Work order shunting method, system and equipment and storage medium

Abstract

The invention discloses a work order shunting method, system and device, and a storage medium. The method comprises the following steps: obtaining a work order set; wherein the work order set comprises a plurality of target work orders to be shunted; extracting attribute information of each target work order in multiple dimensions; according to the attribute information, performing feature extraction on each target work order to obtain feature data corresponding to each target work order; clustering each target work order according to the feature data, and determining a clustering category to which each target work order belongs; and distributing a service channel corresponding to each target work order according to the clustering category to which the target work order belongs. According to the work order shunting method provided by the invention, the multi-dimensional feature extraction and intelligent clustering technology is introduced, so that the work order tasks can be conveniently identified and shunted, the probability that the work orders are overstocked is reduced, and the work order processing efficiency and the automation level are remarkably improved. The method can be widely applied to the technical field of work order shunting.

Description

Technical Field

[0001] The present application relates to the technical field of work order diversion, and in particular to a work order diversion method, system, device and storage medium. Background Art

[0002] Currently, within telecom operators' service systems, ticket diversion technology for Customer Relationship Management (CRM) and Operation Support Systems (OSS) systems is primarily based on business logic. The specific process typically involves: 1) The CRM system issues sales-related work orders via a message queue (such as Kafka or RabbitMQ); 2) The work order data is parsed and stored in a database. The OSS system triggers the process engine based on the incoming work order data; 3) The diversion phase relies on real-time parsing of the incoming data, matching business rules (such as product type, region code, and other fields) to differentiate the work order scenarios, thereby initiating the corresponding service activation process.

[0003] In related technologies, the work order flow is completed by serially connecting message queue consumption, data persistence, rule engine decision-making and other links, forming a fixed linear processing link. However, this implementation method has certain defects: the process efficiency bottleneck is prominent, and the work order must go through the message queue consumption, message parsing, database writing, rule matching and other links in sequence. Each step is executed serially and there is I / O delay, resulting in low diversion efficiency and difficulty in meeting the needs of batch work order processing during peak hours. Moreover, this method lacks automated fault tolerance. When work orders are backlogged, there is a lack of a dynamic resource allocation mechanism, and manual intervention is required to handle data parsing or process jams.

[0004] In summary, the problems existing in related technologies need to be solved urgently. Summary of the Invention

[0005] The purpose of this application is to solve one of the technical problems existing in the related art to at least a certain extent.

[0006] To this end, one purpose of the embodiments of the present application is to provide a work order diversion method, system, device and storage medium.

[0007] In order to achieve the above technical objectives, the technical solutions adopted in the embodiments of the present application include:

[0008] In one aspect, an embodiment of the present application provides a work order diversion method, the method comprising:

[0009] Obtaining a work order collection; wherein the work order collection includes multiple target work orders to be diverted;

[0010] Extracting attribute information of each target work order in multiple dimensions;

[0011] Extract features of each target work order based on the attribute information to obtain feature data corresponding to each target work order;

[0012] Clustering each of the target work orders according to the characteristic data to determine the cluster category to which each of the target work orders belongs;

[0013] According to the cluster category to which it belongs, the business channel corresponding to each target work order is allocated.

[0014] In addition, the work order diversion method according to the above embodiment of the present application may also have the following additional technical features:

[0015] Furthermore, in one embodiment of the present application, the dimensions include region dimension, product dimension, channel acceptance dimension, action dimension and log dimension.

[0016] Furthermore, in one embodiment of the present application, performing feature extraction on each target work order based on the attribute information to obtain feature data corresponding to each target work order includes:

[0017] Filtering the attribute information corresponding to the target work order through a mapping-reduction model to obtain valid project data under each dimension;

[0018] According to the valid project data under each of the dimensions, characteristic data corresponding to the target work order is obtained.

[0019] Furthermore, in one embodiment of the present application, clustering the target work orders according to the feature data to determine the cluster category to which each target work order belongs includes:

[0020] Extracting feature vectors corresponding to each feature data;

[0021] Select the feature vector corresponding to any target work order as the initial cluster centroid vector to establish an initial cluster cluster;

[0022] Calculating the similarity between each feature vector to be clustered and the initial cluster centroid vector; wherein the feature vector to be clustered is the feature vector corresponding to the target work order that is not added to the cluster;

[0023] When the similarity between the feature vector to be clustered and the initial cluster center vector is greater than or equal to a preset threshold, the target work order corresponding to the feature vector to be clustered is added to the initial cluster cluster corresponding to the initial cluster center vector, and the cluster center coordinates of the initial cluster cluster are updated;

[0024] When the similarity between the feature vector to be clustered and any of the initial cluster centroid vectors is less than the preset threshold, the feature vector to be clustered is used as the initial cluster centroid vector to additionally establish a new initial cluster.

[0025] Furthermore, in one embodiment of the present application, the calculating of the similarity between each feature vector to be clustered and the initial cluster centroid vector includes:

[0026] Determining a first length of the feature vector to be clustered and a second length of the initial cluster centroid vector;

[0027] Obtaining a first value according to the product of the first length and the second length, and obtaining a second value according to the inner product of the feature vector to be clustered and the initial cluster center vector;

[0028] The similarity is obtained according to the quotient of the second value and the first value.

