A method and device for determining intervention characteristics

Abstract

The present application provides a method and device for determining intervention features, the method comprising: obtaining service information about multiple stages of a target dual-end service; performing causal analysis on the service information to select intervention features that have a causal relationship with order conversion indicators from the service information; for each type of intervention feature, determining the degree of influence of the type of intervention feature on service order conversion; and selecting target intervention features from the intervention features according to the degree of influence of different types of intervention features on service order conversion. According to the determination method and device, more accurate tendency features that affect user experience can be obtained.

Description

Technical Field

[0001] The present application relates to the field of dual-end service technology, and in particular to a method and device for determining intervention characteristics. Background Art

[0002] With the rapid development of Internet technology, two-end services have been widely used by the public. Whether the commercial value of users can be retained to improve the transaction rate of service orders is one of the key factors to measure whether the two-end services can develop stably. The most direct quantitative reflection of the commercial value of users is the user experience indicator, that is, user loyalty. High-quality user loyalty is conducive to the promotion of service orders.

[0003] At present, the general way to measure user loyalty of two-end services is to use reference factors that are correlated with user loyalty to make judgments. For example, when user loyalty is user retention rate, the user retention rate of the next month can be judged based on the number of active days that users use the two-end services in the current month. However, since the reference factors that affect whether users are loyal to two-end services are often diverse, and the path for each reference factor to affect user loyalty is extremely complex, the accuracy of the reference factors that affect user loyalty determined only based on correlation is not high. In two-end services, reference factors with low accuracy often lead to a decrease in the accuracy of the service providers recommended to the service requester, thereby affecting the transaction rate of service orders. Summary of the invention

[0004] In view of this, the purpose of the present application is to provide a method and device for determining intervention features, which can quantitatively determine the intervention features that have a causal relationship with order conversion indicators related to user experience through causal inference, and determine the tendency characteristics that affect user experience, and screen the service providers that can provide target dual-end services based on the tendency characteristics, thereby improving the accuracy of the service providers recommended to the service requester and increasing the possibility of completing the service order.

[0005] This application mainly includes the following aspects:

[0006] In a first aspect, an embodiment of the present application provides a method for determining intervention features, the determination method comprising: obtaining service information about multiple stages of a target two-end service, wherein the service information of the multiple stages is used to characterize the execution status of service orders when historical service orders of the target two-end service are executed to different stages; performing a causal analysis on the service information to select intervention features that have a causal relationship with order conversion indicators from the service information; for each type of intervention feature, determining the degree of influence of the type of intervention feature on the service order conversion; and selecting a target intervention feature from the intervention features according to the degree of influence of different types of intervention features on the service order conversion.

[0007] In a possible implementation manner, the target dual-end service is a travel service, and the determination method further includes:

[0008] In response to receiving a service order for the travel service placed by the service requester, using the target intervention feature as a screening condition to screen out service providers that provide the travel service to the service requester;

[0009] The screened service providers are pushed to the service requester.

[0010] In a possible implementation manner, the target dual-end service is a takeaway delivery service, and the determination method further includes:

[0011] In response to receiving a service request for the food delivery service from a service requester, using the target intervention feature as a screening condition to screen out service providers that provide the delivery service to the delivery service requester;

[0012] The screened service providers are pushed to the service requester.

[0013] In a possible implementation, the step of performing causal analysis on the service information to select intervention features having a causal relationship with order conversion indicators from the service information includes:

[0014] Constructing a causal relationship graph that represents the causal relationship between the service information and the order conversion indicator, wherein the causal relationship graph includes a plurality of nodes and a plurality of edges, two nodes with a causal relationship are connected by an edge, the edge points to indicate that information corresponding to the end node of the edge is generated depending on information of the start node of the edge, and the plurality of nodes include nodes for representing the service information or for representing the order conversion indicator;

[0015] Obtain service information corresponding to the starting node connected to the edge of the ending node corresponding to the order conversion indicator;

[0016] The obtained service information is determined as an intervention feature having a causal relationship with the order conversion indicator.

[0017] In a possible implementation, the step of determining the degree of influence of such intervention features on service order conversion includes:

[0018] Determine the direct causal path in the causal relationship diagram by which the intervention feature directly affects the order conversion indicator;

[0019] Determine the indirect causal path in the causal relationship diagram by which the type of intervention feature indirectly affects the order conversion indicator;

[0020] Based on the weights of the edges corresponding to the direct causal path and the weights of the edges corresponding to the indirect causal path, the degree of influence of this type of intervention feature on the service order conversion is determined.

[0021] In a possible implementation, the step of determining the degree of influence of such intervention features on service order conversion includes:

[0022] Constructing a first initial linear model, a second initial linear model and a third initial linear model, wherein the first initial linear model represents the linear relationship between the order conversion index and the corresponding constraint condition; the second initial linear model represents the linear relationship between the intervention feature and the corresponding constraint condition; the third initial linear model represents the linear relationship between the intervention feature, the order conversion index and the corresponding constraint condition;

[0023] Obtain historical service information about multiple stages of the target dual-end service;

[0024] Substituting the historical service information into the first initial linear model, the second initial linear model and the third initial linear model respectively to obtain a first linear model, a second linear model and a third linear model,

[0025] Based on the model coefficients of the first linear model, the second linear model and the third linear model, the degree of influence of this type of intervention features on the service order conversion is determined.

[0026] In a possible implementation, the step of determining the degree of influence of such intervention features on service order conversion includes:

[0027] Constructing a first initial nonlinear model and a second initial nonlinear model, wherein the first initial nonlinear model represents the nonlinear relationship between the intervention feature and the corresponding constraint condition; and the second initial nonlinear model represents the nonlinear relationship between the intervention feature, the corresponding constraint condition and the order conversion index;

[0028] Obtain historical service information for multiple stages of a target dual-end service;

[0029] Substituting the historical service information into the first initial nonlinear model and the second initial nonlinear model respectively to obtain a first nonlinear model and a second nonlinear model;

[0030] The degree of influence of such intervention features on service order conversion is determined based on the first nonlinear model and the second nonlinear model.

[0031] In a possible implementation, the step of selecting a target intervention feature from the intervention features according to the degree of influence of different types of intervention features on the service order conversion includes:

[0032] Sort the impact of different types of intervention features on service order conversion in order of magnitude;

[0033] The intervention feature corresponding to the degree of impact under the specified ranking is selected as the target intervention feature.

[0034] In a possible implementation, the order conversion index includes any one of the following items: user retention rate, user satisfaction, and user net recommendation value.

[0035] In a second aspect, an embodiment of the present application further provides a device for determining an intervention feature, the device comprising:

[0036] An acquisition module, which acquires service information of multiple stages of a target dual-end service, wherein the service information of multiple stages is used to represent the execution status of service orders when historical service orders of the target dual-end service are executed to different stages;

[0037] an analysis module, performing causal analysis on the service information to select intervention features having a causal relationship with order conversion indicators from the service information;

[0038] A determination module determines, for each type of intervention feature, the degree of influence of the intervention feature on the conversion of service orders;

[0039] The screening module selects a target intervention feature from the intervention features according to the degree of influence of different types of intervention features on the service order conversion.

[0040] In a possible implementation manner, the determining device further includes:

[0041] a service provider determination module, in response to receiving a service order placed by a service requester for the travel service, using the target intervention feature as a screening condition to screen out service providers that provide the travel service to the service requester;

[0042] A push module pushes the screened service providers to the service requester.

