A method, device, electronic device and storage medium for order distribution

Abstract

The present application provides a method, device, electronic device and storage medium for order allocation, wherein the method includes: receiving a service request sent by a service request end, and generating a current order based on the service request; extracting the current order content in the current order, and determining the conditional probability of each order category label appearing under the condition of the current order content appearing based on the current order content and a pre-built Bayesian network model; determining the order category label matching the current order according to the conditional probability corresponding to each order category label; determining the service provider matching the current order according to the order category label matching the current order, and allocating the current order to the service provider. Through the above method, the efficiency of order allocation and the utilization rate of service resources can be improved.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a method, device, electronic device and storage medium for order distribution. Background Art

[0002] In some service scenarios, the server in the service platform can receive a service request sent by a service requester, generate a service order based on the service request, and then assign the service order to a service provider for execution.

[0003] At present, when allocating service orders, the server often selects service providers that meet the preset order allocation conditions. For example, taking the service order as a ride order, the order allocation area can be determined according to the starting position recorded in the ride order, and the service providers in the order allocation area are regarded as service providers that meet the order allocation conditions. However, if the service providers that can dispatch orders in the current order allocation area are relatively scarce, while the service providers that can dispatch orders in other areas outside the order allocation area are relatively sufficient, in this case, if the above-mentioned order allocation method is adopted, on the one hand, the order allocation efficiency will be low due to the inability to allocate service orders in a timely manner, and on the other hand, the service resources of service providers in other areas will be wasted. Summary of the invention

[0004] In view of this, the purpose of the present application is to provide a method, device, electronic device and storage medium for order allocation to improve the efficiency of order allocation and the utilization rate of service resources.

[0005] In a first aspect, an embodiment of the present application provides a method for allocating an order, comprising:

[0006] Receive a service request sent by a service request end, and generate a current order based on the service request;

[0007] Extracting the current order content in the current order, and determining the conditional probability of each order category label appearing under the condition that the current order content appears based on the current order content and a pre-built Bayesian network model;

[0008] Determine the order category label that matches the current order according to the conditional probability corresponding to each order category label;

[0009] According to the order category tag matched by the current order, a service provider matching the current order is determined, and the current order is allocated to the service provider.

[0010] In a possible implementation, determining the order category label that matches the current order according to the conditional probability corresponding to each order category label includes:

[0011] The order category label with the highest conditional probability is selected as the order category label matched by the current order.

[0012] In a possible implementation, the Bayesian network model is constructed according to the following method:

[0013] Extract order content and corresponding order category label from each historical order contained in the historical order set;

[0014] Construct a Bayesian network model based on the extracted order content and the corresponding order category labels;

[0015] The Bayesian network model includes an n-level network, each node in each level of the network except the n-th level network is used to identify order contents belonging to the same type but with different values, and each node in the n-th level network is used to identify different order category labels; nodes in different levels of the network are connected by directed edges;

[0016] The p-th node in the n-th level network corresponds to at least one historical conditional probability, and the p-th node passes through at least one connected path formed by connecting nodes in different levels of networks; the historical conditional probability corresponding to the p-th node is used to represent the conditional probability of the order category label identified by the p-th node appearing when the order content identified by each node on the connected path formed between the p-th node appears; n is a positive integer greater than 1, and p is a positive integer.

[0017] In a possible implementation manner, the order content includes at least one of the following:

[0018] The identification information of the service requester, the time information of the service request, and the address information of the service request.

[0019] In a possible implementation manner, the order category tag is used to indicate the urgency of the service requested by the service requester.

[0020] In a possible implementation, the historical conditional probability corresponding to the p-th node is calculated according to the following method:

[0021] For each upper-level node on the connected path formed with the p-th node, a first joint probability between the order contents identified by each upper-level node is calculated, and a second joint probability between the order contents identified by each upper-level node and the order category label identified by the p-th node is calculated;

[0022] Based on the calculated first joint probability and the second joint probability, the historical conditional probability corresponding to the p-th node is calculated.

[0023] In a possible implementation, the first joint probability between the order contents identified by each upper-level node is calculated in the following manner:

[0024] Determine a first number of historical orders in the historical order set that contain order contents identified by each upper-level node;

[0025] The ratio of the first quantity to the total quantity of historical orders included in the historical order set is used as the first joint probability.

[0026] In a possible implementation, the second joint probability between the order content identified by each upper-level node and the order category label identified by the p-th node is calculated in the following manner:

[0027] Determine a second number of historical orders in the historical order set that contain order contents identified by each upper-level node and whose order category label is the order category label identified by the p-th node;

[0028] The ratio of the second number to the total number of historical orders included in the historical order set is used as the second joint probability.

[0029] In a possible implementation, after constructing the Bayesian network model according to the extracted order content and the corresponding order category label, the method further includes:

[0030] Receive a new historical order, and extract the order content and the corresponding order category label of the new historical order;

[0031] If any order content of the new historical orders is not included in the Bayesian network model, adding a node identifying the order content in the Bayesian network model;

[0032] Based on other order contents of the new historical order and the corresponding order category label, determining an associated node that has an associated relationship with the newly added node, and constructing a directed edge between the newly added node and the associated node in the Bayesian network model;

[0033] Adjust the historical conditional probabilities corresponding to the nodes in the n-th level network in the Bayesian network model.

[0034] In a possible implementation, after constructing a Bayesian network model based on the extracted order content and the corresponding order category label, the Bayesian network model is adjusted in the following manner:

[0035] Determine the number of other nodes connected to each node in each level of network except the nth level network;

[0036] Determine the weight of each node based on the number of other nodes it connects to;

[0037] Nodes whose weights do not meet the set conditions are screened out, and the directed edges between each screened node and other connected nodes are deleted.