[0029] Furthermore, in one embodiment of the present application, clustering the target work orders according to the feature data to determine the cluster category to which each target work order belongs includes:

[0030] Get pre-set constraint information;

[0031] Clustering is performed on each of the target work orders according to the constraint condition information and the characteristic data, and the cluster category to which each of the target work orders belongs is determined.

[0032] Furthermore, in one embodiment of the present application, obtaining a work order collection includes:

[0033] The work order collection is obtained through a predetermined message middleware.

[0034] On the other hand, an embodiment of the present application provides a work order diversion system, the system comprising:

[0035] An acquisition unit, configured to acquire a work order collection; wherein the work order collection includes a plurality of target work orders to be diverted;

[0036] An extraction unit, configured to extract attribute information of each target work order in multiple dimensions;

[0037] A processing unit, configured to extract features of each target work order based on the attribute information to obtain feature data corresponding to each target work order;

[0038] A clustering unit, configured to cluster each of the target work orders according to the feature data, and determine a cluster category to which each of the target work orders belongs;

[0039] The allocating unit is configured to allocate the service channel corresponding to each target work order according to the cluster category to which it belongs.

[0040] In another aspect, an embodiment of the present application provides an electronic device, including:

[0041] at least one processor;

[0042] at least one memory for storing at least one program;

[0043] When the at least one program is executed by the at least one processor, the at least one processor implements the above-mentioned work order diversion method.

[0044] On the other hand, an embodiment of the present application further provides a computer-readable storage medium, which stores a program executable by a processor. When the program executable by the processor is executed, it is used to implement the above-mentioned work order diversion method.

[0045] The advantages and benefits of this application will be partially given in the following description, and partially become apparent from the following description, or learned through practice of this application:

[0046] The embodiments of the present application disclose a work order diversion method, system, device and storage medium, which obtains a work order collection; wherein the work order collection includes multiple target work orders to be diverted; extracts attribute information of each target work order in multiple dimensions; based on the attribute information, performs feature extraction on each target work order to obtain feature data corresponding to each target work order; based on the feature data, clusters each target work order to determine the cluster category to which each target work order belongs; and allocates the business channel corresponding to each target work order according to the cluster category to which it belongs. The work order diversion method proposed in the present application, by introducing multi-dimensional feature extraction and intelligent clustering technology, can conveniently identify and divert work order tasks, reduce the probability of work order backlogs, and significantly improve the efficiency and automation level of work order processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following introduction is made to the drawings of the embodiments of the present application or the related technical solutions in the prior art. It should be understood that the drawings introduced below are only for the convenience of clearly describing some embodiments of the technical solutions of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without any creative work.

[0048] Figure 1 A schematic diagram of an implementation environment of a work order diversion method provided in an embodiment of the present application;

[0049] Figure 2 A flowchart of a work order diversion method provided in an embodiment of the present application;

[0050] Figure 3 A schematic diagram of a process for extracting features from a target work order provided in an embodiment of the present application;

[0051] Figure 4 A schematic diagram of a process for determining the cluster category to which a target work order belongs provided in an embodiment of the present application;

[0052] Figure 5 A schematic diagram of a process for determining similarity provided in an embodiment of the present application;

[0053] Figure 6 This is a schematic diagram of another process for determining the cluster category to which a target work order belongs, provided in an embodiment of the present application;

[0054] Figure 7 A schematic diagram of a business architecture provided in an embodiment of the present application;

[0055] Figure 8 A schematic diagram of an application flow of work order diversion provided in an embodiment of the present application;

[0056] Figure 9 This is a structural diagram of a work order diversion system provided in an embodiment of the present application;

[0057] Figure 10 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0058] The present application is further described below in conjunction with the accompanying drawings and specific embodiments. The described embodiments should not be considered as limiting the present application. All other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0059] In the following description, reference is made to “some embodiments”, which describes a subset of all possible embodiments, but it will be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0060] Currently, within telecom operators' service systems, ticket diversion technology between Customer Relationship Management (CRM) and Operations Support Systems (OSS) systems is primarily based on business logic. The specific process typically involves: 1) The CRM system issues sales-related work orders via a message queue (such as Kafka or RabbitMQ); 2) The work order data is parsed and stored in a database. The OSS system triggers the process engine based on the incoming work order data; 3) The diversion phase relies on real-time parsing of the incoming data, matching business rules (such as product type, region code, and other fields) to differentiate the work order scenarios, thereby initiating the corresponding service activation process.

[0061] In related technologies, the work order flow is completed by serially connecting message queue consumption, data persistence, rule engine decision-making and other links, forming a fixed linear processing link. However, this implementation method has certain defects: the process efficiency bottleneck is prominent, and the work order must go through the message queue consumption, message parsing, database writing, rule matching and other links in sequence. Each step is executed serially and there is I / O delay, resulting in low diversion efficiency and difficulty in meeting the needs of batch work order processing during peak hours. Moreover, this method lacks automated fault tolerance. When work orders are backlogged, there is a lack of a dynamic resource allocation mechanism, and manual intervention is required to handle data parsing or process jams.

[0062] In view of this, a work order diversion method, system, device and storage medium are provided in an embodiment of the present application to obtain a work order collection; wherein the work order collection includes multiple target work orders to be diverted; the attribute information of each target work order in multiple dimensions is extracted; based on the attribute information, feature extraction is performed on each target work order to obtain feature data corresponding to each target work order; based on the feature data, each target work order is clustered to determine the cluster category to which each target work order belongs; and according to the cluster category to which it belongs, the business channel corresponding to each target work order is allocated. The work order diversion method proposed in the present application, by introducing multi-dimensional feature extraction and intelligent clustering technology, can conveniently identify and divert work order tasks, reduce the probability of work order backlogs, and significantly improve the efficiency and automation level of work order processing.

