Logistics exception work order processing method, device and equipment and storage medium

Abstract

The application discloses a kind of logistics exception work order processing method, device, equipment and storage medium, for the abnormal work order that appears in logistics distribution, so that the problem that logistics order cannot be normally completed distribution, by obtaining the trigger event of work order exception, the trigger event is classified and stored, and the exception work order dataset is obtained;According to the single rule of pre-configuration, the exception work order in exception work order dataset is distributed, so that exception work order is handled;And real-time update the processing state of exception work order, configure query interface, for user to obtain the relevant information of exception work order.It is convenient to analyze the reason of exception work order distribution failure, to carry out targeted improvement, improve the processing efficiency of exception work order.

Description

Technical Field

[0001] This invention belongs to the technical field of logistics management, and in particular relates to a method, apparatus, equipment and storage medium for processing abnormal logistics work orders. Background Technology

[0002] With the continuous development of internet technology, more and more people are gradually adopting online shopping as their main way of shopping in daily life, which in turn promotes the development of the logistics industry.

[0003] The typical flow of express parcels in the logistics chain is pickup → transit → delivery. This process involves multiple logistics nodes, which are the junctions connecting logistics routes in the logistics network. Logistics nodes include, but are not limited to, warehouses, distribution centers, last-mile delivery points, Cainiao stations, etc. Each logistics node can transit, collect, distribute, and store the goods corresponding to the order.

[0004] Because express parcels need to pass through multiple distribution centers or outlets, problems such as improper operation and low transit efficiency often occur during the transfer process, which can lead to delays, loss, and other abnormal situations. This reduces the service quality of logistics nodes, causes customer complaints, and affects the service quality and brand image of logistics companies. Summary of the Invention

[0005] The purpose of this invention is to provide a method, apparatus, equipment, and storage medium for processing abnormal work orders in logistics, to obtain the cause of abnormality in abnormal work orders, to reasonably allocate abnormal work orders, thereby completing the processing of abnormal work orders, facilitating the analysis of the reasons for failure of abnormal work orders, and improving the processing efficiency of abnormal work orders.

[0006] To solve the above problems, the technical solution of the present invention is as follows:

[0007] A method for handling abnormal logistics work orders includes:

[0008] The server obtains the triggering events of the work order exception, classifies and stores the triggering events, and obtains the exception work order dataset;

[0009] According to the pre-configured allocation rules, abnormal work orders in the abnormal work order dataset are allocated so that the abnormal work orders can be processed; the pre-configured allocation rules include work order allocation based on source, work order allocation based on overflow source, and work order allocation based on priority.

[0010] Call the pre-trained exception handling model, input the exception work order into the exception handling model, and perform processing strategy matching to handle the exception work order;

[0011] The system updates the processing status of abnormal work orders in real time and configures a query interface to allow users to obtain relevant information about the processing of abnormal work orders.

[0012] According to an embodiment of the present invention, the process of obtaining triggering events for work order anomalies, classifying and storing the triggering events further includes:

[0013] Obtain the trigger events to be classified, wherein the trigger events are used to describe the triggering operation;

[0014] Obtain the user's identification information and the attribute information of the triggering operation from the triggering event;

[0015] The first feature parameter used to describe the abnormal work order is obtained based on the user's identification information;

[0016] The second characteristic parameter used to describe the abnormal work order is determined based on the attribute information of the triggering operation;

[0017] The first feature parameter and the second feature parameter are input into a preset classification model, wherein the classification model uses the first feature parameter and the second feature parameter as classification parameters;

[0018] Obtain the classification results output by the classification model and save them in the abnormal work order dataset.

[0019] According to an embodiment of the present invention, obtaining the triggering event to be classified further includes:

[0020] Receive an online trigger stream and extract the trigger event from the trigger stream;

[0021] Alternatively, obtain the log of the triggering operation and extract the triggering event from the operation log.

[0022] According to an embodiment of the present invention, allocating abnormal work orders in the abnormal work order dataset according to pre-configured work order allocation rules further includes:

[0023] After detecting the triggering event of work order abnormality, initiate a work order splitting task and obtain the abnormal work order data corresponding to the work order splitting task;

[0024] Based on the type of abnormal work order, determine the processing method for the sub-task, and execute the sub-task according to the processing method; wherein, the correspondence between abnormal work order types and processing methods is stored in advance.

