Work order assignment method and system

Abstract

This application relates to the field of business information system technology, and in particular to a work order allocation method and system. The method includes the following steps: constructing an engineer experience timeline knowledge base, storing the work experience of each engineer in chronological order; obtaining the technical historical feature description information of work order requests; using a preset work order allocation large language model to retrieve relevant engineer experiences from the engineer experience timeline knowledge base by matching engineer experiences with the technical historical feature description information; using the work order allocation large language model to analyze the experience process of each relevant engineer during the corresponding technology's popularity period, and thereby assessing its suitability with the current work order's technical background; and generating a work order allocation scheme based on the assessment results of the suitability of all relevant engineers using the work order allocation large language model. This application provides a work order allocation method and system that can improve the accuracy of work order allocation.

Description

Technical Field

[0001] This application relates to the field of business information system technology, and in particular to a work order allocation method and system. Background Technology

[0002] The work order allocation system is an important part of IT companies' customer service functions.

[0003] However, traditional work order allocation systems often rely solely on matching engineers with corresponding skills based on technical keywords. For example, they typically use keyword extraction to identify technical tags in the current customer's work order request, and then search for and allocate the work order to engineers with the corresponding skill tags.

[0004] However, existing technologies still have some problems. Different engineers with the same skill tags and skill levels exhibit significant differences in efficiency when handling problems of the corresponding skill tag type. Therefore, existing work order allocation methods still have issues, leading to inaccurate work order allocation. Summary of the Invention

[0005] To solve the above-mentioned technical problems, or at least partially solve them, this application provides a work order allocation method and system that can improve the accuracy of work order allocation.

[0006] Firstly, this application provides a work order allocation method, which includes the following steps:

[0007] Build an engineer experience timeline knowledge base to store each engineer's work experience in chronological order;

[0008] Obtain the technical history description information of the work order request;

[0009] By using a pre-defined work order allocation language model, the engineer's experience is matched with the technical history feature description information, and the engineer's experience timeline knowledge base is retrieved to obtain relevant engineer experiences.

[0010] The work order allocation language model is used to analyze the experience of each relevant engineer during the corresponding technology popularization period, and to evaluate its adaptability to the current work order technology background.

[0011] Based on the assessment results of all relevant engineer suitability, a work order allocation scheme is generated using the aforementioned work order allocation language model.

[0012] Optionally, obtaining the technical history feature description information of the work order request includes the following steps:

[0013] The technology component features are identified from the error logs of the current work order request using a pre-defined technology popularity period identification language model;

[0014] The technology popularity period identification big language model is used to match the technology component features in a pre-built technology popularity period knowledge graph to determine the popularity period interval of the technology component corresponding to the technology component features and extract the technology background information corresponding to the popularity period interval. The technology background information includes the technology popularity time range, typical technology combinations and / or problem patterns.

[0015] The technology background information is converted into a natural language form by using the technology popularity period identification large language model to describe the technology history features corresponding to the current work order request.

[0016] Optionally, the process of using the technology popularity period identification large language model to match the technology component features in a pre-built technology popularity period knowledge graph, and generating technology historical feature description information corresponding to the current work order request based on the matching results, includes the following steps:

[0017] The pre-set technology popularity period analysis prompt word template guides the large language model for technology popularity period identification to perform a complete reasoning process, including technology component matching, temporal information extraction, and descriptive text generation, generating natural language form of technology historical feature description information including the time range of technology popularity, typical technology combinations, and / or problem patterns.

[0018] Optionally, the technology popularity period analysis prompt template includes the following guidance:

[0019] The technology component matching guide instructs the technology popularity period identification big language model to match the corresponding technology component in the technology popularity period knowledge graph based on the technology component features identified from the error log;

[0020] The temporal information extraction guide is used to instruct the technology popularity period identification big language model to extract the technology popularity time range, typical technology combinations and / or problem patterns from the technology popularity period knowledge graph based on the matched technology components.

[0021] Contextual reasoning guidance is used to instruct the large language model for identifying the popularity of technology to deduce the technological background corresponding to the current work order request based on the extracted temporal information;

[0022] Description generation guidelines are used to instruct the large language model to generate a natural language form of description information of the historical features of the technology based on the inferred technological background during the period of technological popularity.

