Government affair hotline work order management method, system and equipment based on large model and medium

Abstract

The invention provides a government affair hotline work order management method, system and device based on a large model and a medium, and belongs to the technical field of large model application. The method comprises the following steps: recording an incoming call of a government affair hotline, and converting the incoming call into a dialogue text; generating cue words by using the cue word template, and inputting the cue words into the large model to generate a work order abstract; extracting keywords based on the dialogue text, generating an index in combination with a work order abstract, retrieving business knowledge in a government affair hotline business knowledge base, recording the business knowledge in a knowledge list, and retrieving similar work orders in a historical work order knowledge base; according to the knowledge list and the similar work orders, whether work order assignment is carried out or not is judged; if the work order needs to be dispatched, retrieving the department knowledge base and the historical work order knowledge base in sequence according to the work order abstract and the knowledge list, generating a work order distribution suggestion, executing a work order element extraction process, and generating and inputting a work order; and if the work order does not need to be allocated, executing a work order element extraction process, and generating and inputting the work order.

Description

Technical Field

[0001] The present invention belongs to the field of large model application technology, and more specifically relates to a government hotline work order management method, system, equipment and medium based on a large model. Background Art

[0002] In the government hotline service system, the hotline handles a large number of calls for consultations, complaints, suggestions, and more. However, in actual operation, agents face numerous challenges, stemming not only from the sheer workload but also from the complexity and professionalism of problem handling.

[0003] First, agents are required to maintain a high level of concentration in a long, high-intensity work environment, which severely tests their physical and mental health. During the call process, agents must quickly assess the caller's intent and decipher the conversation's topic and purpose. This requires keen observation and analytical skills.

[0004] Secondly, agents need to answer users' questions promptly and accurately during the conversation. Because the government hotline covers a wide range of issues, agents need to possess comprehensive knowledge and strong professionalism to ensure accurate and authoritative responses.

[0005] However, in practice, agents often struggle to provide satisfactory responses to all inquiries. When encountering highly specialized issues or those requiring cross-departmental collaboration, agents often need to transfer calls to specialized agents or refer work tickets to the relevant departments for resolution. This process not only increases the agent's workload but can also extend customer wait times, negatively impacting the service experience.

[0006] To address these challenges, a large-scale model-based intelligent agent text processing method has emerged. This method leverages text processing capabilities to categorize questions and tasks and match them with specialized agents to provide answers. While this approach improves the targeting and efficiency of problem handling to a certain extent, it still has limitations in practice. For example, it often only redirects questions based on specific areas or task types, ignoring the actual need for agents to directly handle a large number of questions. Summary of the Invention

[0007] In response to the above problems, the purpose of the present invention is to provide a government hotline work order management method, system, equipment and medium based on a big model. The conversation content summary capability, work order element extraction capability, knowledge content retrieval capability, work order transfer capability, etc. provided by the big model greatly improves the direct problem handling ability of the seat personnel, reduces unnecessary work order transfers, and alleviates the pressure on professional seats.

[0008] To achieve the above-mentioned purpose, the present invention is implemented through the following technical solutions: In a first aspect, an embodiment of the present application provides a government hotline work order management method based on a large model, comprising: Record incoming calls to government hotlines in real time and convert them into text using voice recognition technology; Use prompt word templates to generate prompt words based on the conversation text, input the prompt words into the big model, and generate a work order summary; Keywords are extracted from the conversation text and combined with the work order summary to generate an index. Business knowledge is retrieved from the government hotline business knowledge base, recorded in the knowledge list, and similar work orders are retrieved from the historical work order knowledge base. Determine whether to assign a work order based on the knowledge list and similar work orders; If work order assignment is required, the department knowledge base and historical work order knowledge base will be searched in turn based on the work order summary and knowledge list. Work order assignment suggestions will be generated based on the search results. After the work order assignment is completed, the work order element extraction process will be executed to generate and enter the work order. If work order allocation is not required, the work order element extraction process is executed to generate and enter the work order.

[0009] In an optional embodiment, the real-time recording of incoming calls to the government hotline and converting the incoming calls into text using speech recognition technology includes: Record incoming calls to government hotlines in real time and generate corresponding text using speech recognition technology; Based on the generated text, text segmentation technology is used to extract the text of citizens' speeches as dialogue text.

[0010] In an optional embodiment, the method of generating prompt words based on the conversation text using a prompt word template and inputting the prompt words into a large model to generate a work order summary includes: Extract role and task information, output format information, constraint information, and example information from the conversation text based on the work order summary prompt word template; Generate work order summary prompt words based on the extracted content, input the work order summary prompt words into the large language model to generate the work order summary.

