Rule extraction method and apparatus for work order processing model, device and medium

Abstract

The present application relates to the technical field of communications. Provided are a rule extraction method and apparatus for a work order processing model, a device and a medium. The method comprises: acquiring a work order processing model, the work order processing model comprising: decision trees, each decision tree being used for describing a corresponding rule set, the rule set comprising a plurality of rule nodes, the rule nodes representing logic units for processing work orders, and at least some of the rule nodes forming a rule path (S101); processing the decision trees to obtain a target rule file, the target rule file comprising: at least one rule path (S102); and, according to the at least one rule path, determining target rule information (S103).

Description

Rule extraction method, device, equipment and medium for work order processing model

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims priority to Chinese patent application No. 202310934072.2 filed in China on July 27, 2023, the entire contents of which are incorporated herein by reference. Technical Field

[0003] The present application relates to the field of communication technology, and in particular to a rule extraction method, apparatus, device and medium for a work order processing model. Background Art

[0004] In the field of mobile communications, it may be necessary to compare some complaints or optimization data work orders, manually analyze relevant indicators and parameters, and provide adjustment measures after a series of considerations and subdivisions. The conventional approach is to characterize these data and then use tree models, such as XGBoost or lightGBM, to train a classification model with optimization measures as the goal. The trained model is directly used to recommend optimization measures for new work orders.

[0005] In related technologies, after training the data, the model is directly used for prediction and regression, but less attention is paid to the model itself, resulting in poor interpretability of the model.

[0006] Summary of the Invention

[0007] To this end, the purpose of this application is to propose a rule extraction method, device, computer equipment and storage medium for a work order processing model, which can aggregate the rules of multiple decision trees to improve the support of the rules and the confidence of the prediction results, and improve the interpretability of the model.

[0008] To achieve the above-mentioned objectives, the rule extraction method for a work order processing model proposed in the first embodiment of the present application includes:

[0009] Obtaining a work order processing model, wherein the work order processing model includes: a decision tree, each decision tree is used to describe a corresponding rule set, the rule set includes a plurality of rule nodes, the rule nodes represent logical units for processing work orders, and at least some of the rule nodes form a rule path;

[0010] processing the decision tree to obtain a target rule file, wherein the target rule file includes: at least one rule path; and

[0011] Target rule information is determined according to the at least one rule path.

[0012] To achieve the above-mentioned purpose, the second embodiment of the present application proposes a rule extraction device for a work order processing model, comprising:

[0013] an acquisition module, configured to acquire a work order processing model, wherein the work order processing model includes: a decision tree, each decision tree is used to describe a corresponding rule set, the rule set includes a plurality of rule nodes, each rule node represents a logical unit for processing a work order, and at least some of the rule nodes form a rule path;

[0014] a processing module, configured to process the decision tree to obtain a target rule file, wherein the target rule file includes: at least one rule path; and

[0015] The first determining module is configured to determine target rule information according to the at least one rule path.

[0016] The computer device proposed in the third embodiment of the present application includes: a memory, a processor, and a computer program stored in the memory and runnable on the processor. When the processor executes the program, it implements the rule extraction method for the work order processing model proposed in the first embodiment of the present application.

[0017] The fourth aspect embodiment of the present application proposes a non-temporary computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, it implements the rule extraction method for the work order processing model proposed in the first aspect embodiment of the present application.

[0018] The fifth embodiment of the present application proposes a computer program product. When the instructions in the computer program product are executed by a processor, the rule extraction method for the work order processing model proposed in the first embodiment of the present application is executed.

[0019] The present application provides a rule extraction method, apparatus, computer equipment, and storage medium for a work order processing model, which obtains a work order processing model, wherein the work order processing model includes: a decision tree, each decision tree is used to describe a corresponding rule set, the rule set includes multiple rule nodes, the rule nodes represent logical units for processing work orders, and at least some of the rule nodes form a rule path. The decision tree is processed to obtain a target rule file, wherein the target rule file includes: at least one rule path, and according to at least one rule path, the target rule information is determined. Thus, the rules of multiple decision trees can be aggregated to improve the support of the rules and the confidence of the prediction results, thereby improving the interpretability of the model.