[0023] Optionally, the technical component features include exception class paths extracted from error logs, technical framework package name structure, version identifier, configuration file format, programming interface call pattern, and / or error stack hierarchy.

[0024] The pre-built knowledge graph of technology popularity records the temporal evolution information of each technology component, including the initial release time, mainstream adoption period, decline period, relationship with other technology components, version evolution path, and typical problem patterns in each popularity stage.

[0025] Optionally, the process of retrieving relevant engineer experiences from the engineer experience timeline knowledge base by matching engineer experience with the technical history feature description information using a preset work order allocation large language model includes the following steps:

[0026] The preset work order allocation comprehensive prompt word template guides the work order allocation big language model to perform engineer experience retrieval and obtain engineer experience related to the technical background of the current work order;

[0027] The comprehensive prompt template for work order assignment includes the following guidelines:

[0028] Experience retrieval guidance is used to instruct the work order allocation big language model to retrieve relevant engineer experiences from the engineer experience timeline knowledge base based on the time range and technical background in the technical history feature description information.

[0029] Optionally, the process of using the work order allocation large language model to analyze the experience of each relevant engineer during the corresponding technology's popularity period, and thereby assessing its suitability to the current work order technology context, includes the following steps:

[0030] The work order allocation comprehensive prompt word template guides the work order allocation big language model to perform a fit evaluation and generate an engineer evaluation result containing fit reasons;

[0031] The comprehensive prompt template for work order allocation also includes the following guidance:

[0032] The fit assessment guide is used to instruct the work order assignment language model to analyze the specific work content and problem-solving experience of the retrieved engineers based on their experience during the corresponding technology popularization period, and to assess the fit with the current work order technology background.

[0033] Optionally, based on the evaluation results of all relevant engineer suitability, the process of generating a work order allocation scheme using the aforementioned work order allocation language model includes the following steps:

[0034] The work order allocation prompt template guides the work order allocation language model to generate an allocation scheme, and outputs a work order allocation scheme that includes the recommended engineer identifier and the allocation reason.

[0035] The comprehensive prompt template for work order allocation also includes the following guidance:

[0036] The assignment scheme generation guide instructs the work order assignment language model to determine the optimal engineer selection based on the suitability evaluation results of all engineers and generate a detailed assignment scheme containing the reasons for the recommendation.

[0037] Optionally, the engineer experience timeline knowledge base includes engineer identifiers and their corresponding work experiences. Each engineer's work experience is recorded by year, and each work experience includes a description of the project participated in, the technical components used, the specific work content, and the key technical issues and solutions handled.

[0038] Secondly, this application proposes a work order allocation system, including a processor and a memory, wherein the memory stores any of the work order allocation methods described in the first aspect, and is loaded and executed by the processor.

[0039] The technical solution provided in this application has the following advantages compared with the prior art:

[0040] Its beneficial effects and working principle are as follows:

[0041] In actual operation and maintenance, it was found that even when a conventional work order allocation system assigns senior engineers with the corresponding skill tags to handle customer work order requests, the efficiency of these engineers often decreases when they work in high-security data centers that require disconnection from the outside world. This decrease in efficiency is not due to insufficient engineer skills, but rather because different data centers contain systems deployed at different times. When engineers can communicate with the outside world, they can contact engineers more familiar with specific historical systems via phone or remote assistance to inquire about or request assistance in resolving various system-related issues. However, this assistance mechanism fails in environments where engineers are required to disconnect, leading to decreased efficiency.

[0042] Furthermore, by comparing engineers who can resolve issues relatively quickly even in disconnected environments, it was found that they typically have work experience in the same era as the technical system they are currently dealing with. Those with this experience participated in related projects during the technology's heyday, accumulating practical experience and developing problem-solving skills appropriate for that era's technology.

[0043] Traditional work order allocation methods fail to recognize the compatibility between the problem-solving mindset developed by engineers within a specific technological era and the technical problems arising from that era. Therefore, they rely solely on skill-tag-based matching, focusing only on whether an engineer possesses a particular skill when allocating work orders. This approach leads to engineers lacking the corresponding problem-solving mindset for the technical issues at hand, often requiring them to spend considerable time trying different approaches, thus impacting problem-solving efficiency.