[0038] In a second aspect, an embodiment of the present application provides an order allocation device, including:

[0039] A receiving module, used for receiving a service request sent by a service request end, and generating a current order based on the service request;

[0040] A first determination module is used to extract the current order content in the current order, and determine the conditional probability of each order category label appearing under the condition that the current order content appears based on the current order content and a pre-built Bayesian network model;

[0041] A second determination module, configured to determine the order category label matched by the current order according to the conditional probability corresponding to each order category label;

[0042] The order allocation module is used to determine the service provider that matches the current order according to the order category label that matches the current order, and allocate the current order to the service provider.

[0043] In a possible design, when the first determination module determines the order category label matching the current order according to the conditional probability corresponding to each order category label, it is specifically used to:

[0044] The order category label with the highest conditional probability is selected as the order category label matched by the current order.

[0045] In one possible design, the device further includes:

[0046] The model building module is used to build the Bayesian network model in the following manner:

[0047] Extract order content and corresponding order category label from each historical order contained in the historical order set;

[0048] Construct a Bayesian network model based on the extracted order content and the corresponding order category labels;

[0049] The Bayesian network model includes an n-level network, each node in each level of the network except the n-th level network is used to identify order contents belonging to the same type but with different values, and each node in the n-th level network is used to identify different order category labels; nodes in different levels of the network are connected by directed edges;

[0050] The p-th node in the n-th level network corresponds to at least one historical conditional probability, and the p-th node passes through at least one connected path formed by connecting nodes in different levels of networks; the historical conditional probability corresponding to the p-th node is used to represent the conditional probability of the order category label identified by the p-th node appearing when the order content identified by each node on the connected path formed between the p-th node appears; n is a positive integer greater than 1, and p is a positive integer.

[0051] In one possible design, the order content includes at least one of the following:

[0052] The identification information of the service requester, the time information of the service request, and the address information of the service request.

[0053] In one possible design, the order category tag is used to indicate the urgency of the service requested by the service requester.

[0054] In a possible design, the model building module, when calculating the historical conditional probability corresponding to the p-th node, is specifically used to:

[0055] For each upper-level node on the connected path formed with the p-th node, a first joint probability between the order contents identified by each upper-level node is calculated, and a second joint probability between the order contents identified by each upper-level node and the order category label identified by the p-th node is calculated;

[0056] Based on the calculated first joint probability and the second joint probability, the historical conditional probability corresponding to the p-th node is calculated.

[0057] In a possible design, the model building module, when calculating the first joint probability between the order contents identified by each upper-level node, is specifically used to:

[0058] Determine a first number of historical orders in the historical order set that contain order contents identified by each upper-level node;

[0059] The ratio of the first quantity to the total quantity of historical orders included in the historical order set is used as the first joint probability.

[0060] In a possible design, the model building module, when calculating the second joint probability between the order content identified by each upper-level node and the order category label identified by the p-th node, is specifically used to:

[0061] Determine a second number of historical orders in the historical order set that contain order contents identified by each upper-level node and whose order category label is the order category label identified by the p-th node;

[0062] The ratio of the second number to the total number of historical orders included in the historical order set is used as the second joint probability.

[0063] In a possible design, after constructing the Bayesian network model according to the extracted order content and the corresponding order category label, the model building module is further used to:

[0064] Receive a new historical order, and extract the order content and the corresponding order category label of the new historical order;

[0065] If any order content of the new historical orders is not included in the Bayesian network model, adding a node identifying the order content in the Bayesian network model;

[0066] Based on other order contents of the new historical order and the corresponding order category label, determining an associated node that has an associated relationship with the newly added node, and constructing a directed edge between the newly added node and the associated node in the Bayesian network model;

[0067] Adjust the historical conditional probabilities corresponding to the nodes in the n-th level network in the Bayesian network model.

[0068] In a possible design, after the model building module builds the Bayesian network model according to the extracted order content and the corresponding order category label, it adjusts the Bayesian network model according to the following method:

[0069] Determine the number of other nodes connected to each node in each level of network except the nth level network;

[0070] Determine the weight of each node based on the number of other nodes it connects to;

[0071] Nodes whose weights do not meet the set conditions are screened out, and the directed edges between each screened node and other connected nodes are deleted.

[0072] 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 when the machine-readable instructions are executed by the processor, the steps of the order allocation method described in the above-mentioned first aspect, or any possible implementation of the first aspect are performed.

[0073] 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 order allocation method described in the above-mentioned first aspect, or any possible implementation of the first aspect are executed.

[0074] In an embodiment of the present application, after receiving a service request sent by a service request end, a current order is generated based on the service request, and based on the current order content in the current order and a pre-built Bayesian network model, the conditional probability of each order category label appearing under the condition that the current order content appears is determined, and based on the conditional probability corresponding to each order, the service provider end that matches the current order is determined.

[0075] In this way, service providers can be screened according to the current order content and the pre-built Bayesian network model, thereby improving the efficiency of order allocation. Order allocation no longer relies solely on preset rules, allowing idle service resources to be utilized, thereby improving the utilization rate of service resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] 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.

[0077] Figure 1 A schematic diagram of the architecture of a service system provided by an embodiment of the present application is shown;

[0078] Figure 2 A schematic diagram of a process for constructing a Bayesian network model provided in an embodiment of the present application is shown;

[0079] Figure 3 A schematic diagram of the structure of a possible Bayesian network model provided in an embodiment of the present application is shown;

[0080] Figure 4 A schematic diagram of a process flow of a historical conditional probability calculation method provided by an embodiment of the present application is shown;

[0081] Figure 5 A schematic diagram for comparing a Bayesian network model provided in an embodiment of the present application is shown;

[0082] Figure 6 A schematic diagram of the structure of a possible Bayesian network model provided in an embodiment of the present application is shown;

[0083] Figure 7 A schematic diagram of the structure of a possible Bayesian network model provided in an embodiment of the present application is shown;

[0084] Figure 8 A schematic diagram of the structure of a possible Bayesian network model provided in an embodiment of the present application is shown;

[0085] Fig. 9 A schematic diagram of a flow chart of an order allocation method provided in an embodiment of the present application is shown;

[0086] Fig.10 A schematic diagram of the architecture of an order distribution device provided in an embodiment of the present application is shown;

[0087] Fig.11 A schematic structural diagram of an electronic device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0088] 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 can add one or more other operations to the flowchart under the guidance of the content of the present application, or remove one or more operations from the flowchart.