[0020] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0022] FIG1 is a flow chart of a rule extraction method for a work order processing model proposed in one embodiment of the present application;

[0023] FIG2 is a diagram showing a graphical representation of a target rule file proposed in this application;

[0024] FIG3 is a schematic diagram of a graphical representation of another target rule file proposed in this application;

[0025] FIG4 is a schematic diagram of a graphical representation of another target rule file proposed in this application;

[0026] FIG5 is a schematic diagram of a graphical representation of another target rule file proposed in this application;

[0027] FIG6 is a flow chart of a rule extraction method for a work order processing model proposed in another embodiment of the present application;

[0028] FIG7 is a schematic diagram of generating a first rule path according to the present application;

[0029] FIG8 is a flow chart of a rule extraction method for a work order processing model proposed in another embodiment of the present application;

[0030] FIG9 is a schematic diagram of summing up the rule description content according to the present application;

[0031] FIG10 is a schematic diagram of a flow chart of performing intersection processing on rule description contents according to the present application;

[0032] FIG11 is a schematic structural diagram of a rule extraction device for a work order processing model proposed in an embodiment of the present application;

[0033] FIG12 is a schematic structural diagram of a rule extraction device for a work order processing model proposed in another embodiment of the present application;

[0034] FIG13 shows a block diagram of an exemplary computer device suitable for implementing embodiments of the present application. DETAILED DESCRIPTION

[0035] The embodiments of the present application are described in detail below. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended only to explain the present application and are not to be construed as limiting the present application. On the contrary, the embodiments of the present application include all variations, modifications, and equivalents that fall within the spirit and scope of the appended claims.

[0036] FIG1 is a flow chart of a rule extraction method for a work order processing model proposed in an embodiment of the present application.

[0037] Among them, it should be noted that the executor of the rule extraction method for the work order processing model in this embodiment is the rule extraction device for the work order processing model, which can be implemented by software and / or hardware. The device can be configured in a computer device, and the computer device can include but is not limited to a terminal, a server side, etc. For example, the terminal can be a mobile phone, a handheld computer, etc.

[0038] As shown in FIG1 , the rule extraction method for the work order processing model includes:

[0039] S101: Obtain a work order processing model, wherein the work order processing model includes: a decision tree, each decision tree is used to describe a corresponding rule set, the rule set includes multiple rule nodes, the rule node represents a logical unit for processing a work order, and at least some of the rule nodes form a rule path.

[0040] The work order processing model refers to a model used for work order processing, such as a LightGBM model, and there is no restriction on this.

[0041] The rule set refers to a node set consisting of multiple rule nodes.

[0042] For example, in the embodiment of the present application, when obtaining the work order processing model, it can be based on the following steps:

[0043] Massive work orders are processed. After the work orders are processed, the solved problems are marked as positive samples, and the unsolved problems are marked as negative samples. The positive and negative sample columns are used as targets, and the problem is classified. LightGBM is trained on the feature columns to find the best division method, that is, the training model.

[0044] During the training process, LightGBM and catboost can be used to train the sample data separately, and the model performance can be determined based on indicators such as F-score to determine whether it can be used directly.

[0045] S102: Process the decision tree to obtain a target rule file, wherein the target rule file includes at least one rule path.

[0046] The target rule file refers to a file obtained by processing a decision tree and used to indicate rule information, and may be, for example, a file in the gv (Graphviz DOT File) format, without limitation.

[0047] For example, the rule set R described by the decision tree can be expressed as follows: R = {(P1∧P2∧...P n )->A|P i ∈{P},A∈{Action Set}}

[0048] Where A is the set of optimization measures from Action; n represents the number of rule points;

[0049] Pi is a rule point, which is a Boolean expression that satisfies the following conditions; i＝ {(P i .NθP i .V)}|θ∈(=,≠,>,<...),P i ∈(True,False)

[0050] Where Pi.N is the name of the feature column, and Pi.V is the value corresponding to this feature column.

[0051] The number of rule trees contained in a trained work order processing model is indefinite, and each tree corresponds to a rule set R. Using the LGBM library function, these trees can be output one by one as gv (Graphviz DOT File) format files, that is, target rule files.

[0052] In an embodiment of the present application, a work order processing model can output multiple target rule files. For example, as shown in Figures 2, 3, 4, and 5, Figure 2 is a schematic diagram of a graphical representation of a target rule file proposed in the present application, Figure 3 is a schematic diagram of a graphical representation of another target rule file proposed in the present application, Figure 4 is a schematic diagram of a graphical representation of another target rule file proposed in the present application, and Figure 5 is a schematic diagram of a graphical representation of another target rule file proposed in the present application.