[0044] Therefore, this application constructs a timeline knowledge base storing each engineer's work experience in a time sequence, utilizes a large language model for identifying technology popularity periods, and extracts historical technology feature descriptions from work order requests. Then, using a large language model for work order allocation, it retrieves engineer experiences from the knowledge base that match the current technological context based on this historical technology feature description information. Finally, it analyzes and evaluates the engineers found to be best suited to the current work order's technological background. This ensures that engineers with the best historical context fit are assigned to the corresponding work orders, thereby improving the accuracy of work order allocation and increasing operational efficiency.

[0045] Therefore, this application provides a work order allocation method and system that can improve the accuracy of work order allocation. Attached Figure Description

[0046] Figure 1 A flowchart of the work order allocation method provided in the embodiments of this application;

[0047] Figure 2 The system architecture diagram of the large-scale workflow of the work order allocation method provided in the embodiments of this application is shown. Detailed Implementation

[0048] The technical solutions in this application will now be described with reference to the accompanying drawings.

[0049] Many specific details are set forth in the following description to provide a thorough understanding of this application, but this application may also be implemented in other ways different from those described herein; obviously, the embodiments in the specification are only some embodiments of this application, and not all embodiments. It should be noted that, unless otherwise specified, the embodiments of this application and the features in the embodiments can be combined with each other.

[0050] Reference Figures 1-2 As shown in the figure, this application provides a work order allocation method, the overall architecture of which consists of a large model workflow. In this application embodiment, the development platform selected is the N8N platform. In other embodiments, the Dify platform or the Coze platform can also be selected. The work order allocation method includes the following steps:

[0051] S1: Build an engineer experience timeline knowledge base, storing each engineer's work experience in chronological order.

[0052] Specifically, the engineer experience timeline knowledge base includes engineer identifiers and their corresponding work experiences. Each engineer's work experience is recorded by year, and each work experience includes a description of the project participated in, the technical components used, the specific work content, and the key technical issues and solutions handled.

[0053] For example, in this embodiment of the application, the data structure of the engineer's experience timeline knowledge base is as follows:

[0054] For example, one entry in this knowledge base is:

[0055] Engineer ID: ENG_001_Zhang San;

[0056] Work experience timeline:

[0057] Year: 2008;

[0058] Project Description: Maintenance project for a bank's core transaction system;

[0059] The technical components used are: Oracle 9i database, WebLogic 8.1 application server, Java 1.4, and Apache Commons DBCP 1.2 connection pool.

[0060] Specific job responsibilities: Responsible for handling database connection timeout issues and optimizing SQL query performance;

[0061] Key technical issues and solutions addressed: The connection leakage problem caused by improper DBCP connection pool configuration was resolved. By adjusting the maxActive and removeAbandoned parameters and enabling the connection leakage detection mechanism, the number of connection timeout errors was reduced from 20 per day to less than 2.

[0062] Year: 2015;

[0063] Project Description: Microservice architecture transformation project for an e-commerce platform;

[0064] The technologies used are: Spring Boot 1.2, Docker 1.6, MySQL 5.6, and Redis 3.0.

[0065] Specific tasks include: breaking down monolithic applications into microservices, designing a distributed data access layer, and implementing inter-service communication mechanisms;

[0066] Key technical issues and solutions: The Saga pattern was used to solve the distributed transaction consistency problem of order payment. Through compensation mechanisms and state machine design, data consistency in complex business processes was ensured.

[0067] For example, another entry in this knowledge base is:

[0068] Engineer ID: ENG_002_Li Si

[0069] Work experience timeline:

[0070] Year: 2018;

[0071] Project Description: Cloud-Native Containerization Platform Construction Project

[0072] The technologies used are: Kubernetes 1.10, Istio 1.0, Prometheus monitoring, Grafana visualization, and HikariCP.

[0073] Specific job responsibilities: Design a microservice monitoring system, implement distributed tracing, and optimize container resource allocation;

[0074] Key technical issues and solutions addressed: The circular dependency problem between services was located through Jaeger distributed tracing, the service call chain was redesigned, and the average response time was optimized from 2 seconds to 200 milliseconds.

