Service order review method and system, medium and computer system

Abstract

The present disclosure provides a service order review method, including: obtaining service order data of a service order to be reviewed; obtaining an automatic order review rule set; matching the service order data of the service order to be reviewed with the rules in the automatic order review rule set; and, if the match is successful, inputting the service order data of the service order to be reviewed into the automatic order review process for review. The present disclosure provides a service order review system, a computer system and a computer-readable storage medium.

Description

Technical Field

[0001] The present disclosure relates to the field of Internet technology, and more specifically, to a service order review method, a service order review system, a computer system and a computer-readable storage medium. Background Art

[0002] The service order review occurs during the after-sales stage of the product. When a customer has after-sales needs for a product, he or she first submits an after-sales service application form (referred to as the service order). Only after the review is passed will the after-sales process officially begin.

[0003] The application for a service order usually needs to be accompanied by a text description of the product problem and unstructured data materials such as pictures. The machine learning algorithm has limited analysis capabilities for unstructured data, which is far from the analysis and identification capabilities of humans. Therefore, the automatic review system based on unstructured data analysis algorithms is currently difficult to implement. Currently, the review of service orders is still mainly handled manually.

[0004] In the process of realizing the concept of the present disclosure, the inventors found that there are at least the following problems in the related technologies: manual review consumes a lot of manpower, and in order to meet a certain processing time, it is often necessary to equip more manpower than the daily average processing volume to cope with the peak sales season, when the number of service orders also increases; in addition, the manual review of service orders is complicated, and customers usually need to wait for a long time to get the processing results. In short, the disadvantages of manual review are high labor costs and poor user experience. Summary of the invention

[0005] In view of this, the present disclosure provides a service order review method and system that can save labor costs to a certain extent and improve user experience.

[0006] One aspect of the present disclosure provides a service order review method, including: obtaining service order data of a service order to be reviewed; obtaining an automatic order review rule set; matching the service order data of the service order to be reviewed with the rules in the automatic order review rule set; and if the match is successful, inputting the service order data of the service order to be reviewed into the automatic order review process for review.

[0007] According to an embodiment of the present disclosure, before obtaining the automatic order review rule set, the method also includes generating the automatic order review rule set, which operation includes: obtaining original samples, wherein the original samples include positive samples and negative samples, the positive samples are samples of service orders that have passed the review, and the negative samples are samples of service orders that have been reviewed but failed; using the original samples to train a rule decision tree; after the training is completed, filtering out internal nodes whose positive sample purity values ​​meet preset conditions from the rule decision tree; and determining the order review rules corresponding to the internal nodes whose positive sample purity values ​​meet the preset conditions as automatic order review rules and generating the corresponding automatic order review rule set.

[0008] According to an embodiment of the present disclosure, after the training is completed, the internal nodes whose positive sample purity values ​​meet the preset conditions are screened out from the above-mentioned rule decision tree, including: after the training is completed, counting the total number of samples and the total number of positive samples entering each internal node in the above-mentioned decision tree; calculating the positive sample purity value of each internal node by the ratio of the total number of positive samples entering each node to the total number of samples; screening out the internal nodes whose positive sample purity values ​​are greater than or equal to the cost balance value from the above-mentioned rule decision tree as the internal nodes that meet the above-mentioned preset conditions, wherein the above-mentioned cost balance value represents the balance relationship between the average cost loss brought about by the automatic review of each service order and the average manual review cost.

[0009] According to an embodiment of the present disclosure, the above cost balance value is calculated according to the following formula: Where p represents the above cost balance value, C r represents the average manual review cost of a service order, C s represents the average after-sales cost of an object, C a Represents the average object loss cost.

[0010] According to an embodiment of the present disclosure, after filtering out the internal nodes whose positive sample purity values ​​are greater than or equal to the cost balance value from the above-mentioned rule decision tree, the above-mentioned method also includes: finding out the nodes that have inheritance relationships with each other and the nodes that do not have inheritance relationships with each other among the internal nodes whose positive sample purity values ​​are greater than or equal to the cost balance value; for the above-mentioned nodes that have inheritance relationships with each other, discarding the child nodes and retaining the parent nodes; and using the above-mentioned nodes that do not have inheritance relationships with each other and the parent nodes retained among the above-mentioned nodes that have inheritance relationships with each other as internal nodes that meet the above-mentioned preset conditions.

[0011] According to an embodiment of the present disclosure, the above-mentioned obtaining of service order data of the service order to be reviewed includes: for the above-mentioned service order to be reviewed, obtaining a user portrait and / or object portrait associated therewith.

[0012] According to an embodiment of the present disclosure, the above method further includes: in case of a matching failure, inputting the service order data of the above service order to be reviewed into a manual order review process for review.

[0013] Another aspect of the present disclosure provides a service order review system, including: a first acquisition module, used to obtain service order data of a service order to be reviewed; a second acquisition module, used to obtain an automatic order review rule set; a matching module, used to match the service order data of the above-mentioned service order to be reviewed with the rules in the above-mentioned automatic order review rule set; and an automatic review module, used to input the service order data of the above-mentioned service order to be reviewed into the automatic order review process for review when the match is successful.

[0014] According to an embodiment of the present disclosure, the system further includes a generation module for generating the automatic order review rule set before obtaining the automatic order review rule set, and the generation module includes: an acquisition unit for obtaining original samples, wherein the original samples include positive samples and negative samples, the positive samples are samples of service orders that have passed the review, and the negative samples are samples of service orders that have been reviewed but failed; a training unit for training a rule decision tree using the original samples; a first screening unit for screening out internal nodes whose positive sample purity values ​​meet preset conditions from the rule decision tree after the training is completed; and a first determination unit for determining the order review rules corresponding to the internal nodes whose positive sample purity values ​​meet the preset conditions as automatic order review rules and generating a corresponding automatic order review rule set.