[0053] The four GV graphs in Figures 2-5 above are the direct output files from the model calculations. The following is our own code to extract the rules from each GV file. This is because a GV graph is actually a decision tree, and the path from the root node to each leaf node is a rule.

[0054] S103: Determine target rule information according to at least one rule path.

[0055] The target rule information refers to the rule information obtained based on at least one rule path.

[0056] In an embodiment of the present application, when determining the target rule information based on at least one rule path, it can be based on a pre-trained machine learning model, or it can be based on at least one rule path using any other possible method, without limitation.

[0057] In this embodiment, a work order processing model is obtained, wherein the work order processing model includes: a decision tree, each decision tree is used to describe a corresponding rule set, the rule set includes multiple rule nodes, the rule node represents a logical unit for processing a work order, at least some of the rule nodes form a rule path, and the decision tree is processed to obtain a target rule file, wherein the target rule file includes: at least one rule path, and according to at least one rule path, the target rule information is determined, thereby, the rules of multiple decision trees can be aggregated to improve the support of the rules and the confidence of the prediction results, thereby improving the interpretability of the model.

[0058] FIG6 is a flow chart of a rule extraction method for a work order processing model proposed in another embodiment of the present application.

[0059] As shown in FIG6 , the rule extraction method for the work order processing model includes:

[0060] S601: Obtain a work order processing model, wherein the work order processing model includes: a decision tree, each decision tree is used to describe a corresponding rule set, the rule set includes multiple rule nodes, the rule node represents a logical unit for processing a work order, and at least some of the rule nodes form a rule path.

[0061] S602: Process the decision tree to obtain a target rule file, wherein the target rule file includes at least one rule path.

[0062] The description of S601 and S602 can be found in the above embodiments, which will not be repeated here.

[0063] S603: Determine a data processing result corresponding to each rule path, wherein the data processing result has a corresponding sample type.

[0064] The data processing result may refer to the processing result of the work order corresponding to the rule path, for example.

[0065] For example, when the data processing result indicates that the problem involved in the corresponding work order has been resolved, the sample type corresponding to the data processing result is a positive sample; when the data processing result indicates that the problem involved in the corresponding work order has not been resolved, the sample type corresponding to the data processing result is a negative sample.

[0066] S604: Select a first rule path from at least one rule path according to the sample type.

[0067] The first rule path may refer to a path branch in the decision tree that leads to more positive samples.

[0068] Optionally, in some embodiments, when selecting the first rule path from at least one rule path based on the sample type, the rule path corresponding to the data processing result whose sample type is a positive sample type can be used as the first rule path. This can effectively improve the practicality of the obtained first rule path in the rule extraction process.

[0069] For example, as shown in Figure 7, which is a schematic diagram of generating the first rule path proposed in this application, in this embodiment, from a business perspective, only the path branches that lead to more positive samples, such as leaf 1 / 2 / 3 / 4 in Figure 7, which lead to positive samples, are the optimization rules that need to be extracted and can be used as the first rule path.

[0070] S605: Determine target rule information according to the first rule path.

[0071] In an embodiment of the present application, when determining the target rule information according to the first rule path, it can be based on a pre-trained machine learning model, or it can also be based on a third-party rule information determination device, without limitation.

[0072] Optionally, in some embodiments, when determining the target rule information based on the first rule path, the rule nodes included in the first rule path can be determined, the rule description content corresponding to each rule node can be determined, and the target rule information can be determined based on the rule description content of the rule nodes included in the first rule path. Thus, the reliability of the target rule information determination process can be effectively improved based on the rule description content of the rule nodes.

[0073] The rule description content refers to the content specified by the corresponding rule point.

[0074] That is to say, in the embodiment of the present application, after processing the decision tree to obtain the target rule file, the data processing result corresponding to each rule path can be determined, wherein the data processing result has a corresponding sample type, and according to the sample type, the first rule path is selected from at least one rule path, and according to the first rule path, the target rule information is determined. Thus, the first rule path can be accurately and quickly selected from at least one rule path based on the sample type, which can effectively improve the practicality of the obtained target rule information.

