Customer service work order management method and device, equipment and storage medium

Through the customer demand identification model and the LSTM work order classification model, the problems of manual intervention and information consistency in the customer service process are solved, and the rapid identification and automatic classification of work orders are realized, processing efficiency and accuracy are improved, and transparency and customers are ensured. Satisfaction.

CN118941302APending Publication Date: 2024-11-12SHANGHAI DONGPU INFORMATION TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411168768.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-23
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

In the customer service process, the existing technology has problems such as many manual interventions, information loss or inconsistent, resulting in low efficiency and poor accuracy in work order processing, and the inability to effectively solve complex or specific customer needs.

Method used

Identify customer needs through the customer demand identification model, and determine whether it can be solved through the online question-and-answer system. If it cannot be solved, the initial work order information will be extracted and the merchant work orders will be classified and assigned using the LSTM work order classification model, and the work order status will be monitored in real time and Generate feedback information.

Benefits of technology

It improves the efficiency and accuracy of work order processing, reduces manual processing time, avoids allocation errors and delays, ensures the transparency and traceability of the processing process, and improves customer satisfaction.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118941302A_ABST
    Figure CN118941302A_ABST
Patent Text Reader

Abstract

The invention relates to the field of work order processing, and discloses a customer service work order management method, device and equipment and a storage medium, the method is used for identifying customer demands through a customer demand identification model to create merchant work orders, classifying the merchant work orders and then assigning the classified merchant work orders to corresponding processing departments, and the efficiency and accuracy are improved. The method comprises the following steps: analyzing a question input by a client by using a client demand identification model to obtain a client demand, and judging whether the client demand can be solved through an online question and answer system; if the customer demand cannot be solved through the online question-answering system, initial work order information is extracted based on the customer demand; creating a merchant work order based on the initial work order information; classifying merchant work orders by using a work order classification model based on a long short-term memory (LSTM) network, and generating dispatching information according to a classification result; and monitoring the state of the merchant work order, obtaining feedback data corresponding to the merchant work order after the merchant work order is completed, and generating feedback information according to the feedback data.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of work order processing, and in particular to a customer service work order management method, device, equipment and storage medium. Background Art

[0002] In the customer service process, we often encounter some complex or specific problems that cannot be directly solved by a simple online automatic question-and-answer system. At this point, the importance of establishing an efficient work order processing process is highlighted. Specifically, when customers encounter such problems and express their detailed demands to the merchant, the merchant needs to create a targeted form in its internal system to record the customer's specific problems and needs in detail. This form will then be submitted to the customer service work order system, which will intelligently or manually assign the work order to the most relevant or most suitable department for processing based on the nature and content of the problem described in the form. After the work order is transferred to the relevant department, a series of processing processes will be triggered, including problem investigation, solution formulation, execution and verification steps. The entire processing process will be recorded in detail by the work order system to ensure that each step is traceable for subsequent tracking and auditing. When the problem is properly resolved, the processing results will be fed back to the merchant-side system through the work order system again, and finally communicated to the customer.

[0003] However, while creating work orders in internal systems to respond to detailed customer requests is effective in tracking and managing customer needs, there are also some potential drawbacks or challenges:

[0004] Frequent manual intervention: Creating a work order usually requires manual intervention, including reading customer requests, determining the type of problem, filling out forms, etc. This not only increases labor costs, but may also lead to errors or delays due to human factors (such as negligence and misunderstanding).

[0005] Missing or inconsistent information: During the manual work order filling process, information may be missing, wrong, or inconsistent.

[0006] Therefore, the existing technology still needs to be improved and developed. Summary of the invention

[0007] The present invention provides a customer service work order management method, device, equipment and storage medium, which are used to identify customer needs through a customer need identification model to create a merchant work order, and classify the merchant work order and assign it to the corresponding processing department, thereby improving efficiency and accuracy.

[0008] A first aspect of the present invention provides a customer service work order management method, which includes: obtaining questions input by customers, using a customer demand identification model to analyze the questions input by customers to obtain customer needs, and determining whether the customer needs can be solved by an online question-and-answer system; if the customer needs cannot be solved by the online question-and-answer system, extracting initial work order information based on the customer needs; creating a merchant work order based on the initial work order information; classifying the merchant work order using a work order classification model based on a long short-term memory network (LSTM), and generating dispatch information based on the classification results; monitoring the status of the merchant work order, and when the merchant work order is completed, obtaining feedback data corresponding to the merchant work order, and generating feedback information based on the feedback data.

[0009] Optionally, in a first implementation method of the first aspect of the present invention, the steps of obtaining questions input by customers, analyzing the questions input by customers using a customer demand identification model to obtain customer needs, and determining whether the customer needs can be resolved through an online question-and-answer system include: constructing a customer demand identification model based on natural language processing technology; using a customer demand identification model to analyze the questions input by customers to obtain customer needs; calculating the degree of similarity between the customer needs and questions in a preset question-and-answer database; if the degree of similarity is lower than a preset threshold, determining that the customer needs cannot be resolved through the online question-and-answer system; otherwise, determining that the customer needs can be resolved through the online question-and-answer system.