[0011] In an optional embodiment, the process of extracting keywords based on the conversation text, generating an index based on the work order summary, retrieving business knowledge from the government hotline business knowledge base, recording it in a knowledge list, and retrieving similar work orders from the historical work order knowledge base includes: generating a first index based on the work order summary; Use the BERT model to extract a preset number of keywords from the conversation text as the second index; Searching the government hotline business knowledge base using the first index and the second index; Obtain the retrieved business knowledge, sort the business knowledge in descending order of matching scores, select the three business knowledge with the highest matching scores, and record them in the knowledge list; Searching in a historical work order knowledge base using the first index and the second index; Get the retrieved historical work orders and sort them in descending order of similarity scores.

[0012] In an optional embodiment, determining whether to assign a work order based on the knowledge list and similar work orders includes: Determine whether the highest matching score of business knowledge and the similarity score of historical work orders are both higher than the corresponding score threshold; If yes, then no work order assignment is required; if no, then work order assignment is required.

[0013] In an optional embodiment, searching the department knowledge base and the historical work order knowledge base in sequence based on the work order summary and the knowledge list, and generating work order assignment suggestions based on the search results, includes: Based on the work order summary, the BERT model is used to match the corresponding unit department in the department knowledge base; Based on the matched unit departments, knowledge lists and work order summaries, a work order assignment suggestion prompt word template is used to generate a work order assignment suggestion prompt word, which is then input into the large language model to generate a work order assignment suggestion.

[0014] In an optional embodiment, the work order element extraction process includes: Based on the work order element prompt word template, the BERT model is used to extract the name, phone number, work order title, request content, incident location, incident time, emotional state, urgency, and work order type from the conversation text, and generate the work order element prompt word; Enter the work order element prompt words into the large model to generate and enter the work order.

[0015] In a second aspect, the embodiment of the present application further provides a government hotline work order management system based on a large model, including: The text conversion module is used to record the incoming calls of the government hotline in real time and convert them into text through speech recognition technology; Summary extraction module, which uses prompt word templates to generate prompt words based on the conversation text, inputs the prompt words into the large model, and generates a work order summary; The retrieval module is used to extract keywords based on the conversation text, generate an index based on the work order summary, retrieve business knowledge from the government hotline business knowledge base, record it in the knowledge list, and retrieve similar work orders from the historical work order knowledge base; The dispatch determination module is used to determine whether to dispatch a work order based on the knowledge list and similar work orders; The allocation suggestion generation module is used to search the department knowledge base and the historical work order knowledge base in turn according to the work order summary and knowledge list if work order allocation is required, and generate work order allocation suggestions based on the search results; The work order generation module is used to execute the work order element extraction process, generate and enter work orders.

[0016] In a third aspect, an embodiment of the present application further provides an electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the steps of the government hotline work order management method based on a large model as described in any one of the above items are implemented.

[0017] In a fourth aspect, an embodiment of the present application further provides a storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the government hotline work order management method based on a large model as described in any one of the above items are implemented.

[0018] It can be seen from the above technical solutions that the present invention has the following advantages: The government hotline work order management method based on a large model provided in this application realizes an intelligent upgrade of the entire process of government hotline work order management by integrating advanced technologies such as speech recognition, natural language processing and large language models: from real-time recording of incoming calls and converting them into conversation text, to accurately generating work order summaries using prompt word templates; constructing a dual index by combining keyword extraction and summaries, efficiently retrieving relevant knowledge and similar work orders in the business knowledge base and historical work order library, and providing data support for decision-making; intelligently judging whether work orders need to be assigned based on the search results, and generating scientific allocation suggestions by integrating multi-source information during assignment; finally, automatically extracting key information and generating standardized work orders through preset work order element templates, ensuring the accuracy, timeliness and standardization of work order processing, and significantly improving government service efficiency and public satisfaction.

[0019] This application automatically converts incoming calls into text using voice recognition technology and leverages a large model to quickly generate work order summaries, reducing the time and cost of manual recording and organization. A systematic keyword extraction and index retrieval mechanism significantly improves the efficiency of knowledge matching and similar work order searches, significantly shortening the work order processing cycle.

[0020] Based on dual search results from conversation text and a knowledge base, this application allows the system to dynamically determine whether a work order needs to be assigned and generate accurate assignment recommendations. For example, by matching departmental knowledge bases and analyzing large models, work orders can be directly assigned to the most appropriate department, eliminating the subjectivity and errors of manual assignments and improving the accuracy of work order flow.

[0021] This application can search for relevant knowledge and similar cases in the business knowledge base and historical work order database, forming a knowledge list and work order reference, providing rich background information for work order processing. This not only helps staff quickly understand citizens' demands, but also promotes the accumulation and reuse of knowledge and experience, reducing the cost of handling repetitive issues.