[0075] In this embodiment, by determining the data processing result corresponding to each rule path, wherein the data processing result has a corresponding sample type, a first rule path is selected from at least one rule path according to the sample type, and the target rule information is determined according to the first rule path. Thus, the first rule path can be accurately and quickly selected from at least one rule path based on the sample type, which can effectively improve the practicality of the obtained target rule information. By determining the rule nodes included in the first rule path, the rule description content corresponding to each rule node is determined, and the target rule information is determined according to the rule description content of the rule nodes included in the first rule path. Thus, the reliability of the target rule information determination process can be effectively improved based on the rule description content of the rule nodes. The rule path corresponding to the data processing result whose sample type is a positive sample type is used as the first rule path. Thus, the practicality of the obtained first rule path in the rule extraction process can be effectively improved.

[0076] FIG8 is a flow chart of a rule extraction method for a work order processing model proposed in another embodiment of the present application.

[0077] As shown in FIG8 , the rule extraction method for the work order processing model includes:

[0078] S801: Obtain a work order processing model, wherein the work order processing model includes: a decision tree, each decision tree is used to describe a corresponding rule set, the rule set includes multiple rule nodes, the rule node represents a logical unit for processing a work order, and at least some of the rule nodes form a rule path.

[0079] S802: Process the decision tree to obtain a target rule file, wherein the target rule file includes at least one rule path.

[0080] S803: Determine a data processing result corresponding to each rule path, wherein the data processing result has a corresponding sample type.

[0081] S804: Select a first rule path from at least one rule path according to the sample type.

[0082] S805: Determine the rule nodes included in the first rule path.

[0083] The description of S801-S805 can be found in the above embodiment and will not be repeated here.

[0084] S806: Determine the initial description content corresponding to each rule node.

[0085] The initial description content refers to the description content corresponding to the rule node in the initial state.

[0086] In the embodiment of the present application, when the initial description content corresponding to each rule node is determined, reliable reference information can be provided for the subsequent generation of rule description content.

[0087] S807: Determine the content type corresponding to the initial description content.

[0088] The content type may refer to an indication form of the initial description content.

[0089] Optionally, in some embodiments, the content type may be a numerical type, or an enumeration type, etc., which is not limited.

[0090] It is understandable that initial description content of different content types may be suitable for different processing methods. Therefore, in the embodiment of the present application, when determining the content type corresponding to the initial description content, it can provide a reliable selection basis for selecting the corresponding processing method in the subsequent initial description content processing process.

[0091] S808: Process the initial description content according to the content type to obtain the rule description content.

[0092] That is, in the embodiment of the present application, after determining the content type corresponding to the initial description content, the initial description content can be processed according to the content type to obtain the rule description content.

[0093] Optionally, in some embodiments, the content type is a numerical type. When processing the initial description content according to the content type to obtain the rule description content, one or more initial numerical intervals corresponding to at least one numerical type can be parsed from the initial description content. If the number of initial numerical intervals is one, the initial numerical interval is used as the rule description content. If the number of initial numerical intervals is multiple, the intersection of the multiple initial numerical intervals is taken to obtain the target numerical interval corresponding to the corresponding numerical type, and the target numerical interval is used as the rule description content. Thus, when the content type is a numerical type, the target numerical interval can be obtained as the rule description content, thereby effectively improving the indication effect of the rule description content.

[0094] The initial numerical value interval may refer to the numerical value range indicated by the initial description content.

[0095] The target numerical interval refers to the intersection interval of multiple initial numerical intervals.

[0096] Optionally, in some embodiments, the content type is an enumeration type. When processing the initial description content according to the content type to obtain the rule description content, the enumeration content corresponding to at least one enumeration type can be parsed from the initial description content, and the content interval to which the enumeration content belongs can be determined. At least one enumeration content and the content interval to which each enumeration content belongs can be used as the rule description content. In this way, the clarity of the indication of the enumeration content by the obtained rule description content can be effectively improved.