[0015] According to an embodiment of the present disclosure, the above-mentioned first screening unit includes: a statistical subunit, which is used to count the total number of samples and the total number of positive samples entering each internal node in the above-mentioned decision tree after the training is completed; a calculation subunit, which is used to calculate the positive sample purity value of each internal node by the ratio of the total number of positive samples entering each node to the total number of samples; a screening subunit, which is used to screen out internal nodes with positive sample purity values ​​greater than or equal to the cost balance value from the above-mentioned rule decision tree as internal nodes that meet the above-mentioned preset conditions, wherein the above-mentioned cost balance value represents the balance relationship between the average cost loss brought about by the automatic review of each service order and the average manual review cost.

[0016] According to an embodiment of the present disclosure, the above cost balance value is calculated according to the following formula: Where p represents the above cost balance value, C r represents the average manual review cost of a service order, C s represents the average after-sales cost of an object, C a Represents the average object loss cost.

[0017] According to an embodiment of the present disclosure, the above-mentioned generation module also includes: a search unit, which is used to find out the nodes that have inheritance relationships with each other and the nodes that do not have inheritance relationships with each other among the internal nodes whose positive sample purity values ​​are greater than or equal to the cost balance value after filtering out the internal nodes whose positive sample purity values ​​are greater than or equal to the cost balance value from the above-mentioned rule decision tree; a second screening unit, which is used to discard the child nodes and retain the parent nodes for the above-mentioned nodes that have inheritance relationships with each other; and a second determination unit, which is used to use the above-mentioned nodes that do not have inheritance relationships with each other and the parent nodes retained among the above-mentioned nodes that have inheritance relationships with each other as internal nodes that meet the above-mentioned preset conditions.

[0018] According to an embodiment of the present disclosure, the first acquisition module is further used to: acquire, for the service order to be reviewed, a user portrait and / or an object portrait associated therewith.

[0019] According to an embodiment of the present disclosure, the system further includes: a manual review module, which is used to input the service order data of the service order to be reviewed into a manual review process for review when a match fails.

[0020] Another aspect of the present disclosure provides a computer system, comprising: one or more processors; and a memory for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the above-mentioned method.

[0021] Another aspect of the present disclosure provides a computer-readable storage medium storing computer-executable instructions, which are used to implement the above method when executed.

[0022] Another aspect of the present disclosure provides a computer program, which includes computer executable instructions, and the instructions are used to implement the method described above when executed.

[0023] According to the embodiments of the present disclosure, a technical means is adopted to screen service orders that meet the automatic review conditions and enter the automatic review process for review based on a portrait model and the cost balance principle, and to screen service orders that do not meet the automatic review conditions and enter the manual review process for review. Therefore, the technical problem that only manual review can be performed in the related technology, resulting in high labor costs and poor user experience, is at least partially overcome, thereby achieving the technical effect of saving high labor costs and improving user experience by automatically reviewing some or all service orders. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The above and other objects, features and advantages of the present disclosure will become more apparent through the following description of the embodiments of the present disclosure with reference to the accompanying drawings, in which:

[0025] Figure 1 An exemplary system architecture to which the service order review method and system disclosed in the present invention can be applied is schematically shown;

[0026] Figure 2 A flowchart of a service order review method according to an embodiment of the present disclosure is schematically shown;

[0027] Figure 3 A schematic diagram schematically shows a portrait model according to an embodiment of the present disclosure;

[0028] Figure 4 A flowchart for generating an automatic order review rule set according to an embodiment of the present disclosure is schematically shown;

[0029] Figure 5 A flowchart of an inner node in a screening rule decision tree according to an embodiment of the present disclosure is schematically shown;

[0030] Figure 6 and Figure 7 A schematic diagram of generating an automatic order review rule set according to an embodiment of the present disclosure is schematically shown;

[0031] Figure 8 A block diagram of a service order review system according to an embodiment of the present disclosure is schematically shown;

[0032] Fig. 9 A block diagram of a service order review system according to another embodiment of the present disclosure is schematically shown;

[0033] Fig.10 A block diagram schematically shows a first screening unit according to an embodiment of the present disclosure; and

[0034] Fig.11 A block diagram of a computer system suitable for implementing a service order review method and system according to an embodiment of the present disclosure is schematically shown. DETAILED DESCRIPTION

[0035] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the present disclosure. In the following detailed description, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of the embodiments of the present disclosure. However, it is apparent that one or more embodiments may also be implemented without these specific details. In addition, in the following description, descriptions of known structures and technologies are omitted to avoid unnecessary confusion of the concepts of the present disclosure.

[0036] The terms used herein are only for describing specific embodiments and are not intended to limit the present disclosure. The terms "comprise", "include", etc. used herein indicate the existence of the features, steps, operations and / or components, but do not exclude the existence or addition of one or more other features, steps, operations or components.

[0037] All terms (including technical and scientific terms) used herein have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification, and should not be interpreted in an idealized or overly rigid manner.

[0038] In the case of using expressions such as "at least one of A, B, and C, etc.", it should generally be interpreted in accordance with the meaning of the expression generally understood by those skilled in the art (for example, "a system having at least one of A, B, and C" should include but not be limited to a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, C, etc.). In the case of using expressions such as "at least one of A, B, or C, etc.", it should generally be interpreted in accordance with the meaning of the expression generally understood by those skilled in the art (for example, "a system having at least one of A, B, or C" should include but not be limited to a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, C, etc.). Those skilled in the art should also understand that any transitional conjunctions and / or phrases that substantially indicate two or more optional items, whether in the specification, claims, or drawings, should be understood to give the possibility of including one of these items, either of these items, or both of these items. For example, the phrase "A or B" should be understood to include the possibilities of "A" or "B", or "A and B".