[0022] This application uses a pre-set work order template and the BERT model to automatically extract key information (such as name, phone number, and location) and generate standardized work orders. This process avoids omissions or errors that may occur during manual entry, ensures the integrity and standardization of work order information, and provides a reliable data foundation for subsequent work order processing and statistical analysis.

[0023] This application can quickly respond to citizen requests and reduce wait times. Furthermore, through precise knowledge matching and work order assignment, citizens' issues can be handled more professionally, improving the efficiency and quality of problem solving. Furthermore, standardized work order recording and circulation mechanisms help citizens track the progress of work order processing, enhancing the transparency and credibility of government services. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the technical solution of the present invention, the following is a brief introduction to the drawings required for the description. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0025] Figure 1 A flowchart of the government hotline work order management method based on a large model provided for this application.

[0026] Figure 2 A structural diagram of the government hotline work order management system based on a large model provided for this application.

[0027] Figure 3 This is a schematic diagram of the structure of the electronic device provided in this application. DETAILED DESCRIPTION

[0028] The specific steps of the government hotline work order management method based on the big model will be described in detail below, and various embodiments of the present disclosure will be described more comprehensively. The present disclosure can have various embodiments, and adjustments and changes can be made therein. However, it should be understood that there is no intention to limit the various embodiments of the present disclosure to the specific embodiments disclosed herein, but rather that the present disclosure should be understood to cover all adjustments, equivalents, and / or alternatives that fall within the spirit and scope of the various embodiments of the present disclosure.

[0029] Hereinafter, the terms "include" or "may include" as used in various embodiments of the present disclosure indicate the presence of disclosed functions, operations, or elements, and do not limit the addition of one or more functions, operations, or elements. In addition, as used in various embodiments of the present disclosure, the terms "include," "have," and their cognates are intended only to indicate specific features, numbers, steps, operations, elements, components, or combinations of the foregoing, and should not be understood as excluding the presence of one or more other features, numbers, steps, operations, elements, components, or combinations of the foregoing, or the possibility of adding one or more features, numbers, steps, operations, elements, components, or combinations of the foregoing.

[0030] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0031] See also Figure 1 FIG. 1 is a flowchart of a method for managing government hotline work orders based on a large model in a specific embodiment. The method includes: S1: Record incoming calls to the government hotline in real time and convert them into text using speech recognition technology.

[0032] Specifically, the incoming calls to the government hotline are first recorded in real time, and the corresponding text is generated through speech recognition technology; then, based on the generated text, text segmentation technology is used to extract the text of the citizens' speeches as the dialogue text.

[0033] For example, first obtain the call recording (including the conversation between the citizen and the agent), then extract the citizen's speech content through speech recognition or text segmentation (filtering the agent's words), then convert the citizen's speech content into UTF-8 encoded plain text, correct typos and output the pure conversation text (for example: the sewer at No. 12, XX Road, XX District has been blocked for three days without being repaired).

[0034] It should be noted that since the subsequent steps of this method require the use of multiple knowledge bases for retrieval, in the specific implementation, the knowledge bases need to be sorted before this step. The details are as follows: The knowledge base primarily consists of the government hotline business knowledge base (KN1), the historical work order knowledge base (KN2), and the department knowledge base (KN3). The KN1 knowledge base includes policy documents related to current government hotline issues at all levels, frequently asked questions, and other accumulated government hotline knowledge. This knowledge base needs to be continuously updated and evolved over time. The KN2 knowledge base, on the other hand, is derived from historical work order records accumulated in the current government hotline database and also needs to be updated and evolved over time. Both knowledge bases require the support of a knowledge base management system.

[0035] The KN1 knowledge base can be maintained daily using a knowledge base management system. This includes basic operations such as adding, deleting, modifying, and querying the current KN1 knowledge base, adding new policy provisions and frequently used business Q&A experience records, deleting (usually logically deleting) or hiding outdated / inapplicable knowledge, and updating knowledge records and creating revision records.

[0036] For the KN2 knowledge base, the organization process includes: inputting the work order record table as the source table, and selecting specified fields (including work order title, work order content, keywords, reply records, etc.) during the synchronization process for full or incremental updates.

[0037] For the KN3 knowledge base, this knowledge base mainly includes department unit information, department unit responsibilities, etc. The organization process is the same as the above knowledge base.

[0038] S2: Generate prompt words based on the conversation text using the prompt word template, input the prompt words into the big model, and generate a work order summary.

[0039] In this step, ticket summarization is one of the fundamental capabilities provided by this invention. Its purpose is to provide agents with a real-time summary of incoming calls and understand the caller's intent. Generating ticket summaries requires the text processing capabilities of a large language model. Through the Prompt project, constraints are placed on the large model's processing of text content, resulting in a summary.