[0089] 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.

[0090] It should be noted that the term "comprising" will be used in the embodiments of the present application to indicate the existence of the features declared thereafter, but does not exclude the addition of other features.

[0091] First, the application scenarios to which this application is applicable are introduced. The method provided in this application is applicable to the scenario in which a service platform dispatches service orders.

[0092] The service requester sends a service request to the service platform. The service platform can generate a service order based on the service request and dispatch the service order to the service provider.

[0093] In the prior art, when dispatching service orders, screening is performed based on certain conditions. Taking the online car-hailing scenario as an example, when a passenger makes a car-hailing service request through an online car-hailing service platform, the online car-hailing service platform can generate a service order based on the service request, and then can screen the driver based on the passenger's starting position in the service order, and then dispatch the service order to the driver.

[0094] In actual applications, when selecting drivers, online ride-hailing service platforms generally select drivers who are within a preset distance from the starting location of the passenger in the service order. However, this situation is prone to the following two situations:

[0095] 1. There are few drivers whose distance to the passenger's starting position in the distance service order is within the preset distance, and the passenger's service order cannot be dispatched in time.

[0096] Second, there are many drivers whose distance from the starting location of the passengers in the service order exceeds the preset distance, and the drivers at these locations cannot receive enough service orders.

[0097] The above two situations will result in low efficiency in delivering service orders of the service platform on the one hand, and will also result in insufficient utilization of the service resources of the service platform on the other hand.

[0098] Based on this, the present application provides an order allocation method, which can screen service providers based on the order content of the service providers, thereby improving the order allocation efficiency and the utilization rate of service resources.

[0099] Embodiment 1

[0100] Figure 1 1 is a schematic diagram of the architecture of a service system 100 provided in an embodiment of the present application. For example, the service system 100 may be an online transportation service platform for transportation services such as taxis, chauffeur services, express trains, carpooling, bus services, driver rentals, or shuttle services, or any combination thereof. The service system 100 may include one or more of a server 110, a network 120, a service requester terminal 130, a service provider terminal 140, and a database 150.

[0101] In some embodiments, the server 110 may include a processor. The processor may process information and / or data related to the service request to perform one or more functions described in the present application. For example, the processor may generate a current order based on a service request obtained from the service requester terminal 130. In some embodiments, the processor may include one or more processing cores (e.g., a single-core processor (S) or a multi-core processor (S)). By way of example only, the processor may include a central processing unit (CPU), an application specific integrated circuit (ASIC), an application specific instruction set processor (ASIP), a graphics processing unit (GPU), a physical processing unit (PPU), a digital signal processor (DSP), a field programmable gate array (FPGA), a programmable logic device (PLD), a controller, a microcontroller unit, a reduced instruction set computer (RISC), or a microprocessor, or any combination thereof.

[0102] In some embodiments, the device types corresponding to the service requester terminal 130 and the service provider terminal 140 may be mobile devices, such as smart home devices, wearable devices, smart mobile devices, virtual reality devices, or augmented reality devices, etc., or they may be tablet computers, laptop computers, or built-in devices in motor vehicles, etc.

[0103] In some embodiments, the database 150 may be connected to the network 120 to communicate with one or more components in the service system 100 (e.g., the server 110, the service requester terminal 130, the service provider terminal 140, etc.). One or more components in the service system 100 may access data or instructions stored in the database 150 via the network 120. In some embodiments, the database 150 may be directly connected to one or more components in the service system 100, or the database 150 may also be part of the server 110.

[0104] The terms "passenger" and "service requester" in this application are used interchangeably to refer to an individual who can request or order a service. The terms "driver" and "service provider" in this application are used interchangeably to refer to an individual who can provide a service. The terms "service requester terminal" and "service request end" in this application are used interchangeably to refer to an entity or tool that requests a service. The terms "service provider terminal" and "service providing end" in this application are used interchangeably to refer to an entity or tool that provides a service.

[0105] The terms "service order" and "order" in this application are used interchangeably to refer to a request initiated by a passenger, a service requester, a driver, a service provider, etc., or any combination thereof. A service request may be charged or free.

[0106] The following is combined with the above Figure 1 The content described in the service system 100 shown is a detailed description of the order allocation method provided in the embodiment of the present application.

[0107] The method provided in the present application generates a current order based on the service request after receiving the service request sent by the service request end, and determines the conditional probability of each order category label appearing under the condition that the current order content appears based on the current order content and the pre-built Bayesian network model in the current order, and then determines the service provider that matches the current order based on the conditional probability corresponding to each order.

[0108] Embodiment 2

[0109] In this embodiment, the construction method of the Bayesian network model used in this application will be first introduced.

[0110] See also Figure 2 FIG. 1 is a flow chart of a Bayesian network model construction method provided in this embodiment. The method can be executed by the server 110 in the service system 100, and includes the following steps:

[0111] Step 201: extract order content and corresponding order category label from each historical order included in the historical order set.

[0112] The order content includes at least one of the following:

[0113] The identification information of the service requester, the time information of the service request, and the address information of the service request.

[0114] Taking the application scenario of online car-hailing as an example, the identification information of the service requester can be the identification information of the passenger (such as the passenger’s mobile phone number); the time information of the service request can be the time information when the passenger initiates the car-hailing request; the address information of the service request can be the starting location information and the end location information entered by the passenger.

[0115] The order category label of the above historical order is used to indicate the urgency of the service requested by the service requester.

