Work order management methods, devices, equipment, and storage media based on AI analysis

Abstract

This invention proposes an AI-based work order management method, device, equipment, and storage medium. The method includes: creating a first work order based on work order description information and a first work order type; determining a first property management line based on the first work order type; obtaining first and second description information through analysis using a first NLP model; determining a second property management line based on the second description information; analyzing a first and second emotion category using a second NLP model; determining corresponding first and second target tags; assigning a first work order to a first account based on the first target tag; and creating and assigning a second work order to a second account based on the second target tag. This method can analyze emotion categories from description information using a second NLP model, achieve accurate matching of response accounts based on target tags corresponding to emotion categories, and automatically create and assign second work orders without manual work order transfer, thus improving work order management efficiency.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, and in particular to a work order management method, apparatus, equipment, and storage medium based on AI analysis. Background Technology

[0002] Property management work orders are an important tool for property management companies to record, track, and process various work-related matters. Common property management work orders include complaint work orders and maintenance work orders initiated by owners, as well as various internal work orders initiated by management personnel according to the approval process. In property management companies, due to the large number of work order types and management personnel, it is necessary to quickly identify the respondent for each work order in order to improve management efficiency.

[0003] When creating a property management work order, the work order type is usually selected. Artificial intelligence technology can be used to analyze the work order type to determine the respondent and automate the work order allocation. However, even for the same type of property management work order, there are usually multiple managers who can respond. If the initiator of the work order has specific requirements for the respondent—for example, a maintenance work order might require a highly specialized repairman or a repairman with fast response time—or an internal work order might require skipping multiple internal process nodes, and a respondent is randomly assigned based on the work order type using relevant technology, with the respondent then forwarding the work order or creating a new work order according to the initiator's requirements—human intervention is still required in the work order management process, resulting in low management efficiency. Summary of the Invention

[0004] This invention aims to at least solve one of the technical problems existing in the prior art. To this end, this invention proposes a work order management method, device, equipment, and storage medium based on AI analysis, which can achieve accurate management of property work orders through AI model analysis and improve the management efficiency of property work orders.

[0005] In a first aspect, embodiments of the present invention provide an AI-based work order management method applied to a work order management system. The work order management system pre-configures a first NLP model, a second NLP model, and multiple available property lines. Each available property line is associated with multiple available accounts and multiple selectable work order types. Each available account is associated with multiple response capability tags. The method includes:

[0006] A first work order is created based on the work order description information and the first work order type, and a first property line is determined based on the first work order type, wherein the work order description information is used to indicate the root cause of the creation of the first work order;

[0007] The work order description information and the available property lines are input into the first NLP model. When the first NLP model extracts first description information and at least one second description information from the work order description information, it obtains the second property lines identified based on each of the second description information. The first description information can be used to identify the first property line.

[0008] The first emotion category is identified from the first description information based on the second NLP model, and the second emotion category is identified from the second description information.

[0009] Based on the first emotion category and the first mapping table, at least one first target label is determined, and based on the first target label and the first property line, a first account is determined. The first work order is assigned to the first account, wherein the first mapping table records the mapping relationship between optional emotion categories and the response capability label.

[0010] Based on any of the second description information, at least one second target label is determined based on the corresponding second emotion category and the first mapping table. Based on the second target label and the second property line, a second account is determined, and a second work order is created and assigned to the second account.

[0011] According to some embodiments of the present invention, after inputting the work order description information into the first NLP model, the method further includes:

[0012] When the first NLP model outputs the first property line and at least one second property line, the identification log of the first NLP model based on the work order description information is obtained;

[0013] Based on the identification log, the identification words used to identify the first property line are combined into the first description information, and the identification words used to identify the second property line are combined into the second description information, wherein the identification words are obtained by splitting the work order description information.

[0014] According to some embodiments of the present invention, each of the optional work order types is pre-associated with at least one preset operation term, and creating a second work order and assigning it to the second account includes:

[0015] The optional work order types of the second property management line are determined as candidate work order types;

[0016] At least one descriptive operation word is extracted from the second description information, and a second work order type is determined from a plurality of candidate work order types based on the descriptive operation word, wherein the descriptive operation word is semantically the same as or semantically similar to any of the preset operation words of the second work order type;

[0017] When the second description information contains historical work orders, the available account to which the historical work orders are assigned is identified as the third account, and the third account is identified as the work order object;

[0018] The second work order is constructed based on the second work order type and the third account, and the second work order is assigned to the second account.

[0019] According to some embodiments of the present invention, after assigning the second work order to the second account, the method further includes:

[0020] When the second emotion category is used to indicate a negative emotion, a scoring base is determined based on the second emotion category and a preset second mapping table, wherein the second mapping table records the mapping relationship between the selectable emotion categories and the selectable base.

[0021] After the first work order is completed, the first evaluation information based on the feedback of the first work order is obtained, and the first evaluation information is input into the second NLP model to determine the third emotion category.

[0022] When the third emotion category is used to indicate positive emotions, the target assessment score is determined based on a preset correction coefficient and the scoring base.

[0023] The target assessment score is applied to the third account.

[0024] According to some embodiments of the present invention, the optional emotion category is pre-associated with a response priority, and determining a first account based on the first target tag and the first property line, and assigning the first work order to the first account includes:

[0025] When multiple candidate accounts are matched based on the first target tag and the first property line, the work order queue of each candidate account is obtained, wherein the work order queue includes multiple orderly arranged work orders to be processed, and each work order to be processed is associated with a corresponding fourth emotion category.

[0026] Based on the first emotion category, traverse each of the work order queues;

[0027] When the first emotion category matches the same fourth emotion category, the next element of the pending work order corresponding to the matched fourth emotion category is determined as the target sorting;

[0028] When the first emotion category does not match the same fourth emotion category, the response priority corresponding to each fourth emotion category is determined, and the target sorting of the first work order is determined based on the response priority corresponding to the first emotion category.