[0010] Optionally, in a second implementation method of the first aspect of the present invention, the customer demand identification model is constructed based on natural language processing technology, including: constructing an NLP initial model; collecting customer consultation questions, and marking customer needs corresponding to the customer consultation questions to obtain training samples; using the training samples to train and test the NLP initial model to obtain a customer demand identification model.

[0011] Optionally, in a third implementation method of the first aspect of the present invention, if the customer demand cannot be resolved through the online question and answer system, initial work order information is extracted based on the customer demand, including: if the customer demand cannot be resolved through the online question and answer system, extracting key phrases and key entities in the customer demand; processing the extracted key phrases and key entities to obtain processed information; integrating the processed information to generate initial work order information.

[0012] Optionally, in a fourth implementation method of the first aspect of the present invention, creating a merchant work order based on the initial work order information includes: obtaining a customer demand and work order template association table, and determining the work order template number corresponding to the customer demand based on the customer demand and the customer demand and work order template association table; calling the work order template corresponding to the work order template number based on the work order template number; filling the initial work order information into the work order template to create a merchant work order.

[0013] Optionally, in a fifth implementation method of the first aspect of the present invention, the merchant work orders are classified using a work order classification model based on a long short-term memory network (LSTM), and dispatch information is generated based on the classification results, including: collecting historical work order data of merchants, the historical work order data including historical work orders and work order type labels corresponding to the historical work orders; dividing the historical work order data into a training set and a test set, and using a cross-validation method to train a pre-trained long short-term memory network (LSTM) initial model to obtain a work order classification model; using the work order classification model to classify the merchant work orders, and generating dispatch information based on the classification results.

[0014] Optionally, in a sixth implementation method of the first aspect of the present invention, the status of the merchant work order is monitored, and when the merchant work order is completed, feedback data corresponding to the merchant work order is obtained, and feedback information is generated based on the feedback data, including: monitoring the status of the merchant work order, and triggering an event when the status of the merchant work order changes; judging whether the current status of the merchant work order is completed based on the triggering event; if so, obtaining feedback data corresponding to the merchant work order, and sending the feedback data to the merchant through a message queue, and the feedback data includes processing results.

[0015] The second aspect of the present invention provides a customer service work order management device, including: an identification module, used to obtain questions input by customers, use a customer demand identification model to analyze the questions input by customers to obtain customer needs, and determine whether the customer needs can be solved through an online question-and-answer system; an extraction module, used to extract initial work order information based on the customer needs when the customer needs cannot be solved through the online question-and-answer system; a creation module, used to create a merchant work order based on the initial work order information; a classification module, used to classify the merchant work order using a work order classification model based on a long short-term memory network LSTM, and generate dispatch information based on the classification results; a monitoring module, used to monitor the status of the merchant work order, and when the merchant work order is completed, obtain feedback data corresponding to the merchant work order, and generate feedback information based on the feedback data.

[0016] Optionally, in a first implementation method of the second aspect of the present invention, the identification module includes: a construction unit, used to construct a customer demand identification model based on natural language processing technology; an identification unit, used to use the customer demand identification model to analyze the questions input by the customer to obtain the customer needs; a first judgment unit, used to calculate the degree of similarity between the customer needs and the questions in a preset question answering database. If the degree of similarity is lower than a preset threshold, it is determined that the customer needs cannot be solved by the online question and answer system; otherwise, it is determined that the customer needs can be solved by the online question and answer system.

[0017] Optionally, in a second implementation of the second aspect of the present invention, the extraction module includes: an extraction unit, used to extract key phrases and key entities in customer needs when the customer needs cannot be solved through the online question-and-answer system; a processing unit, used to process the extracted key phrases and key entities to obtain processed information; and an integration unit, used to integrate the processed information to generate initial work order information.

[0018] Optionally, in a third implementation method of the second aspect of the present invention, the creation module includes: an acquisition unit, used to acquire a relationship table between customer demand and work order templates, and determine a work order template number corresponding to the customer demand based on the customer demand and the relationship table between customer demand and work order templates; a calling unit, used to call the work order template corresponding to the work order template number based on the work order template number; and a creation unit, used to fill the initial work order information into the work order template to create a merchant work order.

[0019] Optionally, in a fourth implementation of the second aspect of the present invention, the classification module includes: a collection unit, used to collect historical work order data of merchants, the historical work order data including historical work orders and work order type labels corresponding to the historical work orders; a training unit, used to divide the historical work order data into a training set and a test set, and use a cross-validation method to train a pre-trained long short-term memory network LSTM initial model to obtain a work order classification model; a classification unit, used to use the work order classification model to classify the merchant work orders, and generate dispatch information based on the classification results.