[0116] In a possible implementation, the category label of the order may include different levels of emergency labels, such as the category label of the order may be level one emergency, level two emergency, level three emergency, etc. In practical applications, a suitable service provider may be selected for the service request end according to the emergency label of the order. The specific method for selecting the service provider will be described in Example 3 and will not be described here.

[0117] In another possible implementation, the category tag of the order may include a purpose tag for the service request end to send the service request, and the purpose for the service request end to send the service request may be to go home, go to a market, go to a company, etc. In practical applications, a suitable service provider end may be selected for the service request end according to the determined purposes for sending the service request by different service request ends. The specific method for selecting the service provider end will be described in Example 3, and will not be described here for the time being.

[0118] In an example of the present application, the order category label corresponding to each historical order in the historical order set can be manually marked, or after each historical order is completed, the historical order can be sent to the corresponding service request end, and the service request end can mark the historical order. The present application does not limit this.

[0119] In practical applications, considering that the method provided in this application is to construct a Bayesian network model based on the order content in historical orders, in order to ensure the diversity of sample data in constructing the Bayesian network model, the historical orders contained in the historical order set may be historical orders from different user terminals.

[0120] Step 202: construct a Bayesian network model based on the extracted order content and the corresponding order category label.

[0121] The Bayesian network model includes an n-level network, and each node in each level of the network except the n-th level network is used to identify order contents belonging to the same type but with different values. Each node in the n-th level network is used to identify different order category labels, and n is a positive integer greater than 1.

[0122] For example, see Figure 3As shown in the figure, it is a possible schematic diagram of the structure of a Bayesian network model. The Bayesian network model includes a three-level network model. In the third-level network, y1, y2, ..., y n Represent different order category labels, x1, x2, ..., x in the first-level network n They represent the order contents with different values ​​under the first type. In the second-level network, j1, j2, ..., j n They respectively represent the order contents with different values ​​under the second type.

[0123] In actual applications, the types of order content may include identification information of the service requester, time information of the service request, address information of the service request, etc.

[0124] In a possible implementation, the nodes contained in the networks at each level in the Bayesian network model can be connected through directed edges, there is no connection between the nodes contained in the networks at the same level, the nodes contained in the k-th level network can be connected to the nodes contained in the k+1-th level network, and can also be connected to the nodes contained in the k-1-th level network, a node in the k-th level network can be connected to one or more nodes in the k+1-th level network or the k-1-th level network, and there is no common content between the order contents identified by the nodes contained in each level of the network, where k is an integer greater than 1.

[0125] In the nth level network of the Bayesian network model, the pth node corresponds to at least one historical conditional probability, and the pth node passes through at least one connected path formed by connecting nodes in different levels of networks; the historical conditional probability corresponding to the pth node is used to represent the conditional probability of the order category label identified by the pth node appearing when the order content identified by each node on the connected path formed between the pth node appears; n is a positive integer greater than 1, and p is a positive integer.

[0126] For example, the pth node in the nth level network in the Bayesian network model is Figure 3 For example, the node y1 in the Bayesian network model has the following connected paths: x1→j1→y1, x2→j1→y1, …, x n →j1→y1, x1→j2→y1, x2→j2→y1,…,x n →j2→y1,x1→j n →y1,x2→j n →y1,…,x n → n →y1, and then based on the node identifiers on the different connected paths that node y1 passes through, the conditional probability of the order category label identified by node y1 in each connected path can be determined.

[0127] Specifically, when calculating the historical conditional probability corresponding to the pth node, we can use Figure 4 The calculation is performed according to the flowchart of the historical conditional probability calculation method shown in the figure, and the calculation includes the following steps:

[0128] Step 401: For each upper-level node on the connection path formed between the p-th node and the upper-level node, calculate the first joint probability between the order contents identified by each upper-level node.

[0129] Specifically, when calculating the first joint probability between the order contents identified by each upper-level node, the calculation may be performed in the following manner:

[0130] First, a first quantity of historical orders of order contents identified by each upper node in the historical order set is determined, and then a first joint probability is calculated according to the first quantity and the total quantity of historical orders in the historical order set.

[0131] In a possible implementation, the ratio between the first quantity corresponding to each node and the total quantity of historical orders included in the historical order set may be used as the occurrence probability of the order content identified by the node.

[0132] Exemplarily, if the connectivity paths corresponding to the order category label identified by node S1 include: connectivity path 1: a1→b2→c3→S1; connectivity path 2: a1→b1→c2→S1; connectivity path 3: a2→b2→c2→S1, the historical order set contains 100 historical orders, and the number of historical orders passing through the path a1→b2→c3 included in the historical order set is 10, the number of historical orders passing through the path a1→b1→c2 is 20, and the number of historical orders passing through the path a2→b2→c2 is 30, then the first joint probability of connectivity path 1 is 10÷100=0.1, the first joint probability of connectivity path 2 is 20÷100=0.2, and the first joint probability of connectivity path 3 is 30÷100=0.3.

[0133] Step 402: Calculate the second joint probability between the order content identified by each upper-level node and the order category label identified by the p-th node.

[0134] Specifically, when calculating the second joint probability between the order content identified by each upper node and the order category label of the p-th node, the calculation can be performed in the following manner:

[0135] First, determine the second number of historical orders in the historical order set that contain order contents identified by each upper node and whose order category label is the order category label identified by the p-th node, and then calculate the second joint probability based on the second number and the total number of historical orders included in the historical order set.

[0136] In a possible implementation, the ratio between the second quantity and the total quantity of historical orders included in the historical order set may be used as the second joint probability between the order content identified by each upper-level node and the order category label of the p-th node.

[0137] Exemplarily, if the connectivity paths corresponding to the order category labels identified by node S2 include: connectivity path 1: a1→b2→c3→S2; connectivity path 2: a1→b1→c2→S2; connectivity path 3: a2→b2→c2→S2, the historical order set contains 100 historical orders, and the total number of orders in the historical orders contained in the historical order set that pass through connectivity path 1 or connectivity path 2 or connectivity path 3 is 60, then the second joint probability corresponding to node S2 is determined to be 60÷100=0.6.