[0039] The embodiments of the present disclosure provide a service order review method that can save labor costs and improve user experience to a certain extent, and a service order review system that can implement the method. The method includes obtaining service order data of the service order to be reviewed; obtaining an automatic order review rule set; matching the service order data of the service order to be reviewed with the rules in the automatic order review rule set; and if the match is successful, inputting the service order data of the service order to be reviewed into the automatic order review process for review.

[0040] Figure 1 The exemplary system architecture to which the service ticket review method and system disclosed in the present invention can be applied is schematically shown. It should be noted that: Figure 1What is shown is merely an example of a system architecture to which the embodiments of the present disclosure can be applied, in order to help those skilled in the art understand the technical content of the present disclosure, but it does not mean that the embodiments of the present disclosure cannot be used in other devices, systems, environments or scenarios.

[0041] like Figure 1 As shown, the system architecture 100 according to this embodiment may include terminal devices 101, 102, 103, a network 104 and a server 105. The network 104 is used to provide a medium for communication links between the terminal devices 101, 102, 103 and the server 105. The network 104 may include various connection types, such as wired and / or wireless communication links, etc.

[0042] Users can use terminal devices 101, 102, 103 to interact with server 105 through network 104 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 101, 102, 103, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients and / or social platform software, etc. (only as examples).

[0043] The terminal devices 101 , 102 , and 103 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, and desktop computers.

[0044] The server 105 may be a server that provides various services, such as a background management server (only an example) that provides support for websites browsed by users using the terminal devices 101, 102, and 103. The background management server may analyze and process the received data such as user requests, and feed back the processing results (such as web pages, information, or data obtained or generated according to user requests) to the terminal device.

[0045] It should be noted that the service order review method provided in the embodiment of the present disclosure can generally be executed by the server 105. Accordingly, the service order review system provided in the embodiment of the present disclosure can generally be set in the server 105. The service order review method provided in the embodiment of the present disclosure can also be executed by a server or server cluster that is different from the server 105 and can communicate with the terminal devices 101, 102, 103 and / or the server 105. Correspondingly, the service order review system provided in the embodiment of the present disclosure can also be set in a server or server cluster that is different from the server 105 and can communicate with the terminal devices 101, 102, 103 and / or the server 105. Alternatively, the service order review method provided in the embodiment of the present disclosure can also be executed by the terminal devices 101, 102, or 103, or can also be executed by other terminal devices different from the terminal devices 101, 102, or 103. Correspondingly, the service order review system provided by the embodiment of the present disclosure may also be set in the terminal device 101, 102, or 103, or in other terminal devices different from the terminal device 101, 102, or 103.

[0046] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to the implementation requirements.

[0047] In addition, the service order review method and system provided by the present invention can be applied to after-sales service scenarios of goods of Internet trading entities and non-Internet trading entities. Specifically, it can be used to review exchange application forms, return application forms, repair application forms and installation application forms, etc., without limitation here.

[0048] It should be noted that the above scenarios are only examples of application scenarios in which the embodiments of the present disclosure can be applied, so as to help those skilled in the art understand the technical content of the present disclosure, but it does not mean that the embodiments of the present disclosure cannot be used in other environments or scenarios.

[0049] Figure 2 The flowchart of the service order review method according to the embodiment of the present disclosure is schematically shown.

[0050] like Figure 2 As shown, the method includes operations S210 to S240.

[0051] In operation S210, service order data of the service order to be reviewed is obtained.

[0052] Specifically, the service order to be reviewed is an order submitted by a consumer, ie, a user, for one or several after-sales services. The service order data includes, but is not limited to, user information, object information (ie, product information), and service order attributes.

[0053] In operation S220, an automatic order review rule set is obtained.

[0054] The automatic order review rule set includes one or more automatic order review rules, which can be rules about one or more of the following information: user information, product information, service order attributes, etc.

[0055] In operation S230, the service order data of the service order to be reviewed is matched with the rules in the automatic order review rule set.

[0056] In operation S240, when the matching is successful, the service order data of the service order to be reviewed is input into the automatic order review process for review.

[0057] When matching the service order data of the service order to be reviewed with the rules in the automatic order review rule set, one or more automatic order review rules may be successfully matched. In the present disclosed embodiment, as long as there are matching automatic order review rules, the service order to be reviewed can be considered to meet the automatic order review conditions, and the corresponding service order data can be input into the automatic order review process for review.

[0058] Further, as an optional embodiment, the method may also include: in the event of a matching failure, inputting the service order data of the service order to be reviewed into a manual order review process for review.

[0059] Furthermore, in order to make the service order review method provided by the present disclosure more complete, before matching the rules in the automatic order review rule set, the method may also perform the following processing:

[0060] First, determine whether the service order to be reviewed is an unconventional service order. The judgment is based on the attributes of the service order. The judgment logic is that unconventional service orders are pushed to the manual review process for review, and regular service orders need to be judged in the next step. Specific implementation examples include but are not limited to this: For example, the attribute description of a service order is that its service content is to apply for a repair of a certain item purchased by the customer. This is the second time to submit a repair application for this item, which means that the previous after-sales service did not solve the customer's problem. In order not to affect customer satisfaction, the review of the service order needs to be handled carefully, so it needs to be pushed to the manual review process for review.

[0061] Secondly, for regular service orders, it is necessary to further determine whether the account submitting the service order is a risky user. The judgment is based on the identity label in the user portrait. The judgment logic is that the label shows that the account is a risky user. In order to prevent huge losses, the review of the service order also needs to be handled carefully. Therefore, it also needs to be pushed to the manual review process for review. Non-risk users need to make the next judgment. Specific implementation examples include but are not limited to this: For example, if the identity label of the account applying for the service order shows that it is a vest account, an after-sales blacklist user, etc., then the service order can be pushed to the manual review process for review, otherwise the next judgment is made.