[0043] In a possible implementation manner, the determining device further includes:

[0044] a service provider determination module, in response to receiving a service request from a service requester for the food delivery service, using the target intervention feature as a screening condition to screen out service providers that provide the delivery service to the delivery service requester;

[0045] A push module pushes the screened service providers to the service requester.

[0046] In a possible implementation, the analysis module is specifically used to:

[0047] Constructing a causal relationship graph that represents the causal relationship between the service information and the order conversion indicator, wherein the causal relationship graph includes a plurality of nodes and a plurality of edges, two nodes with a causal relationship are connected by an edge, the edge points to indicate that information corresponding to the end node of the edge is generated depending on information of the start node of the edge, and the plurality of nodes include nodes for representing the service information or for representing the order conversion indicator;

[0048] Obtain service information corresponding to the starting node connected to the edge of the ending node corresponding to the order conversion indicator;

[0049] The obtained service information is determined as an intervention feature having a causal relationship with the order conversion indicator.

[0050] In a possible implementation manner, the determining module is specifically configured to:

[0051] Determine the direct causal path in the causal relationship diagram by which the intervention feature directly affects the order conversion indicator;

[0052] Determine the indirect causal path in the causal relationship diagram by which the type of intervention feature indirectly affects the order conversion indicator;

[0053] Based on the weights of the edges corresponding to the direct causal path and the weights of the edges corresponding to the indirect causal path, the degree of influence of this type of intervention feature on the service order conversion is determined.

[0054] In a possible implementation manner, the determining module is specifically configured to:

[0055] Constructing a first initial linear model, a second initial linear model and a third initial linear model, wherein the first initial linear model represents the linear relationship between the order conversion index and the corresponding constraint condition; the second initial linear model represents the linear relationship between the intervention feature and the corresponding constraint condition; the third initial linear model represents the linear relationship between the intervention feature, the order conversion index and the corresponding constraint condition;

[0056] Obtain historical service information for multiple stages of a target dual-end service;

[0057] Substituting the historical service information into the first initial linear model, the second initial linear model and the third initial linear model respectively to obtain a first linear model, a second linear model and a third linear model,

[0058] Based on the model coefficients of the first linear model, the second linear model and the third linear model, the degree of influence of this type of intervention features on the service order conversion is determined.

[0059] In a possible implementation manner, the determining module is specifically configured to:

[0060] Constructing a first initial nonlinear model and a second initial nonlinear model, wherein the first initial nonlinear model represents the nonlinear relationship between the intervention feature and the corresponding constraint condition; and the second initial nonlinear model represents the nonlinear relationship between the intervention feature, the corresponding constraint condition and the order conversion index;

[0061] Obtain historical service information for multiple stages of a target dual-end service;

[0062] Substituting the historical service information into the first initial nonlinear model and the second initial nonlinear model respectively to obtain a first nonlinear model and a second nonlinear model;

[0063] The degree of influence of such intervention features on service order conversion is determined based on the first nonlinear model and the second nonlinear model.

[0064] In a possible implementation, the screening module is specifically used to:

[0065] Sort the impact of different types of intervention features on service order conversion in order of magnitude;

[0066] The intervention feature corresponding to the degree of impact under the specified ranking is selected as the target intervention feature.

[0067] In a possible implementation, the order conversion index includes any one of the following items: user retention rate, user satisfaction, and user net recommendation value.

[0068] In a third aspect, an embodiment of the present application further provides an electronic device, comprising: a processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate through the bus, and the machine-readable instructions are executed by the processor to execute the steps of the method for determining the intervention characteristics described in the first aspect or any possible implementation scheme of the first aspect.

[0069] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method for determining the intervention characteristics described in the first aspect or any possible implementation scheme of the first aspect are executed.

[0070] The method for determining intervention features provided in the embodiment of the present application performs causal analysis on the service information when obtaining service information of multiple stages of the target two-end service, so as to select intervention features with causal relationship with order conversion indicators from the service information of multiple stages, and then determines the degree of influence of each type of intervention feature on the service order conversion, and finally selects the target intervention feature from the intervention features according to the degree of influence of different types of intervention features on the service order conversion. In this way, more accurate tendency features that affect user experience can be obtained.

[0071] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are specifically cited below and described in detail with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0073] Figure 1 A flow chart showing a method for determining an intervention feature provided in an embodiment of the present application is shown;

[0074] Figure 2 A flowchart showing the steps of selecting intervention features provided by an embodiment of the present application;

[0075] Figure 3 An example of a cause-effect relationship diagram provided by an embodiment of the present application is shown;

[0076] Figure 4 A schematic diagram showing the structure of a device for determining an intervention feature provided in an embodiment of the present application is shown;

[0077] Figure 5 A schematic structural diagram of an electronic device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0078] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. It should be understood that the drawings in the present application only serve the purpose of explanation and description and are not used to limit the scope of protection of the present application. In addition, it should be understood that the schematic drawings are not drawn in real proportion. The flowchart used in this application shows the operations implemented according to some embodiments of the present application. It should be understood that the operations of the flowchart can be implemented out of sequence, and the steps without logical context can be reversed in order or implemented simultaneously. In addition, those skilled in the art, under the guidance of the content of the present application, can add one or more other operations to the flowchart, or remove one or more operations from the flowchart.

[0079] In addition, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.

[0080] In order to enable those skilled in the art to use the content of this application, the following implementation is provided in conjunction with the specific application scenario "determination of intervention features". For those skilled in the art, the general principles defined herein can be applied to other embodiments and application scenarios without departing from the spirit and scope of this application.

[0081] The following methods, devices, electronic devices or computer-readable storage media of the embodiments of the present application can be applied to any scenario where intervention features need to be determined. The embodiments of the present application are not limited to specific application scenarios. Any scheme using the methods and devices for determining intervention features provided by the embodiments of the present application are within the scope of protection of the present application.

[0082] It should be noted that the two-end service refers to the service provided by the designated service provider to the service requester by the cloud service platform assigning the service order issued by the service requester. Here, as an example, the two-end service may include but is not limited to the services provided in the application installed on any terminal device, such as travel services (i.e., online taxi services), food delivery services, etc. In addition, the two-end service may also be a sub-service under the service provided in the comprehensive application installed on any terminal device, such as carpooling services and ride-sharing services under travel services. Here, the term "service requester" in this application may refer to an individual, entity or tool that requests or orders a two-end service. The term "service provider" in this application refers to an individual, entity or tool that can provide a two-end service. The terms "service request" and "service order" in this application are used interchangeably to refer to requests initiated by passengers, service requesters, drivers, service providers, or suppliers, etc., or any combination thereof. The "service request" or "service order" may be accepted by passengers, service requesters, drivers, service providers, or suppliers, etc., or any combination thereof. Service requests can be fee-based or free.

[0083] Typically, the execution process of a two-end service is as follows: a cloud service platform establishes an association relationship between a service provider and a service requester; the cloud service platform receives a service order request issued by the service requester to the cloud service platform; in response to the service order request, the cloud service platform screens preferred service providers that can provide services for the service requester, and allocates the service order to the screened preferred service provider, which then provides services to the service requester to complete the service order.

[0084] It is worth noting that before the present application was proposed, in the field of two-end services, in order to improve the transaction rate of service orders, a reference factor or a combination of multiple reference factors is usually selected as a screening condition to screen out the preferred service provider that provides two-end services to the service requester. Usually, the screening condition is obtained by building an evaluation system model on the cloud service platform. For example, the evaluation system model can use the factors that are correlated with user loyalty (i.e., user experience) as reference factors. However, since the reference factors that affect whether users are loyal to two-end services are often diverse, and the path for each reference factor to affect user loyalty is extremely complex, the accuracy of the reference factors that affect user loyalty determined only based on correlation is not high. In two-end services, reference factors with low accuracy often lead to a decrease in the accuracy of the preferred service providers recommended to the service requester, thereby affecting the transaction rate of service orders.