[0020] Optionally, in a fifth implementation of the second aspect of the present invention, the monitoring module includes: a monitoring unit, used to monitor the status of the merchant work order, and trigger an event when the status of the merchant work order changes; a second judgment unit, used to judge whether the current status of the merchant work order is completed based on the triggering event; a feedback unit, used to obtain feedback data corresponding to the merchant work order when the current status of the merchant work order is completed, and send the feedback data to the merchant through a message queue, the feedback data including processing results.

[0021] A third aspect of the present invention provides a customer service work order management device, comprising: a memory and at least one processor, wherein the memory stores computer-readable instructions, and the memory and the at least one processor are interconnected via a line; the at least one processor calls the computer-readable instructions in the memory so that the customer service work order management device executes each step of the customer service work order management method as described above.

[0022] A fourth aspect of the present invention provides a computer-readable storage medium, in which computer-readable instructions are stored. When the computer-readable storage medium is run on a computer, the computer executes each step of the customer service work order management method as described above.

[0023] In the technical solution provided by the present invention, customer needs are identified through a customer need identification model, and customer problems or needs can be accurately captured. When customer needs cannot be solved through an online question-and-answer system, a merchant work order is created, and the merchant work order is classified using an LSTM-based work order classification model. Dispatch information is generated based on the classification results, which facilitates the dispatch of merchant work orders to corresponding processing departments, thereby achieving rapid identification of customer needs and automatic classification of work orders, greatly reducing manual processing time and improving work efficiency. Work orders are automatically dispatched to corresponding processing departments based on classification results, avoiding possible allocation errors or delays in traditional methods, and further improving processing efficiency. In addition, the work order status is monitored in real time to ensure the transparency and traceability of the processing process. Feedback information is generated in a timely manner after the work order is completed, so that merchants can clearly see the problem-solving process and results, further improving customer satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 A first flow chart of the customer service work order management method provided by an embodiment of the present invention;

[0025] Figure 2 A second flow chart of the customer service work order management method provided by an embodiment of the present invention;

[0026] Figure 3 A third flow chart of the customer service work order management method provided by an embodiment of the present invention;

[0027] Figure 4 A fourth flow chart of the customer service work order management method provided by an embodiment of the present invention;

[0028] Figure 5 A fifth flow chart of the customer service work order management method provided by an embodiment of the present invention;

[0029] Figure 6 A sixth flow chart of the customer service work order management method provided by an embodiment of the present invention;

[0030] Figure 7 A schematic diagram of the structure of a customer service work order management device provided by an embodiment of the present invention;

[0031] Figure 8 A schematic diagram of the structure of a customer service work order management device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0032] The embodiment of the present invention provides a customer service work order management method, device, equipment and storage medium, which is used to identify customer needs through a customer need identification model to create a merchant work order, and classify the merchant work order and assign it to the corresponding processing department, thereby improving efficiency and accuracy. The method includes: obtaining customer needs, identifying customer needs based on a customer need identification model, and determining whether the customer needs can be solved through an online question-and-answer system; if the customer needs cannot be solved through the online question-and-answer system, extracting initial work order information based on the customer needs; creating a merchant work order based on the initial work order information; using a work order classification model based on a long short-term memory network LSTM to classify the merchant work order, and generating dispatch information based on the classification results; monitoring the status of the merchant work order, and when the merchant work order is completed, obtaining feedback data corresponding to the merchant work order, and generating feedback information based on the feedback data.

[0033] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0034] For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1 A first embodiment of a customer service work order management method in an embodiment of the present invention includes:

[0035] S101. Obtain customer needs, identify customer needs based on a customer needs identification model, and determine whether customer needs can be solved through an online question-and-answer system.

[0036] It is understandable that the execution subject of the present invention may be a customer service work order management device, or a terminal or a server, which is not limited here. The embodiment of the present invention is described by taking the server as the execution subject as an example.

[0037] In this embodiment, the customer demand identification model is trained with a large amount of customer inquiry and corresponding demand data, and learns how to extract key information from the customer's text description, such as problem type, specific symptoms, urgency, etc., which can significantly improve the accuracy and efficiency of demand identification and provide a solid foundation for subsequent steps.

[0038] In this embodiment, the customer demand identification model uses a pre-trained NLP model (such as BERT, RoBERTa, or GPT series models) to understand customer questions and determine customer needs.

[0039] S102: If the customer demand cannot be solved through the online question-and-answer system, extract initial work order information based on the customer demand.

[0040] In this embodiment, the customer demand is matched with the knowledge base of the online question-and-answer system to find out whether there are similar answers or solutions. If so, it means that the customer demand can be solved by the online question-and-answer system. If not, it means that the customer demand cannot be solved by the online question-and-answer system.