[0063] Please refer to Figure 1 , Figure 1 The following is a schematic diagram of an implementation environment of a work order diversion method provided in an embodiment of the present application. In this implementation environment, the main hardware and software entities involved include a terminal device 110 and a backend server 120. The terminal device 110 and the backend server 120 are in communication connection with each other.

[0064] The work order diversion method provided in the embodiments of the present application can be configured on the backend server 120 side and executed based on the data interaction between the terminal device 110 and the backend server 120. Specifically, the terminal device 110 can be used for users to initiate relevant business requests, and the backend server 120 can receive and process the business requests initiated by the users and feedback the relevant processing results to the users. On the backend server 120 side, when a business request initiated by a user is received, a relevant work order can be generated. By issuing the work order, it can be given to the corresponding business personnel for processing, and the relevant processing results are summarized and fed back to the terminal device 110. Among them, the terminal device 110 in the above embodiment can include a mobile phone, a computer, a smart wearable device, an intelligent voice interaction device, an in-vehicle terminal, etc., but is not limited to this. The backend server 120 can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers. It can also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms.

[0065] The terminal device 110 and the backend server 120 may establish a communication connection via a wireless network or a wired network. The wireless network or wired network uses standard communication technologies and / or protocols, and the network may be the Internet or any other network, including but not limited to a local area network (LAN), a metropolitan area network (MAN), a wide area network (WAN), a mobile, wired or wireless network, a private network or any combination of a virtual private network.

[0066] Of course, it is understandable that Figure 1 The implementation environment is just some optional application scenarios of the work order diversion method provided in the embodiment of this application. The actual application is not fixed. Figure 1 The hardware and software environment shown.

[0067] Below, in combination with the introduction of the aforementioned implementation environment, a work order diversion method provided in an embodiment of the present application is introduced and explained.

[0068] Please refer to Figure 2 , Figure 2 Schematic diagram of a work order diversion method provided in an embodiment of the present application. Specifically, the work order diversion method provided in an embodiment of the present application includes but is not limited to:

[0069] Step 210: Obtain a collection of work orders; wherein the collection of work orders includes multiple target work orders to be diverted;

[0070] Step 220: extract attribute information of each target work order in multiple dimensions;

[0071] Step 230: Extract features of each target work order based on the attribute information to obtain feature data corresponding to each target work order;

[0072] Step 240: Cluster the target work orders according to the feature data to determine the cluster category to which each target work order belongs;

[0073] Step 250: Allocate a business channel corresponding to each target work order according to the cluster category to which it belongs.

[0074] In an embodiment of the present application, a work order diversion method is provided, which can be applied in related enterprises, such as a telecommunications operator system, which is responsible for managing tasks such as the installation, maintenance, and billing of broadband and mobile services. In traditional telecommunications operator systems, CRM (Customer Relationship Management) systems and OSS (Operation Support Systems) systems are generally configured. The method in the embodiment of the present application can be implemented based on the above-mentioned systems. Of course, it can be understood that the method in the embodiment of the present application is not limited to the above-mentioned application scenarios, and it can be deployed in various application scenarios according to needs, and the present application does not impose any restrictions on this.

[0075] The method in the embodiment of the present application is a work order diversion method based on artificial intelligence technology. This method designs how to perform multi-dimensional diversion in the scenario of large-scale work order data based on actual work order processing requirements and standards, so as to achieve the ability to efficiently, accurately and reliably allocate work orders.

[0076] In step 210, a work order collection is obtained, where the work order collection includes multiple target work orders to be diverted.

[0077] In the embodiment of the present application, when diverting work orders, work orders to be diverted can be obtained in batches from upstream business systems (such as CRM systems or order management platforms, etc.), and these work orders are recorded as target work orders. It is understood that these target work orders can come from orders submitted by customers themselves, orders placed on behalf of customers by customer service, or business requests imported in batches, and this application does not impose any restrictions on this.

[0078] In an embodiment of the present application, a batch of target work orders can be obtained at one time, and these target work orders constitute a work order collection. Generally speaking, the work order collection can exist in the form of structured data, such as a message in JSON / XML format or a record in a database table. The data in each target work order can include some basic fields (such as work order ID, creation time, customer number) and business information (such as product type, service address, priority identifier), etc. This application does not impose any restrictions on this.

[0079] When acquiring a work order collection, you can use a high-concurrency interface or message middleware (such as Kafka or RabbitMQ) to collect work order data in real time or periodically pull it. This ensures stable data reception during peak traffic periods. A work order collection can contain any number of target work orders, and this application does not impose any restrictions on this.

[0080] It should be noted that in some embodiments, in order to improve the efficiency of subsequent processing, the system can perform preliminary cleaning on the original work orders obtained, for example, it can eliminate duplicate data or invalid work orders with incorrect formats, and at the same time supplement the missing default values ​​(such as automatically marking it as a normal level when no priority is specified), and use the original work orders after cleaning as the actual target work orders to be diverted, thereby forming a standardized, high-integrity work order collection, providing high-quality input for subsequent feature extraction and intelligent clustering. In addition, in the embodiments of the present application, the work order collection can be obtained through a parallelized data loading and preprocessing mechanism, which effectively avoids the I / O bottleneck problem in the traditional serial processing mode and can improve the efficiency of work order diversion processing.