[0029] The candidate account corresponding to the highest target ranking is determined as the first account, and the first work order is inserted into the work order queue of the first account based on the target ranking.

[0030] According to some embodiments of the present invention, after creating a second work order and assigning it to the second account, the method further includes:

[0031] A first capability tag and a second capability tag are determined from the plurality of response capability tags associated with the first account, and a third capability tag and a fourth capability tag are determined from the plurality of response capability tags associated with the second account, wherein the first capability tag and the third capability tag are used to indicate business capabilities, and the second capability tag and the fourth capability tag are used to indicate work order acceptance capabilities.

[0032] The second capability tag is updated based on the updated work order queue of the first account, and the fourth capability tag is updated based on the updated work order queue of the second account.

[0033] Obtain the second evaluation information in response to the second work order, and input the second evaluation information into the second NLP model to determine the fifth emotion category;

[0034] At least one third target label is determined based on the third emotion category and the first mapping table, and at least one fourth target label is determined based on the fifth emotion category and the first mapping table;

[0035] The first capability label is updated based on the third target label, and the third capability label is updated based on the fourth target label.

[0036] According to some embodiments of the present invention, before assigning the first work order to the first account, the method further includes:

[0037] A first work order process is determined based on the first work order type, wherein the first work order process includes multiple first process nodes;

[0038] The first description information and each of the first process nodes are input into the first NLP model, and the target node corresponding to the first description information is identified by the first NLP model.

[0039] Configure the first work order based on the target node.

[0040] Secondly, embodiments of the present invention provide an AI-based work order management device, including at least one control processor and a memory for communicatively connecting with the at least one control processor; the memory stores instructions executable by the at least one control processor, the instructions being executed by the at least one control processor to enable the at least one control processor to perform the AI-based work order management method as described in the first aspect above.

[0041] Thirdly, embodiments of the present invention provide an electronic device including an AI-based work order management device as described in the second aspect above.

[0042] Fourthly, embodiments of the present invention provide a computer-readable storage medium storing computer-executable instructions for executing the AI-based work order management method as described in the first aspect above.

[0043] The AI-based work order management method according to embodiments of the present invention has at least the following beneficial effects: A first work order is created based on work order description information and a first work order type; a first property line is determined based on the first work order type, wherein the work order description information is used to indicate the root cause of the creation of the first work order; the work order description information and the available property lines are input into a first NLP model; when the first NLP model extracts first description information and at least one second description information from the work order description information, it obtains second property lines identified based on each of the second description information, wherein the first description information can be used to identify the first property line; based on the second NLP model... The system identifies a first emotion category from the first description information and a second emotion category from the second description information. Based on the first emotion category and a first mapping table, it determines at least one first target tag, and based on the first target tag and a first property management line, it determines a first account and assigns the first work order to the first account. The first mapping table records the mapping relationship between selectable emotion categories and the response capability tags. Based on any second description information, based on the corresponding second emotion category and the first mapping table, it determines at least one second target tag, and based on the second target tag and the second property management line, it determines a second account, creates a second work order, and assigns it to the second account. According to the technical solution of this embodiment, at least one second description information can be obtained by analyzing work order description information using a first NLP model, and an emotion category can be analyzed from the description information using a second NLP model. A target tag matching the emotion category is determined, and an account is accurately matched using the target tag. After identifying the second account, a second work order is automatically created and assigned, eliminating the need for manual work order transfer and improving work order management efficiency. Attached Figure Description

[0044] Figure 1 This is a schematic diagram of a work order management system provided in one embodiment of the present invention;

[0045] Figure 2 This is a flowchart of a work order management method based on AI analysis provided in another embodiment of the present invention;

[0046] Figure 3 This is a complete flowchart of a work order management method based on AI analysis provided in another embodiment of the present invention;

[0047] Figure 4 This is a structural diagram of an AI-based work order management device provided in another embodiment of the present invention. Detailed Implementation

[0048] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0049] In the description of this invention, it should be understood that the orientation descriptions, such as up, down, front, back, left, right, etc., are based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limiting this invention.

[0050] In the description of this invention, "several" means one or more, "more than" means two or more, "greater than," "less than," and "exceeding" are understood to exclude the stated number, while "above," "below," and "within" are understood to include the stated number. The use of "first" and "second" in the description is merely for distinguishing technical features and should not be construed as indicating or implying relative importance, or implicitly indicating the number of indicated technical features, or implicitly indicating the order of the indicated technical features.

[0051] In the description of this invention, unless otherwise explicitly defined, terms such as "set up," "install," and "connect" should be interpreted broadly, and those skilled in the art can reasonably determine the specific meaning of the above terms in this invention in conjunction with the specific content of the technical solution.

[0052] The following is an explanation of the technical terms used in this application:

[0053] Property Management Division: Commonly used in the property management or related industries, this refers to a department or business area within a property management company that is specifically responsible for specific property management matters based on function or business division. The property management division covers multiple aspects of property management, including but not limited to customer service, facilities management, security management, environmental maintenance, and financial management.

[0054] Natural Language Processing (NLP) models: NLP models are an important direction in computer science and the branch of artificial intelligence. They study various theories and methods that enable effective communication between humans and computers using natural language. Simply put, the goal of NLP is to enable computers to "understand" human language, comprehend its meaning, and respond accordingly or perform specific tasks.