[0041] In this embodiment, when customer needs cannot be solved through the online question-and-answer system, key information is extracted from the customer needs, such as basic customer information, problem description, urgency, etc., to form initial work order information, providing the necessary information basis for creating a merchant work order.

[0042] S103. Create a merchant work order based on the initial work order information.

[0043] In this embodiment, a work order management system can be used to automatically generate a merchant work order based on the initial work order information and assign a unique work order number.

[0044] In this embodiment, the specific content of the merchant work order may include information such as the work order number, customer requirements, customer information, priority, deadline, and attachments.

[0045] S104. Use a work order classification model based on a long short-term memory network (LSTM) to classify merchant work orders, and generate dispatch information based on the classification results.

[0046] In this embodiment, the text data of the merchant work order is converted into a text vector, and then the text vector is converted into a bit sequence form. Usually, the text is converted into a series of vectors according to the word segmentation results, and then deep learning technologies such as LSTM are used to automatically classify the merchant work order to identify the problem type or business field to which it belongs. Finally, based on the output of the work order classification model, the merchant work order is assigned to the corresponding processing department to ensure that the work order can be handled in a timely manner by professional personnel or teams.

[0047] S105. Monitor the status of the merchant work order. When the merchant work order is completed, obtain feedback data corresponding to the merchant work order, and generate feedback information based on the feedback data.

[0048] In this embodiment, the automation function of the work order management system or the pre-setting can be used to monitor the status changes of the merchant's work order in real time, including the processing progress, estimated completion time, actual completion status, etc. When the work order status is updated to "completed", feedback data is obtained, and feedback information is generated based on the feedback data. The feedback information is fed back to the merchant through email, SMS or system notification, and the feedback data includes information such as the work order number, problem description, processing results, processing time and subsequent suggestions.

[0049] The present embodiment provides a customer service work order management method, which identifies customer needs through a customer need identification model, can accurately capture customer problems or needs, and when customer needs cannot be solved through an online question-and-answer system, creates a merchant work order, and uses an LSTM-based work order classification model to classify the merchant work order, and generates dispatch information based on the classification results, so as to facilitate the dispatch of the merchant work order to the corresponding processing department, thereby realizing rapid identification of customer needs and automatic classification of work orders, greatly reducing manual processing time, and improving work efficiency, and automatically dispatching work orders to corresponding processing departments based on the classification results, avoiding possible allocation errors or delays in traditional methods, and further improving processing efficiency. In addition, real-time monitoring of the work order status ensures the transparency and traceability of the processing process, and timely generates feedback information after the work order is completed, so that the merchant can clearly see the problem solving process and results, further improving customer satisfaction.

[0050] See also Figure 2 , a second embodiment of the customer service work order management method in an embodiment of the present invention includes:

[0051] S201. Build a customer demand identification model based on natural language processing technology.

[0052] In this embodiment, a customer demand identification model is constructed based on natural language processing technology, specifically including: constructing an NLP initial model; collecting customer consultation questions and marking customer needs corresponding to the customer consultation questions to obtain training samples; using the training samples to train and test the NLP initial model to obtain a customer demand identification model.

[0053] In this embodiment, the core challenges of the NLP initial model include language understanding (understanding the meaning, context, and intent of human language) and language generation (generating natural language text in a way that is understandable to humans).

[0054] The NLP initial model can be a pre-trained BERT, RoBERTa or GPT series model. The NLP model architecture includes but is not limited to Naive Bayes, Maximum Entropy, Support Vector Machine (SVM), etc.

[0055] Recurrent Neural Networks (RNNs) and their variants, such as Long Short-Term Memory (LSTM) and Gated Recurrent Units (GRU), are suitable for processing sequential data, such as text.

[0056] Although Convolutional Neural Networks (CNNs) are mainly used for image processing, they are also used in tasks such as text classification.

[0057] Transformer models, such as BERT (Bidirectional Encoder Representations from Transformers) and GPT (Generative Pre-trained Transformer) series, are popular choices in the current NLP field, and they have achieved excellent performance in many tasks.

[0058] In this embodiment, the collected data is cleaned to remove noise and irrelevant information, and text preprocessing operations such as word segmentation, part-of-speech tagging, and stop word removal are performed to facilitate subsequent feature extraction and model training.

[0059] S202: Use the customer demand identification model to analyze the questions input by the customer to obtain the customer demand.

[0060] In this embodiment, when using the customer demand recognition model to analyze the questions input by the customer to obtain the customer needs, the questions input by the customer can be preprocessed first, including word segmentation, part-of-speech tagging, etc., and then the processed text data is input into the customer demand recognition model.

[0061] In this embodiment, if the question input by the customer is audio data, the audio data needs to be converted into text data.

[0062] S203, calculating the similarity between the customer demand and the questions in the preset question answering database. If the similarity is lower than a preset threshold, it is determined that the customer demand cannot be solved by the online question answering system. Otherwise, it is determined that the customer demand can be solved by the online question answering system.