[0025] According to an embodiment of the present invention, the step of invoking a pre-trained exception handling model, inputting the exception work order into the exception handling model, and performing processing strategy matching to process the exception work order further includes:

[0026] Obtain the abnormal work order dataset and processing strategy as the training sample set, and derive the confidence of the abnormal processing model based on the confidence function;

[0027] The confidence level of the anomaly handling model is compared with a preset confidence threshold, and the handling strategy for the anomaly work order with high confidence is output and added to the training sample set to obtain an expanded training sample set.

[0028] The anomaly handling model is corrected based on the confidence score output by the anomaly handling model, and an optimized anomaly handling model is generated through self-training.

[0029] According to an embodiment of the present invention, the real-time update of the processing status of abnormal work orders further includes:

[0030] Create a task status update listener;

[0031] The task status update listener is used to listen for task status update events. When the task status update listener detects a task status update event, the status of the current abnormal work order processing task is updated to the status result of the task status update event.

[0032] A device for processing abnormal logistics work orders, comprising:

[0033] The classification module is used by the server to obtain the triggering events of work order anomalies, classify and store the triggering events, and obtain an abnormal work order dataset.

[0034] The work order allocation module is used to allocate abnormal work orders in the abnormal work order dataset according to pre-configured work order allocation rules, so that the abnormal work orders can be processed; the pre-configured work order allocation rules include work order allocation based on source, work order allocation based on overflow source, and work order allocation based on priority.

[0035] The processing module calls a pre-trained exception handling model, inputs the exception work order into the exception handling model, and performs processing strategy matching to process the exception work order.

[0036] The query module is used to update the processing status of abnormal work orders in real time and configure a query interface for users to obtain relevant information about abnormal work orders.

[0037] According to one embodiment of the present invention, the classification module further includes:

[0038] An event acquisition unit is used to acquire trigger events to be classified, wherein the trigger events are used to describe trigger operations;

[0039] A data acquisition unit is used to acquire the user's identification information and the attribute information of the triggering operation from the triggering event;

[0040] The first feature extraction unit is used to obtain a first feature parameter for describing the abnormal work order based on the user's identification information;

[0041] The second feature extraction unit is used to determine the second feature parameters for describing the abnormal work order based on the attribute information of the triggering operation.

[0042] A classification unit is used to input the first feature parameter and the second feature parameter into a preset classification model, wherein the classification model uses the first feature parameter and the second feature parameter as classification parameters;

[0043] The storage unit is used to obtain the classification results output by the classification model and save them in the abnormal work order dataset.

[0044] A device for processing abnormal logistics work orders includes a memory and a processor. The memory stores computer-readable instructions, which, when executed by the processor, cause the processor to perform the steps in the method for processing abnormal logistics work orders according to an embodiment of the present invention.

[0045] A storage medium storing computer-readable instructions, which, when executed by one or more processors, cause the one or more processors to perform steps in a method for processing logistics exception work orders according to an embodiment of the present invention.

[0046] Because of the above technical solutions, this invention has the following advantages and positive effects compared with the prior art:

[0047] This invention discloses a method for handling abnormal logistics work orders. Addressing the issue of abnormal work orders preventing normal delivery, the method acquires, categorizes, and stores trigger events for work order abnormalities, creating an abnormal work order dataset. Abnormal work orders are then allocated according to pre-configured allocation rules to facilitate processing. The method also updates the processing status of abnormal work orders in real time and configures a query interface for users to retrieve relevant information. This facilitates analysis of the reasons for delivery failures and allows for targeted improvements, thereby increasing the efficiency of abnormal work order processing. Attached Figure Description

[0048] Figure 1 This is a flowchart illustrating a method for handling abnormal logistics work orders in one embodiment of the present invention.

[0049] Figure 2 This is a flowchart of a method for classifying triggering events in one embodiment of the present invention;

[0050] Figure 3 This is a block diagram of a logistics anomaly work order processing device according to an embodiment of the present invention.

[0051] Figure 4 This is a schematic diagram of a logistics anomaly work order processing device according to an embodiment of the present invention. Detailed Implementation

[0052] The following detailed description, in conjunction with the accompanying drawings and specific embodiments, provides a method, apparatus, device, and storage medium for processing abnormal logistics work orders according to the present invention. The advantages and features of the present invention will become clearer from the following description and claims.

[0053] Example 1

[0054] This embodiment addresses the problem of abnormal work orders in logistics delivery, which prevents the normal completion of logistics orders. It provides a method for handling abnormal work orders. By obtaining the cause of the abnormality of the work order, the abnormal work orders are reasonably allocated, thereby completing the processing of abnormal work orders. This facilitates the analysis of the reasons for the failure of abnormal work order delivery and improves the processing efficiency of abnormal work orders.