[0055] This invention provides an AI-based work order management method, apparatus, device, and storage medium. The AI-based work order management method includes: creating a first work order based on work order description information and a first work order type; determining a first property line based on the first work order type; wherein the work order description information is used to indicate the root cause of the creation of the first work order; inputting the work order description information and the available property lines into a first NLP model; and when the first NLP model extracts first description information and at least one second description information from the work order description information, obtaining second property lines identified based on each of the second description information, wherein the first description information can be used to identify the first property line. Based on the second NLP model, a first emotion category is identified from the first description information, and a second emotion category is identified from the second description information. At least one first target label is determined based on the first emotion category and a first mapping table, and a first account is determined based on the first target label and the first property management line. The first work order is assigned to the first account, wherein the first mapping table records the mapping relationship between selectable emotion categories and the responsiveness label. Based on any second description information, at least one second target label is determined based on the corresponding second emotion category and the first mapping table, and a second account is determined based on the second target label and the second property management line. A second work order is created and assigned to the second account. According to the technical solution of this embodiment, at least one second description information can be obtained by analyzing work order description information through a first NLP model, an emotion category can be analyzed from the description information through a second NLP model, and a target label matching the emotion category can be determined. Accurate account matching is achieved through the target label, and a second work order is automatically created and assigned after identifying the second account, eliminating the need for manual work order transfer and improving work order management efficiency.

[0056] First, refer to Figure 1 , Figure 1 This is a schematic diagram of the work order management system provided in an embodiment of the present invention. The work order management system 10 can communicate with the user terminal 20. The user terminal 20 can be the owner's mobile phone, the manager's office computer, etc. The owner's mobile phone can initiate a work order through the property management APP, and the office computer can directly access the work order management system 10 to initiate a work order. In this embodiment, there is no limitation on the initiating terminal of the first work order.

[0057] The following is based on the appendix Figure 1 The work order management system shown further illustrates the technical solution of this embodiment of the invention.

[0058] Refer to 1 and Figure 2 , Figure 2 The flowchart illustrates an AI-based work order management method provided in this embodiment of the invention. The work order management system of this embodiment pre-defines a first NLP model, a second NLP model, and multiple available property lines. Each available property line is associated with multiple available accounts and multiple selectable work order types. Each available account is associated with multiple response capability tags. This AI-based work order management method includes, but is not limited to, the following steps:

[0059] S10, create a first work order based on the work order description information and the first work order type, and determine the first property line based on the first work order type, wherein the work order description information is used to indicate the root cause of the creation of the first work order;

[0060] S20, input the work order description information and available property lines into the first NLP model. When the first NLP model extracts the first description information and at least one second description information from the work order description information, it obtains the second property lines identified based on each of the second description information. The first description information can be used to identify the first property lines.

[0061] S30, Based on the second NLP model, identify the first emotion category from the first descriptive information, and identify the second emotion category from the second descriptive information;

[0062] S40, determine at least one first target label based on the first emotion category and the first mapping table, determine the first account based on the first target label and the first property line, and assign the first work order to the first account, wherein the first mapping table records the mapping relationship between optional emotion categories and response capability labels;

[0063] S50, based on any second description information, determine at least one second target label based on the corresponding second emotion category and the first mapping table, determine a second account based on the second target label and the second property line, create a second work order and assign it to the second account.

[0064] It should be noted that the "property management line" can refer to various lines within a property management company, such as the cleaning line, maintenance line, customer service line, and finance line. Work orders initiated for cleaning issues can be assigned to the cleaning line for response. For example, if a sanitation problem occurs in a certain area of ​​the community, a work order initiated for damage to property equipment can be assigned to the maintenance line for response.

[0065] It should be noted that the available accounts can be the accounts of managers in each line. Each available account corresponds to a manager who can respond to property work orders. Each available property line can include multiple available accounts, and each available account can correspond to multiple available property lines. For example, senior managers can be associated with the cleaning line and the maintenance line. When a homeowner raises a complaint about cleaning and maintenance, the senior manager will handle it. That is, the available property lines and available accounts in this embodiment can have a many-to-many relationship, which will not be repeated hereafter.

[0066] It should be noted that optional work order types can be preset, such as complaint work orders, repair work orders, and internal process work orders. Each optional work order type corresponds to only one optional property management line, ensuring that the first property management line can be matched based on the first work order type. For example, complaint work orders correspond to the customer service line, repair work orders correspond to the repair line, and internal process work orders correspond to the optional property management line determined by the specific process. For example, an initiated budget approval work order corresponds to the finance line.

[0067] It should be noted that the responsiveness tag is used to characterize the responsiveness of an available account, for example... Figure 1 As shown, response capability tags can include strong professional skills, rapid response, and good service attitude. These tags can be preset according to the actual situation of each available account. For example, the tag of strong professional skills can be added to the available account of an experienced maintenance employee. The response capability tags of each available account can also be dynamically adjusted. For example, when an available account has no work orders being processed, the tag of rapid response can be automatically added to it to indicate that the available account can respond to work orders quickly. The specific response capability tags can be set according to the actual situation, and no further restrictions are imposed here.

[0068] It should be noted that, as Figure 1As shown, when creating the first work order, a visual page can be built on the user terminal. This page displays type options, with selectable work order types as available choices. However, since these selectable work order types involve internal and client processes, after logging into the user account on the terminal, the selectable work order types can be filtered, filling the type options with the types the user account can choose. Additionally, the visual interface can display input boxes where users can enter work order descriptions. Prompt messages can also be displayed in these input boxes to indicate the root cause of the work order creation, such as "Please enter the reason for creating this work order." The prompt message stops displaying once the user begins typing, and the text entered by the user is used as the work order description.

[0069] It should be noted that the work order description information is used to indicate the root cause of the work order creation. Users can describe it with any text, such as "complaining about the xx event", "requesting repair for the xx equipment damage", "complaining about the improper handling of the xx work order and requesting more professional personnel to handle it", "requesting budget disbursement from the approval of the xx work order", etc. In this embodiment, the specific content of the work order description information is not limited, as long as it can represent the root cause of the work order creation and the user's intent.