[0138] Step 403: Calculate the historical conditional probability corresponding to the p-th node based on the calculated first joint probability and the calculated second joint probability.

[0139] Specifically, when calculating the historical conditional probability corresponding to the p-th node based on the calculated first joint probability and the second joint probability, the ratio of the calculated first joint probability to the second joint probability can be used as the historical conditional probability corresponding to the p-th node.

[0140] Exemplarily, if the connectivity paths corresponding to the order category label identified by node S3 include: connectivity path 1: a1→b2→c3→S3; connectivity path 2: a1→b1→c2→S3; connectivity path 3: a2→b2→c2→S3, the historical order set contains 100 historical orders, and the number of historical orders passing through the path a1→b2→c3 in the historical order set is 10, the number of historical orders passing through the path a1→b1→c2 is 20, and the number of historical orders passing through the path a2→b2→c2 is 30, then the first joint probability of connectivity path 1 is 10÷100=0.1, the first joint probability of connectivity path 2 is 20÷100=0.2, the first joint probability of connectivity path 3 is 30÷100=0.3, and the second joint probability corresponding to node S3 is (10+20+30)÷100=0.6. At this time, the historical conditional probability of node S3 under connectivity path 1 is The historical conditional probability of node S3 under connected path 2 is The historical conditional probability of node S3 under connected path 3 is

[0141] Among them, the historical conditional probability of node S3 under connected path 1 indicates that the probability of the order category label marked by S3 appearing when the order content marked by a1, b2, and c3 is satisfied is: The historical conditional probability representation of node S3 under connected path 2 shows that the probability of the order category label marked by S3 appearing when the order content marked by a1, b1, and c2 is satisfied at the same time is: The historical conditional probability representation of node S3 under connected path 3 shows that the probability of the order category label marked by S3 appearing when the order content marked by a2, b2, and c2 is satisfied at the same time is:

[0142] Considering that the content of historical orders may be constantly updated, such as the address information of the requested service may be constantly updated, after constructing the Bayesian network model based on the extracted order content and the corresponding order category labels, the Bayesian network model can also be adjusted.

[0143] Specifically, after receiving a new historical order, the order content of the new historical order and the corresponding order category label can be extracted. If it is detected that any order content of the new historical order is not included in the Bayesian network model, a node representing the order content is added to the Bayesian network model, and then based on other order contents of the new historical order and the corresponding order category label, the associated node that has an association relationship with the newly added node is determined, and a directed edge between the newly added node and the associated node is constructed in the Bayesian network model, and the historical conditional probability corresponding to the node representing the order category label in the Bayesian network model is adjusted.

[0144] For example, see Figure 5 As shown in Figure 5 As shown in the network model on the left, after receiving the new historical order, the new historical order contains the order content j n+1 If it is not included in the Bayesian network model, you need to add node j to the constructed Bayesian network model. n+1 , and construct node j in the Bayesian network model n+1 Directed edges with associated nodes.

[0145] After the newly added node is added to the Bayesian network model, the total number of historical orders contained in the historical order combination changes, the occurrence probability of the order content identified by each node changes, and the occurrence probability of the order category label identified by each node also changes. Therefore, it is necessary to adjust the historical conditional probability corresponding to the last-level node.

[0146] In a possible implementation, after constructing the Bayesian network model based on the extracted order content and the corresponding order category labels, considering that different nodes have different roles in distinguishing order categories, the nodes of the Bayesian network model may also be adjusted.

[0147] Specifically, the number of other nodes connected to each node in each level of the network outside the nth level network (each node in the nth level network identifies the category label of the order) can be determined, and the weight of each node can be determined based on the number of other nodes connected to each node. Then, the nodes whose weights do not meet the set conditions are screened out, and the directed edges between each screened out node and other connected nodes are deleted.

[0148] Exemplarily, when determining the weight of each node based on the number of other nodes connected to each node, taking the determination of the weight of the nth node in the x-th level network as an example, the number c of directed edges connected to the nodes in the x-th level network in the x-1th level network can be determined first, and then the number d of directed edges connected to the nth node in the x-th level network in the x-1th level network can be determined, and then the ratio of the number d to the number c is calculated, and the calculated ratio is used as the weight of the nth node in the x-th level network.

[0149] For example, Figure 6 The number of directed edges connecting the nodes in the second-level network with the nodes in the third-level network is 8, and there are 2 directed edges connecting with node b4 in the third-level network, so the weight of node b4 is 2÷8=0.25.

[0150] In one example of the present application, when filtering out nodes whose weights do not meet set conditions based on the weights of the nodes, the nodes whose weights are not within the preset weight range can be determined as nodes that do not meet the set conditions, and the directed edges between the filtered out nodes and other connected nodes can be deleted.

[0151] In one possible application scenario, Figure 7 As shown in the figure, all nodes in the Bayesian network model are connected to one of the nodes. At this time, the node is not very useful for distinguishing order categories, so the node can be deleted in the Bayesian network model.

[0152] Below, the Bayesian network model construction method provided in this embodiment will be introduced in detail by taking the application scenario of online car-hailing orders as an example.

[0153] First, obtain a collection of online ride-hailing historical orders, which contains at least one online ride-hailing historical order. The content of the online ride-hailing historical order may include passenger identification information, booking time, starting location information, end location information, etc., as well as the order category label corresponding to each online ride-hailing order.

[0154] Then, a Bayesian network model is constructed based on the obtained historical online ride-hailing orders and the order category labels corresponding to each online ride-hailing order.