[0055] Please refer to Figure 1 The handling method for this logistics exception work order includes the following steps:

[0056] S1: The server obtains the triggering events of the work order exception, classifies and stores the triggering events, and obtains the exception work order dataset;

[0057] S2: According to the pre-configured work order allocation rules, the abnormal work orders in the abnormal work order dataset are allocated so that the abnormal work orders can be processed; the pre-configured work order allocation rules include work order allocation based on source, work order allocation based on overflow source, and work order allocation based on priority.

[0058] S3: Call the pre-trained exception handling model, input the exception work order into the exception handling model, and perform processing strategy matching to handle the exception work order.

[0059] S4: Real-time updates on the processing status of abnormal work orders, and configuration of a query interface for users to obtain relevant information on the processing of abnormal work orders.

[0060] Specifically, in step S1, the server obtains the triggering events for work order anomalies, classifies and stores these events, and obtains an abnormal work order dataset. This can be achieved through the following steps; please refer to [link / reference]. Figure 2 :

[0061] S11: Obtain the trigger event to be classified, which is used to describe the trigger operation;

[0062] S12: Obtain the user's identification information and the attribute information of the triggering operation from the triggering event;

[0063] S13: Obtain the first characteristic parameter used to describe the abnormal work order based on the user's identification information;

[0064] S14: Determine the second characteristic parameter used to describe the abnormal work order based on the attribute information of the triggering operation;

[0065] S15: Input the first feature parameter and the second feature parameter into a preset classification model, which uses the first feature parameter and the second feature parameter as classification parameters;

[0066] S16: Obtain the classification results output by the classification model and save them in the abnormal work order dataset.

[0067] In step S11, the server obtains the trigger events to be classified. It can do so by receiving the trigger stream from the online logistics system and separating the trigger events from the trigger stream; or by obtaining the log of the trigger operation and extracting the trigger events from the operation log.

[0068] The triggering events for abnormal work orders in logistics and delivery can be categorized as follows:

[0069] Abnormal scheduled time: After the branch / salesperson confirms the pickup time by phone, the system triggers an SMS. If the customer replies with a question, the order will be automatically marked as abnormal.

[0070] Allocation Anomaly: When the system uses automatic order allocation logic, if an order cannot find a primary branch, an anomaly is triggered immediately.

[0071] Branch Refuses to Accept Orders: For orders assigned to branches by headquarters or branches by the system, if the branch clicks to refuse the order, an exception will be triggered immediately.

[0072] Outlet feedback error: Outlet click error feedback is triggered immediately;

[0073] Customer feedback anomaly: The system triggers an SMS message upon completion of pickup, and the customer's response with questions immediately triggers an anomaly.

[0074] Delivery driver assignment timeout: If a delivery driver is not assigned within half an hour after the branch accepts the order, an error is triggered.

[0075] Pickup timeout exception: An exception is triggered when an order is not picked up before the pickup deadline.

[0076] Customer follow-up requests: An anomaly is triggered when headquarters customer service clicks on a follow-up request in the system and when the customer clicks on the member's side.

[0077] Salesperson feedback error: When a salesperson clicks on the error feedback on the collection terminal and selects the type as "customer-initiated cancellation" and the customer cannot be contacted, the system triggers a call. The customer then replies with questions, triggering the error.

[0078] Branch order cancellation: When a branch cancels an order, the system sends an SMS message. If the customer responds with questions within a specified time, an exception is triggered.

[0079] Failure to respond to customers in a timely manner: Failure to contact customers within the specified time limit triggers an exception;

[0080] Customer cancellation exception: Customers can select the reason for cancellation on the member side, such as no one contacting them to pick up the package, poor service attitude, or slow pickup, which triggers an exception.

[0081] Customer Complaint: A customer filed a complaint on the member side, triggering an anomaly.

[0082] The triggering events for the aforementioned abnormal work orders clearly define the reasons for their occurrence and the corresponding triggering operations.

[0083] In step S12, the user's identification information and the attribute information of the triggering operation are obtained from the triggering event.

[0084] The triggering event can be online data. The server receives the trigger stream from the logistics system and extracts the triggering event from it. For example, if a branch employee clicks to cancel an order through the logistics system, the event agent in the logistics system receives the event trigger signal, processes it accordingly, and generates a corresponding exception work order. When many users operate on the logistics system, the event agent processes the continuous trigger stream. When the event agent needs to process a triggering event, it extracts the triggering event from the trigger stream. The server receives the online trigger stream from the logistics system, extracts the triggering event from the trigger stream through a pre-built information processing module, and processes the triggering event.