[0085] In response to the above problems, the embodiment of the present application performs causal analysis on the service information of multiple stages of the target dual-end service when obtaining service information of the multiple stages, so as to select intervention features that have a causal relationship with the order conversion index from the service information of the multiple stages, and then for each type of intervention feature, determine the degree of influence of the type of intervention feature on the service order conversion, and finally select the target intervention feature from the intervention features according to the degree of influence of different types of intervention features on the service order conversion. In this way, the intervention features that affect the user experience can be quantitatively analyzed by causal inference, and the tendency features that affect the user experience can be determined to obtain more accurate intervention features.

[0086] In response to the above problems, the embodiments of the present application perform causal analysis on the service information when obtaining service information about multiple stages of the target two-end service, so as to select intervention features with causal relationships with order conversion indicators from the service information at multiple stages, and then determine the degree of influence of each type of intervention feature on the service order conversion. Finally, based on the degree of influence of different types of intervention features on the service order conversion, target intervention features are selected from the intervention features. In this way, more accurate tendency features that affect user experience can be obtained.

[0087] To facilitate the understanding of the present application, the technical solution provided by the present application is described in detail below in conjunction with specific embodiments.

[0088] Figure 1 This is a flowchart of a method for determining an intervention feature provided in an embodiment of the present application. The device for executing the method for determining an intervention feature may be a server or a cloud service platform. The following describes the method for determining an intervention feature in the present application with the cloud service platform as the execution subject. Figure 1 As shown, the method for determining the intervention feature provided in the embodiment of the present application includes the following steps:

[0089] S101: Acquire service information of multiple stages of a target dual-end service, wherein the service information of multiple stages is used to represent the execution status of service orders when historical service orders of the target dual-end service are executed to different stages.

[0090] In a specific implementation, the cloud service platform can obtain the execution status of the historical service orders of the target two-end service executed within the past predetermined time period at different stages of the service order execution. Taking the target two-end service as an online taxi-hailing service as an example, the cloud service platform can obtain the execution status of the historical service orders of the online taxi-hailing service at certain stages of execution (for example, the response stage, the pick-up stage, and the delivery stage). Specifically, the service information of multiple stages can be represented by the data generated in the above-mentioned different stages. In this example, the service information of multiple stages can be the data generated by the service order at different stages of execution, for example, the response time data, pre-response cancellation data, etc. generated in the response stage; the pick-up time data, post-response cancellation data, etc. generated in the pick-up stage; the detour time data, driver service score data, etc. generated in the delivery stage.

[0091] S102: Performing causal analysis on the service information to select intervention features having a causal relationship with order conversion indicators from the service information.

[0092] It should be noted that the order conversion index is an indicator related to the conversion of service orders, and is used to reflect the user's loyalty to the two-end service. Here, the service order conversion can represent the completion of the service order, for example, service order transaction, service order cancellation, service order update, etc. The service order conversion can be used to measure the user's recognition of the target two-end service. For example, as many service orders as possible are completed, indicating that users use the two-end service more and have a higher degree of recognition of the two-end service; more service orders are canceled, indicating that users do not recognize the two-end service very much or that there are defects in the service at a certain stage of the two-end service.

[0093] Here, the order conversion indicator may include but is not limited to any one of the following items: user retention rate, user satisfaction and user net recommendation value.

[0094] Specifically, the user retention rate can indicate the probability of users remaining in the dual-end service, and often reflects the quality of the dual-end service and the ability to retain users. For example, the user retention rate for the next month can be determined based on the number of active days that users use the dual-end service in the current month. A high user retention rate indicates that the number of days that the dual-end service is used is large, which in turn reflects that the quality of the dual-end service is high, and the higher the ability to retain users, the greater the possibility of closing the corresponding service order.

[0095] User satisfaction is used to measure the user's satisfaction with the two-end service and the interaction between the two-end services. For example, the user's satisfaction with the two-end service can be obtained through the praise rate of the service orders completed by the service provider in the past. The higher the user's satisfaction, the higher the quality of the two-end service and the higher the ability to retain users. The higher the ability to retain users, the greater the possibility of the corresponding service order being completed.

[0096] The user's net recommendation value indicates the likelihood that a user will recommend a two-end service to others. For example, the user's net recommendation value can be determined based on the success rate of users sharing the installation link of the two-end service. The higher the user's net recommendation value, the higher the quality of the two-end service and the higher the ability to retain users. The higher the ability to retain users, the greater the possibility of closing the corresponding service order.

[0097] Furthermore, the following will combine Figure 2 , the process of selecting intervention features that have a causal relationship with the order conversion indicator is explained, that is, the intervention features that have a causal relationship with the order conversion indicator are determined according to the following steps:

[0098] S201: Construct a causal relationship graph that represents the causal relationship between the service information and the order conversion index. The causal relationship graph includes multiple nodes and multiple edges, two nodes with a causal relationship are connected by an edge, the edge points to the information corresponding to the end node of the edge, and the information of the start node of the edge is generated based on the information of the end node of the edge, and the multiple nodes include nodes for representing the service information or for representing the order conversion index.

[0099] It should be noted that a causal relationship graph including the causal relationship between the service information and the order conversion index can be constructed by any existing method. For example, a causal graph can be learned from data by constructing a Bayesian network structure to establish a causal relationship graph including the causal relationship between the service information and the order conversion index. The above method is only exemplary, and a causal relationship graph can also be established by other methods, and the present invention does not make any limitation here.

[0100] S202: Obtain service information corresponding to the start node connected to the edge of the end node corresponding to the order conversion indicator.

[0101] S203: Determine the acquired service information as an intervention feature having a causal relationship with the order conversion indicator.

[0102] Figure 3An example of a causal relationship graph provided by an embodiment of the present application is shown. In this example, the nodes corresponding to the service information of multiple stages are nodes A, B, C, D, E and F, and the node corresponding to the order conversion indicator is node Y. The edge between the nodes indicates that there is a causal relationship between the two nodes, and the direction of the edge indicates that the information corresponding to the end node depends on the information of the start node of the edge, that is, the end node of the edge is affected by the start node of the edge. Figure 3 It can be seen that node Y is the termination node, and the nodes that have a causal relationship with node Y are node A, node B, node C, and node D, and the nodes that do not have a causal relationship with node Y are node E and node F. Therefore, the service information corresponding to node A, node B, node C, and node D can be determined as intervention features that have a causal relationship with the order conversion indicator.

[0103] Return to reference Figure 1 , S103: For each type of intervention feature, determine the impact of the intervention feature on the service order conversion.

[0104] In specific implementation, there are three ways to measure the impact of intervention characteristics on service order conversion.

[0105] In the first example, the following steps can be used to determine the impact of this type of intervention feature on service order conversion:

[0106] First, determine the direct causal path in the causal relationship diagram by which this type of intervention feature directly affects the order conversion indicator.

[0107] It should be noted that in a causal relationship graph, two nodes are connected by a directed edge. The starting node of the two nodes directly affects the ending node. The path of the edge connecting the two nodes is the direct causal path of the two nodes, that is, the direct causal path between the information corresponding to the two nodes. For example, this type of intervention feature corresponds to Figure 3 Taking node D in the causal relationship graph shown as an example, node D, as the starting node, directly points to the terminal node Y through an edge. Then the intervention feature corresponding to node D directly affects the order conversion indicator, and the path of the edge connecting node D to node Y is the direct causal path from node D to node Y.