[0097] For example, in an embodiment of the present application, the tree structure in the gv file can be parsed to filter out only the branches predicted to be positive samples (the leaf nodes corresponding to label = yes in FIG7 ). When traversing the nodes, the nodes may be numerical (if it is an interval mode of greater than or less than, the rule will calculate the intersection of the intervals and obtain the merged interval as the fusion rule criterion) or enumeration values ​​(for example, the SCENE scene, the actual scene value is numbered, so here the corresponding scene name is obtained based on these values, which is converted to "SCENE = residential area or industrial park"), the value interval is extracted, and stored as a rule, as shown in Table 1 below. Four useful rules are extracted from the decision tree of FIG7 , named leaf1 to leaf4 respectively:

[0098] Table 1

[0099] That is to say, in the embodiment of the present application, after determining the rule nodes included in the first rule path, the initial description content corresponding to each rule node can be determined, the content type corresponding to the initial description content can be determined, and the initial description content can be processed according to the content type to obtain the rule description content. In this way, the adaptability of the initial description content processing process can be effectively improved based on the content type, thereby effectively improving the accuracy of the obtained rule description content.

[0100] S809: Perform fusion processing on the rule description contents of the rule nodes included in at least one first rule path.

[0101] In an embodiment of the present application, when the rule description content of the rule nodes included in at least one first rule path is fused, it can be based on a pre-trained rule fusion model, or it can be based on a method combining numbers and shapes, which is not limited.

[0102] Optionally, in some embodiments, when performing fusion processing on the rule description contents of the rule nodes included in at least one first rule path, the rule description contents of the rule nodes included in at least one first rule path may be summed up.

[0103] For example, as shown in Figure 9, which illustrates the summation of rule descriptions proposed in this application, assume the model training objective is to classify the circles and rectangles in Figure 9. N1 through N3 in Figure 9 represent how the rules partition the dataset, and the dashed lines represent the partitioning method proposed by the model. After the three trees are gradually partitioned, the classification effect for each region is good, but the overall partitioning effect is lacking. By superimposing the partitioning rules of the three trees, a comprehensive partitioning effect, N4, can be obtained. As shown in Figure 9, the classification effect of N4, obtained after the intersection of multiple rules, is even better.

[0104] Optionally, in some embodiments, when performing fusion processing on the rule description contents of the rule nodes included in at least one first rule path, an intersection processing may also be performed on the rule description contents of the rule nodes included in at least one first rule path.

[0105] For example, as shown in Figure 10, Figure 10 is a flow chart illustrating the intersection processing of rule descriptions proposed in this application. If there are N trees, the fusion method is to apply this cascaded approach to the N trees, calculating the intersection one by one. For example, a rule with ID = leafABCD is obtained, which means that the rule is formed by the intersection of rule A in Tree0, B in Tree1, C in Tree2, and D in Tree3.

[0106] In order to achieve this fusion effect, a label column matrix can be established for each work order rule: if a certain condition is met, it is marked as 1, otherwise it is 0; and the rules of each tree are strung together in a cascade manner to form more rules.

[0107] Therefore, corresponding fusion processing methods can be flexibly adopted for personalized application scenarios, thereby effectively improving the fusion processing effect of multiple rule description contents.

[0108] S810: Determine target rule information based on the rule description content obtained through the fusion process.

[0109] In the embodiment of the present application, when the target rule information is determined based on the rule description content obtained by the fusion processing, the rules can be subdivided, the confidence of individual rules can be higher, and the model classification effect can be improved.

[0110] That is to say, in the embodiment of the present application, there are multiple first rule paths. After processing the initial description content according to the content type to obtain the rule description content, the rule description content of the rule nodes included in at least one first rule path can be fused, and the target rule information can be determined based on the rule description content obtained by the fusion processing. In this way, the description accuracy of the obtained target rule information for the content of multiple rule nodes can be effectively improved.

[0111] In this embodiment, by determining the initial description content corresponding to each rule node, determining the content type corresponding to the initial description content, and processing the initial description content according to the content type to obtain the rule description content, the adaptability of the initial description content processing process can be effectively improved based on the content type, thereby effectively improving the accuracy of the obtained rule description content. By parsing the initial description content to obtain one or more initial numerical intervals corresponding to at least one numerical type, if the number of initial numerical intervals is one, the initial numerical interval is used as the rule description content, and if the number of initial numerical intervals is multiple, the intersection of the multiple initial numerical intervals is taken to obtain the target numerical interval corresponding to the corresponding numerical type, and the target numerical interval is used as the rule description content. Thus, when the content type is a numerical type, the target numerical interval can be obtained as the rule description content, thereby effectively improving the indication effect of the rule description content. By parsing the enumeration content corresponding to at least one enumeration type from the initial description content, determining the content interval to which the enumeration content belongs, and using at least one enumeration content and the content interval to which each enumeration content belongs as the rule description content, the clarity of the indication of the enumeration content of the obtained rule description content can be effectively improved. By fusing the rule descriptions of the rule nodes included in at least one first rule path and determining target rule information based on the resulting fused rule descriptions, the accuracy of the resulting target rule information in describing the contents of multiple rule nodes can be effectively improved. By summing the rule descriptions of the rule nodes included in at least one first rule path, or performing intersection processing on the rule descriptions of the rule nodes included in at least one first rule path, flexible fusion processing methods can be adopted for personalized application scenarios, effectively improving the fusion processing effect of multiple rule descriptions.