[0081] In step 220, after obtaining the work order collection, attribute information of each target work order in multiple dimensions can be extracted. Here, extracting attribute information of the target work order from multiple business dimensions can provide data support for subsequent intelligent diversion.

[0082] Specifically, in an embodiment of the present application, when extracting the attribute information of the target work order, the work order structure can be parsed through a predefined metadata model to identify relevant attribute information. For attribute information, it can be pre-divided into multiple dimensions. In an embodiment of the present application, there is no restriction on the dimensions corresponding to the attribute information, which can be determined in combination with the specific application scenario of the present application. For example, for a business work order in a telecommunications scenario, in some embodiments, the dimensions corresponding to its relevant attribute information may include basic attributes (such as work order ID, creation timestamp, customer level), business attributes (such as product code, service type, package identifier), geographical attributes (such as administrative divisions, grid codes, service outlets) and time attributes (such as promised delivery time limit, urgency identifier) ​​and other dimensional feature fields.

[0083] In some embodiments, the dimensions corresponding to the attribute information may also include regional dimensions, product dimensions, channel acceptance dimensions, action dimensions, and log dimensions. Under the regional dimension, geographical information such as the administrative division code and grid ID corresponding to the target work order can be explicitly marked. The spatial distance between the service address and the resource node can also be calculated in conjunction with the GIS system, and the historical work order density (such as the number of similar work orders in the past week) and resource load conditions (such as the current backlog of tasks for installation and maintenance personnel in the area) of the region can be associated, thereby facilitating a dynamic assessment of regional construction capabilities. The product dimension can record basic information such as the product code and package type corresponding to the target work order, and can further associate derivative indicators such as product knowledge graphs, product technical elements (such as whether fiber-to-the-home equipment is required), and business complexity (such as integrated packages involving cross-platform linkage) to form a three-dimensional feature expression at the product level. The channel acceptance dimension is used to record the source channel of the target work order (such as online self-service ordering, business hall processing, and agent submission). It can further include channel credit (such as the error rate of a certain agent's historical work order) and channel timeliness requirements (such as the priority of VIP customer hotline work orders). The action dimension can identify the business action of the target work order, such as transfer, release, restart, etc. The log dimension, as an auxiliary analysis method, can aggregate system log information (such as error retry records and interface call exceptions) throughout the life cycle of the target work order, which is not restricted by this application.

[0084] In step 230 , after the attribute information is extracted, feature extraction may be performed on the target work order based on the attribute information to obtain feature data corresponding to each target work order.

[0085] In this embodiment of the present application, after completing the collection of work order attribute information, the feature extraction stage can be entered. In this stage, the original attribute information can be converted into a highly discriminative feature representation through multi-level intelligent processing. For each target work order, its corresponding feature data can be obtained.

[0086] Specifically, when determining the feature data corresponding to the target work order based on the attribute information, in some embodiments, the structured attribute information can be deeply encoded. For example, for numerical attribute information (such as the timeliness requirement for work order response), a segmented normalization process based on business knowledge can be used to convert the absolute time value into a relative urgency score in the interval [0, 1], and the score is used as the corresponding feature data; for categorical attribute information (such as product type code), it can be mapped into a low-dimensional dense vector through embedding layer technology, while retaining the business semantics and solving the dimensionality explosion problem of One-Hot encoding. For unstructured data, a domain-adaptive pre-training model based on BERT can be used for semantic vectorization to obtain feature data corresponding to the attribute information, which is not limited in this application.

[0087] In step 240 , after obtaining the characteristic data corresponding to the target work order, the target work order may be clustered to determine the cluster category to which the target work order belongs.

[0088] In an embodiment of the present application, based on the multi-dimensional feature data extracted in the early stage, an intelligent clustering algorithm can be used to automatically classify the target work orders to identify groups of work orders with similar business characteristics. For example, in some embodiments, the importance of different types of work order attributes can be evaluated by feature weighting, so as to dynamically construct a distance measurement model that adapts to different business scenarios. In an embodiment of the present application, there is no limitation on the specific clustering algorithm used. For example, in some embodiments, an improved density clustering algorithm (such as OPTICS or HDBSCAN, etc.) can be used to perform core clustering analysis.

[0089] In the embodiment of the present application, an unsupervised learning method is adopted to form an intelligent and automated work order diversion system based on a clustering algorithm. Within the calibration range, the execution target of the redundant data set is set according to the processing status of the data, and it is associated with the program for intelligent data classification and processing, gradually expanding the classification scope of the data set to form a complete and dynamic diversion mechanism to deal with new situations that may arise.

[0090] In step 250 , after determining the cluster category to which the target work order belongs, a business channel corresponding to each target work order may be allocated.

[0091] In an embodiment of the present application, after completing the intelligent clustering of the target work orders, the dynamic allocation stage of the business channel can be entered. In an embodiment of the present application, the clustering results and the actual needs of business operations are deeply integrated. Specifically, according to the actual needs of the business, a number of business channels can be established. In an embodiment of the present application, for different application scenarios, the same business channel can be automatically allocated to similar target work orders. In this way, the processing and circulation efficiency of work orders can be greatly improved, and target work orders with similar characteristics can be efficiently circulated in the designated distribution path, avoiding the phenomenon of mixed queuing for various types of work orders, thereby improving the overall business processing efficiency.