[0070] It should be noted that, as Figure 1 As shown, after selecting the first work order type and entering the work order description information, the first work order is created. The first work order can be created by the user terminal and sent to the work order management system, or it can be created by the work order management system after obtaining the above information. According to the description of the above embodiment, each optional work order type corresponds to only one optional property line. After the user selects the first work order type, the optional property line corresponding to the first work order type is determined as the first property line.

[0071] For example, taking a work order management system with customer service, cleaning, maintenance, finance, and human resources lines as an example, the optional work order types for the customer service line include complaint work orders and inquiry work orders; the optional work order types for the cleaning line include cleaning work orders; the optional work order types for the maintenance line include maintenance work orders, equipment maintenance work orders, and inspection work orders; the optional work order types for the finance line include budget approval work orders, expense reimbursement work orders, and compensation payment work orders; and the optional work order types for the human resources line include recruitment work orders and personnel appointment work orders. Figure 1 As shown, if the first work order type selected by the user terminal 20 is a complaint work order, the customer service line corresponding to the complaint work order is determined as the first property management line. If the first work order type is a repair work order, the repair line corresponding to the repair work order is taken as the first property management line, and so on.

[0072] It should be noted that the first NLP model in this embodiment can identify multiple property lines by using the optional property lines as the classification basis. This embodiment does not involve the improvement of the NLP model. The first NLP model and the second NLP model are pre-trained models. Those skilled in the art know how to configure NLP models, and will not repeat the details below.

[0073] It should be noted that if the first NLP model only identifies one first property management line, it can be determined that the first work order is a newly created work order without any other related work orders. For example, there is no further complaint due to improper handling of the previous work order, or it is not a secondary approval initiated based on a prior process work order. In this case, the work order description information can be directly input into the second NLP model for sentiment recognition, and the first target label can be determined and assigned. If the first NLP model identifies a second property management line other than the first property management line, then step S50 is executed to create a second work order for each second property management line. It is worth noting that if the first NLP model identifies multiple first property lines, such as mentioning the word "complaint" multiple times in the work order description information, thereby identifying multiple customer service lines, this embodiment does not need to create multiple work orders. Instead, the second NLP model can filter out the corresponding first account from the customer service line when determining the emotion category. Therefore, in this embodiment, the first property line and the second property line are different optional property lines. The second property line is any optional property line other than the first property line. That is, there can be multiple second property lines, as long as each second property line is different from the others.

[0074] For example, referring to the various examples of the optional property management lines mentioned above, when the first work order type is selected as a complaint work order, the first property management line is the customer service line, and the work order description information is "xx equipment has been poorly repaired for a long time, complain about the repair personnel, continue repair". The second property management line identified by the first NLP model is the maintenance line.

[0075] It should be noted that the first NLP model typically splits the input text into multiple segments and combines them during recognition to identify the category to which each group of associated segments belongs. In this embodiment, the optional property category is used as the classification basis. When part of the work order description information is used to identify the second property category, the corresponding text is determined as the second description text. Since the work order description information will definitely identify the first property category, this embodiment excludes the first property category from the output results of the first NLP model and uses the remaining identified optional property categories as the second property category. For example, in the example above, the customer service category is identified through "complaint". Since the customer service category has already been identified as the first property category by the first work order type, it will not be counted. The repair category is identified through "repair", and the repair category is determined as the second property category.

[0076] It should be noted that, as Figure 1 As shown, after identifying the second property management line, this embodiment does not acquire the first property management line identified by the first NLP model. However, it can obtain the first and second description information based on the work order description information. The text that identifies the first property management line is determined as the first description information, and the text that identifies the second property management line is determined as the second description information. It is worth noting that this embodiment does not split the work order description information into the first and second description information. There can be duplicate text between the first and second description information. For example, in the example above, the first description information could be "xx equipment has been poorly maintained for a long time, complain about the maintenance personnel," and the second description information could be "xx equipment has been poorly maintained for a long time, continue maintenance." Of course, if multiple second property management lines are identified, the corresponding second description information can be acquired separately. For example, in the example above, if the work order description information is "xx equipment has been poorly maintained for a long time, complain about the maintenance personnel, continue maintenance, claim related expenses," then in addition to the maintenance line being identified as a second property management line, the finance line can also be identified as a second property management line based on the second description information "xx equipment has been poorly maintained for a long time, claim related expenses."

[0077] It should be noted that, as Figure 1 As shown, in this embodiment, the first and second descriptive information are input into the second NLP model for emotion recognition. Those skilled in the art are familiar with how to use NLP models to classify emotions, and the specific algorithm will not be described in detail here. The optional emotion categories in this embodiment can be common emotion categories in the property management field, such as "anxious," "angry," "peaceful," and "picky," which can represent the emotions of owners or managers when initiating work orders. This embodiment records the mapping relationship of response capability tags corresponding to each optional emotion category through a first mapping table, thereby determining which response personnel with which capabilities are needed for each emotion, thus achieving accurate matching of work orders. Therefore, after determining the first and second emotion categories, this embodiment determines the corresponding first and second target tags according to the first mapping table, and determines the first account from each optional account in the first property management line according to the first target tag. The number of first target tags can be multiple, and the first account only needs to include all the first target tags. The same applies to the second account.

[0078] For example, such as Figure 1As shown, continuing with the example above, when the first description is "The xx equipment has been difficult to repair for a long time, I am complaining about the repair personnel," the first emotional category can be determined as "loss of patience." Based on the first mapping table, the first target tags are determined as "strong professional skills" and "rapid response." From the customer service line, select an account that simultaneously possesses both of these first target tags as the first account. If there are multiple first accounts, any one can be chosen; no further restrictions are placed here. Similarly, the second emotional category can also include multiple categories. For example, when the second description is "The xx equipment has been difficult to repair for a long time, I am claiming related expenses," the second emotional category can be determined as "anger" and "loss of patience." The corresponding second target tags are "strong professional skills," "rapid response," and "good service attitude." Match the corresponding second account based on these three tags.