[0040] First, based on the work order summary prompt template, role and task information, output format information, constraint information, and example information are extracted from the conversation text. The specific functions of the above information are as follows: Roles and tasks: Set the role identity and tasks to be performed by the large model. Here, the model is a government hotline information analyst who is proficient in natural language and needs to summarize the content of the incoming calls.

[0041] Output format, specify the content format of the summary.

[0042] Constraints: Requirements that need to be considered or avoided when generating content, or goals that need to be achieved.

[0043] Example: Provide call content and summary for large model reference, omitted here due to business privacy.

[0044] Then, a work order summary prompt word is generated based on the extracted content, and the work order summary prompt word is input into the large language model to generate a work order summary.

[0045] For example, in this step, the Prompt project hopes to correctly extract the caller's name, the content of the request, the time and location of the request. When generating the summary, it avoids the creative bias that may occur in large models and imposes some generated content constraints. The following is a prompt example: You are an intelligent information analyst for a government hotline, expert in natural language processing. Your responsibility is to improve the efficiency and quality of hotline services. Leveraging advanced natural language processing, you will conduct in-depth analysis and summarization of call transcripts, accurately identifying citizen conversations, ignoring agent content, and generating ticket summaries detailing the issues handled.

[0046] Output format: where and what happened, the subject, subject, and location of the incident (provinces, cities, districts, streets, communities, roads, buildings, offices, hotels, and other location entities should be as detailed as possible), and the user's demands or requests.

[0047] constraint: 1. Be sure to extract according to the content. Do not make up names of people, places, or times, and do not add irrelevant content information.

[0048] 2. Do not use keywords such as "reflect", "appeal", "expectation", "residents", "citizens", "however", "therefore", "although", "he", "she", "it", "concern", "incoming call", "outgoing call", "relevant departments", "relevant units", etc.

[0049] 3. Directly describe the problem in the conversation and try not to use words such as "consultation" and "suggestion".

[0050] 4. Carefully read the call transcript, distinguishing between the citizen's and the agent's statements. Avoid including the agent's own. Extract key information from the citizen's statement: the event, location, and demands. Describe the event in as much detail as possible, ensuring coherence and correcting any typos.

[0051] 5. Locations can only be extracted from the content. Do not generate them arbitrarily. Do not use content from examples, and do not make them up! 6. Do not output the content in the example. Strictly follow the output format and do not include similar statements such as "relevant department reply" at the end of the output.

[0052] 7. Avoid using too many subjects in your summary. The total word count should not exceed 100.

[0053] Please follow the prompts above to ensure that each work order summary meets the requirements and accurately reflects the demands of citizens.

[0054] The following is a sample output, including the ticket content and a standard answer. Please refer to the standard answer format for output.

[0055] Here is an example: Example 1: xxx Answer 1: xxx End of Example Please summarize the work order according to {} and output it in the format of the standard answer.

[0056] In the above prompt example, "{}" is the part that needs to be replaced with the actual call content.

[0057] S3: Extract keywords based on the conversation text, generate an index based on the work order summary, retrieve business knowledge from the government hotline business knowledge base, record it in the knowledge list, and retrieve similar work orders from the historical work order knowledge base.

[0058] First, the first index is generated based on the work order summary. The BERT model is used to extract a preset number of keywords from the conversation text as the second index.

[0059] Then, the first and second indices are used to search the government hotline business knowledge base. The retrieved business knowledge is sorted in descending order of matching scores, and the three business knowledge with the highest matching scores are selected and recorded in the knowledge list. The returned knowledge includes the title, content, and publishing department.

[0060] At the same time, the first index and the second index are used to search in the historical work order knowledge base; the retrieved historical work orders are obtained, and the historical work orders are sorted in descending order of similarity scores.

[0061] Because knowledge base searches are real-time, ticket summaries and keyword extractions change as calls progress, leading to changes in the knowledge base search results. Therefore, the retrieved knowledge needs to be managed. Retrieved knowledge is organized into lists and sorted by the time it was retrieved. If knowledge retrieved later still appears in the list, it is discarded. Users can manage the knowledge list, including sorting and deleting currently retrieved knowledge.

[0062] S4: Determine whether to assign a work order based on the knowledge list and similar work orders; if work order assignment is required, execute step S5; if work order assignment is not required, execute step S6.

[0063] In a specific implementation, it is first determined whether the highest matching score of the business knowledge and the similarity score of the historical work orders are both higher than the corresponding score thresholds; if so, no work order allocation is required; if not, work order dispatch is required.

[0064] In this step, by evaluating the matching score and similarity score, if both scores are relatively high, it means that the current agent has the knowledge to directly handle the current caller's problem; otherwise, it means that the current agent is unable to directly handle the current caller's problem, and the work order will be assigned to other handling departments or professional agents.

