Information completion method and device for business work order, equipment and storage medium

Abstract

The invention provides a business work order information completion method and device, equipment and a storage medium, and is suitable for the financial science and technology field, the medical service field and the artificial intelligence field. The method comprises the following steps: acquiring a to-be-complemented business work order and supplementary audio data for the to-be-complemented business work order; performing information identification on the supplementary audio data based on a pre-constructed voice identification model to obtain a supplementary dialogue text; generating an information completion prompt instruction based on the to-be-completed business work order, the supplementary dialogue text and a preset prompt word template; inputting the information completion prompt instruction into a pre-constructed information completion model to enable the information completion model to generate structured completion content according to the information completion prompt instruction; and updating the to-be-completed business work order based on the structured completion content to obtain a target business work order. According to the invention, automatic and intelligent information completion of the business work order can be realized, so that the work order processing quality and efficiency are improved.

Description

Technical Field

[0001] This application applies to the fields of financial technology, medical services, and artificial intelligence, and particularly relates to a method, apparatus, equipment, and storage medium for completing information on business work orders. Background Technology

[0002] In industries such as fintech and healthcare, work orders are the core vehicle for handling and driving event processing, service requests, and problem tracking. For example, in fintech, handling abnormal transactions or troubleshooting system failures requires generating work orders for full traceability; similarly, in healthcare, maintenance records for sophisticated medical equipment or tracking patient service processes rely on work orders for management. However, in these scenarios, work orders are often incomplete when initially created, with missing key details, preventing personnel from immediately and accurately assessing the severity of the event. Complete information must be obtained through subsequent communication.

[0003] In related technologies, information completion mainly relies on two methods. The first is manual updating of work orders after human communication. This method is not only inefficient but also prone to errors when transcribing complex information. The second is using meeting recording transcription tools to generate text minutes; however, these tools only provide unstructured dialogue transcripts and cannot understand the context or fill in the specific structured fields of the work order. Processing personnel still need to read the entire text for secondary refinement and manual input, failing to achieve automated and intelligent completion of existing work orders. Summary of the Invention

[0004] The main objective of this application is to propose a method, apparatus, device, and storage medium for completing information on business work orders, aiming to achieve automated and intelligent information completion of business work orders, thereby improving the quality and efficiency of work order processing.

[0005] To achieve the above objectives, a first aspect of this application proposes a method for completing information on a work order, the method comprising:

[0006] Obtain the work order to be completed and the supplementary audio data for the work order to be completed;

[0007] Information recognition is performed on the supplementary audio data based on a pre-built speech recognition model to obtain supplementary dialogue text;

[0008] Based on the work order to be completed, the supplementary dialogue text, and the preset prompt word template, an information completion prompt instruction is generated;

[0009] The information completion prompt instruction is input into the pre-built information completion model so that the information completion model generates structured completion content based on the information completion prompt instruction;

[0010] The work order to be completed is updated based on the structured completion content to obtain the target work order.

[0011] In some embodiments, the pre-construction process of the speech recognition model includes:

[0012] For each of the aforementioned work orders to be completed, acquire the professional terminology data of the business domain corresponding to the work order to be completed, in order to construct a domain knowledge base;

[0013] Based on historical identification error data in the aforementioned business domain, an identification error correction model is constructed;

[0014] The domain knowledge base and the error correction model are configured in the basic speech recognition model to obtain the speech recognition model.

[0015] In some embodiments, generating an information completion prompt instruction based on the work order to be completed, the supplementary dialogue text, and a preset prompt word template includes:

[0016] Extract initial record information from the work order to be completed, and use it as the first instruction content;

[0017] The supplementary dialogue text shall be used as the content of the second instruction;

[0018] The preset prompt word template is used as the third instruction content; wherein, the third instruction content is used to specify the output format of the structured completion content;

[0019] The first instruction content, the second instruction content, and the third instruction content are concatenated according to a preset concatenation order to generate the information completion prompt instruction.

[0020] In some embodiments, before generating the information completion prompt instruction, the method further includes:

[0021] Extract one or more related entity information from the work order to be completed;

[0022] Based on the associated entity information, relevant historical background information is retrieved from one or more domain knowledge bases;

[0023] The historical background information is then appended to the information completion prompt instruction.

[0024] In some embodiments, the pre-construction process of the information completion model includes:

[0025] Obtain a training dataset containing multiple sets of training samples; each set of training samples includes an initial work order record, a piece of associated dialogue text, and the corresponding standard completion content;

[0026] The initial work order record and the associated dialogue text are combined and input into the initial completion model to obtain the original generated content;

[0027] The initial completion model is supervised and fine-tuned by minimizing the difference between the original generated content and the standard completed content to obtain the pre-built information completion model.

[0028] In some embodiments, before generating an information completion prompt instruction based on the work order to be completed, the supplementary dialogue text, and a preset prompt word template, the method further includes:

[0029] Obtain the business type and processing stage of the work order to be completed;

[0030] Based on the business type and the processing stage, a corresponding prompt word template is determined from a pool of multiple alternative prompt word templates.

[0031] In some embodiments, before updating the work order to be completed, the method further includes:

[0032] The structured completion content is displayed on a preset interactive interface;

[0033] Obtain the confirmation or modification command input by the object through the interactive interface;

[0034] When the confirmation instruction is received, the structured completion content is used to update the work order to be completed;

[0035] When the modification instruction is received, the structured completion content is adjusted according to the modification instruction, and the adjusted content is used to update the work order to be completed.