[0075] S2: Obtain the technical history characteristics description information of the work order request;

[0076] Specifically, obtaining the technical history characteristics description information of a work order request includes the following steps:

[0077] The technology component features are identified from the error logs of the current work order request using a pre-defined technology popularity period identification language model;

[0078] Specifically, the technical component features include exception class paths extracted from error logs, technical framework package name structure, version identifiers, configuration file formats, programming interface call patterns, and / or error stack hierarchy.

[0079] For example, in this embodiment of the application, an error log feature recognition knowledge base is constructed, and retrieval augmented generation (RAG) technology, consistent with the above process, is used to identify the features of technical components. The specific construction and recognition process is as follows:

[0080] A large number of historical error log samples were collected and labeled, and a mapping relationship between error log features and technical component features was established to build an error log feature recognition knowledge base. Its structure includes:

[0081] Error log sample: contains raw error stack information

[0082] Corresponding technical component characteristics include: exception classpath, technical framework package name structure, version identifier, configuration file format, programming interface call pattern, and / or error stack hierarchy.

[0083] For example, one of the entries in the error log feature recognition knowledge base is:

[0084] Error log sample:

[0085] "java.sql.SQLException: Cannot get a connection, pool error Timeoutwaiting for idle object;

[0086] at org.apache.commons.dbcp.SQLNestedException.getCause(...);

[0087] at org.apache.commons.dbcp.PoolingDataSource.getConnection(...)".

[0088] Corresponding technical component characteristics:

[0089] Exception class path: org.apache.commons.dbcp.SQLNestedException;

[0090] Technical framework package name structure: org.apache.commons.dbcp;

[0091] Version identifier: Based on the package structure, it is inferred to be DBCP version 1.x;

[0092] Configuration file format: XML configuration format (Spring configuration method);

[0093] Programming interface call pattern: PoolingDataSource.getConnection;

[0094] Error stack hierarchy: a three-tiered call structure of SQL exception, connection pool exception, and Spring data source tools.

[0095] The process of identifying the features of technical components is as follows:

[0096] The pre-defined large language model for identifying the technology popularity period is based on the DeepSeek series of large language models, the Doubao series of large language models, and / or the ChatGPT series of models.

[0097] It should be noted that the large language model for identifying the technology popularity period described in the embodiments of this application, as well as the large language model for work order allocation mentioned below, are based on existing large language models, but different prompt words or optional fine-tuning of the large model are adopted. The naming differences are for ease of description and understanding, and to distinguish the different functions implemented.

[0098] For example, in this embodiment of the application, a portion of the technical component feature recognition prompt word template is:

[0099] "You are a professional technical error log analysis expert. Please accurately identify the characteristics of technical components such as exception class paths, technical framework package name structure, version identifiers, configuration file formats, programming interface call patterns, and error stack hierarchy from the provided error logs..."

[0100] Upon receiving a work order request, the large language model used during the technology popularization period extracts the accompanying error logs. The identification process is as follows:

[0101] The identification process is as follows:

[0102] Step 1: During the technology boom, the large language model receives error logs and prompt word templates, and initiates the feature recognition process;

[0103] Step 2: Convert the current error log into a vector representation, and perform similarity retrieval in the error log feature recognition knowledge base based on RAG technology to obtain the most relevant historical error log samples and their corresponding technical component features;

[0104] Step 3: Based on the retrieved relevant samples, the large language model accurately identifies the technical component features from the current error log through semantic reasoning, according to the requirements of the technical component feature recognition prompt word template.

[0105] Step 4: Output the recognition results.

[0106] The technology popularity period identification big language model is used to match the technology component features in a pre-constructed technology popularity period knowledge graph to determine the popularity period interval of the technology component corresponding to the technology component features and extract the technology background information corresponding to the popularity period interval. The technology background information includes the technology popularity time range, typical technology combinations and / or problem patterns.

[0107] The pre-built knowledge graph of technology popularity records the temporal evolution information of each technology component, including the initial release time, mainstream adoption period, decline period, relationship with other technology components, version evolution path, and typical problem patterns in each popularity stage.

[0108] For example, the knowledge graph of technology popularity records detailed evolutionary information of various technological components in different historical periods. One of its entries is shown below:

[0109] Technical components: Apache Commons DBCP;

[0110] First published: 2001;

[0111] Mainstream adoption period: 2004-2010; peak period: 2006-2008.