[0085] The trigger event can also be offline data. The server retrieves the trigger event from the operation logs. For example, clicking a log entry will extract the trigger event from the operation logs, which will then be processed.

[0086] The user's identification information and the attribute information of the triggering operation can be obtained from the triggering event. The triggering event data can be segmented into words using a text-based word segmentation model, and the segmented information can be classified.

[0087] The triggering events of the above-mentioned abnormal work orders show that these events can be triggered by the customer (e.g., disputes over agreed-upon time, customer requests for pickup, abnormal customer feedback, etc.), by the branch salesperson (e.g., refusal to accept orders, abnormal feedback, etc.), or automatically by the system (e.g., abnormal allocation, failure to respond to customers in a timely manner, pickup timeout, etc.). Therefore, user identification information, such as the customer's ID, the branch salesperson's ID, and the system administrator information, can be obtained from the triggering events of abnormal work orders.

[0088] The attribute information that triggers the operation can be timeout, rejection, abnormal feedback, complaint, doubt, etc.

[0089] The word segmentation model can extract user identifiers and attribute information of triggering operations from the triggering event information. This model uses the LAC segmentation tool to segment and label the triggering event information, then removes redundant and weakly defined words based on importance weights, and finally performs synonym replacement on the remaining segments to obtain the text to be processed. LAC, short for Lexical Analysis of Chinese, is a joint lexical analysis tool developed by Baidu's Natural Language Processing Department, which enables Chinese word segmentation, part-of-speech tagging, and proper noun recognition.

[0090] The TF-IDF algorithm is used to calculate the similarity between the text to be processed and the corpus in a pre-set database to obtain the matching results. Similarity calculation is performed by consulting the TF-IDF table, where the TF-IDF table is a weighted table calculated after word segmentation of all text in the corpus. The similarity is calculated by cosine similarity between the processed text and all texts in the database that have undergone the same processing. The pre-set database contains corpus data related to logistics management processes, including logistics clients, logistics order systems, route scheduling systems, timeliness management systems, and exception handling systems.

[0091] When the obtained similarity score is lower than the matching threshold, the BERT model is used to classify the text to be processed, and the matching result is determined based on the prediction value of the BERT model. For example, if the matching threshold is set to 0.97, after performing the above similarity calculation on all data in the database, the top 5 (sorted from highest to lowest similarity) are selected. If only the top 1 is greater than 0.97, an accurate matching result is returned. If multiple values ​​in the top 5 are greater than 0.97, a result list is returned. Conversely, if the value of the top 1 is less than 0.97, the text to be processed is input into the pre-trained BERT model for re-judgment. If the maximum score obtained by the BERT model is greater than 0.6, and its classification result is in the top 5, the result corresponding to that score is returned. If the maximum score obtained by the BERT model is less than 0.6, a fixed value (e.g., no match) is returned.

[0092] This word segmentation model can also use other algorithms, such as LSTM-based word segmentation + CRF classification.

[0093] In step S13, the first feature parameter used to describe the abnormal work order is obtained based on the user's identification information. That is, the user identification information (such as customer ID, branch salesperson ID, system administrator information) obtained in step S12 is used to identify the user identification information and obtain the corresponding role (such as customer, branch salesperson, system) as the first feature parameter of the abnormal work order.

[0094] In step S14, a second characteristic parameter is determined based on the attribute information of the triggering operation to describe the abnormal work order. This second characteristic parameter may include timeout (pickup timeout, assignment timeout), order rejection, abnormal feedback, complaint, or ambiguity.

[0095] In step S15, the first feature parameter and the second feature parameter are input into a preset classification model, which uses the first feature parameter and the second feature parameter as classification parameters.

[0096] This classification model can be text-based, concatenating the first and second feature parameters to output the corresponding type. For example, if the first feature parameter is "customer" and the second feature parameter is "complaint," the classification model outputs "customer complaint." Similarly, if the first feature parameter is "branch salesperson" and the second feature parameter is "refusal to accept order," the classification model outputs "branch salesperson refuses to accept order."

[0097] In step S16, the classification results output by the classification model are obtained and saved in the abnormal work order dataset. The classification model is trained on the triggering events of the above-mentioned types of abnormal work orders to ensure that the classification accuracy meets the preset requirements.