[0108] Then, the indirect causal path in the causal relationship diagram through which this type of intervention feature indirectly affects the order conversion indicator is determined.

[0109] It should be noted that when two nodes in a causal relationship graph are connected by at least two directed edges, the starting node of the two nodes indirectly affects the ending node, and the path of the edge connecting the two nodes is the indirect causal path of the two nodes, that is, the indirect causal path of the information corresponding to the two nodes. Figure 3 Taking node D in the causal relationship graph shown as an example, node D as the starting node directly points to node A as the ending node through an edge, and node A as the starting node directly points to node Y as the ending node through an edge. It can be seen that the intervention feature corresponding to node D indirectly affects the order conversion indicator. The path of the edge connecting node D to node Y is the edge from node D to node A and the edge from node A to node Y. Then the edge from node D to node A and the edge from node A to node Y together serve as the indirect causal path from node D to node Y.

[0110] Finally, based on the weights of the edges corresponding to the direct causal path and the weights of the edges corresponding to the indirect causal path, the degree of influence of this type of intervention feature on the conversion of service orders is determined.

[0111] It should be noted that the weight of the edge is the probability of the event corresponding to the information corresponding to the end node occurring under the intervention of the information corresponding to the start node. For example, the weight of the edge between two nodes can be determined by the quantitative data between the information corresponding to the nodes. Taking the target two-end service as an online taxi service as an example, the information corresponding to the start node is "pick-up time", and the information corresponding to the end node is "user retention rate". The weight of the edge connecting the two nodes can be determined by estimating the probability of the event that the user continues to use the online taxi service when the "user retention rate" Y is equal to the reference value y1 of the user retention rate under the condition that the "pick-up time" is A.

[0112] It should be noted that after obtaining the direct causal path and the indirect causal path, the weights of the edges corresponding to the direct causal path and the weights of the edges corresponding to the indirect causal path can be directly aggregated, and the weights obtained after the aggregation process are used as the degree of influence of this type of intervention feature on the conversion of service orders. Here, since the weight of the edge is the probability of the event corresponding to the information corresponding to the end node occurring under the intervention of the information corresponding to the start node, the weight obtained after the aggregation process (i.e., the degree of influence of this type of intervention feature on the conversion of service orders) is also a quantitative specific value.

[0113] In the second example, the following steps are used to determine the impact of this type of intervention feature on service order conversion:

[0114] First, a first initial nonlinear model and a second initial nonlinear model are constructed, wherein the first initial nonlinear model characterizes the nonlinear relationship between the intervention feature and the corresponding constraint condition; the second initial nonlinear model characterizes the nonlinear relationship between the intervention feature, the corresponding constraint condition and the order conversion index.

[0115] In a specific implementation, the first initial linear model, the second initial linear model and the third initial linear model are constructed as follows:

[0116] First initial linear model:

[0117] Y i =α 1i +β 1i X i +ε 1i ,

[0118] Among them, Y i represents the order conversion index for user i, α 1i , β 1i and ε 1i represents the model coefficients of the first initial linear model, X i represents the constraint condition for user i, where i represents the user.

[0119] Second initial linear model:

[0120] M i =α 2i +β 2i X i +ε 2i ,

[0121] Among them, M i represents the intervention feature for user i, α 2i , β 2i and ε 2i represents the model coefficients of the second initial linear model, X i represents the constraint condition for user i, where i represents the user.

[0122] The third initial linear model:

[0123] Y i =α 3i +β 3i X i +γM i +ε 3i ,

[0124] Among them, Y i represents the order conversion index for user i, M i represents the intervention feature for user i, α 3i , β 3i , γ and ε 3i represents the model coefficients of the third initial linear model, X i represents the constraint condition for user i, where i represents the user.

[0125] Then, historical service information about multiple stages of the target dual-end service is obtained.

[0126] Next, the historical service information is substituted into the first initial linear model, the second initial linear model and the third initial linear model respectively to obtain a first linear model, a second linear model and a third linear model.

[0127] In a specific implementation, after constructing the first initial linear model, the second initial linear model and the third initial linear model, historical service information about multiple stages of the target two-end service can be obtained. The historical service information may include historical service information corresponding to each user in the target user group set using the target two-end service. Intervention features, order conversion indicators and constraints are screened out from the historical service information of each user. For different service orders of any user, the quantitative values ​​corresponding to the intervention feature order conversion indicators and constraints screened out from the different service orders are substituted into the first initial linear model, the second initial linear model and the third initial linear model, respectively. Using the least squares method, the model coefficients of the first initial linear model, the second initial linear model and the third initial linear model for the user can be obtained. The obtained model coefficients are substituted into the corresponding initial models, and the true models of the first initial linear model, the second initial linear model and the third initial linear model, that is, the first linear model, the second linear model and the third linear model, can be obtained.

[0128] Finally, based on the model coefficients of the first linear model, the second linear model and the third linear model, the degree of influence of this type of intervention features on the service order conversion is determined.

[0129] In the specific implementation, β is obtained in the above implementation process. 1i and β 3i In the case of 1i -β 3i The value of can be used to obtain the influence of intervention feature M on service order conversion for user i. Correspondingly, the influence of intervention feature M on service order conversion for each user in the target user group set can be obtained in the same way. The influence of intervention feature M on service order conversion obtained for all users in the target user group set is averaged to obtain the influence of intervention feature M on service order conversion for the target user group set.

[0130] In the third example, the following steps are used to determine the impact of this type of intervention feature on service order conversion:

[0131] First, a first initial nonlinear model and a second initial nonlinear model are constructed, wherein the first initial nonlinear model characterizes the nonlinear relationship between the intervention feature and the corresponding constraint condition; the second initial nonlinear model characterizes the nonlinear relationship between the intervention feature, the corresponding constraint condition and the order conversion index.

[0132] In a specific implementation, the first initial nonlinear model and the second nonlinear model are constructed as follows:

[0133] First initial nonlinear model:

[0134] M i (T i )=Pr(M i =1|T i , X i )=logit -1 (α 4i +β 4i X i +ε 4i ),

[0135] Among them, M i represents the intervention feature for user i, M i is a binary variable, M i =1 means intervention on user i, M i = 0 means no intervention is performed on user i, Pr(M i =1|T i , X i ) means that under the Pr function, the variable is T i , X i About M i The nonlinear model, X i represents the constraints for user i, T i Indicates whether user i is affected by intervention feature M i Intervention, T i is a binary variable, T i =1 indicates that user i is interfered with by feature M i Intervention, T i =0 means user i is not subject to intervention feature M i Intervention, i represents the user.

[0136] Second initial nonlinear model:

[0137] Y i (T i ,M(T i ))=Pr(Y i =1|T i , M i , X i) = log it -1 (α 5i +β 5i X i +γM i +ε 5i ),

[0138] Among them, M i represents the intervention feature for user i, M i is a binary variable, M i =1 means intervention on user i, M i = 0 means no intervention is performed on user i, Pr(Y i =1|T i , M i , X i ) means that under the Pr function, the variable is T i , X i , M i About Y i The nonlinear model, X i represents the constraints for user i, T i Indicates whether user i is affected by intervention feature M i Intervention, T i is a binary variable, T i =1 indicates that user i is subject to intervention feature M i Intervention, T i =0 means user i is not subject to intervention feature M i Intervention, Y i represents the order conversion index for user i, where i represents the user.