[0112] FIG11 is a schematic structural diagram of a rule extraction device for a work order processing model proposed in an embodiment of the present application.

[0113] As shown in FIG11 , the rule extraction device 110 for the work order processing model includes:

[0114] Acquisition module 1101 is used to acquire a work order processing model, wherein the work order processing model includes: a decision tree, each decision tree is used to describe a corresponding rule set, the rule set includes multiple rule nodes, and the rule node represents a logical unit for processing a work order. At least some of the rule nodes form a rule path;

[0115] The processing module 1102 is configured to process the decision tree to obtain a target rule file, wherein the target rule file includes: at least one rule path; and

[0116] The first determining module 1103 is configured to determine target rule information according to at least one rule path.

[0117] In some embodiments of the present application, as shown in FIG12 , FIG12 is a schematic structural diagram of a rule extraction device for a work order processing model proposed in another embodiment of the present application. The device 110 further includes:

[0118] A second determining module 1104 is configured to determine a data processing result corresponding to each rule path, wherein the data processing result has a corresponding sample type;

[0119] A third determining module 1105 is configured to select a first rule path from at least one rule path according to the sample type;

[0120] The first determining module 1103 is specifically configured to:

[0121] According to the first rule path, target rule information is determined.

[0122] In some embodiments of the present application, the third determining module 1105 is specifically configured to:

[0123] The rule path corresponding to the data processing result whose sample type is a positive sample type is used as the first rule path.

[0124] In some embodiments of the present application, the first determining module 1103 is further configured to:

[0125] Determining rule nodes included in the first rule path;

[0126] Determine the rule description content corresponding to each rule node;

[0127] Target rule information is determined according to the rule description content of the rule nodes included in the first rule path.

[0128] In some embodiments of the present application, the first determining module 1103 is further configured to:

[0129] Determine the initial description content corresponding to each rule node;

[0130] Determining the content type corresponding to the initial description content;

[0131] The initial description content is processed according to the content type to obtain the rule description content.

[0132] In some embodiments of the present application, the content type includes at least one of the following:

[0133] Numeric types;

[0134] Enumeration type.

[0135] In some embodiments of the present application, the content type is a numerical type; wherein the first determining module 1103 is further configured to:

[0136] Parsing the initial description content to obtain one or more initial numerical ranges corresponding to at least one numerical type;

[0137] If the number of initial value intervals is one, the initial value interval is used as the rule description content;

[0138] If there are multiple initial numerical intervals, the intersection of the multiple initial numerical intervals is taken to obtain a target numerical interval corresponding to the corresponding numerical type, and the target numerical interval is used as the rule description content.

[0139] In some embodiments of the present application, the content type is an enumeration type; wherein the first determining module 1103 is further configured to:

[0140] Parsing the initial description content to obtain enumeration content corresponding to at least one enumeration type;

[0141] Determine the content interval to which the enumerated content belongs;

[0142] At least one enumerated content and the content range to which each enumerated content belongs are used as rule description content.

[0143] In some embodiments of the present application, there are multiple first rule paths; wherein the first determining module 1103 is further configured to:

[0144] Performing fusion processing on the rule description contents of the rule nodes included in at least one first rule path;

[0145] The target rule information is determined based on the rule description content obtained through fusion processing.

[0146] In some embodiments of the present application, the first determining module 1103 is further configured to:

[0147] performing summation processing on the rule description contents of the rule nodes included in at least one first rule path; and / or

[0148] An intersection process is performed on the rule description contents of the rule nodes included in at least one first rule path.