[0098] The trained classification model is used to classify the triggering events to be classified, and the classification results of the abnormal work orders are obtained and saved in the abnormal work order dataset. This abnormal work order dataset includes at least the work order number, the time of abnormal work order generation, and the reason for the abnormality.

[0099] In step S2, abnormal work orders in the abnormal work order dataset are allocated according to the pre-configured work order allocation rules so that the abnormal work orders can be processed.

[0100] The pre-configured work order allocation rules include work order allocation based on source, work order allocation based on overflow source, and work order allocation based on priority.

[0101] Source-based work order allocation refers to work orders that fall under the category of network-related issues such as abnormal scheduled delivery times, abnormal branch feedback, delayed courier assignment, delayed pickup, or failure to respond to customers in a timely manner, allowing for the identification of corresponding personnel for handling. When these abnormal work orders occur, they can be assigned to the appropriate personnel based on their source, ensuring that the abnormal work orders are processed.

[0102] Overflowing work order allocation refers to abnormal work orders that fall under categories such as allocation anomalies, branch refusal to accept orders, branch order cancellations, failure to respond to customers in a timely manner, abnormal customer cancellations, and customer complaints. If no corresponding personnel can be found to handle the work order, the order grabbing strategy will be implemented.

[0103] For abnormal work orders originating from overflow sources, configure order-grabbing priorities, with higher priority orders being allocated first.

[0104] If priorities are the same, orders with older generation times will be assigned priority. The system displays configuration items for all business provinces. For a specific business province, you can select one or more provinces. After saving, abnormal work orders from the selected provinces can be processed by that province. You can also select exception types, and selected types can be used to accept orders. Finally, you can select partners, and selected partners can accept orders.

[0105] Work order workflow logic: The system generates an abnormal work order; it checks if the province to which the corresponding order belongs has a provincial order-grabbing configuration. If so, the work order is added to the provincial company's abnormal work order pending allocation list; otherwise, it is directly added to the headquarters' pending allocation list. If the provincial company does not allocate the work order within 60 minutes, it automatically returns to the headquarters' pending processing list, and a new record is generated in the order operation log starting from the work order generation time.

[0106] Among them, order cancellations by branches, failure to respond to customers in a timely manner, overdue assignment of couriers, overdue pickup of packages, and abnormal feedback from branches are considered low priority.

[0107] Allocation anomalies and branch refusal to accept orders are classified as medium priority.

[0108] Customer complaints, abnormal customer cancellations, abnormal salesperson feedback, abnormal customer requests for additional orders, abnormal customer reviews, and abnormal scheduled times are all considered high priority.

[0109] The abnormal work orders in the abnormal work order dataset are allocated according to the pre-configured allocation rules, which can be achieved in the following ways:

[0110] Method 1: After detecting a trigger event for a work order exception, initiate a work order sub-task and obtain the exception work order data corresponding to the sub-task; determine the processing method for the sub-task based on the type of exception work order, and execute the sub-task according to the processing method; wherein, the correspondence between exception work order types and processing methods is pre-stored.

[0111] After detecting a trigger event for an abnormal work order, the server initiates an order allocation task to the logistics system. This allocation task distributes orders in different ways according to pre-configured allocation rules. These pre-configured allocation rules include the aforementioned source-based work order allocation rules and the overflow source-based order grabbing rules.

[0112] Method 2: Call the pre-trained abnormal work order processing model, input the abnormal work order data into the abnormal work order processing model, and match the processing method to process the abnormal work order.

[0113] That is, the above-mentioned order allocation rules are configured in the abnormal work order processing model, and the corresponding processing method is matched according to the type of triggering event of the abnormal work order, and the abnormal work order is allocated to handle the abnormal work order.

[0114] The pre-trained abnormal work order handling model can be a classification model based on a deep neural network or a classification model based on a convolutional neural network.

[0115] In step S3, a pre-trained exception handling model is invoked. The exception work order is input into the exception handling model, and a handling strategy is matched to process the exception work order. Specifically:

[0116] Obtain the abnormal work order dataset and processing strategy as the training sample set, and derive the confidence of the abnormal processing model based on the confidence function;

[0117] The confidence level of the exception handling model is compared with the preset confidence threshold. The handling strategy for the exception work order with high confidence is output and added to the training sample set to obtain the expanded training sample set.

[0118] Based on the confidence results output by the anomaly handling model, the anomaly handling model is corrected, and an optimized anomaly handling model is generated through self-training.

[0119] The abnormal work order to be processed is input into the optimized abnormal handling model, and the processing strategy is matched to output the processing strategy for the abnormal work order in order to process the abnormal work order.