[0063] In this embodiment, in this embodiment, by calculating the semantic similarity between the customer demand and the questions in the preset question answering database, it is determined whether the customer demand can be solved by the online question answering system.

[0064] If the semantic similarity is lower than the set threshold, the question is considered complex and requires further processing, that is, it is determined that the customer demand cannot be solved by the online question-answering system.

[0065] For example, assuming that the customer question is Q, the FAQ set is F = (f1, f2....fn), for each fi belonging to the set F, Si = Sim (Q, fi), where SimSim is a function for calculating the semantic similarity between two sentences, using methods such as cosine similarity and Jaccard similarity. If all Si < threshold, it means that the customer question cannot be directly solved by the online automatic question-answering system.

[0066] In this embodiment, NLP technology can be used to automatically understand, analyze and identify the needs behind the questions input by customers without manual intervention, which greatly improves processing efficiency and reduces the burden on manual customer service.

[0067] See also Figure 3 A third embodiment of a customer service work order management method in an embodiment of the present invention includes:

[0068] S301. If the customer demand cannot be solved by the online question-answering system, extract key phrases and key entities in the customer demand.

[0069] In this embodiment, algorithms such as TF-IDF, TextRank, or statistical methods are used to extract key phrases related to the core of the problem from customer needs, such as the type of problem, the products or services involved, etc.

[0070] In this embodiment, the NER technology in natural language processing is applied to identify named entities in the text, such as names of people, places, institutions, products, etc. These entities are often the key to understanding customer needs and subsequent processing.

[0071] S302: Process the extracted key phrases and key entities to obtain processed information.

[0072] In this embodiment, the extracted key phrases and key entities are de-duplicated to avoid duplicate information. At the same time, related phrases or entities are merged according to the context and semantic relationship.

[0073] In this embodiment, the extracted key phrases and key entities may also be standardized, such as unifying the format, correcting spelling errors, etc., to ensure the accuracy and consistency of the information.

[0074] In this embodiment, the processed information may also be classified and labeled for subsequent processing and query.

[0075] S303: Integrate the processed information to obtain initial work order information.

[0076] In this embodiment, the processed key phrases and key entities are integrated as initial work order information.

[0077] In this embodiment, the initial work order information generally includes the customer's problem description, key information (such as product type, error code, etc.) and priority, etc.

[0078] In this embodiment, by extracting and processing customer needs in an automated or semi-automated manner, initial work order information can be quickly generated, reducing the time and cost of manual intervention.

[0079] See also Figure 4 A fourth embodiment of a customer service work order management method in an embodiment of the present invention includes:

[0080] S401. Obtain a table of associations between customer requirements and work order templates, and determine a work order template number corresponding to the customer requirement based on the customer requirement and the table of associations between customer requirements and work order templates.

[0081] S402. Call the work order template corresponding to the work order template number according to the work order template number.

[0082] In this embodiment, different work order templates are pre-configured for different customer needs, and a work order template library is constructed, in which each work order template presets the fields and formats required for the type of work order.

[0083] In this embodiment, work order templates for different customer needs can be assigned work order template numbers, and a customer need and work order template association table can be formed. After the customer needs are determined, the work order template number of the work order template can be obtained from the association table, and then the corresponding work order template can be called from the work order template library.

[0084] S403: Fill the initial work order information into the work order template to create a merchant work order.

[0085] In this embodiment, according to the requirements of the work order template fields, the initial work order information (such as customer name, contact information, problem description, product information, etc.) is inserted into the corresponding position, and variable replacement or automated scripts can be used to complete the information filling. The filled work order template is exported to a standard format to obtain a merchant work order.

[0086] The following is an example of a merchant work order:

[0087] Work Order Number: [Unique Number]

[0088] Customer Information: [Customer Name, Phone Number]

[0089] Problem description: [A brief description of the problem or requirement reported by the customer]

[0090] Type of question: [such as technical support, after-sales service, complaints, suggestions, etc.]

[0091] Key information: [Name of product or service involved, order number]

[0092] Priority: [priority level, such as urgent, high, medium, low, etc.]

[0093] Creation Time: [date and time the ticket was created]

[0094] Processing status: [Current processing status, such as pending, processing, completed, etc.]

[0095] In this embodiment, the problem type of the work order is confirmed based on the initial work order information, and the work order template corresponding to the problem type of the work order is called according to the problem type of the work order; the key information is filled into the work order template, and a merchant work order is created. By automatically confirming the problem type of the work order and calling the template, the time for manual judgment and manual input is reduced, and the speed of work order creation is improved. In addition, the use of predefined work order templates can ensure that each work order contains the necessary information fields, reducing processing delays or repeated problems caused by missing or incorrect information.

[0096] See also Figure 5 A fifth embodiment of a customer service work order management method in an embodiment of the present invention includes:

[0097] S501. Collect historical work order data of merchants, where the historical work order data includes historical work orders and work order type tags corresponding to the historical work orders.