[0092] In particular, in some embodiments, a flexible allocation mechanism can also be set up for business channels in the system: when a sudden surge in a certain type of work order is detected (such as regional batch failures caused by heavy rain), the capacity expansion strategy of the business channel can be automatically triggered to divert pressure by temporarily enabling a backup channel or dynamically adjusting the business channel weight; for work orders with fuzzy cluster boundaries, a cross-channel collaborative processing mode can be started, and the work order features can be pushed to multiple related channels at the same time, and the channel manager can dynamically determine the optimal route based on the real-time load situation.

[0093] It can be understood that a work order diversion method provided in an embodiment of the present application obtains a work order collection; wherein the work order collection includes multiple target work orders to be diverted; extracts attribute information of each target work order in multiple dimensions; based on the attribute information, performs feature extraction on each target work order to obtain feature data corresponding to each target work order; based on the feature data, clusters each target work order to determine the cluster category to which each target work order belongs; and allocates the business channel corresponding to each target work order according to the cluster category to which it belongs. The work order diversion method proposed by this method, by introducing multi-dimensional feature extraction and intelligent clustering technology, can conveniently identify and divert work order tasks, reduce the probability of work order backlogs, and significantly improve the efficiency and automation level of work order processing.

[0094] Specifically, in some embodiments, please refer to Figure 3 , performing feature extraction on each of the target work orders based on the attribute information to obtain feature data corresponding to each of the target work orders, including:

[0095] Filtering the attribute information corresponding to the target work order through a mapping-reduction model to obtain valid project data under each dimension;

[0096] According to the valid project data under each of the dimensions, characteristic data corresponding to the target work order is obtained.

[0097] In an embodiment of the present application, in the feature extraction stage, a distributed feature engineering framework based on a map-reduce (MapReduce) model can be used to deeply screen and transform the attribute information of the target work order, thereby obtaining the corresponding feature data. Under this process, data purification can be achieved through the multi-level attribute filter in the mapping stage: for example, in the business dimension, invalid or illegal attribute values ​​can be eliminated through a predefined business rule dictionary (such as a product whitelist, a regional service range table); in the statistical dimension, an anomaly detection algorithm based on a box plot and a Z-Score can be used to identify and correct outliers of numerical attributes; in the semantic dimension, regular expressions and a domain keyword library can be used to standardize and reconstruct text attributes. The filtering operation of each dimension is executed by parallel computing nodes to form an intermediate result in the form of a key-value pair (such as <dimension type, purified data>), which significantly improves the processing efficiency of massive work orders. In an embodiment of the present application, the purified data can be recorded as valid project data.

[0098] During the reduction phase, intermediate results can be compressed and refined using a feature importance assessment model. For valid numerical project data, feature correlation analysis based on the MIC (maximum information coefficient) is used to retain feature data that is strongly correlated with work order processing objectives (such as diversion accuracy). For valid categorical project data, a chi-square test is used to screen for enumeration values ​​with significant discrimination. In this way, corresponding feature data can be obtained based on valid project data combinations.

[0099] For example, in the embodiment of the present application, the Map function of MapReduce can be used to process each line of work order feature data, thereby filtering out valid project data. For intuitive expression, the process can be expressed as<key,value> The key-value pairing and processing logic are expressed as follows:

[0100] Map(key i ,value i )->list <key j ,value j >

[0101] In the formula, key i Represents a valid attribute name of attribute information under a certain dimension, value i For its corresponding attribute value, use the mapperAPI function to filter and process, and you can get <key j ,value j >For the valid project data that needs to be output, all feature sets are traversed to finally obtain a more stable and effective set of valid project data, and the corresponding feature data is summarized.

[0102] Specifically, in some embodiments, referring to Figure 4 , clustering each of the target work orders according to the characteristic data to determine the cluster category to which each of the target work orders belongs, including:

[0103] Extracting feature vectors corresponding to each feature data;

[0104] Select the feature vector corresponding to any target work order as the initial cluster centroid vector to establish an initial cluster cluster;

[0105] Calculating the similarity between each feature vector to be clustered and the initial cluster centroid vector; wherein the feature vector to be clustered is the feature vector corresponding to the target work order that is not added to the cluster;

[0106] When the similarity between the feature vector to be clustered and the initial cluster center vector is greater than or equal to a preset threshold, the target work order corresponding to the feature vector to be clustered is added to the initial cluster cluster corresponding to the initial cluster center vector, and the cluster center coordinates of the initial cluster cluster are updated;

[0107] When the similarity between the feature vector to be clustered and any of the initial cluster centroid vectors is less than the preset threshold, the feature vector to be clustered is used as the initial cluster centroid vector to additionally establish a new initial cluster.

[0108] In the embodiment of the present application, when performing cluster analysis, first, the feature vector corresponding to each feature data can be extracted. The specific implementation method of extracting the feature vector is not limited in this application. After obtaining the feature vector, the feature vector corresponding to any target work order can be selected as the initial cluster centroid vector to establish the initial cluster cluster.