[0062] For non-risk users, it is also necessary to determine whether the account submitting the service ticket is an important customer. The judgment is based on the identity label in the user portrait. The judgment logic is that if the label shows that the account is an important customer, the service ticket will be pushed to the automatic review process for review, otherwise the next step of judgment is required. Specific implementation examples include but are not limited to this: For example, if the identity label of the application account of the service ticket shows that it is a high-value or high-credit user, for the sake of prudence, the service ticket will be pushed to the manual review process for review, otherwise the next step of judgment is required.

[0063] For non-important customers, it is also necessary to determine whether the service order meets the relevant rules in the automatic order review rule set. The judgment is based on the user portrait, product portrait and service order attributes. The judgment logic is that if the submitted service order meets one of the rules in the automatic order review rules, the service order will be pushed to the automatic review process for review, otherwise it will be pushed to the manual review process for review. Specific implementation examples include but are not limited to this: For example, there is a rule as follows: user portrait [age (30-40), education (master's degree or above), account level (diamond), after-sales credit score (>90), user value (>80)], product portrait [repair rate (>0.1), number of complaints (>10)], service order attribute [type (repair)], only when the above requirements are met at the same time, it is considered to meet this rule, among which the user portrait is the user portrait corresponding to the account that submitted the service order, and the product portrait is the product portrait mentioned (or bound) in the service order that needs to be returned / repaired. The above rules are only for example purposes and are not specific rules.

[0064] It should be noted that the rules for distinguishing non-routine service orders, risky users, and important customers can be manually specified based on order review specifications and business requirements.

[0065] In addition, in the embodiments of the present disclosure, the processing results of the manual review process include multiple processing results such as pass, fail, etc., which can be specifically determined by the manual review specifications, and the processing results of the automatic review process are all pass.

[0066] Compared with the prior art which usually involves manual review, resulting in high labor costs and poor consumer experience, the technical solution provided in the embodiments of the present invention screens out orders that meet the automatic review rules for automatic review, and conducts manual review on orders that do not meet the automatic review rules, thereby saving labor costs and improving consumer experience.

[0067] Reference below Figure 3 to Figure 7 , combined with specific embodiments Figure 2 The method shown is further explained.

[0068] As an optional embodiment, obtaining service order data of the service order to be reviewed includes: obtaining, for the service order to be reviewed, a user portrait and / or an object portrait associated therewith.

[0069] like Figure 3 As shown in the figure, the portrait system (also known as the portrait model) has a three-layer structure of portrait layer, module layer and label layer, which specifically includes user portrait and product portrait. Among them, the user portrait specifically includes the basic model, credit model, value model and identity model, each model has its own label set, and the label set consists of a series of related labels; the product portrait specifically includes the product model, category model and brand model, each model has its own label set, and the label set consists of a series of related labels.

[0070] That is, user portraits are divided into four categories: user basic information, user credit information, user value information, and user identity information, which respectively include tags describing user basic information, tags describing user credit information, tags describing user value information, and tags describing user identity information; product portraits include tags describing basic product information and tags related to after-sales service. The modeling method and data used for each tag vary depending on the meaning of the tag.

[0071] Specifically, in the process of obtaining the service order data of the service order to be reviewed, the corresponding portrait model can be associated based on the service order, such as associating the user portrait of the service order and the relevant label attributes of the product portrait through the account number of the service order and the product number bound to the service order, to obtain a service order with extended attributes.

[0072] In order to solve the problems of high labor costs and poor user experience in manual order review, the present disclosed embodiment proposes a service order review method based on a portrait model. The method uses user portraits, product portraits, service order attributes and a complete set of judgment rules to filter service orders layer by layer, screen out some service orders for automatic review, so as to reduce labor costs and increase user experience.

[0073] As an optional embodiment, in addition to operations S210 to S240, the method may also include an operation of generating an automatic order review rule set before obtaining the automatic order review rule set in operation S220, such as Figure 4 As shown, the operation may include operations S310 to S340. For the sake of ease of description, the description of operations S210 to S240 is omitted here. Among them:

[0074] Operation S410: obtaining original samples, wherein the original samples include positive samples and negative samples, the positive samples are samples of service orders that have passed the review, and the negative samples are samples of service orders that have been reviewed but not passed.

[0075] In addition, the original sample also includes the portrait system and its label information.

[0076] Operation S420, training a rule decision tree using the original samples;

[0077] Operation S430, after the training is completed, selecting internal nodes whose purity values ​​of positive samples meet preset conditions from the rule decision tree; and

[0078] In operation S440, the order review rules corresponding to the internal nodes whose purity values ​​of the positive samples meet the preset conditions are determined as automatic order review rules and a corresponding automatic order review rule set is generated.

[0079] It should be noted that the preset conditions can be set according to user needs and are not limited here.

[0080] Since each internal node in the decision tree corresponds to a review rule, only the rules corresponding to all retained internal nodes (i.e., internal nodes whose positive sample purity values ​​meet the preset conditions) can be extracted to form an automatic review rule set R.

[0081] Through the embodiments of the present disclosure, automatic review rules are generated based on the labels of the portrait system, which requires judging whether all possible rules meet the preset conditions. Moreover, since there are many labels in the portrait system, traversing all possible rules requires a large amount of computing resources, which is not feasible in engineering. The present disclosure adopts a decision tree model to extract rules, which has lower algorithm complexity and fast calculation speed, and is an atypical usage of the decision tree model.

[0082] Further, as an optional embodiment, Figure 5 As shown, operation S330, i.e., after the training is completed, filtering out the inner nodes whose positive sample purity values ​​meet the preset conditions from the rule decision tree may include:

[0083] Operation S510, after the training is completed, counting the total number of samples entering each internal node in the decision tree and the total number of positive samples;

[0084] Operation S520, calculating the positive sample purity value of each internal node by the ratio of the total number of positive samples entering each node to the total number of samples;

[0085] Operation S530, filtering out internal nodes whose positive sample purity values ​​are greater than or equal to the cost balance value from the rule decision tree as internal nodes that meet preset conditions, wherein the cost balance value represents the balance relationship between the average cost loss caused by the automatic review of each service order and the average manual review cost.