[0098] In this embodiment, a large amount of work order data is collected from the merchant's historical records, and each work order should contain a detailed text description (such as problem description, customer feedback, etc.) and a corresponding work order type label (such as technical support, after-sales service, complaints, suggestions, etc.).

[0099] In this embodiment, the collected data is cleaned, including removing duplicates, processing missing values, formatting text (such as unifying upper and lower case, removing special characters, etc.), etc.

[0100] S502, dividing the historical work order data into a training set and a test set, using a cross-validation method to train a pre-trained long short-term memory network LSTM initial model to obtain a work order classification model.

[0101] In this embodiment, since the LSTM model cannot process text data directly, it is necessary to convert the text into numerical form, which can be achieved by embedding words extracted by pre-trained models such as Bag of Words, TF-IDF, Word2Vec or BERT.

[0102] In this embodiment, a deep learning framework such as TensorFlow or PyTorch is used to define a long short-term memory network LSTM initial model. The long short-term memory network LSTM initial model includes an input layer, one or more LSTM layers, and an output layer. The input layer receives word embedding as input, the LSTM layer is used to capture long-term dependencies in the text, and the output layer is used to output the probability distribution of the merchant work order type.

[0103] In this embodiment, before training the pre-trained long short-term memory network LSTM initial model, hyperparameters such as the number of units, learning rate, batch size, and training rounds of the LSTM layer are set.

[0104] In this embodiment, in order to obtain more stable model performance, a cross-validation method (such as k-fold cross-validation) is used to train the model. In each iteration, the data set is divided into k subsets, k-1 subsets are used as training sets, and the remaining subset is used as a test set.

[0105] S503: Use the work order classification model to classify the merchant work orders, and generate dispatch information according to the classification results.

[0106] In this embodiment, a merchant work order category and processing department association table is pre-configured. After the type of the merchant work order is determined, it is matched with the association table to determine its corresponding processing department, and then the merchant work order is assigned to the corresponding department.

[0107] In this embodiment, the merchant work order is input into the work order classification model, and the work order classification model classifies the merchant work order and outputs the predicted probability of each category or directly gives the most likely category label.

[0108] In this embodiment, the LSTM-based work order classification model has good adaptability to different types of merchant work orders, and can continuously optimize the classification strategy according to historical data. The trained work order classification model can quickly classify newly submitted merchant work orders, realize real-time processing, and provide customers with more timely services. The distribution information is generated based on the classification results, which helps to accurately distribute merchant work orders to the corresponding processing departments, thereby improving work efficiency and avoiding errors and delays in manual allocation.

[0109] See also Figure 6 A sixth embodiment of a customer service work order management method in an embodiment of the present invention includes:

[0110] S601. Monitor the status of the merchant work order, and trigger an event when the status of the merchant work order changes.

[0111] In this embodiment, a status listener is set in the work order management system to monitor the status field of the merchant work order in the database in real time. Whenever the status of the work order changes, the status listener captures the change and triggers an event. This event contains the unique identifier (such as ID) of the work order and the new status information.

[0112] In this embodiment, monitoring can be achieved through database triggers, polling queries, or a publish / subscribe mode using a message queue.

[0113] S602: Determine whether the current status of the merchant work order is completed based on the triggering event.

[0114] In this embodiment, the new status of the work order is compared with the preset "completed" status value of the merchant. If the two match, it indicates that the current status of the merchant's work order is completed.

[0115] S603: If yes, obtain feedback data corresponding to the merchant work order, and generate feedback information according to the feedback data.

[0116] In this embodiment, when it is confirmed that the merchant work order has been completed, the feedback data of the merchant work order is obtained, and the feedback data includes work order details, processing process summary, processing results, processing personnel information and other information.

[0117] In this embodiment, the feedback information can be sent to the merchant using a message queue, and specifically, a message queue such as RabbitMQ, Kafka, ActiveMQ, etc. can be selected. When the feedback information is sent to the merchant through the message queue, the feedback information is encapsulated into a message and sent to the specified queue using the message queue API. During the sending process, parameters such as the message priority, persistence options, and timeout period can be specified.

[0118] In this embodiment, before the feedback information is sent to the message queue, necessary data verification and cleaning are performed to ensure the accuracy and completeness of the data.

[0119] In this embodiment, by monitoring the status changes of the merchant work order in real time and triggering events immediately when the status is updated, the scheme ensures the real-time processing. Once the work order is completed, the feedback data corresponding to the merchant work order is obtained, and feedback information is generated based on the feedback data, which reduces waiting time and improves overall processing efficiency.