[0109] Then, for each target work order whose cluster category has not been determined, its corresponding feature vector is recorded as the feature vector to be clustered. Based on the feature vector to be clustered, the similarity between the feature vector to be clustered and the initial cluster center vector can be determined. It can be understood that the similarity between the feature vector to be clustered and the initial cluster center vector can reflect the similarity between the two target work orders. The greater the similarity, the closer the target work order and the corresponding category of the initial cluster cluster are. Therefore, in an embodiment of the present application, a similarity threshold can be set in advance, recorded as the preset threshold. If the similarity between the feature vector to be clustered and a certain initial cluster center vector is greater than or equal to the preset threshold, it means that the target work order corresponding to the feature vector to be clustered and the target work order in the initial cluster cluster corresponding to the initial cluster center vector are sufficiently similar. Therefore, at this time, the target work order corresponding to the feature vector to be clustered can be added to the initial cluster cluster corresponding to the initial cluster center vector, and the cluster center coordinates of the initial cluster cluster are updated once.

[0110] Conversely, if the similarity between the feature vector to be clustered and all the established initial cluster centroids is less than a preset threshold, this indicates that the target work order corresponding to the feature vector to be clustered is dissimilar to the target work orders in the initial clusters corresponding to the initial cluster centroids. In this case, the target work order corresponding to the feature vector to be clustered needs to be divided into an additional category. That is, a new initial cluster is established using the feature vector to be clustered as the initial cluster centroid. The above process is repeated until the coordinates of the cluster centroids of each initial cluster no longer change or change very little. Clustering can then be stopped. In this way, the target work order can be classified.

[0111] In the embodiments of the present application, there is no restriction on the numerical representation of similarity. For example, in some embodiments, a percentage can be used to represent the magnitude of similarity, and 100% can be set as the maximum value of similarity. When the similarity between the feature vector to be clustered and the initial cluster centroid vector reaches 100%, it means that the two are completely consistent. Generally speaking, commonly used similarity algorithms include cosine similarity algorithm, Jaccard similarity algorithm, Hamming distance algorithm, etc. In the embodiments of the present application, there is no restriction on the specific type of algorithm used.

[0112] Specifically, in some embodiments, referring to Figure 5 , the calculating of the similarity between each feature vector to be clustered and the initial cluster centroid vector includes:

[0113] Determining a first length of the feature vector to be clustered and a second length of the initial cluster centroid vector;

[0114] Obtaining a first value according to the product of the first length and the second length, and obtaining a second value according to the inner product of the feature vector to be clustered and the initial cluster center vector;

[0115] The similarity is obtained according to the quotient of the second value and the first value.

[0116] In an embodiment of the present application, the similarity between the feature vector to be clustered and the initial cluster center vector can be determined by a cosine similarity algorithm. First, the vector length of the feature vector to be clustered can be determined, recorded as the first length, and the length of the initial cluster center vector can be recorded as the second length. Then, the product of the first length and the second length is calculated as the first value, and the inner product of the feature vector to be clustered and the initial cluster center vector is calculated as the second value, and then the quotient of the first value and the second value is calculated as the similarity between the feature vector to be clustered and the initial cluster center vector.

[0117] Specifically, in some embodiments, referring to Figure 6 , clustering each of the target work orders according to the characteristic data to determine the cluster category to which each of the target work orders belongs, including:

[0118] Get pre-set constraint information;

[0119] Clustering is performed on each of the target work orders according to the constraint condition information and the characteristic data, and the cluster category to which each of the target work orders belongs is determined.

[0120] In an embodiment of the present application, during the work order clustering stage, intelligent classification that complies with business rules can be achieved by integrating business-related constraint information and data features in a dual-drive mechanism. Specifically, first, predefined constraint information can be loaded. These constraint information can be flexibly set according to business needs and can include hard constraints (such as "work orders for government and enterprise customers must be clustered independently") and soft constraints (such as "work orders for similar faults are recommended to be clustered in the time dimension"). Through the constraint satisfaction problem (CSP) modeling framework, they are converted into constraints of the clustering algorithm.

[0121] In specific implementation, for the aforementioned hard constraints, the feature spaces of government, enterprises and individual customers can be forcibly isolated when initializing the clustering centers, and constraint violation penalties can be set during the iteration process; for soft constraints, they can be converted into regularization terms of the objective function (such as adding a time smoothing term to make similar work orders in adjacent time periods more easily classified into the same cluster).

[0122] The clustering process dynamically balances the influence of data features and business rules: initially, feature similarity is emphasized to discover the natural distribution of data. As the clustering profile stabilizes in the mid-term, the constraints are gradually strengthened. In some embodiments, an interpretable decision log can be generated during clustering iterations, recording the satisfaction of constraints for analysis and adjustment by operators; this is not a limitation of this application.

[0123] Below, the technical solutions in the embodiments of this application are introduced and explained in conjunction with some specific application examples.