[0120] In this embodiment, the confidence level can be calculated based on association rules using the following formula: Divide by support(X), association rules The confidence score represents the probability that a dataset D contains items of itemset X and also contains items of itemset Y. It is equal to the ratio of the support of itemset X∪Y to ​​the support of itemset X.

[0121] In this embodiment, the specific confidence level setting is a function of the sample size and the range of numerical result fluctuations. This can be understood as the obtained results fluctuating around a specific value, with the confidence level decreasing as the range increases. For example, assuming the average score is 50, the probability of a score between 45 and 55 is obviously smaller than that between 35 and 65, meaning the confidence level is lower. However, the confidence level for a score between 0 and 100 is 100%, because the entire range is limited. Furthermore, a larger sample size generally improves the confidence level; that is, the more individual samples, the more reliable the results.

[0122] When estimating the confidence level, for example, to obtain a 95% confidence level from 10,000 samples, at least 600 samples are needed. The formula for calculating the survey sample size is as follows:

[0123] n=Z[(2×S)2 / d]2

[0124] Where: N: represents the required sample size, Z: the Z-statistic at the confidence level (e.g., the Z-statistic at the 95% confidence level is 1.96), S: the population standard deviation, and d: half of the confidence interval.

[0125] This embodiment can also test the effectiveness of the anomaly handling model based on confidence level.

[0126] This paper applies confidence scores to the calibration of anomaly handling models, using them to improve self-training methods in GCNs. Self-training refers to predicting pseudo-labels for unlabeled nodes and then adding some high-confidence nodes along with the pseudo-labels to the training set, thereby expanding the training set and improving model performance. First, the confidence scores of the GCN output are calibrated, and then the calibrated confidence scores are used to select unlabeled nodes, thus better utilizing correct low-confidence predictions. Compared with other self-training methods, higher output values ​​indicate better performance.

[0127] Obtain abnormal work orders, determine the confidence level of candidate processing strategies through the abnormal handling model, and select the target processing strategy for the abnormal work order based on the confidence level of the candidate processing strategies.

[0128] In this embodiment, the confidence level can be tracked using a generative method, which has a similarity measurement function. Alternatively, a discriminative method can be used, where the discriminator's response represents the classification confidence of a sample. A correlation filter method, which belongs to the discriminative class, can also be used as the tracking confidence level for correlation filter algorithms.

[0129] In step S4, the processing status of abnormal work orders is updated in real time.

[0130] In practical applications, after the server assigns an abnormal work order, it is necessary to track the processing status of the abnormal work order. For example, when an abnormal work order is received, its status is "pending assignment"; after the server assigns the abnormal work order, its status is "processing"; and when the server receives the completion information of the abnormal work order, its status is "closed".

[0131] The following methods can be used to update the status of abnormal work orders in real time:

[0132] Create a task status update listener and use it to listen for task status update events. When the task status update listener detects the task status update event, it updates the status of the current abnormal work order's processing task to the status result of the task status update event.

[0133] The task status update listener can be a low-level process that listens for the sub-tasks published by the server, as shown below:

[0134]

[0135]

[0136] It inherits from chord and adds task_id to the options dictionary in the body, so that when using this class as the default celery.chord, the task_id can be obtained.

[0137] header=[task.s(url=item['href'],page=item['page'],total=self.total,filename=self.filename)for itemin items]

[0138] callback = templink.s(1)

[0139] task = progress_chord(group(header))(callback) #callback is a callback Celerytask task.

[0140] In the task class, use self.request.chord['options']['task_id'] to get the id and use self.update_state(task_id=task_id,state=state,meta=meta) to update it.

[0141] Configure a query interface to allow users to retrieve information about abnormal work orders.

[0142] After the server detects the triggering event of an abnormal work order, it can obtain the relevant data of the abnormal work order and allow users to query the abnormal work order by configuring a query interface.

[0143] Based on the time the abnormal work order was generated, by default, abnormal work order data from day t-14 to day t can be queried;

[0144] Precise queries can be performed based on individual order numbers and abnormal work order numbers;

[0145] Based on the region, it supports five levels of filtering conditions: region, business province, district / county, first-level outlet, and second-level outlet.

[0146] Based on the timeout status: if a salesperson is assigned a timeout or an order pickup is timeout, relevant abnormal work order data can be read from the timeout flag.

[0147] The query results can be sorted by time: work order creation time ascending, work order creation time descending, order placement time ascending, and order placement time descending.