[0139] Then, historical service information about multiple stages of the target dual-end service is obtained.

[0140] Next, the historical service information is substituted into the first initial nonlinear model and the second initial nonlinear model respectively to obtain a first nonlinear model and a second nonlinear model.

[0141] In a specific implementation, after constructing the first initial nonlinear model and the second initial nonlinear model, historical service information about multiple stages of the target two-end service can be obtained. The historical service information may include historical service information corresponding to each user in the target user group set using the target two-end service. Intervention features, order conversion indicators and constraints are screened out from the historical service information of each user. For different service orders of any user, the quantitative values ​​corresponding to the intervention feature order conversion indicators and constraints screened out from the different service orders are substituted into the first initial nonlinear model and the second initial nonlinear model, respectively. Using the gradient descent method, the model coefficients of the first initial nonlinear model and the second initial nonlinear model for the user can be obtained. The obtained model coefficients are substituted into the corresponding models, and the true models of the first initial nonlinear model and the second initial nonlinear model, that is, the first nonlinear model, the second nonlinear model and the third nonlinear model, can be obtained.

[0142] Finally, the influence of such intervention features on the service order conversion is determined based on the first nonlinear model and the second nonlinear model.

[0143] In the specific implementation, M is obtained in the above implementation process. i (T i ) and Y i (T i ,M(T i )) In the case of i (T i )=Y i (T i ,M(1))-Y i (T i , M(0)), we can get the intervention feature M for user i i Impact on service order conversionδ i (T i ). Accordingly, the same method can be used to obtain the influence of the intervention feature M on the service order conversion for each user in the target user group set. The influence of the intervention feature M on the service order conversion obtained for all users in the target user group set is averaged to obtain the influence of the intervention feature M on the service order conversion.

[0144] Return to reference Figure 1 , S104: Select a target intervention feature from the intervention features according to the degree of influence of different types of intervention features on the service order conversion.

[0145] In a specific implementation, the influence of different types of intervention features on the service order conversion can be ranked in order of size, and then the intervention feature corresponding to the influence degree under the specified ranking is selected as the target intervention feature.

[0146] For example, after sorting the influences of different types of intervention features on the service order conversion in order of magnitude, the intervention feature corresponding to the highest influence may be used as the target intervention feature.

[0147] In an embodiment of the present application, when obtaining service information about multiple stages of a target two-end service, a causal analysis is performed on the service information to select intervention features with causal relationships with order conversion indicators from the service information at multiple stages, and then for each type of intervention feature, the degree of influence of the type of intervention feature on the service order conversion is determined. Finally, based on the degree of influence of different types of intervention features on the service order conversion, a target intervention feature is selected from the intervention features. In this way, more accurate tendency features that affect user experience can be obtained.

[0148] In addition, after selecting the target intervention feature, in a preferred example of the present application, in addition to S101, S102, S103 and S104, after S104, when the target dual-end service is a travel service, the method for determining the intervention feature further includes the following steps (in Figure 1 Not shown):

[0149] Step (A): in response to receiving a service order for the travel service issued by the service requester, the target intervention feature is used as a screening condition to screen out service providers that provide the travel service to the service requester.

[0150] In the specific implementation, if the service requester sends a travel service request for travel services to the cloud service platform, the cloud service platform will use the selected target intervention features as screening conditions to screen the service providers that can provide travel services after receiving the travel service request, and screen out the most preferred service provider that can provide travel services to the service requester.

[0151] Step (B): Pushing the screened service providers to the service requester.

[0152] In specific implementation, the screened service providers may be pushed to the service requester to complete the service order.

[0153] For example, when the travel service is an online taxi service and the order conversion indicator is the user retention rate, the cloud service platform can obtain service information about multiple stages of the online taxi service, such as the response time data and cancellation data before response generated in the response stage; the pick-up time data and cancellation data after response generated in the pick-up stage; the detour time data and driver service score data generated in the delivery stage. The cloud service platform performs causal analysis on the above service information, and selects the intervention features "response time" and "pick-up time" that have a causal relationship with "user retention rate", and determines the influence of "response time" on "user retention rate" and the influence of "pick-up time" on "user retention rate". Among them, after comparison, the influence of "response time" on "user retention rate" is greater than the influence of "pick-up time" on "user retention rate". At this time, "response time" can be determined as the target interference feature. The terminal device (service requester) generates a service order in response to the passenger inputting the departure place, destination, departure time, number of people in the taxi-hailing interface of the travel application installed on the terminal device, and sends the service order to the cloud service platform. The cloud service platform selects online-hailing cars (i.e., service providers) that can complete the service order in response to the service order. During the screening process, the "response time" can be used as one of the screening conditions to screen online-hailing cars that can complete the service order. For example, the online-hailing car with the shortest "response time" can be selected as the most preferred online-hailing car for providing travel services. At the same time, the cloud service platform can push the online-hailing car with the shortest "response time" to the passenger to complete the travel order.

[0154] In addition, after selecting the target intervention feature, in another preferred example of the present application, in addition to S101, S102, S103 and S104, after S104, when the target dual-end service is a food delivery service, the method for determining the intervention feature further includes the following steps (in Figure 1 Not shown):

[0155] Step (a): in response to receiving a service request for the food delivery service from a service requester, using the target intervention feature as a screening condition to screen out service providers that provide the delivery service to the delivery service requester.

[0156] In the specific implementation, if the service requester sends a travel service request for a food delivery service to the cloud service platform, after receiving the food delivery request, the cloud service platform will use the selected target intervention features as screening conditions to screen the service providers that can provide food delivery services, and select the most preferred service provider that can provide food delivery services to the service requester.

[0157] Step (b): Pushing the screened service providers to the service requester.

[0158] In specific implementation, the screened service providers may be pushed to the service requester to complete the service order.

[0159] For example, when the travel service is a takeaway delivery service and the order conversion indicator is the user retention rate, the cloud service platform can obtain service information about multiple stages of the takeaway delivery service, such as the response time data and cancellation data before response generated in the response stage; the order time data, cancellation data after response, and distance from the merchant generated in the order receiving stage; and the delivery time data and deliveryman service rating data generated in the delivery stage. The cloud service platform performs causal analysis on the above service information and selects the intervention features "order receiving time" and "cancellation after response" that have a causal relationship with "user retention rate" to determine the influence of "order receiving time" on "user retention rate" and the influence of "cancellation after response" on "user retention rate". Among them, after comparison, the influence of "order receiving time" on "user retention rate" is greater than the influence of "cancellation after response" on "user retention rate". At this time, "order receiving time" can be determined as the target interference feature. The terminal device (service requester) generates a service order in response to the user inputting the selected goods, delivery location, delivery time and other information in the takeaway ordering interface of the takeaway application installed on the terminal device and then paying the order, and sends the service order to the cloud service platform. In response to the service order, the cloud service platform selects the takeaway deliveryman (i.e., the service provider) who can complete the service order. During the screening process, the "order acceptance time" can be used as one of the screening conditions to select the takeaway deliveryman who can complete the service order. For example, the takeaway deliveryman with the shortest "order acceptance time" can be selected as the most preferred takeaway deliveryman for providing travel services. At the same time, the cloud service platform can push the takeaway deliveryman with the shortest "order acceptance time" to the user to complete the travel order.

[0160] In this way, the intervention characteristics with causal relationships in order conversion indicators related to user experience can be quantitatively determined by causal inference, and the tendency characteristics that affect user experience can be determined. Service providers that can provide target dual-end services can be screened based on the tendency characteristics, thereby improving the accuracy of service providers recommended to service requesters and increasing the possibility of completing service orders.