[0120] The above describes the customer service work order management method in the embodiment of the present invention. The following describes the device in the embodiment of the present invention. Figure 7 , the implementation method of the customer service work order management device in the embodiment of the present invention includes:

[0121] Identification module 701, used to obtain the question input by the customer, analyze the question input by the customer using the customer demand identification model to obtain the customer demand, and determine whether the customer demand can be solved by the online question-answering system;

[0122] An extraction module 702, configured to extract initial work order information based on the customer demand when the customer demand cannot be solved by the online question-answering system;

[0123] A creation module 703, used to create a merchant work order based on the initial work order information;

[0124] A classification module 704 is used to classify the merchant work orders using a work order classification model based on a long short-term memory network (LSTM), and generate dispatch information according to the classification results;

[0125] The monitoring module 705 is used to monitor the status of the merchant work order. When the merchant work order is completed, feedback data corresponding to the merchant work order is obtained, and feedback information is generated according to the feedback data.

[0126] In this embodiment, the identification module 701 includes: a construction unit 7011, which is used to construct a customer demand identification model based on natural language processing technology; an identification unit 7012, which is used to use the customer demand identification model to analyze the questions input by the customer to obtain the customer needs; a first judgment unit 7013, which is used to calculate the similarity between the customer needs and the questions in the preset question answering database. If the similarity is lower than the preset threshold, it is determined that the customer needs cannot be solved by the online question and answer system. Otherwise, it is determined that the customer needs can be solved by the online question and answer system.

[0127] In this embodiment, the extraction module 702 includes: an extraction unit 7021, which is used to extract key phrases and key entities in customer needs when the customer needs cannot be solved through the online question-and-answer system; a processing unit 7022, which is used to process the extracted key phrases and key entities to obtain processed information; and an integration unit 7023, which is used to integrate the processed information to generate initial work order information.

[0128] In this implementation, the creation module 703 includes: an acquisition unit 7031, used to obtain the customer demand and work order template association relationship table, and determine the work order template number corresponding to the customer demand based on the customer demand and the customer demand and work order template association relationship table; a calling unit 7032, used to call the work order template corresponding to the work order template number based on the work order template number; a creation unit 7033, used to fill the initial work order information into the work order template to create a merchant work order.

[0129] In this implementation, the classification module 704 includes: a collection unit 7041, which is used to collect the merchant's historical work order data, and the historical work order data includes historical work orders and work order type labels corresponding to the historical work orders; a training unit 7042, which is used to divide the historical work order data into a training set and a test set, and use a cross-validation method to train a pre-trained long short-term memory network LSTM initial model to obtain a work order classification model; a classification unit 7043, which is used to use the work order classification model to classify the merchant's work orders, and generate dispatch information based on the classification results.

[0130] In this implementation, the monitoring module 705 includes: a monitoring unit 7051, which is used to monitor the status of the merchant work order and trigger an event when the status of the merchant work order changes; a second judgment unit 7052, which is used to judge whether the current status of the merchant work order is completed based on the triggering event; a feedback unit 7053, which is used to obtain feedback data corresponding to the merchant work order when the current status of the merchant work order is completed, and generate feedback information based on the feedback data, wherein the feedback data includes processing results.

[0131] In this embodiment, customer needs are identified through a customer need identification model, which can accurately capture customer problems or needs. When customer needs cannot be solved through the online question-and-answer system, a merchant work order is created, and the merchant work order is classified using an LSTM-based work order classification model. Dispatch information is generated based on the classification results to facilitate the dispatch of merchant work orders to corresponding processing departments, thereby achieving rapid identification of customer needs and automatic classification of work orders, greatly reducing manual processing time and improving work efficiency. Work orders are automatically dispatched to corresponding processing departments based on the classification results, avoiding possible allocation errors or delays in traditional methods, and further improving processing efficiency. In addition, the work order status is monitored in real time to ensure the transparency and traceability of the processing process. Feedback information is generated in a timely manner after the work order is completed, so that merchants can clearly see the problem-solving process and results, further improving customer satisfaction.

[0132] above Figure 7 The customer service work order management device in the embodiment of the present invention is described in detail from the perspective of modular functional entities. The customer service work order management device in the embodiment of the present invention is described in detail from the perspective of hardware processing.

[0133] Figure 8 It is a structural diagram of a customer service work order management device provided by an embodiment of the present invention. The device 800 may have relatively large differences due to different configurations or performances, and may include one or more processors (central processing units, CPU) 810 (for example, one or more processors) and a memory 820, and one or more storage media 830 (for example, one or more mass storage devices) storing application programs 833 or data 832. Among them, the memory 820 and the storage medium 830 can be short-term storage or permanent storage. The program stored in the storage medium 830 may include one or more modules (not shown), and each module may include a series of instruction operations in the device 800. Furthermore, the processor 810 can be configured to communicate with the storage medium 830 to execute a series of instruction operations in the storage medium on the device 800.

[0134] The device 800 may also include one or more power supplies 840, one or more wired or wireless network interfaces 850, one or more input and output interfaces 860, and / or one or more operating systems 831, such as Windows Serve, Mac OS X, Unix, Linux, FreeBSD, etc.