[0148] In addition, the sender's mobile phone only supports precise search, while the pickup address supports fuzzy search.

[0149] Example 2

[0150] This embodiment provides a device for processing abnormal logistics work orders. Please refer to [link / reference]. Figure 3 The device for processing abnormal logistics work orders includes:

[0151] Classification module 1 is used by the server to obtain the triggering events of work order anomalies, classify and store the triggering events, and obtain the abnormal work order dataset;

[0152] The task allocation module 2 is used to allocate abnormal work orders in the abnormal work order dataset according to pre-configured task allocation rules so that the abnormal work orders can be processed; wherein, the pre-configured task allocation rules include work order allocation based on source, work order allocation based on overflow source, and work order allocation based on priority.

[0153] Processing module 3 calls the pre-trained exception handling model, inputs the exception work order into the exception handling model, performs processing strategy matching, and processes the exception work order.

[0154] Query module 4 is used to update the processing status of abnormal work orders in real time and configure a query interface for users to obtain relevant information about abnormal work orders.

[0155] The classification module 1 further includes:

[0156] The event acquisition unit is used to acquire trigger events to be classified. The trigger events are used to describe the triggering operation.

[0157] The data acquisition unit is used to obtain the user's identification information and the attribute information of the triggering operation from the triggering event;

[0158] The first feature extraction unit is used to obtain the first feature parameters for describing the abnormal work order based on the user's identification information;

[0159] The second feature extraction unit is used to determine the second feature parameters for describing the abnormal work order based on the attribute information of the triggering operation.

[0160] A classification unit is used to input the first feature parameter and the second feature parameter into a preset classification model, which uses the first feature parameter and the second feature parameter as classification parameters.

[0161] The storage unit is used to retrieve the classification results output by the classification model and save them in the abnormal work order dataset.

[0162] The functions and implementation methods of the above-mentioned classification module 1, order sorting module 2, processing module 3, and query module 4 are as described in the above embodiment 1, and will not be repeated here.

[0163] Example 3

[0164] This embodiment provides a device for processing abnormal logistics work orders. Please refer to... Figure 4 The processing device 500 for the logistics exception work order can vary significantly due to differences in configuration or performance. It may include one or more central processing units (CPUs) 510 (e.g., one or more processors) and memory 520, and one or more storage media 530 (e.g., one or more mass storage devices) for storing application programs 533 or data 532. The memory 520 and storage media 530 may be temporary or persistent storage. The program stored in the storage media 530 may include one or more modules (not shown in the diagram), each module may include a series of instruction operations on the logistics exception work order processing device 500.

[0165] Furthermore, the processor 510 can be configured to communicate with the storage medium 530 and execute a series of instruction operations in the storage medium 530 on the processing device 500 for logistics exception work orders.

[0166] The equipment 500 for processing abnormal logistics work orders may also include one or more power supplies 540, one or more wired or wireless network interfaces 550, one or more input / output interfaces 560, and / or one or more operating systems 531, such as Windows Server, Vista, etc.

[0167] Those skilled in the art will understand that Figure 4 The illustrated equipment structure for processing logistics exception work orders does not constitute a limitation on the equipment for processing logistics exception work orders. It may include more or fewer components than illustrated, or combine certain components, or have different component arrangements.

[0168] Another embodiment of the present invention also provides a computer-readable storage medium.

[0169] The computer-readable storage medium can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. The computer-readable storage medium stores instructions that, when executed on a computer, cause the computer to perform the steps of the logistics anomaly work order processing method in Embodiment 1.

[0170] If the method for handling abnormal logistics work orders is implemented in the form of program instructions and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this embodiment, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in software. This computer software is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this disclosure. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0171] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific identification content executed by the system and device described above can be referred to the corresponding process in the foregoing method embodiments.

[0172] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings, but the present invention is not limited to the above embodiments. Even if various changes are made to the present invention, if these changes fall within the scope of the claims of the present invention and their equivalents, they shall still fall within the protection scope of the present invention.

Claims

1. A method for handling abnormal logistics work orders, characterized in that, include: The server obtains the triggering events of the work order exception, classifies and stores the triggering events, and obtains the exception work order dataset; According to the pre-configured allocation rules, abnormal work orders in the abnormal work order dataset are allocated so that the abnormal work orders can be processed; the pre-configured allocation rules include work order allocation based on source, work order allocation based on overflow source, and work order allocation based on priority. Call the pre-trained exception handling model, input the exception work order into the exception handling model, and perform processing strategy matching to handle the exception work order; The system updates the processing status of abnormal work orders in real time and configures a query interface to allow users to obtain relevant information about the processing of abnormal work orders.