[0161] Based on the same application concept, the embodiments of the present application also provide a device for determining intervention features corresponding to the method for determining intervention features provided in the above embodiments. Since the principle of solving the problem by the device in the embodiments of the present application is similar to the method for determining intervention features in the above embodiments of the present application, the implementation of the device can refer to the implementation of the method, and the repeated parts will not be repeated.

[0162] Figure 4 A schematic diagram of the structure of a device for determining intervention features provided in an embodiment of the present application is shown.

[0163] like Figure 4 As shown, the determining device 400 includes:

[0164] The acquisition module 401 acquires service information of multiple stages of a target dual-end service, wherein the service information of multiple stages is used to represent the execution status of service orders of historical service orders of the target dual-end service when they are executed to different stages.

[0165] In a specific implementation, the acquisition module 401 may acquire the service order execution status of historical service orders of the target dual-end service executed within a predetermined time period in the past at different stages of the service order execution.

[0166] The analysis module 402 performs causal analysis on the service information to select intervention features that have a causal relationship with order conversion indicators from the service information.

[0167] It should be noted that the order conversion index is an indicator related to the conversion of service orders, and is used to reflect the user's loyalty to the two-end service. Here, the service order conversion can represent the completion of the service order, for example, service order transaction, service order cancellation, service order update, etc. The service order conversion can be used to measure the user's recognition of the target two-end service. For example, as many service orders as possible are completed, indicating that users use the two-end service more and have a higher degree of recognition of the two-end service; more service orders are canceled, indicating that users do not recognize the two-end service very much or that there are defects in the service at a certain stage of the two-end service.

[0168] In a possible implementation, the order conversion index may include, but is not limited to, any one of the following items: user retention rate, user satisfaction, and user net recommendation value.

[0169] In a possible implementation, the analysis module 402 is specifically configured to:

[0170] Construct a causal relationship graph that characterizes the causal relationship between the service information and the order conversion indicator, wherein the causal relationship graph includes multiple nodes and multiple edges, two nodes with a causal relationship are connected by an edge, the direction of the edge is used to indicate that the information corresponding to the end node of the edge depends on the information of the start node of the edge, and the multiple nodes include nodes for representing the service information or for representing the order conversion indicator.

[0171] It should be noted that a causal relationship graph including the causal relationship between the service information and the order conversion index can be constructed by any existing method. For example, a causal graph can be learned from data by constructing a Bayesian network structure to establish a causal relationship graph including the causal relationship between the service information and the order conversion index. The above method is only exemplary, and a causal relationship graph can also be established by other methods, and the present invention does not make any limitation here.

[0172] Obtain service information corresponding to the starting node connected to the edge of the terminating node corresponding to the order conversion indicator.

[0173] The obtained service information is determined as an intervention feature having a causal relationship with the order conversion indicator.

[0174] The determination module 403 determines, for each type of intervention feature, the degree of influence of the type of intervention feature on the service order conversion.

[0175] In a possible implementation manner, the determining module 403 is specifically configured to:

[0176] Determine the direct causal path in the causal relationship diagram by which such intervention feature directly affects the order conversion indicator.

[0177] It should be noted that two nodes in the causal relationship graph are connected by a directed edge, the starting node of the two nodes directly affects the ending node, and the path of the edge connecting the two nodes is the direct causal path of the two nodes, that is, the direct causal path between the information corresponding to the two nodes.

[0178] Determine the indirect causal path in the causal relationship diagram by which this type of intervention feature indirectly affects the order conversion indicator.

[0179] It should be noted that when two nodes in a causal relationship graph are connected by at least two directed edges, the starting node of the two nodes indirectly affects the ending node, and the path of the edges connecting the two nodes is the indirect causal path of the two nodes, that is, the indirect causal path of the information corresponding to the two nodes.

[0180] Based on the weights of the edges corresponding to the direct causal path and the weights of the edges corresponding to the indirect causal path, the degree of influence of this type of intervention feature on the service order conversion is determined.

[0181] It should be noted that the weight of the edge is the probability of the event corresponding to the information corresponding to the end node occurring under the intervention of the information corresponding to the start node. After obtaining the direct causal path and the indirect causal path, the weight of the edge corresponding to the direct causal path and the weight of the edge corresponding to the indirect causal path can be directly aggregated, and the weight obtained after the aggregation process is used as the degree of influence of this type of intervention feature on the conversion of service orders. Here, since the weight of the edge is the probability of the event corresponding to the information corresponding to the end node occurring under the intervention of the information corresponding to the start node, the weight obtained after the aggregation process (that is, the degree of influence of this type of intervention feature on the conversion of service orders) is also a quantitative specific value.

[0182] In a possible implementation manner, the determining module 403 is specifically configured to:

[0183] A first initial linear model, a second initial linear model and a third initial linear model are constructed, wherein the first initial linear model represents the linear relationship between the order conversion index and the corresponding constraint conditions; the second initial linear model represents the linear relationship between the intervention feature and the corresponding constraint conditions; and the third initial linear model represents the linear relationship between the intervention feature, the order conversion index and the corresponding constraint conditions.

[0184] Obtain historical service information for multiple stages of a target dual-end service;

[0185] Substituting the historical service information into the first initial linear model, the second initial linear model and the third initial linear model respectively to obtain a first linear model, a second linear model and a third linear model,

[0186] Based on the model coefficients of the first linear model, the second linear model and the third linear model, the degree of influence of this type of intervention features on the service order conversion is determined.

[0187] In a possible implementation manner, the determining module 403 is specifically configured to:

[0188] Constructing a first initial nonlinear model and a second initial nonlinear model, wherein the first initial nonlinear model represents the nonlinear relationship between the intervention feature and the corresponding constraint condition; and the second initial nonlinear model represents the nonlinear relationship between the intervention feature, the corresponding constraint condition and the order conversion index;

[0189] Obtain historical service information for multiple stages of a target dual-end service;

[0190] Substituting the historical service information into the first initial nonlinear model and the second initial nonlinear model respectively to obtain a first nonlinear model and a second nonlinear model;

[0191] The degree of influence of such intervention features on service order conversion is determined based on the first nonlinear model and the second nonlinear model.

[0192] The screening module 404 selects a target intervention feature from the intervention features according to the degree of influence of different types of intervention features on the service order conversion.

[0193] In a possible implementation, the screening module 404 is specifically configured to:

[0194] Sort the impact of different types of intervention features on service order conversion in order of magnitude;

[0195] The intervention feature corresponding to the degree of impact under the specified ranking is selected as the target intervention feature.

[0196] In a possible implementation manner, the determining device 400 further includes:

[0197] a service provider determination module, in response to receiving a service order placed by a service requester for the travel service, using the target intervention feature as a screening condition to screen out service providers that provide the travel service to the service requester;

[0198] A push module pushes the screened service providers to the service requester.

[0199] In a possible implementation manner, the determining device 400 further includes:

[0200] a service provider determination module, in response to receiving a service request from a service requester for the food delivery service, using the target intervention feature as a screening condition to screen out service providers that provide the delivery service to the delivery service requester;

[0201] A push module pushes the screened service providers to the service requester.

[0202] In an embodiment of the present application, when obtaining service information about multiple stages of a target two-end service, a causal analysis is performed on the service information to select intervention features with causal relationships with order conversion indicators from the service information at multiple stages, and then for each type of intervention feature, the degree of influence of the type of intervention feature on the service order conversion is determined. Finally, based on the degree of influence of different types of intervention features on the service order conversion, a target intervention feature is selected from the intervention features. In this way, more accurate tendency features that affect user experience can be obtained.