[0155] Specifically, we can first determine the order content type identified by each level of the network in the Bayesian network model, see Figure 8 As shown, it is a structural schematic diagram of a possible Bayesian network model provided in an embodiment of the present application, where the first-level network is used to identify the passenger's identification information, the second-level network is used to identify the booking time, the third-level network is used to identify the starting point location information, and the fourth-level network is used to identify the destination location information. Then, based on the content of the acquired online car-hailing historical orders and the category label corresponding to each online car-hailing order, the connection path corresponding to each online car-hailing historical order is determined.

[0156] If the order content of an online car-hailing order is {1, 12:00, place A, place B}, the order category label corresponding to the online car-hailing order is going home, the ath node in the first-level network in the Bayesian network model identifies the passenger's identification information as 1, the bth node in the second-level network in the Bayesian network model identifies the car-hailing time as 11:00-12:00, the cth node in the third-level network in the Bayesian network model identifies the starting location information as place A, the dth node in the fourth-level network in the Bayesian network model identifies the ending location information as place B, and the eth node in the fifth-level network in the Bayesian network model identifies the order category label as going home, then the connectivity path of the online car-hailing order is a→b→c→d→e.

[0157] After determining the connection path corresponding to each online ride-hailing historical order in the online ride-hailing historical order set according to the above method, the probability of the order category label corresponding to each connection path can be determined according to the number of online ride-hailing historical orders contained in each connection path.

[0158] In actual applications, if a new online car-hailing order is obtained, when it is necessary to determine the order category label of the newly obtained online car-hailing order, it is necessary to first determine the values ​​of the newly obtained online car-hailing order in the first-level network, second-level network, third-level network, and fourth-level network of the Bayesian network model, and then determine the connection path of the newly obtained online car-hailing order except the last-level network, and then determine the probability of each online car-hailing order category label under the connection path, and use the online car-hailing order category label with the required probability as the order category label of the newly obtained online car-hailing order.

[0159] The Bayesian network model constructed by the Bayesian network model construction method provided in this embodiment can judge the order category label corresponding to the order according to the order content when the order is received, and then select different service providers according to the different order category labels, thereby improving the efficiency of order allocation and the utilization of service resources.

[0160] Embodiment 3

[0161] Reference Fig. 9 FIG. 1 is a flow chart of a method for allocating orders provided in an embodiment of the present application, which comprises Figure 1 The server 110 in the service system shown executes the following steps:

[0162] Step 901: Receive a service request sent by a service request end, and generate a current order based on the service request.

[0163] In a specific implementation, when the service request end sends a service request, the service request may include the identification information of the service requester, the time information of the requested service, the starting location information of the requested service, the end location information of the requested service, etc. After receiving the service request sent by the service request end, the server may generate the current order based on the above content included in the service request.

[0164] Step 902: extract the current order content in the current order, and determine the conditional probability of each order category label appearing under the condition that the current order content appears based on the current order content and the pre-built Bayesian network model.

[0165] In a possible implementation, the current order content extracted from the current order may be input into a pre-built Bayesian network model, and the conditional probability of each order category label appearing under the condition that the current order content appears may be output.

[0166] Specifically, we can first determine the nodes corresponding to the current order content in each level of the network except the last level of the network in the pre-built Bayesian network model based on the current order content, and then determine the connection path corresponding to the current order content based on the nodes corresponding to the current order content in each level of the network, and then determine the conditional probability of each order category label under the connection path based on the connection path.

[0167] Step 903: Determine the order category label that matches the current order based on the conditional probability corresponding to each order category label.

[0168] In a possible implementation, the order category label with the highest conditional probability may be selected as the order category label matched by the current order.

[0169] Step 904: Determine the service provider that matches the current order based on the order category tag that matches the current order, and assign the current order to the service provider.

[0170] In a specific implementation, different service provider screening strategies can be configured for different order category tags. For example, if the order category tags include a first-level emergency tag, a second-level emergency tag, and a third-level emergency tag, the higher the emergency tag level, the higher the urgency. For orders with a third-level emergency tag, a screening strategy with looser screening conditions can be configured for this type of order to achieve rapid allocation of orders of this urgency. For orders with a first-level emergency tag, since the service requester is not particularly eager for order allocation, a screening strategy with stricter screening conditions can be configured for this type of order.

[0171] Taking the application scenario of online car-hailing as an example, in the scenario of online car-hailing, order category tags may include going home, going to work, entertainment, etc. In a possible implementation, if after receiving the service request of the passenger, it is determined that the order category tag of the order corresponding to the passenger is going home, then an online car-hailing driver can be selected for the passenger within a first preset distance range from the passenger's current location; if it is determined that the order category tag of the order corresponding to the passenger is going to work, then an online car-hailing driver can be selected for the passenger within a second preset distance range from the passenger's current location, wherein the first preset distance is less than the second preset distance.

[0172] In another possible implementation, different types of online taxi drivers can be assigned to orders with different order category labels. For example, in the passenger's current location area, there are fewer taxi drivers and more private car drivers. For the same driving distance, the fees charged by private car drivers are higher than those of taxi drivers. If the order category label corresponding to the passenger's order is going to work, the order is an emergency order. At this time, when screening online taxi drivers for the passenger, not only taxi drivers can be selected, but taxi drivers and private car drivers can also be selected at the same time.

[0173] In specific applications, when obtaining the current location of the passenger, it can be based on the Global Positioning System (GPS), Global Navigation Satellite System (GLONASS), Compass Navigation System (COMPASS), Galileo Positioning System, Quasi-Zenith Satellite System (QZSS), Wireless Fidelity (WiFi) positioning technology, etc., or any combination thereof.

[0174] In this way, service providers can be screened according to the current order content and the pre-built Bayesian network model, thereby improving the efficiency of order allocation. Order allocation no longer relies solely on preset rules, allowing idle service resources to be utilized, thereby improving the utilization rate of service resources.