2. The method for handling abnormal logistics work orders as described in claim 1, characterized in that, The process of obtaining and classifying the triggering events of work order anomalies, and storing them, further includes: Obtain the trigger events to be classified, wherein the trigger events are used to describe the triggering operation; Obtain the user's identification information and the attribute information of the triggering operation from the triggering event; The first feature parameter used to describe the abnormal work order is obtained based on the user's identification information; The second characteristic parameter used to describe the abnormal work order is determined based on the attribute information of the triggering operation; The first feature parameter and the second feature parameter are input into a preset classification model, wherein the classification model uses the first feature parameter and the second feature parameter as classification parameters; Obtain the classification results output by the classification model and save them in the abnormal work order dataset.

3. The method for handling abnormal logistics work orders as described in claim 2, characterized in that, The acquisition of the triggering events to be classified further includes: Receive an online trigger stream, extract the trigger event from the trigger stream, or, Obtain the log of the triggered operation, and extract the triggered event from the operation log.

4. The method for handling abnormal logistics work orders as described in claim 1, characterized in that, The allocation of abnormal work orders in the abnormal work order dataset according to the pre-configured allocation rules further includes: After detecting the triggering event of work order abnormality, initiate a work order splitting task and obtain the abnormal work order data corresponding to the work order splitting task; Based on the type of abnormal work order, determine the processing method for the sub-task, and execute the sub-task according to the processing method; wherein, the correspondence between abnormal work order types and processing methods is stored in advance.

5. The method for handling abnormal logistics work orders as described in claim 1, characterized in that, The step of invoking a pre-trained exception handling model, inputting the exception work order into the exception handling model, and performing processing strategy matching to handle the exception work order further includes: Obtain the abnormal work order dataset and processing strategy as the training sample set, and derive the confidence of the abnormal processing model based on the confidence function; The confidence level of the anomaly handling model is compared with a preset confidence threshold, and the handling strategy for the anomaly work order with high confidence is output and added to the training sample set to obtain an expanded training sample set. The anomaly handling model is corrected based on the confidence score output by the anomaly handling model, and an optimized anomaly handling model is generated through self-training.

6. The method for handling abnormal logistics work orders as described in claim 1, characterized in that, The real-time update of the processing status of abnormal work orders further includes: Create a task status update listener; The task status update listener is used to listen for task status update events. When the task status update listener detects a task status update event, the status of the current abnormal work order processing task is updated to the status result of the task status update event.

7. A device for processing abnormal logistics work orders, characterized in that, include: The classification module is used by the server to obtain the triggering events of work order anomalies, classify and store the triggering events, and obtain an abnormal work order dataset. The work order allocation module is used to allocate abnormal work orders in the abnormal work order dataset according to pre-configured work order allocation rules, so that the abnormal work orders can be processed; the pre-configured work order allocation rules include work order allocation based on source, work order allocation based on overflow source, and work order allocation based on priority. The processing module calls a pre-trained exception handling model, inputs the exception work order into the exception handling model, and performs processing strategy matching to process the exception work order. The query module is used to update the processing status of abnormal work orders in real time and configure a query interface for users to obtain relevant information about abnormal work orders.

8. The device for processing abnormal logistics work orders as described in claim 7, characterized in that, The classification module further includes: An event acquisition unit is used to acquire trigger events to be classified, wherein the trigger events are used to describe trigger operations; A data acquisition unit is used to acquire the user's identification information and the attribute information of the triggering operation from the triggering event; The first feature extraction unit is used to obtain a first feature parameter for describing the abnormal work order based on the user's identification information; The second feature extraction unit is used to determine the second feature parameters for describing the abnormal work order based on the attribute information of the triggering operation. A classification unit is used to input the first feature parameter and the second feature parameter into a preset classification model, wherein the classification model uses the first feature parameter and the second feature parameter as classification parameters; The storage unit is used to obtain the classification results output by the classification model and save them in the abnormal work order dataset.

9. A device for processing abnormal logistics work orders, characterized in that, include: A memory and a processor, wherein the memory stores computer-readable instructions, which, when executed by the processor, cause the processor to perform the steps in the method for processing logistics exception work orders as described in any one of claims 1 to 6.

10. A storage medium storing computer-readable instructions, characterized in that, When the computer-readable instructions are executed by one or more processors, the one or more processors perform the steps in the method for processing logistics exception work orders as described in any one of claims 1 to 6.

Images