[0149] It should be noted that the aforementioned explanation of the rule extraction method for the work order processing model is also applicable to the rule extraction device for the work order processing model of this embodiment, and will not be repeated here.

[0150] In this embodiment, a work order processing model is obtained, wherein the work order processing model includes: a decision tree, each decision tree is used to describe a corresponding rule set, the rule set includes multiple rule nodes, the rule node represents a logical unit for processing a work order, at least some of the rule nodes form a rule path, and the decision tree is processed to obtain a target rule file, wherein the target rule file includes: at least one rule path, and according to at least one rule path, the target rule information is determined, thereby, the rules of multiple decision trees can be aggregated to improve the support of the rules and the confidence of the prediction results, thereby improving the interpretability of the model.

[0151] Figure 13 shows a block diagram of an exemplary computer device suitable for implementing the embodiments of the present application. The computer device 12 shown in Figure 13 is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0152] As shown in Figure 13, computer device 12 is a general-purpose computing device. Components of computer device 12 may include, but are not limited to, one or more processors or processing units 16, system memory 28, and a bus 18 that connects various system components (including system memory 28 and processing unit 16).

[0153] Bus 18 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor, or a local bus using any of a variety of bus architectures. Examples of these architectures include, but are not limited to, the Industry Standard Architecture (ISA) bus, the Micro Channel Architecture (MAC) bus, the Enhanced ISA bus, the Video Electronics Standards Association (VESA) local bus, and the Peripheral Component Interconnection (PCI) bus.

[0154] The computer device 12 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by the computer device 12, including volatile and non-volatile media, removable and non-removable media.

[0155] Memory 28 may include computer-readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. Computer device 12 may further include other removable / non-removable, volatile / non-volatile computer-readable storage media. By way of example only, storage system 34 may be configured to read and write to non-removable, non-volatile magnetic media (not shown in FIG. 13 , and commonly referred to as a "hard drive").

[0156] Although not shown in FIG13 , a disk drive for reading and writing to a removable non-volatile disk (e.g., a floppy disk) and an optical disk drive for reading and writing to a removable non-volatile optical disk (e.g., a Compact Disc Read Only Memory (CD-ROM), a Digital Video Disc Read Only Memory (DVD-ROM), or other optical media) can be provided. In these cases, each drive can be connected to the bus 18 via one or more data medium interfaces. The memory 28 can include at least one program product having a set (e.g., at least one) of program modules configured to perform the functions of the various embodiments of the present application.

[0157] A program / utility 40 having a set (at least one) of program modules 42 may be stored, for example, in memory 28. Such program modules 42 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data, each of which, or some combination thereof, may include an implementation of a network environment. Program modules 42 generally implement the functions and / or methods of the embodiments described herein.

[0158] The computer device 12 can also communicate with one or more external devices 14 (e.g., a keyboard, pointing device, display 24, etc.), one or more devices that enable human interaction with the computer device 12, and / or any device that enables the computer device 12 to communicate with one or more other computing devices (e.g., a network card, a modem, etc.). This communication can occur via an input / output (I / O) interface 22. Furthermore, the computer device 12 can communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network such as the Internet) via a network adapter 20. As shown, the network adapter 20 communicates with the other modules of the computer device 12 via a bus 18. It should be understood that, although not shown, other hardware and / or software modules can be used in conjunction with the computer device 12, including but not limited to microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0159] The processing unit 16 executes various functional applications and data processing by running programs stored in the system memory 28, such as implementing the rule extraction method for the work order processing model mentioned in the above embodiment.

[0160] In order to implement the above embodiments, the present application also proposes a non-temporary computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, it implements the rule extraction method for the work order processing model proposed in the above embodiments of the present application.

[0161] In order to implement the above embodiments, the present application also proposes a computer program product. When the instruction processor in the computer program product is executed, it executes the rule extraction method for the work order processing model proposed in the above embodiments of the present application.

[0162] Those skilled in the art will readily appreciate other embodiments of the present invention after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present invention that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only, and the true scope and spirit of the present application are indicated by the following claims.

[0163] It should be understood that the present application is not limited to the exact structures described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

[0164] It should be noted that, in the description of this application, the terms "first", "second", etc. are used for descriptive purposes only and should not be understood as indicating or implying relative importance. In addition, in the description of this application, unless otherwise specified, the meaning of "plurality" is two or more.