[0086] Specifically, the automatic order review rule is not a single rule, but a rule set R composed of multiple rules, and each element in the set R represents a rule.

[0087] For example, the rules in the automatic order review rule set R need to be determined through the following steps:

[0088] Step 1: Construct an extended attribute sample set. Use relevant historical data (i.e., relevant data of reviewed historical service orders) as the original sample D. Use the profiling system to extend the attributes of the original sample to obtain an extended attribute sample set De. Samples labeled "passed" in the original sample D are defined as positive samples, and samples labeled "failed" are defined as negative samples.

[0089] Step 2: Establish a rule tree T (also called a decision tree). Train the decision tree based on the extended attribute sample set. The variable selection and splitting algorithm of the decision tree adopts C4.5 or CART algorithm until it cannot be split.

[0090] Step 3, such as Figure 6 As shown, the inner nodes of the rule tree T are initially screened. The screened rule tree is marked as T1. The screening rule is to calculate the purity Q of the positive samples on each inner node of the rule tree T. The calculation formula is:

[0091]

[0092] Among them, M is the total number of samples entering the node, m is the number of positive samples entering the node, and then the internal nodes with positive sample purity Q≥p are retained, and the internal nodes with positive sample purity Q<p are discarded, where p is the cost balance value.

[0093] Through the embodiments of the present disclosure, based on the extended attributes such as pass and fail of the reviewed historical service orders, as well as the user portraits, product portraits and service order attributes contained in these service orders themselves, a rule decision tree is trained, and based on the cost balance model, the internal nodes whose positive sample purity values ​​meet the cost balance value are retained to generate an automatic order review rule set. This can save labor costs, improve consumer experience, and minimize the cost losses of merchants.

[0094] Further, as an optional embodiment, specifically, the cost balance model can be used to calculate the cost balance value, and the calculation formula is as follows: Among them, p represents the cost balance value, C r represents the average manual review cost of after-sales service orders, C s represents the after-sales cost of an object (such as a product) per order, C a It represents the average loss cost of an object (such as a product) per order, which is defined as the ratio of the total price loss of the product to the total order quantity. The calculation formula is as follows:

[0095]

[0096] Among them, d i is the ith record in the relevant historical data set, N is the number of all records in the set, L is the commodity loss function, and commodity loss is the loss of commodity price due to the after-sales service of the commodity. For example, if the commodity is scrapped after return, the commodity loss is the commodity price. If it is resold after return, the commodity price loss is the original price minus the resale price.

[0097] Through the disclosed embodiment, a method for calculating the cost balance value using a cost balance model is adopted, and its technical principle is: since there is only one processing result of automatic order review (i.e. "pass"), there is a possibility of misjudgment when using rule judgment to perform automatic order review. Although automatic order review can save manual review costs, if a misjudgment occurs, it will increase product losses and after-sales service costs. In order to resolve this contradiction, in the disclosed embodiment, the selection of rules is determined based on the cost balance principle. In the rule generation step, when the sample purity Q is greater than the cost balance point p, in a statistical sense, the manual review cost saved by this rule will be greater than the sum of the product losses and after-sales service costs caused by its misjudgment, so this rule will reduce the total cost.

[0098] As an optional embodiment, after filtering out the internal nodes whose positive sample purity values ​​are greater than or equal to the cost balance value from the rule decision tree in operation S530, the method may further include: finding out the nodes that have an inheritance relationship with each other and the nodes that do not have an inheritance relationship with each other among the internal nodes whose positive sample purity values ​​are greater than or equal to the cost balance value; for the nodes that have an inheritance relationship with each other, discarding the child nodes and retaining the parent nodes; and using the nodes that do not have an inheritance relationship with each other and the parent nodes retained among the nodes that have an inheritance relationship with each other as internal nodes that meet preset conditions.

[0099] Based on the examples in the above embodiments, Figure 7As shown, on the basis of rule tree T1, the internal nodes are further screened, that is, the rule tree T is screened twice. The screened rule tree is marked as T2. The screening rule is to find out the nodes that have inheritance relationship with each other among all the nodes of T1. For the nodes that have inheritance relationship with each other, only the parent node is retained and the child node is discarded. This can filter out repeated and redundant rules, simplify the rule matching logic, and improve the rule matching efficiency.

[0100] Figure 8 A block diagram of a service ticket review system according to an embodiment of the present disclosure is schematically shown.

[0101] like Figure 8 As shown, the system 800 includes a first acquisition module 810 , a second acquisition module 820 , a matching module 830 and an automatic review module 840 .

[0102] A first acquisition module 810 is used to acquire service order data of a service order to be reviewed;

[0103] The second acquisition module 820 is used to acquire the automatic order review rule set;

[0104] A matching module 830 is used to match the service order data of the service order to be reviewed with the rules in the automatic order review rule set; and

[0105] The automatic review module 840 is used to input the service order data of the service order to be reviewed into the automatic review process for review when the match is successful.

[0106] As an optional embodiment, the system further includes: a manual review module, which is used to input the service order data of the service order to be reviewed into a manual review process for review when the matching fails.

[0107] Compared with the prior art which usually involves manual review, resulting in high labor costs and poor consumer experience, the technical solution provided in the embodiments of the present invention screens out orders that meet the automatic review rules for automatic review, and conducts manual review on orders that do not meet the automatic review rules, thereby saving labor costs and improving consumer experience.

[0108] As an optional embodiment, the first acquisition module is further used to: acquire the user portrait and / or object portrait associated with the service order to be reviewed.