[0112] Decline period: After 2010, it was gradually replaced by C3P0 and HikariCP;

[0113] Compatibility with other technical components:

[0114] From 2001 to 2003: it was often used in conjunction with Struts 1.x, JSP, Servlet, Tomcat 3 / 4, and Oracle 8i / 9i;

[0115] From 2004 to 2007: it was often used in conjunction with Spring 1.x / 2.x, Hibernate 2.x / 3.x, and Tomcat 4 / 5;

[0116] 2008-2010: Often used in conjunction with Spring 2.x, Hibernate 3.x, and Tomcat 5 / 6;

[0117] Version evolution path: 1.0 (2001-2003) to 1.2.x (2004-2008) to 1.4.x (2009-2012);

[0118] Typical problem patterns at each stage of the epidemic:

[0119] 2001-2003: Issues with the implementation of the basic connection pool and compatibility with early application servers;

[0120] 2004-2007: Basic connection pool configuration issues and improper connection count settings;

[0121] 2008-2010: Connection leaks under high load and compatibility issues with newer framework versions.

[0122] It should be noted that the technology popularity period knowledge graph with such a structure is constructed in this application embodiment in order to solve the problem of fuzzy boundaries in the process of technology evolution.

[0123] Matching methods based solely on technology tags cannot capture the differences in usage of the same technology component at different times. For example, Apache Commons DBCP was mainly used with Struts 1.x, JSP, and Servlet from 2001 to 2003, while it was mainly used with Spring 2.x and Hibernate 3.x from 2008 to 2010. The technology tag DBCP alone cannot distinguish this difference in time.

[0124] The combination of technical components is determined by the development period of the project at that time, and has strong characteristics of the times. The combination of technologies in different periods reflects the relatively cutting-edge and stable level of technological development at that time. By analyzing the combination of technologies, we can accurately locate the popular period of the technology and thus locate the typical problem patterns of the corresponding popular stage.

[0125] This overcomes the problem of ambiguous boundaries in the process of technological evolution, avoids the historical context deviation that may occur with simple version number matching, and ensures the accuracy of subsequent generation of information describing the historical features of technology.

[0126] The pre-set technology popularity period analysis prompt word template guides the large language model for technology popularity period identification to perform a complete reasoning process, including technology component matching, temporal information extraction, and descriptive text generation, generating natural language form of technology historical feature description information including the time range of technology popularity, typical technology combinations, and / or problem patterns.

[0127] The technology background information is converted into a natural language form by using the technology popularity period identification large language model to describe the technology history features corresponding to the current work order request.

[0128] Specifically, the technology popularity period analysis prompt template includes the following guidance:

[0129] The technology component matching guide instructs the technology popularity period identification big language model to match the corresponding technology component in the technology popularity period knowledge graph based on the technology component features identified from the error log;

[0130] The temporal information extraction guide is used to instruct the technology popularity period identification big language model to extract the technology popularity time range, typical technology combinations and / or problem patterns from the technology popularity period knowledge graph based on the matched technology components.

[0131] Contextual reasoning guidance is used to instruct the large language model for identifying the popularity of technology to deduce the technological background corresponding to the current work order request based on the extracted temporal information;

[0132] Description generation guidelines are used to instruct the large language model to generate a natural language form of description information of the historical features of the technology based on the inferred technological background during the period of technological popularity.

[0133] For example, in this embodiment of the application, a portion of the technology popularity period analysis prompt is:

[0134] "You are a professional technology popularity period analysis expert. Please complete a comprehensive technology background analysis based on the provided technology component characteristics. First, based on the input anomaly path, technology framework package name structure, version identifier, and other technology component characteristics, match the corresponding technology components in the technology popularity period knowledge graph. During matching, package name structure, anomaly characteristics, and version identifiers should be considered comprehensively. Then, extract the technology popularity time range, typical technology combinations in different periods, and typical problem patterns in each popularity stage from the matched technology component entries. Next, based on this time-series information, deduce the technology background of the system corresponding to the current work order request, infer the technology era of the system based on the technology combination relationships, and analyze whether the current problem is closer to the peak or decline period of the technology. Based on this, determine the essential bias of the problem type: if it is closer to the peak period, the problem may be biased towards large-scale application, high concurrency, and performance optimization; if it is closer to the decline period, the problem may be biased towards version compatibility, system maintenance, and technology migration. Finally, convert the analysis results into a natural language description in the following format:"

[0135] The error pertains to a {technical component} that was widely used during the {time period}. During this period, the {technical component} was typically used in conjunction with {pairing technology}.