[0079] It should be noted that, as Figure 1 As shown, since the first work order initiated by the user is a single work order, but requires responses from multiple departments, it can be determined that resolving the user's issue requires generating multiple work orders. Therefore, this embodiment automatically constructs a second work order for each second property management department. For example, a maintenance work order is constructed for the aforementioned maintenance department, and a claim work order is constructed for the aforementioned finance department. The user only needs to perform information input once, and the work order management system automatically analyzes the user's actual needs through an NLP model, thereby automatically adding work orders without requiring manual operation by work order response personnel, effectively improving work order processing efficiency. At the same time, the system judges the user's emotions through the work order description information, determines the response capability tags of the managers responding to the first or second work order based on the emotion category, and achieves accurate matching of response personnel from the corresponding property management departments through tag identification, effectively improving the accuracy of work order response.

[0080] Additionally, in one embodiment, reference is made to Figure 3 In step S20, after inputting the work order description information into the first NLP model, the following steps are also included, but are not limited to:

[0081] S21, when the first NLP model outputs a first property line and at least one second property line, obtain the identification log of the first NLP model based on the work order description information;

[0082] S22, based on the identification log, the identification words used to identify the first property line are combined into first description information, and the identification words used to identify the second property line are combined into second description information, wherein the identification words are obtained by splitting the work order description information.

[0083] It should be noted that this embodiment needs to use the recognition process of the first NLP model to obtain the first description information related to the first property line and the second description information related to the second property line. The first description information is used as the basis for the subsequent identification of the first emotion category of the first property line, and the second description information is used as the basis for the identification of the second emotion category of the second property line. Therefore, it is necessary to ensure that the first description information is associated only with the first property line as much as possible, and the second description information is associated only with the second property line, so as to avoid misidentification of emotions.

[0084] For example, when the work order description information is "The work order creator entered a large amount of incorrect information, resulting in the budget approval work order being rejected. A complaint is filed against the previous work order creator, and the approval process is re-initiated, requesting expedited processing," and the first description information is "The work order creator entered a large amount of incorrect information, resulting in the budget approval work order being rejected. A complaint is filed against the previous work order creator," then the first emotion category can be determined to be anger. If the first description information is not accurately extracted, and the extracted first description information is "Re-initiated approval, requesting expedited processing," then the corresponding first emotion category is anxiety. Therefore, the first and second description information in this embodiment can have a significant impact on the accuracy of emotion category identification, and it is necessary to ensure the accuracy of extraction.

[0085] It should be noted that, in order to extract accurate first and second description information, this embodiment, after inputting the work order description information into the first NLP model, and obtaining the output results of the first NLP model (i.e., each property management line), uses the output results to query the recognition log of the first NLP model. Taking the first property management line as an example, the recognition log is queried based on the recognition result of the first property management line, and the recognition word segmentation used to obtain the recognition result is determined from the recognition log. The corresponding recognition word segmentation is then used to form the first description information. The second description information is processed similarly, ensuring that the text of the first or second description information is related to the identified property management line to ensure the accuracy of the recognition.

[0086] For example, continuing to refer to the above work order description information, taking the first work order type as budget approval as an example, the corresponding first property management line is the finance line. The first NLP model is used to identify the text for the finance line as "The work order creator entered a large amount of incorrect information, resulting in the budget approval work order being rejected. The approval is being re-initiated, and we request expedited processing." Similarly, the second property management line is the customer service line, and the corresponding second description information is "The work order creator entered a large amount of incorrect information, resulting in the budget approval work order being rejected. We are filing a complaint against the previous work order creator," ensuring the accuracy of the description information.

[0087] Additionally, in one embodiment, reference is made to Figure 3Each optional work order type is pre-associated with at least one preset operation term. In step S50, a second work order is created and assigned to a second account, including but not limited to the following steps:

[0088] S51, the optional work order type of the second property line is determined as the candidate work order type;

[0089] S52, extract at least one descriptive operation word from the second description information, and determine the second work order type from multiple candidate work order types based on the descriptive operation word, wherein the descriptive operation word has the same or similar semantics as any preset operation word of the second work order type;

[0090] S53, when the second description information contains historical work orders, the available account to which the historical work order is assigned is determined as the third account, and the third account is determined as the work order object;

[0091] S54: Create a second work order based on the second work order type and the third account, and assign the second work order to the second account.

[0092] It should be noted that, according to the description of the above embodiments, each optional property line is assigned a unique optional work order type. In order to construct a second work order, it is necessary to identify the optional work order type to which the second work order belongs. In this embodiment, preset operation terms are configured for each optional work order type, which can be used to initially filter the work order types. After determining the second property line, all optional work order types associated with the second property line are first identified as candidate work order types, thus narrowing down the selection range of work order types.

[0093] It should be noted that after determining the second descriptive information, descriptive operation words can be extracted from the second descriptive information through semantic recognition. The descriptive operation words can be preset operation words associated with the candidate work order type, or synonyms of preset operation words. For example, referring to the example above where the second descriptive information is "xx equipment has been poorly repaired for a long time, continue repair", the descriptive operation words are usually verbs in the text. In this example, it is "repair". The preset operation words that can be paired can include the same or similar words such as "repair", "maintain", and "repair".

[0094] It should be noted that after determining the second work order type, since the second work order is a derivative of the first work order, it is necessary to determine whether the second work order has any associated work order objects. If the first work order is a secondary operation initiated based on a historical work order, then the available account corresponding to the historical work order is the third account in this embodiment. The third account is then identified as the work order object, and a new work order is constructed based on the second work order type as the second work order. The third account is then used as the work order object of the second work order. The second account can be used to conduct relevant assessments of the third account based on the second work order. For example, it can be used to assess customer service personnel who are complained about in a complaint work order, or to assess repair personnel who handled a previous repair work order improperly.

