannot be transferred to the problem to be solved, obtain the information of the problem to be solved, and generate a solution to the problem to be solved according to the problem to be solved and the information of the problem to be solved through a large language model.

[0116] The problem matching device provided by the present application adopts the problem matching method in the above embodiment, which can solve the technical problem of how to effectively improve the retrieval efficiency and accuracy, and then quickly and accurately determine the problem solution, and improve the overall efficiency of project management. Compared with the prior art, the beneficial effects of the problem matching device provided by the present application are the same as the beneficial effects of the problem matching method provided by the above embodiment, and the other technical features in the problem matching device are the same as the features disclosed in the above embodiment method, which will not be repeated here.

[0117] The present application provides a question matching device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the question matching method in the above-mentioned embodiment one.

[0118] Reference below Figure 4 , which shows a schematic diagram of the structure of a question matching device suitable for implementing the embodiment of the present application. The question matching device in the embodiment of the present application may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 4 The problem matching device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0119] like Figure 4 As shown, the question matching device may include a processing device 1001 (e.g., a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage device 1003 to a random access memory (RAM: Random Access Memory) 1004. In RAM1004, various programs and data required for the operation of the question matching device are also stored. The processing device 1001, ROM1002, and RAM1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems can be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the problem matching device to communicate with other devices wirelessly or by wire to exchange data. Although the figure shows a problem matching device with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems can be implemented or have alternatively.

[0120] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

[0121] The problem matching device provided by the present application adopts the problem matching method in the above embodiment, which can solve the technical problem of how to effectively improve the retrieval efficiency and accuracy, and then quickly and accurately determine the problem solution, and improve the overall efficiency of project management. Compared with the prior art, the beneficial effects of the problem matching device provided by the present application are the same as the beneficial effects of the problem matching method provided by the above embodiment, and the other technical features in the problem matching device are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.

[0122] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0123] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0124] The present application provides a computer-readable storage medium having computer-readable program instructions (ie, computer programs) stored thereon, wherein the computer-readable program instructions are used to execute the question matching method in the above-mentioned embodiment.

[0125] The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above.

[0126] The computer-readable storage medium may be included in the question-matching device; or may exist independently without being assembled into the question-matching device.

[0127] The above-mentioned computer-readable storage medium carries one or more programs. When the above-mentioned one or more programs are executed by the problem matching device, the problem matching device: determines the problem to be solved based on the notification information, and vectorizes and extracts labels for the problem to be solved to obtain the problem vector and the problem label; performs label matching and vector retrieval based on the problem label and the problem vector respectively to obtain several historical problems; performs semantic analysis based on the several historical problems to obtain the target historical problem; and determines the solution to the problem to be solved based on the target historical problem.

[0128] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0129] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0130] The modules involved in the embodiments described in this application may be implemented by software or hardware, wherein the name of the module does not constitute a limitation on the unit itself in some cases.

[0131] The readable storage medium provided by the present application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned problem matching method, and can solve the technical problem of how to effectively improve retrieval efficiency and accuracy, and then quickly and accurately determine the solution to the problem, and improve the overall efficiency of project management. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided by the present application are the same as the beneficial effects of the problem matching method provided by the above-mentioned embodiment, and will not be repeated here.

[0132] The present application also provides a computer program product, including a computer program, which implements the steps of the above-mentioned question matching method when executed by a processor.

[0133] The computer program product provided by this application can solve the technical problem of how to effectively improve the efficiency and accuracy of retrieval, thereby quickly and accurately determining the solution to the problem and improving the overall efficiency of project management. Compared with the prior art, the beneficial effects of the computer program product provided by this application are the same as the beneficial effects of the problem matching method provided by the above embodiment, and will not be repeated here.

[0134] The above descriptions are only some embodiments of the present application, and are not intended to limit the patent scope of the present application. All equivalent structural changes made using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect applications in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A question matching method, It is characterized in that The method comprises: Determine the problem to be solved based on the notification information, and vectorize and extract labels for the problem to be solved to obtain a problem vector and a problem label; Based on the labels of the unsolved problems and the vectors of the unsolved problems, label matching and vector retrieval are performed to obtain a number of historical problems; Performing semantic analysis based on the several historical issues to obtain target historical issues; A solution to the problem to be solved is determined based on the target historical problem.

