Order processing method and device, electronic equipment and storage medium

Abstract

The invention provides an order processing method and device, electronic equipment and a storage medium. The method comprises the following steps: receiving a payment instruction for a target order, and obtaining order environment data of multiple modes associated with the target order; searching a target order scene in a plurality of predetermined order scenes according to the order environment data of the plurality of modes; calling a plurality of order intelligent agents to perform order analysis based on the target order scene to obtain a plurality of order analysis results; acquiring historical order payment data of an order initiation object of the target order, and generating a target order payment scheme according to the historical order payment data and the plurality of order analysis results; and recommending the target order payment scheme to the order initiation object. According to the embodiment of the invention, the accuracy of recommending the order payment scheme to the object can be improved.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of data processing, and in particular, to an order processing method and device, electronic equipment and storage medium. BACKGROUND

[0002] With the development of technology, more and more people tend to use mobile devices to pay orders online when purchasing goods or services. Through online payment platforms, order processing efficiency can be greatly improved, and economic development is promoted.

[0003] In order to meet people's demand for online order processing, more and more platforms begin to support online payment, and also provide people with a variety of payment methods. In this way, people can choose one from a variety of order payment schemes. The selection of order payment schemes is related to resource utilization and resource security, etc.

[0004] However, in the related art, the accuracy of recommending order payment schemes to objects is low. SUMMARY

[0005] The embodiments of the present disclosure provide an order processing method and device, electronic equipment and storage medium, which can improve the accuracy of recommending order payment schemes to objects.

[0006] According to an aspect of the present disclosure, an order processing method is provided, comprising:

[0007] receiving a payment instruction for a target order, and obtaining order environment data of multiple modalities associated with the target order;

[0008] finding a target order scene in a plurality of predetermined order scenes according to the order environment data of the multiple modalities;

[0009] calling a plurality of order agents for order analysis based on the target order scene, and obtaining a plurality of order analysis results;

[0010] obtaining historical order payment data of an order initiating object of the target order, and generating a target order payment scheme according to the historical order payment data and the plurality of order analysis results;

[0011] recommending the target order payment scheme to the order initiating object.

[0012] According to an aspect of the present disclosure, an order processing device is provided, comprising:

[0013] an obtaining unit configured to receive a payment instruction for a target order, and obtain order environment data of multiple modalities associated with the target order;

[0014] The searching unit is configured to search for a target order scenario from a plurality of predetermined order scenarios according to the order environment data of the plurality of modalities.

[0015] The analyzing unit is configured to invoke a plurality of order agents for order analysis based on the target order scenario, to obtain a plurality of order analysis results.

[0016] The generating unit is configured to obtain historical order payment data of an order initiator of the target order, and generate a target order payment scheme according to the historical order payment data and the plurality of order analysis results.

[0017] The recommending unit is configured to recommend the target order payment scheme to the order initiator.

[0018] Optionally, in an embodiment, the searching unit is specifically configured to:

[0019] Obtain a multi-modality fusion model.

[0020] Perform data analysis on the order environment data of the plurality of modalities based on the multi-modality fusion model, and search for a target order scenario from a plurality of predetermined order scenarios.

[0021] Optionally, in an embodiment, the searching unit is specifically configured to:

[0022] Obtain order category information of the target order, and search for a plurality of target modalities for data analysis based on the order category information.

[0023] Perform data analysis on the order environment data of the plurality of target modalities based on the multi-modality fusion model, and search for a target order scenario from a plurality of predetermined order scenarios.

[0024] Optionally, in an embodiment, the searching unit is specifically configured to:

[0025] Obtain a plurality of historical order environment data in a predetermined time period.

[0026] Search for a target order scenario from a plurality of predetermined order scenarios according to the order environment data of the plurality of modalities and historical order scenarios corresponding to the plurality of historical order environment data.

[0027] Optionally, in an embodiment, the searching unit is specifically configured to:

[0028] Query, from the plurality of historical order environment data, reference order environment data that has a matching degree reaching a predetermined condition with the order environment data of the plurality of modalities.

[0029] Search for a target order scenario from a plurality of predetermined order scenarios based on the reference order environment data.

[0030] Optionally, in an implementation, the acquisition unit is specifically configured to:

[0031] acquire order environment data of multiple modalities associated with the target order and historical operation data of an order initiator of the target order;

[0032] The search unit is specifically configured to:

[0033] search for a target order scenario from among a plurality of predetermined order scenarios according to the order environment data of the multiple modalities and the historical operation data.

[0034] Optionally, in an implementation, the search unit is specifically configured to:

[0035] query target operation data related to the order environment data from among the historical operation data;

[0036] perform environment analysis on the order environment data of the multiple modalities and the target operation data based on a first neural network, and search for a target order scenario from among a plurality of predetermined order scenarios.