[0175] Embodiment 4

[0176] This embodiment also provides an order distribution device, see Fig.10 As shown, it is a schematic diagram of the architecture of the order allocation device provided in this embodiment, including a receiving module 1001, a first determining module 1002, a second determining module 1003, an order allocation module 1004, and a model building module 1005. Specifically:

[0177] The receiving module 1001 is used to receive a service request sent by a service request end, and generate a current order based on the service request;

[0178] The first determination module 1002 is used to extract the current order content in the current order, and determine the conditional probability of each order category label appearing under the condition that the current order content appears based on the current order content and the pre-built Bayesian network model;

[0179] The second determination module 1003 is used to determine the order category label matched by the current order according to the conditional probability corresponding to each order category label;

[0180] The order allocation module 1004 is used to determine the service provider that matches the current order according to the order category tag that matches the current order, and allocate the current order to the service provider.

[0181] In a possible design, the first determination module 1002, when determining the order category label matched by the current order according to the conditional probability corresponding to each order category label, is specifically configured to:

[0182] The order category label with the highest conditional probability is selected as the order category label matched by the current order.

[0183] In a possible design, the model building module 1005 is used to build the Bayesian network model according to the following method:

[0184] Extract order content and corresponding order category label from each historical order contained in the historical order set;

[0185] Construct a Bayesian network model based on the extracted order content and the corresponding order category labels;

[0186] The Bayesian network model includes an n-level network, each node in each level of the network except the n-th level network is used to identify order contents belonging to the same type but with different values, and each node in the n-th level network is used to identify different order category labels; nodes in different levels of the network are connected by directed edges;

[0187] The p-th node in the n-th level network corresponds to at least one historical conditional probability, and the p-th node passes through at least one connected path formed by connecting nodes in different levels of networks; the historical conditional probability corresponding to the p-th node is used to represent the conditional probability of the order category label identified by the p-th node appearing when the order content identified by each node on the connected path formed between the p-th node appears; n is a positive integer greater than 1, and p is a positive integer.

[0188] In a possible design, the order content includes at least one of the following:

[0189] The identification information of the service requester, the time information of the service request, and the address information of the service request.

[0190] In a possible design, the order category tag is used to indicate the urgency of the service requested by the service requester.

[0191] In a possible design, the model building module 1005, when calculating the historical conditional probability corresponding to the p-th node, is specifically used to:

[0192] For each upper-level node on the connected path formed with the p-th node, a first joint probability between the order contents identified by each upper-level node is calculated, and a second joint probability between the order contents identified by each upper-level node and the order category label identified by the p-th node is calculated;

[0193] Based on the calculated first joint probability and the second joint probability, the historical conditional probability corresponding to the p-th node is calculated.

[0194] In a possible design, the model building module 1005, when calculating the first joint probability between the order contents identified by each upper-level node, is specifically used to:

[0195] Determine a first number of historical orders in the historical order set that contain order contents identified by each upper-level node;

[0196] The ratio of the first quantity to the total quantity of historical orders included in the historical order set is used as the first joint probability.

[0197] In a possible design, the model building module 1005, when calculating the second joint probability between the order content identified by each upper-level node and the order category label identified by the p-th node, is specifically used to:

[0198] Determine a second number of historical orders in the historical order set that contain order contents identified by each upper-level node and whose order category label is the order category label identified by the p-th node;

[0199] The ratio of the second number to the total number of historical orders included in the historical order set is used as the second joint probability.

[0200] In a possible design, after constructing the Bayesian network model according to the extracted order content and the corresponding order category label, the model building module 1005 is further used to:

[0201] Receive a new historical order, and extract the order content and the corresponding order category label of the new historical order;

[0202] If any order content of the new historical orders is not included in the Bayesian network model, adding a node identifying the order content in the Bayesian network model;

[0203] Based on other order contents of the new historical order and the corresponding order category label, determining an associated node that has an associated relationship with the newly added node, and constructing a directed edge between the newly added node and the associated node in the Bayesian network model;

[0204] Adjust the historical conditional probabilities corresponding to the nodes in the n-th level network in the Bayesian network model.

[0205] In a possible design, after constructing the Bayesian network model according to the extracted order content and the corresponding order category label, the model construction module 1005 adjusts the Bayesian network model according to the following method:

[0206] Determine the number of other nodes connected to each node in each level of network except the nth level network;

[0207] Determine the weight of each node based on the number of other nodes it connects to;

[0208] Nodes whose weights do not meet the set conditions are screened out, and the directed edges between each screened node and other connected nodes are deleted.

[0209] Through this device, service providers can be screened according to the current order content and the pre-built Bayesian network model, thereby improving the efficiency of order allocation. Order allocation no longer relies solely on preset rules, allowing idle service resources to be utilized, thereby improving the utilization rate of service resources.

[0210] Embodiment 5

[0211] Based on the same technical concept, the embodiment of the present application also provides an electronic device. Fig.11 As shown, it is a schematic diagram of the structure of the electronic device 1100 provided in the embodiment of the present application, including a processor 1101, a memory 1102, and a bus 1103. Among them, the memory 1102 is used to store execution instructions, including a memory 11021 and an external memory 11022; the memory 11021 here is also called an internal memory, which is used to temporarily store the operation data in the processor 1101, and the data exchanged with the external memory 11022 such as a hard disk. The processor 1101 exchanges data with the external memory 11022 through the memory 11021. When the electronic device 1100 is running, the processor 1101 communicates with the memory 1102 through the bus 1103, so that the processor 1101 executes the following instructions:

[0212] Receive a service request sent by a service request end, and generate a current order based on the service request;

[0213] Extracting the current order content in the current order, and determining the conditional probability of each order category label appearing under the condition that the current order content appears based on the current order content and a pre-built Bayesian network model;

[0214] Determine the order category label that matches the current order according to the conditional probability corresponding to each order category label;

[0215] According to the order category tag matched by the current order, a service provider matching the current order is determined, and the current order is allocated to the service provider.

[0216] The specific processing process performed by the processor 1101 can be found in the description of the above method embodiment, which will not be elaborated here.

[0217] An embodiment of the present application also 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 above-mentioned order allocation method are executed.