[0124] Please refer to Figure 7 and Figure 8 , Figure 7 A schematic diagram of a business architecture provided in an embodiment of the present application is shown. Figure 8 The following is a schematic diagram of an application process of a work order diversion provided in an embodiment of the present application. The method in the embodiment of the present application can be applied in Figure 7 In the architecture shown, the relevant application processes are as follows Figure 8 As shown, this may include:

[0125] 1) CRM acts as a work order producer and adds the work order data to the MQ queue;

[0126] 2) The message distribution OD module built by the OSS orchestration center reads messages from the MQ queue and consumes them;

[0127] 3) The OD module parses the message header, reads attribute information such as region, product, channel acceptance, and action, and enters it into the Redis cache database;

[0128] 4) The OD module's AI model reads real-time information from the Redis cache database and calculates a work order diversion strategy based on attribute information, current system operating status, exception logs, and other content;

[0129] 5) The OD module distributes the work order to the corresponding business channel, batch channel, or other special channel based on the AI ​​output results;

[0130] 6) The core service of the OSS orchestration center obtains work order messages and parses and stores them in the database;

[0131] 7) The OSS orchestration center starts the process based on the parsed information stored in the database to complete the process distribution goal.

[0132] Compared with the traditional work order diversion technology, it takes 8 seconds from CRM production message to OSS service activation and startup process. The technical solution of this application can effectively improve the efficiency of work order diversion. Specifically, based on this application and the traditional solution, the embodiment of this application uses batch work orders of 50 / minute, 200 / minute, 1000 / minute, and 2000 / minute, and mixed work order types to conduct stress testing test results. The relevant results are shown in Table 1:

[0133] Table 1

[0134]

[0135]

[0136] Experimental data shows that as the number of work orders increases, the concurrency capacity of the traditional solution gradually decreases. When the number of work orders increases to more than 200 per minute, it can be clearly seen that the processing efficiency of the technical solution of this application is improved by about 200%-300%.

[0137] Compared with traditional technical solutions, the technical solution of this application has at least the following advantages:

[0138] 1. Different from the traditional mixed construction of work orders, the intelligent diversion mechanism improves the efficiency of business flow;

[0139] 2. The autonomous diversion mechanism does not require human monitoring or intervention. When a large number of work orders are issued, the system can automatically intervene and adjust to ensure smooth business acceptance and reduce complaints.

[0140] 3. Separate channels for construction of businesses with different characteristics can ensure efficient flow of business in similar construction links, avoiding construction waiting and waste of resources.

[0141] Reference Figure 9 In an embodiment of the present application, a work order diversion system is further provided, and the work order diversion system includes:

[0142] The acquisition unit 910 is configured to acquire a work order collection, wherein the work order collection includes a plurality of target work orders to be diverted;

[0143] An extraction unit 920 is configured to extract attribute information of each target work order in multiple dimensions;

[0144] The processing unit 930 is configured to extract features of each target work order according to the attribute information to obtain feature data corresponding to each target work order;

[0145] A clustering unit 940 is configured to cluster each of the target work orders according to the feature data, and determine a cluster category to which each of the target work orders belongs;

[0146] The allocation unit 950 is configured to allocate the business channel corresponding to each target work order according to the cluster category to which it belongs. It is understood that the contents of the above-mentioned method embodiment are applicable to the present system embodiment. The functions implemented by the present system embodiment are the same as those of the above-mentioned method embodiment, and the beneficial effects achieved are also the same as those achieved by the above-mentioned method embodiment.

[0147] An embodiment of the present application further provides an electronic device, including:

[0148] at least one processor;

[0149] at least one memory for storing at least one program;

[0150] When at least one program is executed by at least one processor, the at least one processor implements the above-mentioned work order diversion method.

[0151] The electronic device in the embodiment of the present application may be a terminal device, a computer device or a server device.

[0152] For example, taking the electronic device as a server device as an example, referring to Figure 10The server device 1000 may vary significantly due to different configurations or performance, and may include one or more central processing units 1010 (CPUs for short), a memory 1060, and one or more storage media 1030 (e.g., one or more mass storage devices) storing application programs 1033 or data 1032. The memory 1060 and the storage medium 1030 may be either short-term storage or persistent storage. The program stored in the storage medium 1030 may include one or more units or modules, each of which may include a series of operating instructions for the server device 1000. Furthermore, the central processing unit 1010 may be configured to communicate with the storage medium 1030 and execute the series of operating instructions in the storage medium 1030 on the server device 1000.

[0153] The server device 1000 may further include one or more power supplies 1020 , one or more wired or wireless network interfaces 1040 , one or more input and output interfaces 1050 , and one or more operating systems 1031 .

[0154] The central processing unit 1010 in the server device 1000 can be used to execute the following Figure 2 An embodiment of a work order diversion method is shown.

[0155] Similarly, the contents of the above method embodiments are applicable to the present electronic device embodiment. The functions specifically implemented by the present electronic device embodiment are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0156] An embodiment of the present application further provides a computer-readable storage medium, in which a program executable by the central processing unit 1010 is stored. When the program executable by the central processing unit 1010 is executed by the central processing unit 1010, it is used to execute the above-mentioned work order diversion method.

[0157] Similarly, the contents of the above method embodiments are applicable to the computer-readable storage medium embodiments. The functions specifically implemented by the computer-readable storage medium embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0158] In some optional embodiments, the functions / operations mentioned in the block diagram may not occur in the order mentioned in the operation diagram. For example, depending on the functions / operations involved, the two boxes shown in succession may actually be executed substantially simultaneously or the boxes can sometimes be executed in reverse order. In addition, the embodiments presented and described in the flow chart of the present application are provided in an exemplary manner for the purpose of providing a more comprehensive understanding of the technology. The disclosed method is not limited to the operations and logic flows presented herein. Optional embodiments are contemplated in which the order of the various operations is changed and the sub-operations described as a part of a larger operation are performed independently.