[0037] Optionally, in an implementation, the analysis unit is specifically configured to:

[0038] invoke a financial analysis agent, a product analysis agent, a security analysis agent, and a payment method analysis agent based on the target order scenario to perform order analysis, and obtain a financial analysis result, a product analysis result, a security analysis result, and a payment method analysis result;

[0039] The generation unit is specifically configured to:

[0040] acquire historical order payment data of an order initiator of the target order, and generate a target order payment scheme according to the historical order payment data, the financial analysis result, the product analysis result, the security analysis result, and the payment method analysis result.

[0041] Optionally, in an implementation, the financial analysis result includes first recommendable scores corresponding to a plurality of order payment schemes, the product analysis result includes second recommendable scores corresponding to the plurality of order payment schemes, the security analysis result includes third recommendable scores corresponding to the plurality of order payment schemes, and the payment method analysis result includes fourth recommendable scores corresponding to the plurality of order payment schemes.

[0042] The generation unit is specifically configured to:

[0043] acquire usage scores corresponding to the plurality of order payment schemes based on the historical order payment data;

[0044] The target order payment scheme is generated according to the use score, the first recommendable score, the second recommendable score, the third recommendable score and the fourth recommendable score corresponding to the plurality of order payment schemes.

[0045] Optionally, in an embodiment, the generation unit is specifically configured to:

[0046] obtain a first weight corresponding to the historical order payment data, a second weight corresponding to the financial analysis agent, a third weight corresponding to the product analysis agent, a fourth weight corresponding to the security analysis agent and a fifth weight corresponding to the payment method analysis agent;

[0047] The target order payment scheme is generated from the plurality of order payment schemes by performing weighted calculation on the use score, the first recommendable score, the second recommendable score, the third recommendable score and the fourth recommendable score based on the first weight, the second weight, the third weight, the fourth weight and the fifth weight.

[0048] Optionally, in an embodiment, the generation unit is specifically configured to:

[0049] obtain a second neural network;

[0050] The first weight corresponding to the historical order payment data, the second weight corresponding to the financial analysis agent, the third weight corresponding to the product analysis agent, the fourth weight corresponding to the security analysis agent and the fifth weight corresponding to the payment method analysis agent are generated by performing environment analysis on the order environment data of the plurality of modalities based on the second neural network.

[0051] Optionally, in an embodiment, the analysis unit is specifically configured to:

[0052] a plurality of target order agents are found from a plurality of order agents based on the target order scenario;

[0053] The target order scenario is analyzed by using the plurality of target order agents to obtain a plurality of order analysis results.

[0054] Optionally, in an embodiment, the order processing apparatus further comprises:

[0055] a first display unit configured to display an order payment interface in response to the target order payment scheme recommended to the order initiating object, and display a payment method selection area in the order payment interface, wherein the payment method selection area includes the target order payment scheme;

[0056] The second display unit is configured to display an order payment result in response to a triggering operation on a selected payment scheme in the payment scheme selection area.

[0057] Optionally, in an embodiment, the first display unit is specifically configured to:

[0058] display an order payment interface in response to the target order payment scheme recommended to the order initiator, and display a payment scheme selection area and a payment scenario analysis area in the order payment interface, wherein the payment scenario analysis area comprises a payment adjustment control;

[0059] jump to a payment adjustment pop-up window in response to a triggering operation on the payment adjustment control.

[0060] Optionally, in an embodiment, the second display unit is specifically configured to:

[0061] display a payment result display interface in response to a triggering operation on a selected payment scheme in the payment scheme selection area, wherein the payment result display interface comprises an order payment result and an order payment analysis conclusion.

[0062] In the order processing method of the embodiments of the present disclosure, a payment instruction for a target order is received, and order environment data of multiple modalities associated with the target order is obtained; the target order scene is searched for among multiple predetermined order scenes according to the order environment data of the multiple modalities; the multiple order agents are called based on the target order scene to perform order analysis, and multiple order analysis results are obtained; historical order payment data of an order initiator of the target order is obtained, and a target order payment scheme is generated according to the historical order payment data and the multiple order analysis results; and the target order payment scheme is recommended to the order initiator.

[0063] In this way, in the embodiments of the present disclosure, after receiving the payment instruction for the target order, the target order scene can be searched for based on the order environment data of multiple modalities. The difference in the target order scene will affect the selection of the order payment scheme. After performing order analysis on the multiple order agents corresponding to the target order scene, the target order payment scheme is generated based on the historical order payment data of the order initiator and the order analysis results. That is, the target order payment scheme recommended to the order initiator in the embodiments of the present disclosure is determined based on the scene information of the target order and the historical payment preferences of the object, so that the target order payment scheme can better meet the needs of the order payment scene, and the accuracy of recommending the order payment scheme to the object is improved.