[0036] To achieve the above objectives, a second aspect of this application provides a work order information completion device, the device comprising:

[0037] The acquisition module is used to acquire the business work order to be completed and the supplementary audio data for the business work order to be completed;

[0038] The information recognition module is used to perform information recognition on the supplementary audio data based on a pre-built speech recognition model to obtain supplementary dialogue text;

[0039] The generation module is used to generate information completion prompt instructions based on the work order to be completed, the supplementary dialogue text, and the preset prompt word template;

[0040] The completion module is used to input the information completion prompt instruction into a pre-built information completion model, so that the information completion model generates structured completion content according to the information completion prompt instruction;

[0041] The update module is used to update the work order to be completed based on the structured completion content to obtain the target work order.

[0042] To achieve the above objectives, a third aspect of this application provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the method described in the first aspect.

[0043] To achieve the above objectives, a fourth aspect of the present application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method described in the first aspect.

[0044] The information completion method, apparatus, device, and storage medium for business work orders proposed in this application first acquire the business work order to be completed and supplementary audio data for the business work order to be completed; perform information recognition on the supplementary audio data based on a pre-built speech recognition model to obtain supplementary dialogue text; generate information completion prompt instructions based on the business work order to be completed, the supplementary dialogue text, and a preset prompt word template; input the information completion prompt instructions into a pre-built information completion model so that the information completion model generates structured completion content according to the information completion prompt instructions; finally, update the business work order to be completed based on the structured completion content to obtain the target business work order.

[0045] According to the information completion method for work orders provided by this invention, firstly, by acquiring the work order to be completed and supplementary audio data, it is ensured that subsequent processes can grasp both the initial structured context from the work order and the source of supplementary information to be analyzed. Next, information recognition is performed based on a pre-built speech recognition model to obtain supplementary dialogue text. This step achieves a crucial data modality conversion, transforming unstructured speech signals into machine-readable text format, providing the necessary prerequisite for subsequent deep semantic analysis. Subsequently, based on the work order, dialogue text, and preset prompt word templates, an information completion prompt instruction is generated. This step overcomes the deficiency of existing technologies that can only provide indiscriminate transcription by integrating "existing context," "new dialogue content," and "target output format" into a highly specific instruction. This provides clear and unambiguous guidance for subsequent models to perform complex and targeted completion tasks. Then, this instruction is input into a pre-built information completion model to generate structured completion content. For example, in a financial system failure review meeting, it can automatically generate a structured description of the event from multi-party discussions; or in a telemedicine consultation, it can automatically organize the diagnosis and treatment discussions into a clear and structured record, ensuring high availability of key information during the conversion process. Finally, based on the generated structured content, the business work order to be completed is automatically updated. This step completes the closed loop of the entire business process, accurately and reliably backfilling the extracted high-quality information into the system records, avoiding errors and omissions that may be caused by manual transcription and secondary entry. This achieves automated and intelligent information completion of business work orders, thereby improving the quality and efficiency of work order processing. Attached Figure Description

[0046] Figure 1 This is a flowchart of the information completion method for business work orders provided in the embodiments of this application;

[0047] Figure 2 This is another flowchart of the information completion method for business work orders provided in the embodiments of this application;

[0048] Figure 3 This is another flowchart of the information completion method for business work orders provided in the embodiments of this application;

[0049] Figure 4 This is another flowchart of the information completion method for business work orders provided in the embodiments of this application;

[0050] Figure 5 This is another flowchart of the information completion method for business work orders provided in the embodiments of this application;

[0051] Figure 6 This is another flowchart of the information completion method for business work orders provided in the embodiments of this application;

[0052] Figure 7This is another flowchart of the information completion method for business work orders provided in the embodiments of this application;

[0053] Figure 8 This is a schematic diagram of the structure of the information completion device for business work orders provided in the embodiments of this application;

[0054] Figure 9 This is a schematic diagram of the hardware structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0055] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0056] It should be noted that although functional modules are divided in the device schematic diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first," "second," etc., in the specification, claims, and the aforementioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0057] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0058] In industries such as fintech and healthcare, work orders are the core vehicle for handling and driving event processing, service requests, and problem tracking. For example, in fintech, handling abnormal transactions or troubleshooting system failures requires generating work orders for full traceability; similarly, in healthcare, maintenance records for sophisticated medical equipment or tracking patient service processes rely on work orders for management. However, in these scenarios, work orders are often incomplete when initially created, with missing key details, preventing personnel from immediately and accurately assessing the severity of the event. Complete information must be obtained through subsequent communication.

[0059] In related technologies, information completion mainly relies on two methods. The first is manual updating of work orders after human communication. This method is not only inefficient but also prone to errors when transcribing complex information. The second is using meeting recording transcription tools to generate text minutes; however, these tools only provide unstructured dialogue transcripts and cannot understand the context or fill in the specific structured fields of the work order. Processing personnel still need to read the entire text for secondary refinement and manual input, failing to achieve automated and intelligent completion of existing work orders.

[0060] Based on this, embodiments of this application provide a method, apparatus, device, and storage medium for completing information on business work orders, which can realize automated and intelligent information completion of business work orders, thereby improving the quality and efficiency of work order processing.

[0061] The information completion method, apparatus, device, and storage medium for business work orders provided in this application are specifically described through the following embodiments. First, the information completion method for business work orders in this application embodiment is described.