[0136] Based on the current technology's more recent {peak / decline} phase characteristics, these types of errors typically occur in {use cases}, and the nature of the problem leans towards {peak / decline related issues}, with common causes including {problem causes}. Engineers at this stage typically use {solutions} to handle such problems.

[0137] Please describe the technology flexibly according to its actual stage of popularity, ensuring that the problem type analysis matches the technology life cycle stage.

[0138] It should be noted that in the embodiments of this application, the construction of this prompt word takes into account the correlation between the technology lifecycle stage and the nature of the problem. This is because problems during the peak period (i.e., the mainstream adoption period) usually stem from application pressure issues such as performance bottlenecks and insufficient concurrent processing capabilities exposed when the technology is adopted on a large scale. On the other hand, problems during the decline period mostly stem from transitional issues such as compatibility conflicts, maintenance difficulties, and integration of old and new systems caused by technology aging.

[0139] This application links the problem type to the peak or decline phase of the technology, enabling the large model to perform directional analysis and improving the accuracy of locating the technical problems involved in the error logs of current work order requests.

[0140] S3: Using a pre-defined work order allocation language model, the engineer's experience is matched with the technical history feature description information to retrieve the engineer's experience timeline knowledge base and obtain relevant engineer experiences.

[0141] Specifically, it includes the following steps:

[0142] The system uses a pre-defined comprehensive prompt template for work order allocation to guide the large language model for work order allocation in retrieving engineer experience relevant to the technical background of the current work order.

[0143] The comprehensive prompt template for work order assignment includes the following guidelines:

[0144] Experience retrieval guidance is used to instruct the work order allocation big language model to retrieve relevant engineer experiences from the engineer experience timeline knowledge base based on the time range and technical background in the technical history feature description information.

[0145] S4: Analyze the experience of each relevant engineer during the corresponding technology popularization period using the work order allocation big language model, and evaluate its adaptability to the current work order technology background.

[0146] Specifically, it includes the following steps:

[0147] The work order allocation comprehensive prompt word template guides the work order allocation big language model to perform a fit evaluation and generate an engineer evaluation result containing fit reasons;

[0148] The comprehensive prompt template for work order allocation also includes the following guidance:

[0149] The fit assessment guide is used to instruct the work order assignment language model to analyze the specific work content and problem-solving experience of the retrieved engineers based on their experience during the corresponding technology popularization period, and to assess the fit with the current work order technology background.

[0150] S5: Based on the evaluation results of all relevant engineer suitability, generate a work order allocation scheme using the work order allocation language model.

[0151] Specifically, it includes the following steps:

[0152] The work order allocation comprehensive prompt word template guides the work order allocation language model to generate an allocation scheme, and outputs a work order allocation scheme that includes the recommended engineer identifier and the allocation reason.

[0153] The comprehensive prompt template for work order allocation also includes the following guidance:

[0154] The assignment scheme generation guide instructs the work order assignment language model to determine the optimal engineer selection based on the suitability evaluation results of all engineers and generate a detailed assignment scheme containing the reasons for the recommendation.

[0155] In this embodiment of the application, a portion of the work order allocation comprehensive prompt template is "

[0156] You are a professional work order allocation expert. Please complete the matching and allocation scheme for engineer experience based on the provided technical history feature description information. First, based on the time range and technical background in the technical history feature description information, search the engineer experience timeline knowledge base for relevant engineer experiences, focusing on engineers with relevant work experience during the corresponding technology's popularity period, especially those who have handled similar technical components and problem patterns. Then, analyze the specific work content and problem-solving experience of each retrieved engineer during the corresponding technology's popularity period, assessing their suitability with the current work order's technical background. Evaluation dimensions include: the degree of matching with the era of technology use, the depth of use of specific technical components, the similarity between the types of problems handled and the current work order's problem, and the degree of project participation during the corresponding technology's popularity period. Next, based on the suitability evaluation results of all engineers, determine the optimal engineer selection, prioritizing engineers with in-depth practical experience during the corresponding technology's popularity period and who have handled similar problem types. Finally, generate a detailed work order allocation scheme, in the following format:

[0157] Recommended Engineer: {Engineer Identifier}

[0158] Reasons for allocation:

[0159] Historical context matching: The engineer has practical experience in {the corresponding year if applicable} {the relevant technology if applicable}, which matches / does not match the current work order's technology popularity period;

[0160] Technical experience relevance: Have used {technical components} in {project description} if applicable, and have / do not have practical experience in solving {problem type} if applicable;

[0161] Problem-solving ability: Successfully handled similar problems in {if any}, and the solutions used were similar to the current work order problem;

[0162] Overall compatibility rating: {0-100 rating result};

[0163] Please ensure that the assignment reason is based on the engineer's specific experience and facts, reflecting the importance of matching the era and context, and avoid making recommendations based solely on skill tags.

[0164] The working process of the large language model for work order allocation is as follows:

[0165] Step 1: After receiving the technical history feature description information and the comprehensive prompt word template for work order allocation, the big language model for work order allocation starts the matching process for engineers;

[0166] Step 2: Convert the technology history feature description information into a vector representation, and perform similarity retrieval in the engineer experience timeline knowledge base based on RAG technology. Focus on matching the time range and technical background in the technology history feature description information to obtain engineer experience samples with relevant work experience during the corresponding technology popularity period.

[0167] Step 3: Based on the retrieved relevant engineer experiences, the work order allocation language model, in accordance with the requirements of the work order allocation comprehensive prompt word template, uses semantic reasoning to analyze each engineer's specific work content and problem-solving experience during the corresponding technology's popularity period. It conducts a comprehensive suitability assessment from dimensions such as the matching degree of the era background of technology use, the depth of use of specific technology components, the similarity between the types of problems handled and the current work order problem, and the degree of project participation during the corresponding technology's popularity period.

[0168] Step 4: Based on the fit evaluation results of all engineers, the work order allocation big language model determines the optimal engineer selection and generates a detailed work order allocation plan according to the format specified by the prompt word template. The output includes a complete allocation reason including the recommended engineer's identifier, background matching, technical experience relevance, problem-solving ability, and comprehensive fit score.

[0169] In summary, this application provides a work order allocation method that can improve the accuracy of work order allocation.

[0170] This application also proposes a work order allocation system, including a processor and a memory, wherein the memory stores the work order allocation method described in the above embodiments, and is loaded and executed by the processor.

[0171] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element. Moreover, in the description of the embodiments of this application, unless otherwise stated, " / " signifies "or," for example, A / B can mean A or B; "and / or" in this document is merely a description of the relationship between related objects, indicating that three relationships can exist, for example, A and / or B can represent: A alone, A and B simultaneously, and B alone. Furthermore, in the description of the embodiments of this application, "multiple" refers to two or more.

[0172] The above description is merely a specific embodiment of this application, enabling those skilled in the art to understand or implement this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments described herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A work order allocation method, characterized in that, The work order allocation method includes the following steps: Build an engineer experience timeline knowledge base to store each engineer's work experience in chronological order; Obtain the technical history description information of the work order request; By using a pre-defined work order allocation language model, the engineer's experience is matched with the technical history feature description information, and the engineer's experience timeline knowledge base is retrieved to obtain relevant engineer experiences. The work order allocation language model is used to analyze the experience of each relevant engineer during the corresponding technology popularization period, and to evaluate its adaptability to the current work order technology background. The evaluation dimensions used to assess its compatibility with the current technical background of the work order include: the matching degree of the era in which the technology is used, the depth of use of specific technical components, the similarity between the types of problems handled and the current work order problem, and the degree of project participation during the popular period of the corresponding technology. Based on the evaluation results of all relevant engineer suitability, a work order allocation scheme is generated using the aforementioned work order allocation language model; Obtaining the technical history characteristics description information of a work order request includes the following steps: The technology component features are identified from the error logs of the current work order request using a pre-defined technology popularity period identification language model; The technology popularity period identification big language model is used to match the technology component features in a pre-built technology popularity period knowledge graph to determine the popularity period interval of the technology component corresponding to the technology component features and extract the technology background information corresponding to the popularity period interval. The technology background information includes the technology popularity time range, typical technology combinations and / or problem patterns. The technology background information is converted into a natural language form by using the technology popularity period identification large language model to describe the technology history features corresponding to the current work order request.