[0159] In addition, although the present application is described in the context of functional modules, it should be understood that, unless otherwise stated, one or more of the functions and / or features may be integrated into a single physical device and / or software module, or one or more functions and / or features may be implemented in separate physical devices or software modules. It is also understood that a detailed discussion of the actual implementation of each module is not necessary for understanding the present application. More specifically, given the properties, functions, and internal relationships of the various functional modules in the devices disclosed herein, the actual implementation of the module will be understood within the routine skills of an engineer. Therefore, a person skilled in the art can implement the present application as set forth in the claims using ordinary techniques without undue experimentation. It is also understood that the specific concepts disclosed are merely illustrative and are not intended to limit the scope of the present application, which is determined by the full scope of the appended claims and their equivalents.

[0160] If the function 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 application, or the part that contributes to the prior art, or the 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 and includes several instructions for enabling a device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0161] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0162] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic devices), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disc read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering, or processing in another suitable manner as necessary, and then stored in a computer memory.

[0163] It should be understood that various parts of the present application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used to implement: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0164] In the above description of this specification, reference to the terms "one embodiment / example," "another embodiment / example," or "certain embodiments / examples" means that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any appropriate manner in any one or more embodiments or examples.

[0165] Although the embodiments of the present application have been shown and described, those skilled in the art will appreciate that various changes, modifications, substitutions, and variations may be made to the embodiments without departing from the principles and intent of the present application, and that the scope of the present application is defined by the claims and their equivalents.

[0166] The above is a specific description of the preferred implementation of the present application, but the present application is not limited to the embodiments. Those skilled in the art may make various equivalent modifications or substitutions without violating the spirit of the present application, and these equivalent modifications or substitutions are all included in the scope defined by the claims of the present application.

Claims

1. A work order diversion method, characterized in that: The method comprises: Obtaining a work order collection; wherein the work order collection includes multiple target work orders to be diverted; Extracting attribute information of each target work order in multiple dimensions; Extract features of each target work order based on the attribute information to obtain feature data corresponding to each target work order; Clustering each of the target work orders according to the characteristic data to determine the cluster category to which each of the target work orders belongs; According to the cluster category to which it belongs, the business channel corresponding to each target work order is allocated.

2. A work order diversion method according to claim 1, characterized in that: The dimensions include region dimension, product dimension, channel acceptance dimension, action dimension and log dimension.

3. A work order diversion method according to claim 1 or 2, characterized in that: The step of extracting features of each target work order based on the attribute information to obtain feature data corresponding to each target work order includes: Filtering the attribute information corresponding to the target work order through a mapping-reduction model to obtain valid project data under each dimension; According to the valid project data under each of the dimensions, characteristic data corresponding to the target work order is obtained.

4. A work order diversion method according to claim 1, characterized in that: Clustering the target work orders according to the characteristic data to determine the cluster category to which each target work order belongs includes: Extracting feature vectors corresponding to each feature data; Select the feature vector corresponding to any target work order as the initial cluster centroid vector to establish an initial cluster cluster; Calculating the similarity between each feature vector to be clustered and the initial cluster centroid vector; wherein the feature vector to be clustered is the feature vector corresponding to the target work order that is not added to the cluster; When the similarity between the feature vector to be clustered and the initial cluster center vector is greater than or equal to a preset threshold, the target work order corresponding to the feature vector to be clustered is added to the initial cluster cluster corresponding to the initial cluster center vector, and the cluster center coordinates of the initial cluster cluster are updated; When the similarity between the feature vector to be clustered and any of the initial cluster centroid vectors is less than the preset threshold, the feature vector to be clustered is used as the initial cluster centroid vector to additionally establish a new initial cluster.

5. A work order diversion method according to claim 4, characterized in that: The calculating of the similarity between each feature vector to be clustered and the initial cluster centroid vector includes: Determining a first length of the feature vector to be clustered and a second length of the initial cluster centroid vector; Obtaining a first value according to the product of the first length and the second length, and obtaining a second value according to the inner product of the feature vector to be clustered and the initial cluster center vector; The similarity is obtained according to the quotient of the second value and the first value.

6. A work order diversion method according to claim 1, characterized in that: Clustering the target work orders according to the characteristic data to determine the cluster category to which each target work order belongs includes: Get pre-set constraint information; Clustering is performed on each of the target work orders according to the constraint condition information and the characteristic data, and the cluster category to which each of the target work orders belongs is determined.

7. A work order diversion method according to claim 1, characterized in that: The obtaining of the work order collection includes: The work order collection is obtained through a predetermined message middleware.

8. A work order diversion system, characterized in that: The system comprises: An acquisition unit, configured to acquire a work order collection; wherein the work order collection includes a plurality of target work orders to be diverted; An extraction unit, configured to extract attribute information of each target work order in multiple dimensions; A processing unit, configured to extract features of each target work order based on the attribute information to obtain feature data corresponding to each target work order; A clustering unit, configured to cluster each of the target work orders according to the feature data, and determine a cluster category to which each of the target work orders belongs; The allocating unit is configured to allocate the service channel corresponding to each target work order according to the cluster category to which it belongs.

9. An electronic device, characterized in that: include: at least one processor; at least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements a work order diversion method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a program executable by a processor, characterized in that: The processor-executable program is used to implement a work order diversion method as described in any one of claims 1 to 7 when executed by the processor.

Images