[0203] Based on the same application concept, see Figure 5As shown, it is a schematic diagram of the structure of an electronic device 500 provided in an embodiment of the present application, including: a processor 510, a memory 520 and a bus 530, the memory 520 stores machine-readable instructions executable by the processor 510, when the electronic device 500 is running, the processor 510 and the memory 520 communicate through the bus 530, and the machine-readable instructions are executed by the processor 510 when running to perform the steps of the method for determining the intervention feature as described in any of the above embodiments.

[0204] Specifically, when the machine-readable instruction is executed by the processor 510, the following processing may be performed:

[0205] Acquire service information of multiple stages of the target dual-end service, wherein the service information of the multiple stages is used to represent the execution status of service orders when historical service orders of the target dual-end service are executed to different stages;

[0206] performing causal analysis on the service information to select intervention features from the service information that have a causal relationship with order conversion indicators;

[0207] For each type of intervention feature, determine the degree of influence of the intervention feature on the conversion of service orders;

[0208] According to the degree of influence of different types of intervention features on the service order conversion, a target intervention feature is selected from the intervention features.

[0209] Furthermore, when the machine-readable instructions are executed by the processor 510, the following processing may be performed:

[0210] In response to receiving a service order for the travel service placed by the service requester, using the target intervention feature as a screening condition to screen out service providers that provide the travel service to the service requester;

[0211] The screened service providers are pushed to the service requester.

[0212] Furthermore, when the machine-readable instructions are executed by the processor 510, the following processing may be performed:

[0213] In response to receiving a service request for the food delivery service from a service requester, using the target intervention feature as a screening condition to screen out service providers that provide the delivery service to the delivery service requester;

[0214] The screened service providers are pushed to the service requester.

[0215] Furthermore, when the machine-readable instructions are executed by the processor 510, the following processing may be performed:

[0216] Constructing a causal relationship graph that represents the causal relationship between the service information and the order conversion indicator, wherein the causal relationship graph includes a plurality of nodes and a plurality of edges, two nodes with a causal relationship are connected by an edge, the edge points to indicate that information corresponding to the end node of the edge is generated depending on information of the start node of the edge, and the plurality of nodes include nodes for representing the service information or for representing the order conversion indicator;

[0217] Obtain service information corresponding to the starting node connected to the edge of the ending node corresponding to the order conversion indicator;

[0218] The obtained service information is determined as an intervention feature having a causal relationship with the order conversion indicator.

[0219] Furthermore, when the machine-readable instructions are executed by the processor 510, the following processing may be performed:

[0220] Determine the direct causal path in the causal relationship diagram by which the intervention feature directly affects the order conversion indicator;

[0221] Determine the indirect causal path in the causal relationship diagram by which the type of intervention feature indirectly affects the order conversion indicator;

[0222] Based on the weights of the edges corresponding to the direct causal path and the weights of the edges corresponding to the indirect causal path, the degree of influence of this type of intervention feature on the service order conversion is determined.

[0223] Furthermore, when the machine-readable instructions are executed by the processor 510, the following processing may be performed:

[0224] Constructing a first initial linear model, a second initial linear model and a third initial linear model, wherein the first initial linear model represents the linear relationship between the order conversion index and the corresponding constraint condition; the second initial linear model represents the linear relationship between the intervention feature and the corresponding constraint condition; the third initial linear model represents the linear relationship between the intervention feature, the order conversion index and the corresponding constraint condition;

[0225] Obtain historical service information for multiple stages of a target dual-end service;

[0226] Substituting the historical service information into the first initial linear model, the second initial linear model and the third initial linear model respectively to obtain a first linear model, a second linear model and a third linear model,

[0227] Based on the model coefficients of the first linear model, the second linear model and the third linear model, the degree of influence of this type of intervention features on the service order conversion is determined.

[0228] Furthermore, when the machine-readable instructions are executed by the processor 510, the following processing may be performed:

[0229] Constructing a first initial nonlinear model and a second initial nonlinear model, wherein the first initial nonlinear model represents the nonlinear relationship between the intervention feature and the corresponding constraint condition; and the second initial nonlinear model represents the nonlinear relationship between the intervention feature, the corresponding constraint condition and the order conversion index;

[0230] Obtain historical service information for multiple stages of a target dual-end service;

[0231] Substituting the historical service information into the first initial nonlinear model and the second initial nonlinear model respectively to obtain a first nonlinear model and a second nonlinear model;

[0232] The degree of influence of such intervention features on service order conversion is determined based on the first nonlinear model and the second nonlinear model.

[0233] Furthermore, when the machine-readable instructions are executed by the processor 510, the following processing may be performed:

[0234] Sort the impact of different types of intervention features on service order conversion in order of magnitude;

[0235] The intervention feature corresponding to the degree of impact under the specified ranking is selected as the target intervention feature.

[0236] In an embodiment of the present application, when obtaining service information about multiple stages of a target two-end service, a causal analysis is performed on the service information to select intervention features with causal relationships with order conversion indicators from the service information at multiple stages, and then for each type of intervention feature, the degree of influence of the type of intervention feature on the service order conversion is determined. Finally, based on the degree of influence of different types of intervention features on the service order conversion, a target intervention feature is selected from the intervention features. In this way, more accurate tendency features that affect user experience can be obtained.

[0237] Based on the same application concept, an embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method for determining the intervention feature provided in the above embodiment are executed.

[0238] Specifically, the storage medium can be a general storage medium, such as a mobile disk, a hard disk, etc. When the computer program on the storage medium is run, the above-mentioned method for determining the intervention characteristics can be executed, and the intervention characteristics that affect the user experience can be quantitatively analyzed by causal inference, and the tendency characteristics that affect the user experience can be determined. The operation strategy of the target dual-end service is guided according to the tendency characteristics to achieve the healthy development of the target dual-end business.

[0239] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0240] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of devices or units can be electrical, mechanical or other forms.

[0241] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0242] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0243] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium that is executable by a processor. Based on this understanding, the technical solution of the present application can essentially be embodied in the form of a software product, or in other words, the part that contributes to the prior art or the part of the technical solution. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0244] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present application, which are used to illustrate the technical solutions of the present application, rather than to limit them. The protection scope of the present application is not limited thereto. Although the present application is described in detail with reference to the above-mentioned embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-mentioned embodiments within the technical scope disclosed in the present application, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A method for determining intervention characteristics, characterized in that: include: Acquire service information of multiple stages of the target dual-end service, wherein the service information of the multiple stages is used to represent the execution status of service orders when historical service orders of the target dual-end service are executed to different stages; Constructing a causal relationship graph that characterizes the causal relationship between the service information and the order conversion index, wherein the causal relationship graph includes a plurality of nodes and a plurality of edges, two nodes with a causal relationship are connected by an edge, the edge is directed to indicate that information corresponding to the end node of the edge is generated depending on information of the start node of the edge, and the plurality of nodes include nodes for indicating the service information or for indicating the order conversion index; the order conversion index is an indicator related to service order conversion, and is used to reflect the user's loyalty to the target dual-end service; the service order conversion is used to measure the user's recognition of the target dual-end service; Acquire service information corresponding to a start node connected to an edge of a termination node corresponding to the order conversion indicator; determine the acquired service information as an intervention feature having a causal relationship with the order conversion indicator; For each type of intervention feature, determine the degree of influence of the intervention feature on the conversion of service orders; According to the degree of influence of different types of intervention features on the service order conversion, a target intervention feature is selected from the intervention features.