[0109] In order to solve the problems of high labor costs and poor user experience in manual order review, the present disclosed embodiment proposes a service order review method based on a portrait model. The method uses user portraits, product portraits, service order attributes and a complete set of judgment rules to filter service orders layer by layer, screen out some service orders for automatic review, so as to reduce labor costs and increase user experience.

[0110] As an optional embodiment, Fig. 9 As shown, the system may further include a generation module 910, which is used to generate the automatic order review rule set before obtaining the automatic order review rule set, and the generation module 910 includes: an acquisition unit 911, which is used to obtain original samples, wherein the original samples include positive samples and negative samples, the positive samples are samples of service orders that have passed the review, and the negative samples are samples of service orders that have been reviewed but not passed; a training unit 912, which is used to train a rule decision tree using the original samples; a first screening unit 913, which is used to screen out internal nodes whose positive sample purity values ​​meet preset conditions from the rule decision tree after the training is completed; and a first determination unit 914, which is used to determine the order review rules corresponding to the internal nodes whose positive sample purity values ​​meet the preset conditions as automatic order review rules and generate a corresponding automatic order review rule set.

[0111] Through the embodiments of the present disclosure, automatic review rules are generated based on the labels of the portrait system, which requires judging whether all possible rules meet the preset conditions. Moreover, since there are many labels in the portrait system, traversing all possible rules requires a large amount of computing resources, which is not feasible in engineering. The present disclosure adopts a decision tree model to extract rules, which has lower algorithm complexity and fast calculation speed, and is an atypical usage of the decision tree model.

[0112] As an optional embodiment, Fig.10 As shown, the first screening unit 913 includes: a statistical subunit 9131, which is used to count the total number of samples and the total number of positive samples entering each internal node in the decision tree after the training is completed; a calculation subunit 9132, which is used to calculate the positive sample purity value of each internal node by the ratio of the total number of positive samples entering each node to the total number of samples; a screening subunit 9133, which is used to screen out the internal nodes whose positive sample purity values ​​are greater than or equal to the cost balance value from the rule decision tree as the internal nodes that meet the preset conditions, wherein the cost balance value represents the balance relationship between the average cost loss caused by the automatic review of each service order and the average manual review cost.

[0113] Through the embodiments of the present disclosure, based on the extended attributes such as pass and fail of the reviewed historical service orders, as well as the user portraits, product portraits and service order attributes contained in these service orders themselves, a rule decision tree is trained, and based on the cost balance model, the internal nodes whose positive sample purity values ​​meet the cost balance value are retained to generate an automatic order review rule set. This can save labor costs, improve consumer experience, and minimize the cost losses of merchants.

[0114] As an optional embodiment, the cost balance value is calculated according to the following formula: Where p represents the above cost balance value, C r represents the average manual review cost of a service order, C s represents the average after-sales cost of an object, C a Represents the average object loss cost.

[0115] Through the disclosed embodiment, a method for calculating the cost balance value using a cost balance model is adopted, and its technical principle is: since there is only one processing result of automatic order review (i.e. "pass"), there is a possibility of misjudgment when using rule judgment to perform automatic order review. Although automatic order review can save manual review costs, if a misjudgment occurs, it will increase product losses and after-sales service costs. In order to resolve this contradiction, in the disclosed embodiment, the selection of rules is determined based on the cost balance principle. In the rule generation step, when the sample purity Q is greater than the cost balance point p, in a statistical sense, the manual review cost saved by this rule will be greater than the sum of the product losses and after-sales service costs caused by its misjudgment, so this rule will reduce the total cost.

[0116] As an optional embodiment, the above-mentioned generation module also includes: a search unit, which is used to find out the nodes that have inheritance relationship with each other and the nodes that do not have inheritance relationship with each other among the internal nodes whose positive sample purity value is greater than or equal to the cost balance value after filtering out the internal nodes whose positive sample purity value is greater than or equal to the cost balance value from the above-mentioned rule decision tree; a second screening unit, which is used to discard the child nodes and retain the parent nodes of the above-mentioned nodes that have inheritance relationship with each other; and a second determination unit, which is used to use the above-mentioned nodes that do not have inheritance relationship with each other and the parent nodes retained among the above-mentioned nodes that have inheritance relationship with each other as internal nodes that meet the above-mentioned preset conditions.

[0117] Through the embodiments of the present disclosure, repeated and redundant rules can be filtered out, rule matching logic can be simplified, and rule matching efficiency can be improved.

[0118] According to the modules, units, and sub-units of the embodiments of the present disclosure, any multiple or at least part of the functions of any multiple thereof can be implemented in one module. According to the modules, units, and sub-units of the embodiments of the present disclosure, any one or more thereof can be split into multiple modules for implementation. According to the modules, units, and sub-units of the embodiments of the present disclosure, any one or more thereof can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application-specific integrated circuit (ASIC), or can be implemented by hardware or firmware of any other reasonable way of integrating or packaging the circuit, or implemented in any one of the three implementation modes of software, hardware, and firmware or in any appropriate combination of any of them. Alternatively, according to the modules, units, and sub-units of the embodiments of the present disclosure, one or more thereof can be at least partially implemented as a computer program module, and when the computer program module is run, the corresponding function can be performed.

[0119] For example, any multiple of the first acquisition module 810, the second acquisition module 820, the matching module 830 and the automatic audit module 840 can be combined in one module / unit / sub-unit for implementation, or any one of the modules / units / sub-units can be split into multiple modules / units / sub-units. Alternatively, at least part of the functions of one or more of these modules / units / sub-units can be combined with at least part of the functions of other modules / units / sub-units and implemented in one module / unit / sub-unit. As an optional embodiment, at least one of the first acquisition module 810, the second acquisition module 820, the matching module 830 and the automatic audit module 840 can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application specific integrated circuit (ASIC), or can be implemented by hardware or firmware such as any other reasonable way of integrating or packaging the circuit, or implemented in any one of the three implementation methods of software, hardware and firmware or in a suitable combination of any of them. Alternatively, at least one of the first acquisition module 810, the second acquisition module 820, the matching module 830 and the automatic review module 840 may be at least partially implemented as a computer program module, and when the computer program module is executed, the corresponding function may be executed.