[0095] Additionally, in one embodiment, reference is made to Figure 3 After step S54 is completed, the following steps are included, but are not limited to:

[0096] S55, when the second emotion category is used to indicate negative emotions, the scoring base is determined based on the second emotion category and the preset second mapping table, wherein the second mapping table records the mapping relationship between the optional emotion categories and the optional base;

[0097] S56, after the first work order is completed, obtain the first evaluation information based on the feedback of the first work order, and input the first evaluation information into the second NLP model to determine the third emotion category;

[0098] S57, When the third emotion category is used to indicate positive emotions, the target assessment score is determined based on the preset correction coefficient and the scoring base, and the correction coefficient is less than 1;

[0099] S58 applies the target assessment score to a third account.

[0100] It should be noted that when the second emotion category is used to indicate negative emotions, such as anger, resentment, or anxiety, it indicates that the third account has mishandled historical work orders and needs to be evaluated. This embodiment uses a second mapping table to record the available base values ​​corresponding to each optional emotion category, and determines a scoring base value for subsequent scoring based on the second emotion category. For example, when the second emotion category is impatience, the corresponding available base value is 80 points; when the second emotion category is anger, the corresponding available base value is 100 points. Since the scoring base value is used for punitive evaluation of the third account, the target evaluation score can be used as a subtraction, for example, subtracting the target evaluation score from the third account's monthly evaluation score. Therefore, the higher the target evaluation score, the greater the punishment for the third account. Higher available base values ​​can be set for stronger negative emotions in the second mapping table.

[0101] It should be noted that the first work order is a related work order to the historical work orders. That is, the matters handled by the first work order are related to the historical work orders. For example, after re-initiating an approval work order with incorrect information, the first work order is the re-initiated approval work order, and the historical work order is the previously incorrectly filled approval work order. Similarly, if equipment repair fails and continued repair is requested, the first work order is the re-initiated repair work order, and the historical work order is the previously initiated repair work order. The processing status of the first work order can affect the user's mood when initiating it. If the first evaluation information for the first work order is positive, the scoring base can be lowered to the target assessment score using a correction coefficient less than 1. Determining the third emotion category using the second NLP model can be referred to the description in the above embodiments, and will not be repeated here.

[0102] Additionally, in one embodiment, reference is made to Figure 3 The selectable emotion categories are pre-associated with response priorities. In step S40, the first account is determined based on the first target label and the first property line, and the first work order is assigned to the first account, including:

[0103] S41, when multiple candidate accounts are matched based on the first target label and the first property line, obtain the work order queue of each candidate account, wherein the work order queue includes multiple orderly arranged work orders to be processed, and each work order to be processed is associated with a corresponding fourth emotion category.

[0104] S42, traverse each work order queue based on the first emotion category;

[0105] S43, when the first emotion category matches the same fourth emotion category, the next position of the pending work order corresponding to the matched fourth emotion category is determined as the target sorting;

[0106] S44, when the first emotion category does not match the same fourth emotion category, determine the response priority corresponding to each fourth emotion category, and determine the target order of the first work order based on the response priority corresponding to the first emotion category;

[0107] S45, determine the candidate account corresponding to the highest target ranking as the first account, and insert the first work order into the work order queue of the first account based on the target ranking.

[0108] It should be noted that the first account includes all first target tags. However, when there are many selectable accounts in the first property management line, multiple candidate accounts may be matched based on the first target tags. In this embodiment, the first work order is dynamically allocated according to the work order load and urgency of each candidate account to determine the target ranking of the first work order in each work order queue. Then, the first work order is inserted into the work order queue with the highest target ranking to ensure that the first work order can be responded to as quickly as possible.

[0109] It should be noted that each work order to be processed in the work order queue is obtained based on the method of the above embodiment. Therefore, the fourth emotion category is equivalent to the first emotion category mentioned above, and will not be repeated hereafter.

[0110] It should be noted that since the emotion category represents the emotion of the user initiating the ticket, the stronger the corresponding emotion category, the higher the urgency of the ticket, and the more it needs to be prioritized to ensure user experience. After determining the first emotion category, the ticket queues of each candidate account are traversed based on the first emotion category. If the ticket queues include the same emotion category and the ticket queues are sorted, the first ticket can be inserted after the pending tickets of the corresponding fourth emotion category. For example, if the ticket queue includes 5 pending tickets, and the fourth emotion category corresponding to the 3rd pending ticket is anger, and the first emotion category is also anger, then the target ranking of the first ticket can be determined as 4. If the fourth emotion category corresponding to the 3rd and 4th pending tickets is anger, then the target ranking of the first ticket is 5, and so on. If the ticket queue does not include the same fourth emotion category as the first emotion category, then the pending tickets corresponding to the candidate account are not associated with the same emotion, and the response priority of each fourth emotion category can be determined. The position of the first ticket is determined by comparing the response priorities. Sorting according to priority is a technique well known to those skilled in the art, and will not be elaborated here.

[0111] It is worth noting that this embodiment prioritizes matching based on whether the emotion categories are the same, rather than directly using priority to match all cases. This is because different emotion categories may have the same priority. If sorting is done based on priority alone, work orders with the same emotion will not be continuous in the work order queue, resulting in work orders with the same emotion being processed in an intermittent order. This is not very reasonable when processing multiple work orders. Therefore, this embodiment first compares the emotion categories, and then sorts them by priority if the comparison fails, to ensure that work orders with the same emotion are continuous in the work order queue.

[0112] It should be noted that after determining the target sorting for each work order queue, the account at the top of the target sorting can be designated as the first account.