[0135] An embodiment of the present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions are executed on a computer, the computer executes the steps of the customer service work order management method.

[0136] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described system, device, or unit can refer to the corresponding process in the aforementioned method embodiment and will not be repeated here.

[0137] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program codes.

[0138] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A customer service work order management method, characterized in that: The customer service work order management method comprises: Obtain questions input by customers, analyze the questions input by customers using a customer demand identification model to obtain customer demands, and determine whether the customer demands can be solved by the online question-answering system; If the customer demand cannot be solved through the online question-and-answer system, extracting initial work order information based on the customer demand; Creating a merchant work order based on the initial work order information; Classify the merchant work orders using a work order classification model based on a long short-term memory network (LSTM), and generate dispatch information according to the classification results; Monitor the status of the merchant work order. When the merchant work order is completed, obtain the feedback data corresponding to the merchant work order and generate feedback information based on the feedback data.

2. The customer service work order management method according to claim 1, characterized in that: The obtaining of the question input by the customer, analyzing the question input by the customer using the customer demand identification model to obtain the customer demand, and determining whether the customer demand can be solved by the online question-answering system, includes: Build a customer demand identification model based on natural language processing technology; Use the customer needs identification model to analyze customer input questions to obtain customer needs; The similarity between the customer demand and the questions in the preset question answering database is calculated. If the similarity is lower than a preset threshold, it is determined that the customer demand cannot be solved by the online question answering system. Otherwise, it is determined that the customer demand can be solved by the online question answering system.

3. The customer service work order management method according to claim 2, characterized in that: The customer demand identification model is constructed based on natural language processing technology, including: Build an initial NLP model; Collect customer consultation questions, and mark customer needs corresponding to the customer consultation questions to obtain training samples; The training samples are used to train and test the NLP initial model to obtain a customer demand recognition model.

4. The customer service work order management method according to claim 1, characterized in that: If the customer demand cannot be solved through the online question-and-answer system, extracting initial work order information based on the customer demand includes: If the customer demand cannot be solved by the online question-answering system, extract the key phrases and key entities in the customer demand; Process the extracted key phrases and key entities to obtain processed information; Integrate the processed information and generate initial work order information.

5. The customer service work order management method according to claim 1, characterized in that: The creating a merchant work order based on the initial work order information includes: Obtaining a table of associations between customer requirements and work order templates, and determining a work order template number corresponding to the customer requirement based on the customer requirement and the table of associations between customer requirements and work order templates; Calling the work order template corresponding to the work order template number according to the work order template number; Fill the initial work order information into the work order template to create a merchant work order.

6. The customer service work order management method according to claim 1, characterized in that: The use of a work order classification model based on a long short-term memory network (LSTM) to classify the merchant work order and generating dispatch information according to the classification result includes: Collect historical work order data of merchants, wherein the historical work order data includes historical work orders and work order type labels corresponding to the historical work orders; The historical work order data is divided into a training set and a test set, and a pre-trained long short-term memory network LSTM initial model is trained using a cross-validation method to obtain a work order classification model; The merchant work orders are classified using the work order classification model, and dispatch information is generated according to the classification results.

7. The customer service work order management method according to claim 1, characterized in that: The monitoring of the status of the merchant work order, when the merchant work order is completed, obtaining feedback data corresponding to the merchant work order, and generating feedback information according to the feedback data, includes: Monitor the status of the merchant work order, and trigger an event when the status of the merchant work order changes; Determine whether the current status of the merchant work order is completed according to the triggering event; If so, feedback data corresponding to the merchant work order is obtained, and feedback information is generated according to the feedback data.

8. A customer service work order management device, characterized in that: include: An identification module is used to obtain questions input by customers, analyze the questions input by customers using a customer demand identification model to obtain customer demands, and determine whether the customer demands can be solved by the online question-answering system; An extraction module, used for extracting initial work order information based on the customer demand when the customer demand cannot be solved by the online question-answering system; A creation module, used to create a merchant work order based on the initial work order information; A classification module, used to classify the merchant work orders using a work order classification model based on a long short-term memory network (LSTM), and generate dispatch information according to the classification results; The monitoring module is used to monitor the status of the merchant work order. When the merchant work order is completed, the feedback data corresponding to the merchant work order is obtained, and feedback information is generated based on the feedback data.

9. A customer service work order management device, characterized in that: comprising a memory and at least one processor, wherein the memory has computer-readable instructions stored therein; The at least one processor calls the computer-readable instructions in the memory to execute the various steps of the customer service work order management method as described in any one of claims 1-7.

10. A computer-readable storage medium having computer-readable instructions stored thereon, characterized in that: When the computer-readable instructions are executed by a processor, the various steps of the customer service work order management method as described in any one of claims 1-7 are implemented.