2. The determination method according to claim 1, characterized in that: The target dual-end service is a travel service, and the determination method further includes: In response to receiving a service order for the travel service placed by the service requester, using the target intervention feature as a screening condition to screen out service providers that provide the travel service to the service requester; The screened service providers are pushed to the service requester.

3. The determination method according to claim 1, characterized in that: The target dual-end service is a takeaway delivery service, and the determination method further includes: In response to receiving a service request for the food delivery service from a service requester, using the target intervention feature as a screening condition to screen out service providers that provide the delivery service to the delivery service requester; The screened service providers are pushed to the service requester.

4. The determination method according to claim 1, characterized in that: The step of determining the degree of influence of such intervention characteristics on the service order conversion includes: Determine the direct causal path in the causal relationship diagram by which the intervention feature directly affects the order conversion indicator; Determine the indirect causal path in the causal relationship diagram by which the type of intervention feature indirectly affects the order conversion indicator; Based on the weights of the edges corresponding to the direct causal path and the weights of the edges corresponding to the indirect causal path, the degree of influence of this type of intervention feature on the service order conversion is determined.

5. The determination method according to claim 1, characterized in that: The step of determining the degree of influence of such intervention characteristics on the service order conversion includes: Constructing a first initial linear model, a second initial linear model and a third initial linear model, wherein the first initial linear model represents the linear relationship between the order conversion index and the corresponding constraint condition; the second initial linear model represents the linear relationship between the intervention feature and the corresponding constraint condition; the third initial linear model represents the linear relationship between the intervention feature, the order conversion index and the corresponding constraint condition; Obtain historical service information for multiple stages of a target dual-end service; Substituting the historical service information into the first initial linear model, the second initial linear model and the third initial linear model respectively to obtain a first linear model, a second linear model and a third linear model, Based on the model coefficients of the first linear model, the second linear model and the third linear model, the degree of influence of this type of intervention features on the service order conversion is determined.

6. The determination method according to claim 1, characterized in that: The step of determining the degree of influence of such intervention characteristics on the service order conversion includes: Constructing a first initial nonlinear model and a second initial nonlinear model, wherein the first initial nonlinear model represents the nonlinear relationship between the intervention feature and the corresponding constraint condition; and the second initial nonlinear model represents the nonlinear relationship between the intervention feature, the corresponding constraint condition and the order conversion index; Obtain historical service information for multiple stages of a target dual-end service; Substituting the historical service information into the first initial nonlinear model and the second initial nonlinear model respectively to obtain a first nonlinear model and a second nonlinear model; The degree of influence of such intervention features on service order conversion is determined based on the first nonlinear model and the second nonlinear model.

7. The determination method according to claim 1, characterized in that: The step of selecting a target intervention feature from the intervention features according to the degree of influence of different types of intervention features on the service order conversion includes: Sort the impact of different types of intervention features on service order conversion in order of magnitude; The intervention feature corresponding to the degree of impact under the specified ranking is selected as the target intervention feature.

8. The determination method according to claim 1, characterized in that: The order conversion indicator includes any one of the following items: user retention rate, user satisfaction and user net recommendation value.

9. A device for determining an intervention feature, characterized in that: include: An acquisition module, which acquires service information of multiple stages of a target dual-end service, wherein the service information of multiple stages is used to represent the execution status of service orders when historical service orders of the target dual-end service are executed to different stages; An analysis module is provided to construct a causal relationship graph including a characterization of a causal relationship between the service information and the order conversion index, wherein the causal relationship graph includes a plurality of nodes and a plurality of edges, two nodes having a causal relationship are connected by an edge, the direction of the edge is used to indicate that the information corresponding to the end node of the edge depends on the information of the start node of the edge, and the plurality of nodes include nodes for indicating the service information or for indicating the order conversion index; the order conversion index is an indicator related to the service order conversion, and is used to reflect the user's loyalty to the target two-end service; the service order conversion is used to measure the user's recognition of the target two-end service; the service information corresponding to the start node connected to the edge of the end node corresponding to the order conversion index is obtained; and the obtained service information is determined as an intervention feature having a causal relationship with the order conversion index; A determination module determines, for each type of intervention feature, the degree of influence of the intervention feature on the conversion of service orders; The screening module selects a target intervention feature from the intervention features according to the degree of influence of different types of intervention features on the service order conversion.

10. The determination device according to claim 9, characterized in that: The target dual-end service is a travel service, and the determining device further includes: a service provider determination module, in response to receiving a service order placed by a service requester for the travel service, using the target intervention feature as a screening condition to screen out service providers that provide the travel service to the service requester; A push module pushes the screened service providers to the service requester.

11. The determination device according to claim 9, characterized in that: The target dual-end service is a takeaway delivery service, and the determining device further includes: a service provider determination module, in response to receiving a service request from a service requester for the food delivery service, using the target intervention feature as a screening condition to screen out service providers that provide the delivery service to the delivery service requester; A push module pushes the screened service providers to the service requester.

12. The determination device according to claim 9, characterized in that: The determination module is specifically used for: Determine the direct causal path in the causal relationship diagram by which the intervention feature directly affects the order conversion indicator; Determine the indirect causal path in the causal relationship diagram by which the type of intervention feature indirectly affects the order conversion indicator; Based on the weights of the edges corresponding to the direct causal path and the weights of the edges corresponding to the indirect causal path, the degree of influence of this type of intervention feature on the service order conversion is determined.

13. The determination device according to claim 9, characterized in that: The determination module is specifically used for: Constructing a first initial linear model, a second initial linear model and a third initial linear model, wherein the first initial linear model represents the linear relationship between the order conversion index and the corresponding constraint condition; the second initial linear model represents the linear relationship between the intervention feature and the corresponding constraint condition; the third initial linear model represents the linear relationship between the intervention feature, the order conversion index and the corresponding constraint condition; Obtain historical service information for multiple stages of a target dual-end service; Substituting the historical service information into the first initial linear model, the second initial linear model and the third initial linear model respectively to obtain a first linear model, a second linear model and a third linear model, Based on the model coefficients of the first linear model, the second linear model and the third linear model, the degree of influence of this type of intervention features on the service order conversion is determined.

14. The determination device according to claim 9, characterized in that: The determination module is specifically used for: Constructing a first initial nonlinear model and a second initial nonlinear model, wherein the first initial nonlinear model represents the nonlinear relationship between the intervention feature and the corresponding constraint condition; and the second initial nonlinear model represents the nonlinear relationship between the intervention feature, the corresponding constraint condition and the order conversion index; Obtain historical service information for multiple stages of a target dual-end service; Substituting the historical service information into the first initial nonlinear model and the second initial nonlinear model respectively to obtain a first nonlinear model and a second nonlinear model; The degree of influence of such intervention features on service order conversion is determined based on the first nonlinear model and the second nonlinear model.

15. The determination device according to claim 9, characterized in that: The screening module is specifically used for: Sort the impact of different types of intervention features on service order conversion in order of magnitude; The intervention feature corresponding to the degree of impact under the specified ranking is selected as the target intervention feature.

16. The determination device according to claim 9, characterized in that: The order conversion indicator includes any one of the following items: user retention rate, user satisfaction and user net recommendation value.

17. An electronic device, characterized in that: include: A processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor communicates with the memory via the bus, and the processor executes the machine-readable instructions to perform the steps of the method described in any one of claims 1 to 8.

18. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are executed.

19. A computer program product comprising a computer program or instructions, characterized in that When the computer program or instruction is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images