[0165] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code comprising one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may be performed out of the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present application belong.

[0166] 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-described 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: 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.

[0167] Those skilled in the art will understand that all or part of the steps in the method of the above embodiment can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

[0168] In addition, the functional units in the various embodiments of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into a module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0169] The storage medium mentioned above can be a read-only memory, a magnetic disk or an optical disk, etc.

[0170] Throughout this specification, reference to terms such as "one embodiment," "some embodiments," "examples," "specific examples," or "some examples" means that a specific feature, structure, material, or characteristic described in conjunction with that embodiment or example is included in at least one embodiment or example of the present application. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0171] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A rule extraction method for a work order processing model, the method comprising: Acquire a work order processing model, wherein the work order processing model includes: a decision tree, each decision tree is used to describe a corresponding rule set, the rule set includes a plurality of rule nodes, the rule nodes represent logical units for processing work orders, and at least some of the rule nodes form a rule path; processing the decision tree to obtain a target rule file, wherein the target rule file includes: at least one rule path; and According to the at least one rule path, target rule information is determined.

2. The method of claim 1, wherein: The method further comprises: Determine a data processing result corresponding to each of the rule paths, wherein the data processing result has a corresponding sample type; Selecting a first rule path from at least one rule path according to the sample type; Wherein, determining target rule information according to the at least one rule path includes: The target rule information is determined according to the first rule path.

3. The method of claim 2, wherein: The step of selecting a first rule path from at least one rule path according to the sample type includes: The rule path corresponding to the data processing result whose sample type is a positive sample type is used as the first rule path.

4. The method of claim 2, wherein: The determining the target rule information according to the first rule path includes: Determining the rule nodes included in the first rule path; Determine the rule description content corresponding to each of the rule nodes; The target rule information is determined according to the rule description content of the rule node included in the first rule path.

5. The method of claim 4, wherein: The determining the rule description content corresponding to each rule node includes: Determine the initial description content corresponding to each of the rule nodes; Determining a content type corresponding to the initial description content; The initial description content is processed according to the content type to obtain the rule description content.

6. The method of claim 5, wherein: The content type includes at least one of the following: Numeric types; Enumeration type.

7. The method of claim 6, wherein: The content type is the value type; wherein the processing of the initial description content according to the content type to obtain the rule description content includes: Parsing the initial description content to obtain one or more initial value intervals corresponding to at least one of the value types; If the number of the initial numerical intervals is one, the initial numerical interval is used as the rule description content; If there are multiple initial numerical intervals, an intersection is taken for the multiple initial numerical intervals to obtain a target numerical interval corresponding to the corresponding numerical type, and the target numerical interval is used as the rule description content.

8. The method of claim 6, wherein: The content type is the enumeration type; wherein the processing of the initial description content according to the content type to obtain the rule description content includes: Parsing the initial description content to obtain enumeration content corresponding to at least one of the enumeration types; Determine the content interval to which the enumerated content belongs; At least one of the enumerated contents and the content interval to which each of the enumerated contents belongs are used as the rule description content.

9. The method according to any one of claims 4 to 8, wherein: The number of the first rule paths is multiple; wherein the determining the target rule information according to the rule description content of the rule node included in the first rule path includes: Performing fusion processing on the rule description content of the rule node included in at least one of the first rule paths; The target rule information is determined according to the rule description content obtained through the fusion process.

10. The method of claim 9, wherein: The fusing the rule description content of the rule node included in at least one of the first rule paths includes at least one of the following: performing summation processing on the rule description contents of the rule nodes included in at least one of the first rule paths; An intersection process is performed on the rule description contents of the rule nodes included in at least one of the first rule paths.

11. A rule extraction device for a work order processing model, the device comprising: An acquisition module is used to acquire a work order processing model, wherein the work order processing model includes: a decision tree, each of the decision trees is used to describe a corresponding rule set, the rule set includes a plurality of rule nodes, the rule nodes represent logical units for processing work orders, and at least some of the rule nodes form a rule path; A processing module, configured to process the decision tree to obtain a target rule file, wherein the target rule file includes: at least one rule path; and The first determination module is used to determine target rule information according to the at least one rule path.

12. A computer device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 10.

13. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1 to 10.

14. A computer program product comprising a computer program, which, when executed by a processor, implements the steps of the method according to any one of claims 1 to 10.

Images