[0113] Additionally, in one embodiment, reference is made to Figure 3 After step S50 is completed, the following steps are included, but are not limited to:

[0114] S61, determine the first capability label and the second capability label from the multiple response capability labels associated with the first account, and determine the third capability label and the fourth capability label from the multiple response capability labels associated with the second account, wherein the first capability label and the third capability label are used to indicate business capabilities, and the second capability label and the fourth capability label are used to indicate work order acceptance capabilities.

[0115] S62, update the second capability tag based on the updated first account's work order queue, and update the fourth capability tag based on the updated second account's work order queue;

[0116] S63, obtain the second evaluation information based on the feedback of the second work order, and input the second evaluation information into the second NLP model to determine the fifth emotion category;

[0117] S64, determine at least one third target label based on the third emotion category and the first mapping table, and determine at least one fourth target label based on the fifth emotion category and the first mapping table;

[0118] S65, update the first capability label based on the third target label, and update the third capability label based on the fourth target label.

[0119] It should be noted that, as Figure 1 As shown, response capability tags can include both work order processing capability and business capability. For example, tags representing business capability include technical expertise, proficiency in processes, and rich experience, while tags representing work order processing capability can include fast response speed, no current work orders, and high timeliness. This embodiment divides response capability tags into two dimensions: business capability and work order processing capability. It can dynamically update the response capability tags for each available account based on the work order processing status, increasing the number of assignable optional accounts, improving the descriptive ability of response capability tags for optional accounts, and ensuring the accuracy of work order allocation.

[0120] It should be noted that the second and fourth capability tags are the same, the only difference being the object to which they belong. The second and fourth capability tags can be updated in real time based on the number of work orders in their respective work order queues.

[0121] For example, the response capability tags for optional account A include fast response speed and professional skills. When optional account A is assigned a large number of work orders, the "fast response speed" tag is removed to prevent work orders that require fast response speed from being assigned to optional account A, thus affecting the efficiency of work order processing. Similarly, when optional account A has fewer remaining work orders, the "fast response speed" tag can be added back.

[0122] It should be noted that the first and third capability tags are business capability tags. The third emotion category is obtained based on the first evaluation information of user feedback. The first mapping table records the response capability tags corresponding to each optional emotion category. The corresponding business capability tags can be preset for each optional account and then dynamically updated according to the evaluation information of the work order. The specific update direction can be determined according to the corresponding emotion category. Tags can be added in the case of positive emotions and removed in the case of negative emotions.

[0123] For example, refer to Figure 1Taking the first ability tag as an example, if the third emotion category is "satisfaction with professionalism," and the first mapping table records a response ability tag as "high level of professionalism," then the corresponding response ability tag can be added to the first account. Conversely, if the third emotion category is "anger," and the first account already has the tag "good service attitude," then that tag can be removed, thus achieving dynamic tag updates. Alternatively, updates can be implemented by accumulating occurrences; for example, after accumulating multiple instances of the same negative emotion, the tag can be removed to prevent individual user evaluations from affecting the objectivity of the response ability tags for select accounts.

[0124] Additionally, in one embodiment, reference is made to Figure 3 In step S40, before assigning the first work order to the first account, the following steps are included, but are not limited to:

[0125] S46, determine the first work order process based on the first work order type, wherein the first work order process includes multiple first process nodes;

[0126] S47, input the first description information and each first process node into the first NLP model, and identify the target node corresponding to the first description information through the first NLP model;

[0127] S48, configure the first work order based on the target node.

[0128] It should be noted that each type of work order includes its own work order process. For example, a complaint work order has a corresponding complaint process, and an approval work order has a corresponding approval process. This embodiment can not only achieve accurate allocation of responders, but also automatically identify the target node to which the first work order belongs, ensuring that the first work order can be directly assigned to the corresponding target node for work order flow. After determining the first work order process, this embodiment inputs the first process node into the first NLP model as a classification basis. The first description information is a requirement description for the first work order. Therefore, the first description information can be re-identified to determine the target node and achieve automatic process matching.

[0129] For example, the work order description information is "Approval work order A was rejected during approval in department X, and after modification, it jumped to that department for re-approval". Multiple first-process nodes of the approval work order are input into the first NLP model as classification basis. The above description information is input into the first NLP model for identification. Based on "approval in department X" and "re-approval", "department X" is determined as the target node. The approval work order is configured to the target node, thereby eliminating the preceding nodes. This allows the first account to know the current processing node based on the target node after obtaining the first work order, thereby achieving accurate allocation of work orders.

[0130] like Figure 4 As shown, Figure 4This is a structural diagram of an AI-based work order management device according to an embodiment of the present invention. The present invention also provides an AI-based work order management device, comprising:

[0131] The processor 401 can be implemented using a general-purpose central processing unit (CPU), 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.

[0132] The memory 402 can be implemented as a read-only memory (ROM), static storage device, dynamic storage device, or random access memory (RAM). The memory 402 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 402 and is called and executed by the processor 401 to execute the AI-based work order management method of the embodiments of this application.

[0133] Input / output interface 403 is used to implement information input and output;

[0134] The communication interface 404 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.).

[0135] Bus 405 transmits information between various components of the device (e.g., processor 401, memory 402, input / output interface 403, and communication interface 404);

[0136] The processor 401, memory 402, input / output interface 403 and communication interface 404 are connected to each other within the device via bus 405.

[0137] This application also provides an electronic device, including the AI-based work order management device described above.

[0138] This application embodiment also provides a storage medium, which is a computer-readable storage medium, storing a computer program that, when executed by a processor, implements the above-described AI-based work order management method.

[0139] 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. The device embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separate, and 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.

[0140] It will be understood by those skilled in the art that all or some of the steps and systems in the methods disclosed above can be implemented as software, firmware, hardware, and suitable combinations thereof. Some or all of the physical components can be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software can be distributed on a computer-readable medium, which can include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and is accessible to a computer. Furthermore, as is known to those skilled in the art, communication media typically include computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0141] The above provides a detailed description of the preferred embodiments of the present invention. However, the present invention is not limited to the above embodiments. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention. All such equivalent modifications or substitutions are included within the scope defined by the claims of the present invention.