[0062] The information completion method for business work orders provided in this application is applicable to the fields of fintech and medical services. This method can be applied to a terminal, a server, or software running on either a terminal or a server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, etc.; the server can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms; the software can be an application implementing the information completion method for business work orders, but is not limited to the above forms.

[0063] This application can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0064] It should be noted that in all specific embodiments of this application, when processing data related to user identity or characteristics, such as user information, user behavior data, user historical data, and user location information, user permission or consent is obtained first. Furthermore, the collection, use, and processing of this data comply with relevant laws, regulations, and standards. In addition, when embodiments of this application require access to sensitive personal information of users, separate permission or consent from the user is obtained through pop-ups or redirection to confirmation pages. Only after obtaining the user's separate permission or consent is the necessary user-related data required for the proper functioning of these embodiments acquired.

[0065] Figure 1 This is an optional flowchart of the information completion method for business work orders provided in the embodiments of this application. Figure 1 The method may include, but is not limited to, steps S101 to S105.

[0066] Step S101: Obtain the business work order to be completed and the supplementary audio data for the business work order to be completed.

[0067] Step S102: Based on the pre-built speech recognition model, information recognition is performed on the supplementary audio data to obtain supplementary dialogue text.

[0068] Step S103: Based on the work order to be completed, the supplementary dialogue text, and the preset prompt word template, generate an information completion prompt instruction.

[0069] Step S104: Input the information completion prompt instruction into the pre-built information completion model so that the information completion model generates structured completion content according to the information completion prompt instruction.

[0070] Step S105: Update the work order to be completed based on the structured completion content to obtain the target work order.

[0071] Steps S101 to S105, as illustrated in this embodiment, firstly, by acquiring the work order to be completed and supplementary audio data, it ensures that subsequent processes can grasp both the initial structured context from the work order and the source of supplementary information to be analyzed. Next, information recognition is performed based on a pre-built speech recognition model to obtain supplementary dialogue text. This step achieves crucial data modality conversion, transforming unstructured speech signals into machine-readable text format, providing the necessary prerequisite for subsequent deep semantic analysis. Subsequently, based on the work order, dialogue text, and preset prompt word templates, information completion prompt instructions are generated. This step overcomes the deficiency of existing technologies that can only provide indiscriminate transcription by integrating "existing context," "new dialogue content," and "target output format" into a highly specific instruction. This instruction provides clear and unambiguous guidance for subsequent models to perform complex and targeted completion tasks. Then, the instruction is input into a pre-built information completion model to generate structured completion content. For example, in a financial system failure review meeting, a structured description of the event process can be automatically generated from multi-party discussions; or in a telemedicine consultation, the discussion can be automatically organized into a clear and structured record, ensuring high availability of key information during the conversion process. Finally, the business work order to be completed is automatically updated based on the generated structured content. This step completes the closed loop of the entire business process, accurately and reliably backfilling the extracted high-quality information into the system records, avoiding errors and omissions that may occur due to manual transcription and secondary entry. This achieves automated and intelligent information completion of business work orders, thereby improving the quality and efficiency of work order processing.

[0072] In step S101 of some embodiments, it is first necessary to locate an incomplete work order record based on the specific time. In the financial field, this could be an initial work order about "the user's login failing due to special characters in the entered username," which could exist in JSON format and only contain fields such as ID, title, and preliminary description. In the medical field, this might be an initial maintenance work order about a malfunction in medical imaging equipment, which only records the equipment number and a brief description of the malfunction. Simultaneously, supplementary audio data will be generated to diagnose the problem. In a financial scenario, this could be a recording of a fault debriefing meeting between the development and operations teams; in a medical scenario, it could be a recording of a remote diagnostic call between a hospital biomedical engineer and the original equipment manufacturer's technical support.

[0073] In step S102 of some embodiments, unstructured information such as supplementary audio data needs to be converted into machine-processable text. To ensure conversion quality, the "pre-built speech recognition model" is professionally enhanced, for example, by configuring a domain knowledge base and error correction model on demand to improve transcription accuracy. In the financial field example, the model is optimized to accurately recognize terms such as "interface," "server," and "development process," thereby transcribing the discussions of a troubleshooting review meeting completely into supplementary dialogue text. In the medical field example, the model is configured with a medical device knowledge base containing terms such as "gradient coil," "radio frequency," and "signal-to-noise ratio" to ensure accurate conversion of diagnostic dialogues between engineers and technical support into text.

[0074] Please see Figure 2 In some embodiments, the pre-construction process of the speech recognition model may include, but is not limited to, steps S201 to S203.

[0075] Step S201: For each work order to be completed, obtain the professional terminology data of the business domain corresponding to the work order to be completed, so as to build a domain knowledge base.

[0076] Step S202: Construct an identification and correction model based on historical identification error data in the business domain.

[0077] Step S203: Configure the domain knowledge base and error correction model into the basic speech recognition model to obtain the speech recognition model.

[0078] In step S201 of some embodiments, specialized terminology data can be collected and organized for specific business domains to construct one or more domain knowledge bases. Specialized terminology data is a highly industry-specific vocabulary set outside the general speech recognition model's knowledge system. In the financial field, this data may include transaction and risk control terms such as "financial risk," "circuit breaker mechanism," and "T+1 settlement." In the medical field, it may include names of precision equipment components, such as "ultrasound probe," names of specific diseases or drugs, such as "aspirin," and diagnostic and treatment operation terms, such as "venous puncture." By structuring this specialized terminology data to form a domain knowledge base, basic vocabulary support is provided for subsequently improving the speech recognition model's ability to understand dialogues in specific industries.