[0065] S5: Based on the work order summary and knowledge list, the department knowledge base and the historical work order knowledge base are searched in turn, and work order assignment suggestions are generated based on the search results. After the work order assignment is completed, the work order element extraction process is executed to generate and enter the work order.

[0066] Specifically, this step first provides agents with analysis and recommendations for work order assignments, including the department to which the work order should be assigned, the reason for the assignment, and corresponding handling recommendations. The work order assignment analysis and recommendation phase primarily includes: preparing work order assignment knowledge and providing work order assignment analysis and recommendations.

[0067] First, the BERT model is used to match the corresponding departmental units in the department knowledge base based on the work order summary. Then, based on the matched departmental units, knowledge list, and work order summary, a work order assignment suggestion prompt word template is used to generate work order assignment suggestion prompt words. These work order assignment suggestion prompt words are then input into the large language model to generate work order assignment suggestions.

[0068] Since department assignment is a text classification problem, a smaller model, such as the BERT model, can be used instead of a large one. Department assignment selects up to three matching departments, again sorted by matching score. After determining the assigned department, the capabilities of the large model are used to generate the reasons for selecting that department and recommendations for addressing the current issues facing that department.

[0069] For example, the prompt used in the work order assignment suggestion template is as follows: Please provide three specific and detailed suggestions for each department in the work order assignment department list based on the work order title, work order content, and department functions, and refer to the handling results of similar historical work orders. Do not include any content that is not reflected in the work order title and work order content.

[0070] Please give the department functions and analyze the reasons why this work order is assigned to each department in the work order assignment department list. The output department name must be consistent with the given work order assignment department and cannot be arbitrarily fabricated.

[0071] The departmental functions may be those of the superiors; please pay attention to the jurisdiction.

[0072] Results of similar historical work orders: "{}" Ticket Title: {} Work order content: {} Department Function: Work order assigned department: {} Output requirements: Provide three specific and detailed suggestions (in the form of a string, not a list). The suggestions and reasons should be different between different functional departments. The number of words in each department's suggestions should be around 200.

[0073] The reason for assignment should state what problem the work order involves and what work the department is responsible for.

[0074] The variable information that needs to be provided in the above prompt includes: the processing results of historical similar work orders, the title of the current work order, the content of the current work order, the functions of the assigned department, and the assigned department.

[0075] When a work order assignment suggestion is generated, the relevant information can be transferred to the corresponding agent for processing based on the suggestion, and a work order can be generated by executing the work order element extraction process.

[0076] S6: Execute the work order element extraction process to generate and enter the work order.

[0077] Work order element extraction facilitates agent entry. When entering a work order, agents need to record the caller's name, phone number, request, work order title, location, time of occurrence, work order type (including consultation, complaint, and report), and emotional state. This feature extracts essential work order elements, allowing agents to make adjustments and modifications, improving work order entry efficiency.

[0078] In this specific implementation, the BERT model is first used to extract the name, phone number, work order title, request content, location, time of incident, emotional state, urgency, and work order type from the conversation text based on the work order element prompt template, and then generates the work order element prompt. The work order element prompt is then input into the large model to generate and enter the work order.

[0079] For example, work order element extraction includes name, phone number, work order title, request content, incident location, incident time, emotional state, urgency, and work order type. Work order element extraction occurs in real time as the call progresses. Agents can verify the extraction results at any time, locking any that meet the requirements and modifying and locking any that do not. Once all extracted content is locked, work order element extraction ceases. Extraction of common work order elements is also implemented using the large model's prompt project. The following is an example of a work order element prompt.

[0080] #Characters and Scenes You are the intelligent assistant of the government hotline, responsible for accurately extracting and summarizing the main information from the conversation texts between citizens and agents.

[0081] Skills and Expertise ###Key information extraction Title - a brief summary of the entire call, including location and county information; Call content - extract the caller's problem in detail, including but not limited to the exact time, location, people involved in the incident, detailed events, the caller's request, etc., to ensure the information is objective, accurate and complete, and avoid missing important details; Location of the incident - comprehensive information on the district, county, street, etc. where the problem occurred; Time of the incident - the time when the issue occurred, or vague expressions such as yesterday, the day before yesterday, or the day of last week. If the content does not include the year XX, do not include it; Emotion classification - Please judge the caller's emotion. You can only choose one from the four options: calm, depressed, excited, and out of control; Urgency - Please judge the urgency of the issue. You can only choose one from the two categories: General and Urgent. Call category - Please determine the type of problem the caller is asking. You can only select one from consultation, help, suggestion, thanks, complaint, invalid and report.

[0082] Keywords—Efficiently extract keywords from the conversation to achieve accurate information retrieval and service matching. Return 5 to 10 words, separated by commas, and avoid unnecessary explanations. For example, "provident fund, processing, Shizhong District, withdrawal." Try to include the region, field, and business.