[0218] 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, it can execute the above-mentioned order allocation method, thereby improving the efficiency of order allocation and the utilization rate of service resources.

[0219] The computer program product of the order allocation method provided in the embodiment of the present application includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the method in the previous method embodiment. The specific implementation can be found in the method embodiment, which will not be repeated here.

[0220] Those skilled in the art can clearly understand that, for the convenience and simplicity of description, the specific working process of the system and device described above can refer to the corresponding process in the method embodiment, and will not be repeated in this application. In the several embodiments provided in this application, it should be understood that the disclosed system, device and method can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the modules is only a logical function division. There may be other division methods in actual implementation. For example, multiple modules 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, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0221] The modules described as separate components may or may not be physically separated, and the components shown as modules 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.

[0222] 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.

[0223] 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, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server, or 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 USB flash drives, mobile hard drives, ROM, RAM, magnetic disks, or optical disks.

[0224] The above are only specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A device for order distribution, characterized in that: include: A receiving module, used for receiving a service request sent by a service request end, and generating a current order based on the service request; A first determination module is used to extract the current order content in the current order, and determine the conditional probability of each order category label appearing under the condition that the current order content appears based on the current order content and a pre-built Bayesian network model; wherein the order content includes at least one of the following: identification information of the service requester, time information of the service request, and address information of the service request; A second determination module, configured to determine the order category label matched by the current order according to the conditional probability corresponding to each order category label; An order allocation module, used to determine a service provider matching the current order according to an order category tag matching the current order, and allocate the current order to the service provider; The model building module is used to build the Bayesian network model in the following manner: Extract order content and corresponding order category label from each historical order contained in the historical order set; Construct a Bayesian network model based on the extracted order content and the corresponding order category labels; The Bayesian network model includes an n-level network, each node in each level of the network except the n-th level network is used to identify order contents belonging to the same type but with different values, and each node in the n-th level network is used to identify different order category labels; nodes in different levels of the network are connected by directed edges; The p-th node in the n-th level network corresponds to at least one historical conditional probability, and the p-th node passes through at least one connected path formed by connecting nodes in different levels of networks; the historical conditional probability corresponding to the p-th node is used to represent the conditional probability of the order category label identified by the p-th node appearing when the order content identified by each node on the connected path formed between the p-th node appears; n is a positive integer greater than 1, and p is a positive integer.

2. The device according to claim 1, characterized in that The first determination module, when determining the order category label matched by the current order according to the conditional probability corresponding to each order category label, is specifically used to: The order category label with the highest conditional probability is selected as the order category label matched by the current order.

3. The device according to claim 1, characterized in that The order category tag is used to indicate the urgency of the service requested by the service requester.

4. The device according to claim 1, characterized in that The model building module, when calculating the historical conditional probability corresponding to the p-th node, is specifically used to: For each upper-level node on the connected path formed with the p-th node, a first joint probability between the order contents identified by each upper-level node is calculated, and a second joint probability between the order contents identified by each upper-level node and the order category label identified by the p-th node is calculated; Based on the calculated first joint probability and the second joint probability, the historical conditional probability corresponding to the p-th node is calculated.

5. The device according to claim 1, characterized in that The model building module, after building the Bayesian network model according to the extracted order content and the corresponding order category label, is further used to: Receive a new historical order, and extract the order content and the corresponding order category label of the new historical order; If any order content of the new historical orders is not included in the Bayesian network model, adding a node identifying the order content in the Bayesian network model; Based on other order contents of the new historical order and the corresponding order category label, determining an associated node that has an associated relationship with the newly added node, and constructing a directed edge between the newly added node and the associated node in the Bayesian network model; Adjust the historical conditional probabilities corresponding to the nodes in the n-th level network in the Bayesian network model.

6. The device according to claim 1, characterized in that The model building module, after building the Bayesian network model according to the extracted order content and the corresponding order category label, adjusts the Bayesian network model according to the following method: Determine the number of other nodes connected to each node in each level of network except the nth level network; Determine the weight of each node based on the number of other nodes it connects to; Nodes whose weights do not meet the set conditions are screened out, and the directed edges between each screened node and other connected nodes are deleted.

7. A method for order allocation, characterized in that: include: Receive a service request sent by a service request end, and generate a current order based on the service request; Extract the current order content in the current order, and determine the conditional probability of each order category label appearing under the condition that the current order content appears based on the current order content and the pre-built Bayesian network model; wherein the order content includes at least one of the following: identification information of the service requester, time information of the service request, address information of the service request Determine the order category label that matches the current order according to the conditional probability corresponding to each order category label; Determine a service provider that matches the current order according to the order category tag that matches the current order, and assign the current order to the service provider; The model building module is used to build the Bayesian network model in the following manner: Extract order content and corresponding order category label from each historical order contained in the historical order set; Construct a Bayesian network model based on the extracted order content and the corresponding order category labels; The Bayesian network model includes an n-level network, each node in each level of the network except the n-th level network is used to identify order contents belonging to the same type but with different values, and each node in the n-th level network is used to identify different order category labels; nodes in different levels of the network are connected by directed edges; The p-th node in the n-th level network corresponds to at least one historical conditional probability, and the p-th node passes through at least one connected path formed by connecting nodes in different levels of networks; the historical conditional probability corresponding to the p-th node is used to represent the conditional probability of the order category label identified by the p-th node appearing when the order content identified by each node on the connected path formed between the p-th node appears; n is a positive integer greater than 1, and p is a positive integer.

8. An electronic device, characterized in that: include: A processor, a storage medium and a bus, wherein the storage medium stores machine-readable instructions executable by the processor. When the electronic device is running, the processor and the storage medium communicate via the bus, and the processor executes the machine-readable instructions to perform the steps of the order allocation method as described in claim 7.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which, when executed by a processor, executes the steps of the order allocation method as claimed in claim 7.

Images