[0079] In step S202 of some embodiments, a recognition error correction model is trained and constructed using historical recognition error data accumulated within the business domain. Historical recognition error data typically exists in the form of "(incorrectly recognized text, correct text)" pairs, reflecting typical, high-frequency errors caused by factors such as homophones, accents, speech rate, or background noise when the basic speech recognition model processes audio in a specific domain. For example, in a fast-paced conversation in a financial transaction scenario, "futures" might be frequently misidentified as "rare commodity"; in a medical ward round transcript, "systolic blood pressure" might be misidentified as "systolic ya". The recognition error correction model learns from massive amounts of such error samples, mastering the contextual patterns of these error-prone points, thereby possessing the ability to perform secondary verification and correction after recognition.

[0080] In step S203 of some embodiments, the domain knowledge base constructed in the preceding steps and the recognition and error correction model are configured and integrated together into a basic speech recognition model, thereby obtaining a deeply optimized speech recognition model for the final information completion task. The domain knowledge base is typically used to optimize the language model layer of the basic model by adjusting the output probabilities of words, making the model more inclined to output professional terms that conform to the business domain during recognition. The recognition and error correction model can serve as a post-processing procedure to refine and correct the initial recognition results of the basic model.

[0081] Through steps S201 to S203, this embodiment of the application significantly improves the accuracy and reliability of the speech recognition process through a systematic optimization strategy. When processing professional dialogues, the main bottleneck of the basic speech recognition model lies in specialized vocabulary and contextual information. The domain knowledge base constructed in step S201 solves the problem of insufficient vocabulary, fundamentally reducing recognition errors caused by unknown words. Based on this, the recognition and error correction model constructed in step S202 further addresses the contextual understanding bias caused by pronunciation similarity and colloquial expressions. Finally, step S203 effectively integrates these two aspects, resulting in a final speech recognition model that possesses both specialized vocabulary and error correction capabilities. This process transforms a general-purpose tool into a highly customized professional tool, ensuring the high fidelity of the "supplementary dialogue text" obtained in the subsequent information completion process.

[0082] In step S103 of some embodiments, unlike traditional meeting recording tools, this application embodiment does not record from scratch, but supplements existing content. A script can be used to concatenate information from multiple parties to generate an "information completion prompt." This prompt is structured as follows: First, extract existing information (such as title and preliminary description) from the work order to be completed; then, append the full text of the supplementary dialogue generated in the previous step; finally, based on the business scenario, select a "preset prompt template" and append it to the end. This template guides the structure and style of the model's output content. For example, a timeline format template can be set, such as: "yyyy-mm-dd hh-mm Problem occurrence and phenomenon description, yyyy-mm-dd hh-mm Emergency handling description, yyyy-mm-dd hh-mm Problem handling resolution," to guide the model to organize and output information in chronological order.

[0083] Please see Figure 3 In some embodiments, step S103 may include, but is not limited to, steps S301 to S304.

[0084] Step S301: Extract initial record information from the work order to be completed, and use it as the first instruction content.

[0085] Step S302: Supplement the dialogue text as the second instruction content.

[0086] Step S303: Use the preset prompt word template as the third instruction content; wherein, the third instruction content is used to specify the output format of the structured completion content.

[0087] Step S304: According to the preset splicing order, the first instruction content, the second instruction content, and the third instruction content are spliced ​​together to generate an information completion prompt instruction.

[0088] In step S301 of some embodiments, key initial record information is extracted from the data structure of the work order to be completed, and this information is used as the first instruction content of the information completion prompt instruction. The extracted content is usually fields that can provide core background for subsequent large language models. For example, in a financial work order where "the user name contains special characters, causing the APP customer login to fail," the system will extract the content of its "title" and "desc" fields. In the scenario of a medical equipment maintenance work order, the system will extract the equipment ID, fault code, and preliminary fault phenomenon description entered by front-line personnel. The purpose of this step is to set a clear context for the entire instruction, ensuring that the large language model can focus on the specific problem to be solved in subsequent processing.

[0089] In step S302 of some embodiments, the supplementary dialogue text generated in step S102 is directly adopted and used as a whole as the second component of the instruction, namely the second instruction content. This part is an unedited, complete dialogue transcript, containing a large amount of detailed information, reasoning process, and final conclusion required to resolve the work order problem. The purpose of this step is to provide the large language model with a detailed, unrefined source of raw information for analysis and mining.

[0090] In step S303 of some embodiments, a matching prompt word template is selected from a preset template library based on the type or processing stage of the current business work order, and this template is used as the third component of the instruction, i.e., the third instruction content. The core function of this template is to explicitly specify the format that the final output structured completion content should follow to the large language model. This is the key to enabling customized completion in this solution. As shown in the provided technical proposal application form, the template can be a text format that requires description according to a specific timeline and event nodes. In other scenarios, the template can also require the output of a JSON code containing preset output items such as "root cause," "risk level," and "solution."

[0091] In step S304 of some embodiments, the first, second, and third instruction contents prepared in the previous three steps are sequentially concatenated according to a preset concatenation order to finally generate a complete information completion prompt instruction. This concatenation order usually follows the logic of "context-information source-instruction format" to maximize the understanding efficiency of the large language model. For example, the final generated instruction text will present the following structure: "{team} has an issue with '{title}', the issue description is: {initial description}, based on the following on-site discussion content: {full dialogue text}, refer to the following content style, adjust and complete the issue information: {output format template}". In this way, a clear and unambiguous machine-executable instruction containing background information, raw data, and formatting requirements is constructed and ready to be input into the information completion model.