[0083] # Output display ## Output content requirements: 1. Please extract the title, call content, location, time of incident, call type, emotion classification, and urgency level respectively. If the corresponding key information is not extracted from the text, the corresponding field will be clearly marked as "None".

[0084] ##Output format requirements: 1. Strictly follow the following JSON format to return the results without adding any additional explanations: {{"Title":"...","Call Content":"...","Location":"...","Time":"...","Call Category":"...","Emotional Classification":"...","Emergency":"...","Keywords":"..."}} 2. Do not add any additional explanation at the beginning or end of the returned content. Please strictly follow the output format.

[0085] =========================== Please extract information from the following incoming call conversation text.

[0086] "{}" Additionally, in this step, after extracting work order elements, you can select a work order entry style to enter the work order. This style is often specified by agents when entering work orders, and is primarily reflected in the presentation of the request. The work order entry style is typically designed to better reflect user issues, document the request process, and facilitate work order analysis and management. The work order entry style is also subject to change, depending not only on the type of work order being entered but also on regulatory requirements.

[0087] The selection of work order entry style is also achieved by modifying the Prompt template. The present invention provides agents with a tool for managing style. Agents can propose style requirements, provide style examples, and assemble them in the Prompt template.

[0088] In this embodiment, real-time speech recognition and text segmentation technology are used to accurately record and extract the conversation text of citizen calls. Prompt word templates and large models are used to efficiently generate structured work order summaries to extract core information. A dual-indexing mechanism of keywords and summaries is used to retrieve matching business knowledge and similar work orders from the business knowledge base and historical work order database, forming a knowledge list. Based on intelligent analysis of the knowledge list and similar work orders, work order assignment requirements are dynamically determined. Based on the search results, accurate work order allocation recommendations are generated to optimize resource allocation. Finally, automated work order element extraction and standardized data entry processes ensure information integrity and standardization. This method improves the response speed, processing accuracy, and citizen satisfaction of government hotline services, while providing data support for government decision-making and reducing manual processing costs.

[0089] like Figure 2 As shown, the following is an embodiment of the government hotline work order management system based on a big model provided by the embodiment of the present disclosure. This system and the government hotline work order management method based on a big model in the above-mentioned embodiments belong to the same inventive concept. For details not fully described in the embodiment of the government hotline work order management system based on a big model, please refer to the embodiment of the above-mentioned government hotline work order management method based on a big model.

[0090] A government hotline work order management system based on a large model, including: The text conversion module is used to record incoming calls to the government hotline in real time and convert the calls into text using speech recognition technology.

[0091] The summary extraction module is used to generate prompt words based on the dialogue text using the prompt word template, input the prompt words into the large model, and generate a work order summary.

[0092] The retrieval module is used to extract keywords based on the conversation text, generate an index based on the work order summary, retrieve business knowledge from the government hotline business knowledge base, record it in the knowledge list, and retrieve similar work orders from the historical work order knowledge base.

[0093] The dispatch determination module is used to determine whether to dispatch a work order based on the knowledge list and similar work orders.

[0094] The allocation suggestion generation module is used to search the department knowledge base and the historical work order knowledge base in turn according to the work order summary and knowledge list if work order allocation is required, and generate work order allocation suggestions based on the search results.

[0095] The work order generation module is used to execute the work order element extraction process, generate and enter work orders.

[0096] The government hotline work order management system based on a large model provided in this embodiment realizes accurate recording, intelligent summary, knowledge matching, dynamic dispatch and automatic entry of government hotline work orders through full-process intelligent technology, significantly improving service response speed and processing accuracy, while reducing labor costs and enhancing decision-making data support capabilities.

[0097] Figure 3 A schematic diagram of the hardware structure of an electronic device for implementing various embodiments of the present invention.

[0098] The government hotline work order management method based on a large model provided in the embodiment of the present application can be applied to electronic devices. Those skilled in the art will understand that the electronic device structure involved in the embodiment of the present invention does not constitute a limitation on the electronic device, and the electronic device may include more or fewer components than shown in the figure, or combine certain components, or arrange components differently. In the embodiment of the present invention, the electronic device includes but is not limited to a laptop computer, a desktop computer, a workbench, a personal digital assistant, a server, a blade server, a mainframe computer, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the embodiments of the present application described and / or required herein.

[0099] The electronic device may include a processor, an external memory interface, an internal memory, a universal serial bus (USB) interface, a charging management module, a power management module, a battery, a wireless communication module, an audio module, a speaker, a microphone, a sensor module, a button, a camera, a display, and a SIM card interface, etc.