2. The work order allocation method according to claim 1, characterized in that, The process of using the aforementioned technology popularity period identification large language model to match the technology component features in a pre-built technology popularity period knowledge graph, and generating technology historical feature description information corresponding to the current work order request based on the matching results, includes the following steps: The pre-set technology popularity period analysis prompt word template guides the large language model for technology popularity period identification to perform a complete reasoning process, including technology component matching, temporal information extraction, and descriptive text generation, generating natural language form of technology historical feature description information including the time range of technology popularity, typical technology combinations, and / or problem patterns.

3. The work order allocation method according to claim 2, characterized in that, The technology popularity period analysis prompt template includes the following guidance: The technology component matching guide instructs the technology popularity period identification big language model to match the corresponding technology component in the technology popularity period knowledge graph based on the technology component features identified from the error log; The temporal information extraction guide is used to instruct the technology popularity period identification big language model to extract the technology popularity time range, typical technology combinations and / or problem patterns from the technology popularity period knowledge graph based on the matched technology components. Contextual reasoning guidance is used to instruct the large language model for identifying the popularity of technology to deduce the technological background corresponding to the current work order request based on the extracted temporal information; Description generation guidelines are used to instruct the large language model to generate a natural language form of description information of the historical features of the technology based on the inferred technological background during the period of technological popularity.

4. The work order allocation method according to claim 1, characterized in that, The technical component features include exception class paths extracted from error logs, technical framework package name structure, version identifiers, configuration file formats, programming interface call patterns, and / or error stack hierarchy. The pre-built knowledge graph of technology popularity records the temporal evolution information of each technology component, including the initial release time, mainstream adoption period, decline period, relationship with other technology components, version evolution path, and typical problem patterns in each popularity stage.

5. The work order allocation method according to claim 1, characterized in that, The process of retrieving relevant engineer experiences by using a pre-defined work order allocation language model to match engineer experience with the technical history feature description information and then retrieving the engineer experience timeline knowledge base includes the following steps: The preset work order allocation comprehensive prompt word template guides the work order allocation big language model to perform engineer experience retrieval and obtain engineer experience related to the technical background of the current work order; The comprehensive prompt template for work order assignment includes the following guidelines: Experience retrieval guidance is used to instruct the work order allocation big language model to retrieve relevant engineer experiences from the engineer experience timeline knowledge base based on the time range and technical background in the technical history feature description information.

6. The work order allocation method according to claim 1, characterized in that, The process of analyzing the experience of each relevant engineer during the corresponding technology's popularity period using the aforementioned work order allocation large language model, and assessing its suitability to the current work order technology context, includes the following steps: The work order allocation comprehensive prompt word template guides the work order allocation big language model to perform a fit evaluation and generate an engineer evaluation result containing fit reasons; The comprehensive prompt template for work order allocation also includes the following guidance: The fit assessment guide is used to instruct the work order assignment language model to analyze the specific work content and problem-solving experience of the retrieved engineers based on their experience during the corresponding technology popularization period, and to assess the fit with the current work order technology background.

7. The work order allocation method according to claim 1, characterized in that, Based on the evaluation results of all relevant engineer suitability, the process of generating a work order allocation scheme using the aforementioned work order allocation language model includes the following steps: The work order allocation prompt template guides the work order allocation language model to generate an allocation scheme, and outputs a work order allocation scheme that includes the recommended engineer identifier and the allocation reason. The comprehensive prompt template for work order allocation also includes the following guidance: The assignment scheme generation guide instructs the work order assignment language model to determine the optimal engineer selection based on the suitability evaluation results of all engineers and generate a detailed assignment scheme containing the reasons for the recommendation.

8. The work order allocation method according to claim 1, characterized in that, The engineer experience timeline knowledge base includes engineer identifiers and their corresponding work experience. Each engineer's work experience is recorded by year, and each work experience includes a description of the project participated in, the technical components used, the specific work content, and the key technical issues and solutions handled.

9. A work order allocation system, characterized in that, It includes a processor and a memory, wherein the memory stores the work order allocation method according to any one of claims 1-8, and is loaded and executed by the processor.

Images