[0120] It should be noted that the system part in the embodiments of the present disclosure corresponds to the method part in the embodiments of the present disclosure. The description of the system part specifically refers to the method part and will not be repeated here.

[0121] Fig.11A block diagram of a computer system suitable for implementing a service order review method and system according to an embodiment of the present disclosure is schematically shown. Fig.11 The computer system shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0122] like Fig.11 As shown, the computer system 1100 according to an embodiment of the present disclosure includes a processor 1101, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1102 or a program loaded from a storage part 1108 to a random access memory (RAM) 1103. The processor 1101 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or a related chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 1101 may also include an onboard memory for caching purposes. The processor 1101 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present disclosure.

[0123] In RAM 1103, various programs and data required for the operation of system 1100 are stored. Processor 1101, ROM 1102 and RAM 1103 are connected to each other through bus 1104. Processor 1101 performs various operations of the method flow according to the embodiment of the present disclosure by executing the program in ROM 1102 and / or RAM 1103. It should be noted that the program can also be stored in one or more memories other than ROM 1102 and RAM 1103. Processor 1101 can also perform various operations of the method flow according to the embodiment of the present disclosure by executing the program stored in the one or more memories.

[0124] According to an embodiment of the present disclosure, the system 1100 may further include an input / output (I / O) interface 1105, which is also connected to the bus 1104. The system 1100 may further include one or more of the following components connected to the I / O interface 1105: an input section 1106 including a keyboard, a mouse, etc.; an output section 1107 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 1108 including a hard disk, etc.; and a communication section 1109 including a network interface card such as a LAN card, a modem, etc. The communication section 1109 performs communication processing via a network such as the Internet. A drive 1110 is also connected to the I / O interface 1105 as needed. A removable medium 1111, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 1110 as needed, so that a computer program read therefrom is installed into the storage section 1108 as needed.

[0125] According to an embodiment of the present disclosure, the method flow according to an embodiment of the present disclosure can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a computer-readable storage medium, and the computer program contains a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 1109, and / or installed from the removable medium 1111. When the computer program is executed by the processor 1101, the above-mentioned functions defined in the system of the embodiment of the present disclosure are executed. According to an embodiment of the present disclosure, the system, equipment, device, module, unit, etc. described above can be implemented by a computer program module.

[0126] The present disclosure also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or may exist independently without being assembled into the device / apparatus / system. The above computer-readable storage medium carries one or more programs, and when the above one or more programs are executed, the method according to the embodiment of the present disclosure is implemented.

[0127] According to an embodiment of the present disclosure, the computer-readable storage medium may be a non-volatile computer-readable storage medium. For example, it may include, but is not limited to: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, apparatus, or device.

[0128] For example, according to an embodiment of the present disclosure, the computer-readable storage medium may include the ROM 1102 and / or the RAM 1103 described above and / or one or more memories other than the ROM 1102 and the RAM 1103 .

[0129] The flow charts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each box in the flow chart or block diagram can represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flow chart, and the combination of the boxes in the block diagram or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0130] It will be appreciated by those skilled in the art that the features described in the various embodiments and / or claims of the present disclosure may be combined and / or combined in a variety of ways, even if such combinations and / or combinations are not explicitly described in the present disclosure. In particular, the features described in the various embodiments and / or claims of the present disclosure may be combined and / or combined in a variety of ways without departing from the spirit and teachings of the present disclosure. All of these combinations and / or combinations fall within the scope of the present disclosure.

[0131] The embodiments of the present disclosure are described above. However, these embodiments are only for illustrative purposes and are not intended to limit the scope of the present disclosure. Although the embodiments are described above separately, this does not mean that the measures in the various embodiments cannot be used in combination to advantage. The scope of the present disclosure is defined by the attached claims and their equivalents. Without departing from the scope of the present disclosure, those skilled in the art may make a variety of substitutions and modifications, which should all fall within the scope of the present disclosure.

Claims

1. A service order review method, comprising: Get the service order data of the service order to be reviewed; Get the automatic order review rule set; Matching the service order data of the service order to be reviewed with the rules in the automatic order review rule set; as well as If the match is successful, the service order data of the service order to be reviewed is input into the automatic review process for review; The method for generating the automatic document review rule set includes: Based on the extended attributes of the reviewed historical service orders, as well as the user portraits, product portraits, and service order attributes contained in the historical service orders, a rule decision tree is trained, wherein the rule training tree is trained using original samples, the original samples include positive samples and negative samples, the positive samples are samples of service orders that have passed the review, the negative samples are samples of service orders that have been reviewed but not passed, and the extended attributes include passed and failed; Based on the cost balance model, the inner nodes whose positive sample purity values ​​satisfy the cost balance value in the rule decision tree are screened, wherein the positive sample purity value is determined by calculating the ratio of the total number of positive samples entering each inner node in the decision tree to the total number of samples entering each inner node in the decision tree, and the cost balance value represents the balance relationship between the average cost loss caused by the automatic review of each service order and the average manual review cost; and The automatic order review rule set is generated based on the internal nodes whose positive sample purity values ​​satisfy the cost balance value.

2. The method according to claim 1, wherein: The training rule decision tree based on the extended attributes of the audited historical service orders, as well as the user portrait, product portrait and service order attributes contained in the historical service orders includes: obtaining the original samples; and training the rule decision tree using the original samples; The step of screening the inner nodes whose positive sample purity values ​​satisfy the cost balance value in the rule decision tree based on the cost balance model comprises: after the training is completed, screening the inner nodes whose positive sample purity values ​​satisfy the preset conditions from the rule decision tree; The generating of the automatic order review rule set based on the inner nodes whose positive sample purity values ​​satisfy the cost balance value includes: determining the order review rules corresponding to the inner nodes whose positive sample purity values ​​satisfy the preset conditions as the automatic order review rules and generating the corresponding automatic order review rule set.