Claims

1. A work order management method based on AI analysis, characterized in that, The method is applied to a work order management system, which is pre-configured with a first NLP model, a second NLP model, and multiple available property lines. Each available property line is associated with multiple available accounts and multiple selectable work order types. Each available account is associated with multiple response capability tags. The method includes: A first work order is created based on the work order description information and the first work order type, and a first property line is determined based on the first work order type, wherein the work order description information is used to indicate the root cause of the creation of the first work order; The work order description information and the available property lines are input into the first NLP model. When the first NLP model extracts first description information and at least one second description information from the work order description information, it obtains the second property lines identified based on each of the second description information. The first description information can be used to identify the first property line. The first emotion category is identified from the first description information based on the second NLP model, and the second emotion category is identified from the second description information. Based on the first emotion category and the first mapping table, at least one first target label is determined, and based on the first target label and the first property line, a first account is determined. The first work order is assigned to the first account, wherein the first mapping table records the mapping relationship between optional emotion categories and the response capability label. Based on any of the second description information, at least one second target label is determined based on the corresponding second emotion category and the first mapping table, and a second account is determined based on the second target label and the second property line. A second work order is created and assigned to the second account. After inputting the work order description information into the first NLP model, the method further includes: When the first NLP model outputs the first property line and at least one second property line, the identification log of the first NLP model based on the work order description information is obtained; Based on the identification log, the identification words used to identify the first property line are combined into the first description information, and the identification words used to identify the second property line are combined into the second description information, wherein the identification words are obtained by splitting the work order description information.

2. The work order management method based on AI analysis according to claim 1, characterized in that, Each of the optional work order types is pre-associated with at least one preset operation term. Creating a second work order and assigning it to the second account includes: The optional work order types of the second property management line are determined as candidate work order types; At least one descriptive operation word is extracted from the second description information, and a second work order type is determined from a plurality of candidate work order types based on the descriptive operation word, wherein the descriptive operation word is semantically the same as or semantically similar to any of the preset operation words of the second work order type; When the second description information contains historical work orders, the available account to which the historical work orders are assigned is identified as the third account, and the third account is identified as the work order object; The second work order is constructed based on the second work order type and the third account, and the second work order is assigned to the second account.

3. The work order management method based on AI analysis according to claim 2, characterized in that, After assigning the second work order to the second account, the method further includes: When the second emotion category is used to indicate a negative emotion, a scoring base is determined based on the second emotion category and a preset second mapping table, wherein the second mapping table records the mapping relationship between the selectable emotion categories and the selectable scoring base; After the first work order is completed, the first evaluation information based on the feedback of the first work order is obtained, and the first evaluation information is input into the second NLP model to determine the third emotion category. When the third emotion category is used to indicate positive emotions, the target assessment score is determined based on a preset correction coefficient and the scoring base. The target assessment score is applied to the third account.

4. The work order management method based on AI analysis according to claim 3, characterized in that, The selectable emotion categories are pre-associated with response priorities. Determining a first account based on the first target tag and the first property management line, and assigning the first work order to the first account includes: When multiple candidate accounts are matched based on the first target tag and the first property line, the work order queue of each candidate account is obtained, wherein the work order queue includes multiple orderly arranged work orders to be processed, and each work order to be processed is associated with a corresponding fourth emotion category. Based on the first emotion category, traverse each of the work order queues; When the first emotion category matches the same fourth emotion category, the next element of the pending work order corresponding to the matched fourth emotion category is determined as the target sorting; When the first emotion category does not match the same fourth emotion category, the response priority corresponding to each fourth emotion category is determined, and the target sorting of the first work order is determined based on the response priority corresponding to the first emotion category. The candidate account corresponding to the highest target ranking is determined as the first account, and the first work order is inserted into the work order queue of the first account based on the target ranking.

5. The work order management method based on AI analysis according to claim 4, characterized in that, After creating a second work order and assigning it to the second account, the method further includes: A first capability tag and a second capability tag are determined from the plurality of response capability tags associated with the first account, and a third capability tag and a fourth capability tag are determined from the plurality of response capability tags associated with the second account, wherein the first capability tag and the third capability tag are used to indicate business capabilities, and the second capability tag and the fourth capability tag are used to indicate work order acceptance capabilities. The second capability tag is updated based on the updated work order queue of the first account, and the fourth capability tag is updated based on the updated work order queue of the second account. Obtain the second evaluation information in response to the second work order, and input the second evaluation information into the second NLP model to determine the fifth emotion category; At least one third target label is determined based on the third emotion category and the first mapping table, and at least one fourth target label is determined based on the fifth emotion category and the first mapping table; The first capability label is updated based on the third target label, and the third capability label is updated based on the fourth target label.

6. The work order management method based on AI analysis according to claim 1, characterized in that, Before assigning the first work order to the first account, the method further includes: A first work order process is determined based on the first work order type, wherein the first work order process includes multiple first process nodes; The first description information and each of the first process nodes are input into the first NLP model, and the target node corresponding to the first description information is identified by the first NLP model. Configure the first work order based on the target node.

7. A work order management device based on AI analysis, characterized in that, It includes at least one control processor and a memory for communicatively connecting to the at least one control processor; the memory stores instructions executable by the at least one control processor, which, when executed by the at least one control processor, enable the at least one control processor to perform the AI-based work order management method as described in any one of claims 1 to 6.

8. An electronic device, characterized in that, Includes the work order management device based on AI analysis as described in claim 7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions for causing a computer to perform the AI-based work order management method as described in any one of claims 1 to 6.

Images