[0092] Through steps S301 to S304, this embodiment of the application decomposes an information completion task into three clearly defined information modules (context, original information, and output format), and then merges them into a specific instruction that can be input into the information completion model through orderly concatenation. Step S301 provides the context of "what to do," step S302 provides the basis for "where to find information," and step S303 provides the specification for "in what form to submit the result." This design overcomes the ambiguity and uncertainty brought about by simple questioning methods, ensuring that the output of the information completion model is structured data that can be directly parsed and utilized by business systems.

[0093] Please see Figure 4 In some embodiments, steps S401 to S402 may be included before generating the information completion prompt instruction.

[0094] Step S401: Obtain the business type and processing stage of the work order to be completed.

[0095] Step S402: Based on the business type and processing stage, determine the corresponding prompt word template from the multiple alternative prompt word templates.

[0096] In step S401 of some embodiments, before generating the information completion prompt instruction, the metadata of the work order to be completed is first parsed to obtain its current business type and processing stage. "Business type" refers to the business scope to which the work order belongs, while "processing stage" identifies the specific node in the work order's entire lifecycle. In the financial field, the business type of a work order might be "system failure," "transaction application," or "compliance review"; its processing stage might be "preliminary diagnosis," "root cause analysis," or "review and summary." In the medical field, the business type could be "equipment maintenance" or "patient complaint," and the processing stage could be "fault reporting," "remote troubleshooting," or "on-site repair." This step, by obtaining these two key dimensions of information, provides a decision-making basis for subsequent adaptive operations.

[0097] In step S402 of some embodiments, based on the business type and processing stage obtained in the previous step, the most suitable prompt word template for the current scenario can be accurately determined through a query or matching mechanism from a library containing multiple alternative prompt word templates. This means that this embodiment does not use a single fixed template, but has the ability to dynamically adapt. For example, when a financial "system failure" work order is in the "root cause analysis" stage, the system will select a JSON format template that requires outputting {"direct cause":"","root cause":"","long-term improvement measures":""}. When the same work order flows to the "review and summary" stage, the system will automatically switch to another template that requires outputting {"complete event timeline":"","responsible persons at each stage":"","future prevention items":""}. Similarly, medical "equipment maintenance" work orders will also have prompt word templates with different information structures required in the "remote troubleshooting" stage and the "on-site repair" stage, respectively.

[0098] Through steps S401 and S402, this embodiment of the application first identifies the type and stage of the work order (step S401), and then matches the most suitable output format from the template library accordingly (step S402). This solution ensures that the most accurate and relevant instructions can be generated at any time. This allows the information completion behavior to be deeply understood in the evolution of the business process, improves business-related output, and thus ensures that the final generated target business work order has maximum practical value in every processing stage.

[0099] Please see Figure 5 In some embodiments, steps S501 to S503 may be included before generating the information completion prompt instruction.

[0100] Step S501: Extract one or more related entity information from the work order to be completed.

[0101] Step S502: Based on the associated entity information, retrieve relevant historical background information from one or more domain knowledge bases.

[0102] Step S503: Combine the historical background information into the information completion prompt instruction.

[0103] In step S501 of some embodiments, before the concatenation instruction, an entity recognition scan can be performed on the original work order content to be completed to extract one or more key related entity information. "Related entity information" refers to noun entities explicitly mentioned in the work order text that have unique identifiers or specific referential meanings. In the fintech field, these entities may include specific server names, specific user IDs, or transaction serial numbers. In the healthcare field, these entities may be a medical device serial number, a patient's medical record number, or a specific drug code. The core of this step is to identify the core objects surrounding the dialogue and work order.

[0104] In step S502 of some embodiments, the associated entity information extracted in the previous step is used as a query index to retrieve historical background information related to these entities from one or more relevant domain knowledge bases. Here, "domain knowledge base" is used broadly; it can be an enterprise's configuration management database, historical work order system, electronic medical record system, or equipment maintenance history database, etc. For example, after extracting the server name, the system will automatically query the server's most recent change records, patch history, and performance baseline. After extracting the device serial number, the system will retrieve all past maintenance records and fault history of the device from the maintenance database.

[0105] In step S503 of some embodiments, the historical background information retrieved in the previous step is added as a new independent content module and concatenated into the final generated information completion prompt instruction. That is, the final instruction provided to the information completion model not only includes the initial context of the work order, the complete dialogue text, and the formatted template, but also background information about the core entity. For example, the instruction might explicitly include content such as: "Background information supplement: The server just underwent an operating system kernel update 24 hours ago...", which can provide the information completion model with a deeper level of decision-making basis.

[0106] Through steps S501 to S503, this embodiment of the application injects historical background information into the instruction (step S503) through automated entity recognition (step S501) and historical information retrieval (step S502), enabling the information completion model to discover potential connections between current and historical events. For example, it can determine whether the current failure is a consequence of a system change or whether the failure of a device exhibits a periodic pattern. This capability qualitatively adds analytical and diagnostic capabilities to the information completion model, building upon its existing dialogue record organization capabilities, thereby producing more accurate structured completion content.