2. The method according to claim 1, It is characterized in that The determining of the problem to be solved based on the notification information, and vectorizing and extracting labels for the problem to be solved to obtain a problem vector and a problem label include: Obtaining notification information, and determining whether the notification information is information of interest; When the notification information is concerned information, extracting the data of the unresolved issues in the notification information; Clean and structure the data of the unsolved problems to generate unsolved problems; Vectorize the problem to be solved by using an embedded model to obtain a vector of the problem to be solved; The label of the problem to be solved is extracted by a large language model to obtain a label of the problem to be solved.

3. The method according to claim 1, It is characterized in that The label matching and vector retrieval are performed based on the label of the problem to be solved and the vector of the problem to be solved, respectively, to obtain several historical problems, including: Perform label matching on the historical question vectors in the target question library based on the labels of the questions to be solved, to obtain a historical question vector set; A vector search is performed on the historical problem vector set according to the unsolved problem vector to obtain a plurality of historical problems.

4. The method according to claim 3, It is characterized in that The vector retrieval of the historical problem vector set is performed according to the problem vector to be solved to obtain a number of historical problems, including: Obtaining vector similarities between the to-be-solved problem vector and each historical problem vector in the historical problem vector set; If there is a vector similarity greater than a preset vector similarity threshold, determining a number of target historical question vectors in the historical question vector set based on the vector similarity; A plurality of historical issues are determined according to the plurality of target historical issue vectors and the issue identifier-vector index table.

5. The method according to claim 3, It is characterized in that Before performing label matching on the historical question vectors in the target question library based on the labels of the unsolved questions to obtain the historical question set, the method further includes: Acquire historical question data, and perform data cleaning and structuring on the historical question data to obtain historical questions in a target format; Based on the historical question database in the target format and database vectorization, an initial question database is obtained; Perform semantic analysis based on the historical issues to generate hot spots and graphic elements of the historical issues; A label of each historical question is generated based on the heat map and the graphic element, and the corresponding historical questions in the initial question library are marked according to the label of each historical question to obtain a target question library.

6. The method according to claim 1, It is characterized in that The semantic analysis is performed based on the plurality of historical issues to obtain target historical issues, including: Performing semantic analysis on the several historical issues to obtain semantic similarities between the problem to be solved and each historical issue; If there is a semantic similarity greater than a preset semantic similarity threshold, a target historical question among the plurality of historical questions is determined according to the semantic similarity.

7. The method according to claim 1, It is characterized in that Determining a solution to the problem to be solved based on the target historical problem includes: Obtaining a solution to the target historical problem; Determine whether the solution to the target historical problem can be transferred to the problem to be solved by using a large language model according to preset prompt words; When the solution to the target historical problem can be transferred to the unresolved problem, generating a solution to the unresolved problem according to the solution to the target historical problem; When the solution to the target historical problem cannot be transferred to the unsolved problem, the information of the unsolved problem is obtained, and a solution to the unsolved problem is generated according to the unsolved problem and the information of the unsolved problem through a large language model.

8. A question matching device, It is characterized in that The problem matching device comprises: An extraction module, used to determine the problem to be solved based on the notification information, and vectorize and extract labels for the problem to be solved to obtain a vector of the problem to be solved and a label of the problem to be solved; A retrieval module, used to perform label matching and vector retrieval based on the labels of the unsolved problems and the vectors of the unsolved problems, respectively, to obtain a number of historical problems; An analysis module, used for performing semantic analysis based on the plurality of historical issues to obtain target historical issues; A determination module is used to determine a solution to the problem to be solved based on the target historical problem.

9. A problem matching device, It is characterized in that The question matching device comprises: a memory, a processor, and a question matching program stored in the memory and executable on the processor, wherein the question matching program is configured to implement the question matching method according to any one of claims 1 to 7.

10. A storage medium, It is characterized in that The storage medium stores a question matching program, and when the question matching program is executed by the processor, the question matching method according to any one of claims 1 to 7 is implemented.

Images