3. The method according to claim 2, wherein: After the training is completed, the inner nodes whose purity values ​​of positive samples meet the preset conditions are selected from the rule decision tree, including: After the training is completed, the total number of samples entering each internal node in the decision tree and the total number of positive samples are counted; The purity value of positive samples of each internal node is calculated by the ratio of the total number of positive samples entering each node to the total number of samples; Internal nodes whose positive sample purity values ​​are greater than or equal to the cost balance value are selected from the rule decision tree as internal nodes that meet the preset conditions, wherein the cost balance value represents the balance relationship between the average cost loss caused by the automatic review of each service order and the average manual review cost.

4. The method according to claim 3, wherein: The cost balance value is calculated according to the following formula: ， in, represents the cost balance value, Indicates the average manual review cost of service orders. Indicates the average after-sales cost of an object. Represents the average object loss cost.

5. The method according to claim 3, wherein: After selecting the inner nodes whose positive sample purity values ​​are greater than or equal to the cost balance value from the rule decision tree, the method further includes: Find out nodes that have an inheritance relationship with each other and nodes that do not have an inheritance relationship with each other among the internal nodes whose purity value of the positive sample is greater than or equal to the cost balance value; For the nodes that have an inheritance relationship with each other, discard the child nodes and retain the parent nodes; and The nodes that do not have an inheritance relationship with each other and the parent nodes that are retained among the nodes that have an inheritance relationship with each other are used as internal nodes that meet the preset conditions.

6. The method according to claim 1, wherein: The step of obtaining service order data of the service order to be reviewed includes: For the service order to be reviewed, obtain the user portrait and / or object portrait associated therewith.

7. The method according to claim 1, wherein: The method further comprises: In the event of a matching failure, the service order data of the service order to be reviewed is input into a manual review process for review.

8. A service ticket review system, comprising: The first acquisition module is used to acquire the service order data of the service order to be reviewed; The second acquisition module is used to obtain the automatic order review rule set; A matching module, used for matching the service order data of the service order to be reviewed with the rules in the automatic order review rule set; as well as An automatic review module, for inputting the service order data of the service order to be reviewed into an automatic review process for review if the match is successful; The method for generating the automatic document review rule set includes: Based on the extended attributes of the reviewed historical service orders, as well as the user portraits, product portraits, and service order attributes contained in the historical service orders, a rule decision tree is trained, wherein the rule training tree is trained using original samples, the original samples include positive samples and negative samples, the positive samples are samples of service orders that have passed the review, the negative samples are samples of service orders that have been reviewed but not passed, and the extended attributes include passed and failed; Based on the cost balance model, the inner nodes whose positive sample purity values ​​satisfy the cost balance value in the rule decision tree are screened, wherein the positive sample purity value is determined by calculating the ratio of the total number of positive samples entering each inner node in the decision tree to the total number of samples entering each inner node in the decision tree, and the cost balance value represents the balance relationship between the average cost loss caused by the automatic review of each service order and the average manual review cost; and The automatic order review rule set is generated based on the internal nodes whose positive sample purity values ​​satisfy the cost balance value.

9. The system according to claim 8, wherein: The system further includes a generation module, which is used to generate the automatic order examination rule set before acquiring the automatic order examination rule set, and the generation module includes: An acquisition unit, used for acquiring original samples; A training unit, used for training a rule decision tree using the original samples; A first screening unit is used to screen out internal nodes whose positive sample purity values ​​meet preset conditions from the rule decision tree after the training is completed; and The first determination unit is used to determine the order review rules corresponding to the internal nodes whose purity values ​​of the positive samples meet the preset conditions as automatic order review rules and generate a corresponding automatic order review rule set.

10. The system according to claim 9, wherein: The first screening unit comprises: A statistical subunit, used to count the total number of samples entering each internal node in the decision tree and the total number of positive samples after the training is completed; A calculation subunit, used to calculate the purity value of positive samples of each internal node by the ratio of the total number of positive samples entering each node to the total number of samples; A screening sub-unit is used to screen out internal nodes whose positive sample purity values ​​are greater than or equal to the cost balance value from the rule decision tree as internal nodes that meet the preset conditions, wherein the cost balance value represents the balance relationship between the average cost loss caused by the automatic review of each service order and the average manual review cost.

11. The system according to claim 10, wherein: The cost balance value is calculated according to the following formula: ， in, represents the cost balance value, Indicates the average manual review cost of service orders. Indicates the average after-sales cost of an object. Represents the average object loss cost.

12. The system according to claim 10, wherein: The generating module also includes: A search unit, configured to find out nodes that have an inheritance relationship with each other and nodes that do not have an inheritance relationship with each other among the internal nodes whose positive sample purity values ​​are greater than or equal to the cost balance value after filtering out the internal nodes whose positive sample purity values ​​are greater than or equal to the cost balance value from the rule decision tree; A second screening unit is used to discard child nodes and retain parent nodes for the nodes that have an inheritance relationship with each other; and The second determining unit is used to use the nodes that do not have an inheritance relationship with each other and the parent nodes retained among the nodes that have an inheritance relationship with each other as internal nodes that meet the preset conditions.

13. The system according to claim 8, wherein: The first acquisition module is also used for: For the service order to be reviewed, obtain the user portrait and / or object portrait associated therewith.

14. The system according to claim 8, wherein: The system further comprises: The manual review module is used to input the service order data of the service order to be reviewed into the manual review process for review when the matching fails.

15. A computer system comprising: one or more processors; a memory for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 7.

16. A computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, enables the processor to implement the method according to any one of claims 1 to 7.

Images