[0107] In step S104 of some embodiments, a large language model can be used for semantic understanding and output results. This model will fully understand the complex instructions generated in step S103. In the case of a failed financial login, the model will extract the time of the incident, the emergency response process, and the final handling decision from the discussions between the operations and development teams, and generate a clearly structured text according to the template requirements, such as: "2023-7-8 08:00 Problem occurred... 2023-7-8 09:00 Problem avoidance... 2023-7-8 17:00 Problem handling decision...". In the case of medical equipment maintenance, the model will similarly generate a structured description based on the conversations between engineers, including the precise time of the failure, remote troubleshooting steps, and the final decision to "dispatch an engineer for repair".

[0108] Please see Figure 6 In some embodiments, the pre-construction process of the information completion model may include, but is not limited to, steps S601 to S603.

[0109] Step S601: Obtain a training dataset containing multiple sets of training samples; wherein each set of training samples includes an initial work order record, a piece of associated dialogue text, and the corresponding standard completion content.

[0110] Step S602: Input the initial work order record and associated dialogue text into the initial completion model to obtain the original generated content.

[0111] Step S603: By minimizing the difference between the original generated content and the standard completed content, supervised fine-tuning of the initial completion model is performed to obtain the pre-built information completion model.

[0112] In step S601 of some embodiments, the training dataset consists of multiple sets of training samples. Each training sample contains three parts: an initial work order record with incomplete information, a related dialogue text associated with the work order, and a corresponding "standard completion content". The standard completion content is the expected correct output of the information completion task, which is usually manually compiled and reviewed by business experts in the relevant field to ensure that it meets business requirements in terms of information accuracy and format standardization.

[0113] In step S602 of some embodiments, the initial work order record and associated dialogue text in the training samples are combined, and the combined content is used as input to an initial completion model. After processing, the model generates an original content. The "initial completion model" is a general-purpose, pre-trained large language model with basic language understanding capabilities, but it is not optimized for this specific business. For example, for an abnormal transaction work order in the financial field, the output of the initial model might be a roughly correct text summary, but it differs from the standard completion content in structure and the use of key terms.

[0114] In step S603 of some embodiments, supervised fine-tuning techniques are used to train the initial completion model. During this process, the original generated content from the previous step is compared with the standard completion content given in the training samples, and the difference between the two is calculated. Subsequently, the model's internal parameters are adjusted using an optimization algorithm to minimize this loss value. This process is repeated across the entire training dataset. In this way, the model learns the ability to generate outputs (standard completion content) in a specific format and content from specific inputs (work orders and dialogues), ultimately resulting in a pre-built information completion model specifically designed for this scheme.

[0115] Through steps S601 to S603, this embodiment employs a supervised fine-tuning method to adapt a general language model to a specific business work order completion task. An initial, general model cannot stably output text with a specific structure that meets the requirements of the business system. By training with a dataset validated by experts, the model learns the information extraction rules and output format for that specific scenario. This training process improves the accuracy, reliability, and consistency with the business process of the model's output.

[0116] In step S105 of some embodiments, a preset update script parses the structured completion content generated in the previous step. For example, for a JSON-formatted work order, the script updates the initial work order with the complete description generated by the model, which includes a timeline and processing procedure. In this way, the originally brief work order is transformed into a detailed, clear, and well-defined "target business work order." This process avoids potential omissions in manual recording and the inefficiency of repeated communication, achieving accurate and efficient completion of work order information.

[0117] Please see Figure 7 In some embodiments, before updating the work order to be completed, steps S701 to S704 may be included, but are not limited to.

[0118] Step S701: Display the structured completion content on the preset interactive interface.

[0119] Step S702: Obtain the confirmation command or modification command input by the object through the interactive interface.

[0120] Step S703: When a confirmation instruction is received, the structured completion content is used to update the work order to be completed.

[0121] Step S704: When a modification instruction is received, the structured completion content is adjusted according to the modification instruction, and the adjusted content is used to update the work order to be completed.

[0122] In step S701 of some embodiments, after the information completion model generates structured completion content, it is not immediately written to the work order, but is first displayed on a preset interactive interface. This interactive interface is usually a component of the business work order system, such as a pop-up window for review or a specific panel area. In the work order system of the financial field, the cause analysis and handling suggestions for an abnormal transaction generated by the model will be filled into the interface in the form of editable text for review by the processing personnel. In the medical maintenance scenario, the equipment fault diagnosis process and maintenance decision compiled by the model will also be displayed to the engineer for final confirmation.

[0123] In step S702 of some embodiments, after the content is displayed, an operation instruction from the object (i.e., the user) is awaited and obtained. The operation instructions are mainly divided into two categories: confirmation instructions and modification instructions. If the user reviews the content and deems it accurate, they can issue a confirmation instruction by clicking the "Confirm" or "Accept" button. If deviations, omissions, or further professional judgments are found, the user can directly edit the content in the text box of the interactive interface; this editing and saving action constitutes a modification instruction.

[0124] In step S703 of some embodiments, when a confirmation instruction is received, it indicates that the user fully approves of the generated structured completion content. In this case, the unmodified structured completion content originally generated by the model will be directly used to update the work order to be completed. This process branch maximizes processing efficiency, enabling a rapid flow from information completion to work order update when the model output quality is high and no manual intervention is required.

[0125] In step S704 of some embodiments, when a modification instruction is received, it indicates that the user has manually adjusted the content generated by the model. The system first obtains the final text content edited by the user on the interactive interface. Then, this modified and confirmed content is used to update the work order to be completed. This process branch ensures that when the model output is imperfect or requires expert intervention, the information ultimately recorded in the system is accurate information that has been manually verified and corrected, guaranteeing the reliability of business records.