[0100] A processor may include one or more processing units, such as a central processing unit (CPU), an application processor (AP), a modem processor, a graphics processing unit (GPU), an image signal processor (ISP), a controller, a memory, a video codec, a digital signal processor (DSP), a baseband processor, and / or a neural-network processing unit (NPU). Different processing units may be independent devices or integrated into one or more processors.

[0101] The processor can be the nerve center and command center of the electronic device. The controller can generate operation control signals based on the instruction opcode and timing signal to complete the control of instruction fetching and execution.

[0102] The processor may also include a memory for storing instructions and data. In some embodiments, the memory in the processor is a cache memory. This memory can store instructions or data that the processor has just used or is reusing. If the processor needs to use the instruction or data again, it can directly call it from the memory. This avoids repeated accesses, reduces processor latency, and thus improves system efficiency.

[0103] The external memory interface can be used to connect an external memory card, such as a MicroSD card, to expand the storage capacity of an electronic device. The external memory card communicates with the processor through the external memory interface, enabling data storage. For example, files such as music and videos can be stored on the external memory card.

[0104] Internal memory can be used to store computer-executable program code, which includes instructions. The processor executes the instructions stored in the internal memory to perform various functional applications and data processing of the electronic device. The internal memory can include a program storage area and a data storage area. The internal memory can include high-speed random access memory and non-volatile memory, such as at least one disk storage device, flash memory device, universal flash storage (UFS), etc.

[0105] The wireless communication function of an electronic device can be implemented through an antenna, a wireless communication module, a modem processor, and a baseband processor.

[0106] Wireless communication modules can provide wireless communication solutions for electronic devices, including wireless local area networks (WLAN) (such as wireless fidelity (Wi-Fi) networks), Bluetooth (BT), global navigation satellite system (GNSS), frequency modulation (FM), near field communication (NFC), infrared technology (IR), etc.

[0107] Electronic devices can implement audio functions through audio modules, speakers, receivers, microphones, headphone jacks, and application processors.

[0108] Electronic devices can achieve shooting functions through ISP, camera, video codec, GPU, display and application processor.

[0109] Electronic devices can achieve display functions through GPU, display screen and application processor.

[0110] A GPU is a microprocessor for image processing that connects the display screen to the application processor. It performs mathematical and geometric calculations for graphics rendering. A processor may include one or more GPUs, which execute program instructions to generate or modify display information.

[0111] The display screen is used to display images, videos, etc. The display screen includes a display panel.

[0112] The above-mentioned electronic equipment realizes the government hotline work order management method based on the big model of this application, which realizes the accurate recording and effective dialogue extraction of government hotline call requests through real-time voice recognition and text segmentation technology, and uses prompt word templates and big models to generate structured work order summaries to quickly extract core information and shorten processing time. At the same time, through the multi-dimensional index retrieval mechanism of keywords and summaries, it efficiently matches business knowledge and similar work orders to provide decision-making basis for work order handling, and then dynamically judges the work order dispatch requirements based on the intelligent analysis of knowledge lists and similar work orders to reduce manual intervention errors, and then generates accurate work order allocation suggestions based on the search results to optimize resource allocation and improve problem-solving efficiency. Finally, through automated work order element extraction and standardized entry processes, it ensures information integrity and standardization, supports efficient statistical analysis, and achieves the beneficial effect of significantly improving the response speed, processing accuracy and citizen satisfaction of government hotline services, providing solid support for building an intelligent and transparent government service system.

[0113] The storage medium provided in this application stores a program product that can implement a government hotline work order management method based on a large model.

[0114] The government hotline work order management method based on the big model includes: Record incoming calls to government hotlines in real time and convert them into text using voice recognition technology; Use prompt word templates to generate prompt words based on the conversation text, input the prompt words into the big model, and generate a work order summary; Keywords are extracted from the conversation text and combined with the work order summary to generate an index. Business knowledge is retrieved from the government hotline business knowledge base, recorded in the knowledge list, and similar work orders are retrieved from the historical work order knowledge base. Determine whether to assign a work order based on the knowledge list and similar work orders; If work order assignment is required, the department knowledge base and historical work order knowledge base will be searched in turn based on the work order summary and knowledge list. Work order assignment suggestions will be generated based on the search results. After the work order assignment is completed, the work order element extraction process will be executed to generate and enter the work order. If work order allocation is not required, the work order element extraction process is executed to generate and enter the work order.

[0115] In some possible implementations, the government hotline work order management method based on a big model disclosed herein can be implemented in the form of a program product, which includes a program code. When the program product is run on a terminal device, the program code is used to enable the terminal device to execute the steps described in the above "Exemplary Method" section of this specification according to various exemplary implementations of the present disclosure.

[0116] The storage medium of the present disclosure can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can be, for example, but not limited to, a system, device or component of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination thereof. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof.