[0126] Through steps S701 to S704, this application embodiment provides a human-machine collaborative review mechanism as a security guarantee for a fully automated process. In fields such as finance and healthcare, where information accuracy is extremely critical, relying entirely on automated processing carries certain risks. This solution introduces a human review stage, giving domain experts final control and veto power without sacrificing major automation efficiency. It utilizes a model to handle information organization and structuring, while a concise interactive process transfers professional judgment to the final responsible party, making this technical solution more feasible and secure in practical applications.

[0127] The information completion method, apparatus, device, and storage medium for business work orders proposed in this application first acquire the business work order to be completed and supplementary audio data for the business work order to be completed; perform information recognition on the supplementary audio data based on a pre-built speech recognition model to obtain supplementary dialogue text; generate information completion prompt instructions based on the business work order to be completed, the supplementary dialogue text, and a preset prompt word template; input the information completion prompt instructions into a pre-built information completion model so that the information completion model generates structured completion content according to the information completion prompt instructions; finally, update the business work order to be completed based on the structured completion content to obtain the target business work order.

[0128] According to the information completion method for work orders provided by this invention, firstly, by acquiring the work order to be completed and supplementary audio data, it is ensured that subsequent processes can grasp both the initial structured context from the work order and the source of supplementary information to be analyzed. Next, information recognition is performed based on a pre-built speech recognition model to obtain supplementary dialogue text. This step achieves a crucial data modality conversion, transforming unstructured speech signals into machine-readable text format, providing the necessary prerequisite for subsequent deep semantic analysis. Subsequently, based on the work order, dialogue text, and preset prompt word templates, an information completion prompt instruction is generated. This step overcomes the deficiency of existing technologies that can only provide indiscriminate transcription by integrating "existing context," "new dialogue content," and "target output format" into a highly specific instruction. This provides clear and unambiguous guidance for subsequent models to perform complex and targeted completion tasks. Then, this instruction is input into a pre-built information completion model to generate structured completion content. For example, in a financial system failure review meeting, it can automatically generate a structured description of the event from multi-party discussions; or in a telemedicine consultation, it can automatically organize the diagnosis and treatment discussions into a clear and structured record, ensuring high availability of key information during the conversion process. Finally, based on the generated structured content, the business work order to be completed is automatically updated. This step completes the closed loop of the entire business process, accurately and reliably backfilling the extracted high-quality information into the system records, avoiding errors and omissions that may be caused by manual transcription and secondary entry. This achieves automated and intelligent information completion of business work orders, thereby improving the quality and efficiency of work order processing.

[0129] Please see Figure 8 This application also provides a device for completing information on business work orders, which can implement the above-mentioned method for completing information on business work orders. The device includes:

[0130] The acquisition module is used to acquire the business work order to be completed and the supplementary audio data for the business work order to be completed;

[0131] The information recognition module is used to perform information recognition on the supplementary audio data based on a pre-built speech recognition model to obtain supplementary dialogue text;

[0132] The generation module is used to generate information completion prompts based on the work order to be completed, the supplementary dialogue text, and the preset prompt word templates.

[0133] The completion module is used to input information completion prompts into a pre-built information completion model, so that the information completion model can generate structured completion content based on the information completion prompts.

[0134] The update module is used to update the work order to be completed based on the structured completion content, so as to obtain the target work order.

[0135] The specific implementation of the information completion device for this business work order is basically the same as the specific implementation of the information completion method for the business work order described above, and will not be repeated here.

[0136] This application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the aforementioned information completion method for business work orders. This electronic device can be any smart terminal, including tablet computers, in-vehicle computers, etc.

[0137] Please see Figure 9 , Figure 9 The hardware structure of an electronic device according to another embodiment is illustrated. The electronic device includes:

[0138] The processor 901 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application.

[0139] The memory 902 can be implemented as a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 902 can store the operating system and other applications. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 902, and the processor 901 calls and executes the information completion method for the business work order in the embodiments of this application.

[0140] The input / output interface 903 is used to implement information input and output;

[0141] The communication interface 904 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0142] Bus 905 transmits information between various components of the device (e.g., processor 901, memory 902, input / output interface 903, and communication interface 904);

[0143] The processor 901, memory 902, input / output interface 903, and communication interface 904 are connected to each other within the device via bus 905.

[0144] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for completing information on business work orders.

[0145] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0146] The information completion method, apparatus, device, and storage medium for business work orders proposed in this application first acquire the business work order to be completed and supplementary audio data for the business work order to be completed; perform information recognition on the supplementary audio data based on a pre-built speech recognition model to obtain supplementary dialogue text; generate information completion prompt instructions based on the business work order to be completed, the supplementary dialogue text, and a preset prompt word template; input the information completion prompt instructions into a pre-built information completion model so that the information completion model generates structured completion content according to the information completion prompt instructions; finally, update the business work order to be completed based on the structured completion content to obtain the target business work order.

[0147] According to the information completion method for work orders provided by this invention, firstly, by acquiring the work order to be completed and supplementary audio data, it is ensured that subsequent processes can grasp both the initial structured context from the work order and the source of supplementary information to be analyzed. Next, information recognition is performed based on a pre-built speech recognition model to obtain supplementary dialogue text. This step achieves a crucial data modality conversion, transforming unstructured speech signals into machine-readable text format, providing the necessary prerequisite for subsequent deep semantic analysis. Subsequently, based on the work order, dialogue text, and preset prompt word templates, an information completion prompt instruction is generated. This step overcomes the deficiency of existing technologies that can only provide indiscriminate transcription by integrating "existing context," "new dialogue content," and "target output format" into a highly specific instruction. This provides clear and unambiguous guidance for subsequent models to perform complex and targeted completion tasks. Then, this instruction is input into a pre-built information completion model to generate structured completion content. For example, in a financial system failure review meeting, it can automatically generate a structured description of the event from multi-party discussions; or in a telemedicine consultation, it can automatically organize the diagnosis and treatment discussions into a clear and structured record, ensuring high availability of key information during the conversion process. Finally, based on the generated structured content, the business work order to be completed is automatically updated. This step completes the closed loop of the entire business process, accurately and reliably backfilling the extracted high-quality information into the system records, avoiding errors and omissions that may be caused by manual transcription and secondary entry. This achieves automated and intelligent information completion of business work orders, thereby improving the quality and efficiency of work order processing.

[0148] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

[0149] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this application, and may include more or fewer steps than shown, or combine certain steps, or different steps.

[0150] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0151] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.

[0152] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0153] It should be understood that in this application, "at least one (item)" means one or more, and "more than one" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0154] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of the units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0155] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0156] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0157] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0158] The preferred embodiments of the present application have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and substance of the embodiments of the present application shall be within the scope of the claims of the present application.

Claims

1. A method for completing information on a business work order, characterized in that, The method includes: Obtain the work order to be completed and the supplementary audio data for the work order to be completed; Information recognition is performed on the supplementary audio data based on a pre-built speech recognition model to obtain supplementary dialogue text; Based on the work order to be completed, the supplementary dialogue text, and the preset prompt word template, an information completion prompt instruction is generated; The information completion prompt instruction is input into the pre-built information completion model so that the information completion model generates structured completion content based on the information completion prompt instruction; The work order to be completed is updated based on the structured completion content to obtain the target work order.

2. The method for completing information on a business work order according to claim 1, characterized in that, The pre-construction process of the speech recognition model includes: For each of the aforementioned work orders to be completed, acquire the professional terminology data of the business domain corresponding to the work order to be completed, in order to construct a domain knowledge base; Based on historical identification error data in the aforementioned business domain, an identification error correction model is constructed; The domain knowledge base and the error correction model are configured in the basic speech recognition model to obtain the speech recognition model.

3. The method for completing information on a business work order according to claim 1, characterized in that, The step of generating information completion prompt instructions based on the work order to be completed, the supplementary dialogue text, and the preset prompt word template includes: Extract initial record information from the work order to be completed, and use it as the first instruction content; The supplementary dialogue text shall be used as the content of the second instruction; The preset prompt word template is used as the third instruction content; wherein, the third instruction content is used to specify the output format of the structured completion content; The first instruction content, the second instruction content, and the third instruction content are concatenated according to a preset concatenation order to generate the information completion prompt instruction.

4. The method for completing information on a work order according to claim 3, characterized in that, Before generating information completion prompts, the method further includes: Extract one or more related entity information from the work order to be completed; Based on the associated entity information, relevant historical background information is retrieved from one or more domain knowledge bases; The historical background information is then appended to the information completion prompt instruction.

5. The method for completing information on a business work order according to claim 1, characterized in that, The pre-construction process of the information completion model includes: Obtain a training dataset containing multiple sets of training samples; each set of training samples includes an initial work order record, a piece of associated dialogue text, and the corresponding standard completion content; The initial work order record and the associated dialogue text are combined and input into the initial completion model to obtain the original generated content; The initial completion model is supervised and fine-tuned by minimizing the difference between the original generated content and the standard completed content to obtain the pre-built information completion model.

6. The method for completing information on a business work order according to claim 1, characterized in that, Before generating information completion prompts based on the work order to be completed, the supplementary dialogue text, and the preset prompt word template, the method further includes: Obtain the business type and processing stage of the work order to be completed; Based on the business type and the processing stage, a corresponding prompt word template is determined from a pool of multiple alternative prompt word templates.

7. The method for completing information on a business work order according to claim 1, characterized in that, Before updating the work order to be completed, the method further includes: The structured completion content is displayed on a preset interactive interface; Obtain the confirmation or modification command input by the object through the interactive interface; When the confirmation instruction is received, the structured completion content is used to update the work order to be completed; When the modification instruction is received, the structured completion content is adjusted according to the modification instruction, and the adjusted content is used to update the work order to be completed.

8. A device for completing information on a work order, characterized in that, The device includes: The acquisition module is used to acquire the business work order to be completed and the supplementary audio data for the business work order to be completed; The information recognition module is used to perform information recognition on the supplementary audio data based on a pre-built speech recognition model to obtain supplementary dialogue text; The generation module is used to generate information completion prompt instructions based on the work order to be completed, the supplementary dialogue text, and the preset prompt word template; The completion module is used to input the information completion prompt instruction into a pre-built information completion model, so that the information completion model generates structured completion content according to the information completion prompt instruction; The update module is used to update the work order to be completed based on the structured completion content to obtain the target work order.

9. An electronic device, characterized in that, The electronic device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the information completion method for business work orders as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, characterized in that, when the computer program is executed by a processor, it implements the information completion method for business work orders as described in any one of claims 1 to 7.

Images