[0117] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A government hotline work order management method based on a large model, characterized in that: include: Record incoming calls to government hotlines in real time and convert them into text using voice recognition technology; Use prompt word templates to generate prompt words based on the conversation text, input the prompt words into the big model, and generate a work order summary; Keywords are extracted from the conversation text and combined with the work order summary to generate an index. Business knowledge is retrieved from the government hotline business knowledge base, recorded in the knowledge list, and similar work orders are retrieved from the historical work order knowledge base. Determine whether to assign a work order based on the knowledge list and similar work orders; If work order assignment is required, the department knowledge base and historical work order knowledge base will be searched in turn based on the work order summary and knowledge list. Work order assignment suggestions will be generated based on the search results. After the work order assignment is completed, the work order element extraction process will be executed to generate and enter the work order. If work order allocation is not required, the work order element extraction process is executed to generate and enter the work order.

2. The government hotline work order management method based on a large model according to claim 1 is characterized in that: The real-time recording of incoming calls to the government hotline and the conversion of incoming calls into text using speech recognition technology include: Record incoming calls to government hotlines in real time and generate corresponding text using speech recognition technology; Based on the generated text, text segmentation technology is used to extract the text of citizens' speeches as dialogue text.

3. The government hotline work order management method based on a large model according to claim 1 is characterized in that: The prompt word template is used to generate prompt words according to the dialogue text, and the prompt words are input into the large model to generate a work order summary, including: Extract role and task information, output format information, constraint information, and example information from the conversation text based on the work order summary prompt word template; Generate work order summary prompt words based on the extracted content, input the work order summary prompt words into the large language model to generate the work order summary.

4. The government hotline work order management method based on a large model according to claim 3 is characterized in that: The process extracts keywords based on the conversation text, generates an index based on the work order summary, retrieves business knowledge from the government hotline business knowledge base, records it in the knowledge list, and retrieves similar work orders from the historical work order knowledge base, including: generating a first index based on the work order summary; Use the BERT model to extract a preset number of keywords from the conversation text as the second index; Searching the government hotline business knowledge base using the first index and the second index; Obtain the retrieved business knowledge, sort the business knowledge in descending order of matching scores, select the three business knowledge with the highest matching scores, and record them in the knowledge list; Searching in a historical work order knowledge base using the first index and the second index; Get the retrieved historical work orders and sort them in descending order of similarity scores.

5. The government hotline work order management method based on a large model according to claim 4 is characterized in that: The step of determining whether to assign a work order based on the knowledge list and similar work orders includes: Determine whether the highest matching score of business knowledge and the similarity score of historical work orders are both higher than the corresponding score threshold; If yes, then no work order assignment is required; if no, then work order assignment is required.

6. The government hotline work order management method based on a large model according to claim 5 is characterized in that: The work order summary and knowledge list are searched in sequence in the department knowledge base and the historical work order knowledge base, and work order assignment suggestions are generated based on the search results, including: Based on the work order summary, the BERT model is used to match the corresponding unit department in the department knowledge base; Based on the matched unit departments, knowledge lists and work order summaries, a work order assignment suggestion prompt word template is used to generate a work order assignment suggestion prompt word, which is then input into the large language model to generate a work order assignment suggestion.

7. The government hotline work order management method based on a large model according to claim 6 is characterized in that: The work order element extraction process includes: Based on the work order element prompt word template, the BERT model is used to extract the name, phone number, work order title, request content, incident location, incident time, emotional state, urgency, and work order type from the conversation text, and generate the work order element prompt word; Enter the work order element prompt words into the large model to generate and enter the work order.

8. A government hotline work order management system based on a large model, characterized by: The system adopts the government hotline work order management method based on a large model as described in any one of claims 1 to 7; The system comprises: The text conversion module is used to record the incoming calls of the government hotline in real time and convert them into text through speech recognition technology; Summary extraction module, which uses prompt word templates to generate prompt words based on the conversation text, inputs the prompt words into the large model, and generates a work order summary; The retrieval module is used to extract keywords based on the conversation text, generate an index based on the work order summary, retrieve business knowledge from the government hotline business knowledge base, record it in the knowledge list, and retrieve similar work orders from the historical work order knowledge base; The dispatch determination module is used to determine whether to dispatch a work order based on the knowledge list and similar work orders; The allocation suggestion generation module is used to search the department knowledge base and the historical work order knowledge base in turn according to the work order summary and knowledge list if work order allocation is required, and generate work order allocation suggestions based on the search results; The work order generation module is used to execute the work order element extraction process, generate and enter work orders.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps of the government hotline work order management method based on a large model as described in any one of claims 1 to 7 are implemented.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, the steps of the government hotline work order management method based on a large model as described in any one of claims 1 to 7 are implemented.

Images