[0064] Additional features and advantages of the present disclosure will be set forth in the description that follows, and in part will be apparent from the description, or can be learned by practice of the present disclosure. The objectives and other advantages of the present disclosure will be realized and attained by the structure particularly pointed out in the description and claims. BRIEF DESCRIPTION OF DRAWINGS

[0065] The accompanying drawings are included to provide a further understanding of the technical solutions of the present disclosure, and constitute a part of the specification, and are used to explain the technical solutions of the present disclosure together with the embodiments of the present disclosure, and do not constitute a limitation to the technical solutions of the present disclosure.

[0066] Figure 1 is a framework diagram of a system to which an order processing method according to an embodiment of the present disclosure is applied;

[0067] Figure 2A is a schematic diagram of an online order payment scenario to which an embodiment of the present disclosure is applied;

[0068] Figure 2B is another schematic diagram of an online order payment scenario to which an embodiment of the present disclosure is applied;

[0069] Figure 2C is another schematic diagram of an online order payment scenario to which an embodiment of the present disclosure is applied;

[0070] Figure 3 is a flowchart of an order processing method according to an embodiment of the present disclosure;

[0071] Figure 4 is a schematic diagram of calling a financial analysis agent, a product analysis agent, a security analysis agent, and a payment method analysis agent for parallel order analysis in an embodiment of the present disclosure;

[0072] Figure 5 is a schematic diagram of generating a target order payment scheme by using a multi-objective optimization algorithm in an embodiment of the present disclosure;

[0073] Figure 6 is a schematic diagram of selecting an order payment scheme by an order initiator in an embodiment of the present disclosure;

[0074] Figure 7 is a schematic diagram of displaying a target order payment scheme on an order payment interface in an embodiment of the present disclosure;

[0075] Figure 8A is a schematic diagram of displaying a payment method selection area and a payment scenario analysis area on an order payment interface in an embodiment of the present disclosure;

[0076] Figure 8Bis a schematic diagram of displaying a payment adjustment pop-up window in an order payment interface according to an embodiment of the present disclosure;

[0077] Figure 9 is a schematic diagram of displaying an order payment result and an order payment analysis conclusion in a payment result display interface according to an embodiment of the present disclosure;

[0078] Figure 10 is a system architecture diagram of an order processing method according to an embodiment of the present disclosure;

[0079] Figure 11 is another flowchart of an order processing method according to an embodiment of the present disclosure;

[0080] Figure 12 is a multi-terminal interaction diagram of an order processing method according to an embodiment of the present disclosure;

[0081] Figure 13 is a structural schematic diagram of an order processing device according to an embodiment of the present disclosure;

[0082] Figure 14 is a terminal structure diagram for implementing methods according to an embodiment of the present disclosure;

[0083] Figure 15 is a server structure diagram for implementing methods according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0084] In order to make the objectives, technical solutions and advantages of the present disclosure clearer, the present disclosure will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present disclosure and do not limit the present disclosure.

[0085] Before the present disclosure is further described in detail, the terms and phrases involved in the present disclosure are explained, and the terms and phrases involved in the present disclosure are applicable to the following explanations:

[0086] Transformers: a deep learning architecture based on self-attention mechanism, composed of an encoder and a decoder stack. The encoder can dynamically capture the dependency of positions in the sequence (such as the association between words in a sentence) through a multi-head self-attention layer, and then further extract features through a feedforward neural network, and assist with residual connection and normalization to stabilize training; the decoder adds a self-attention layer and an encoder-decoder attention layer to the encoder to gradually generate the target sequence. Therefore, transformers can be used for fusion analysis of long-distance time series data, which is conducive to accurately fusing the dependency between long-distance data.

[0087] Bidirectional Encoder Representations from Transformers (BERT): BERT uses a bidirectional Transformer encoder that can capture both the left and right context of a text, enabling deep understanding of the text content.

[0088] Convolutional Neural Network (CNN): CNN is a deep learning model designed specifically for processing grid-like data such as images, videos, and speech. It simulates the hierarchical processing mechanism of the biological visual system to achieve efficient feature extraction and pattern recognition.

[0089] Multi-objective optimization algorithm: Multi-objective optimization algorithm is a method used to solve decision-making problems with multiple conflicting objectives. Its core is to find a set of trade-off solutions, rather than a single optimal solution for a single objective.

[0090] In the related art, when receiving an order payment instruction, a payment scheme recommendation rule can be used to recommend a payment scheme to the target object. For example, the most commonly used payment scheme is recommended to the target object, or the newly developed payment function of the order processing platform is recommended. In this way, the payment scheme recommended to the target object may not be the most suitable for the target object and the order payment scene, and using this payment scheme for order payment may result in resource loss and affect the payment experience of the target object. To this end, the present disclosure provides an order processing method to improve the accuracy of recommending an order payment scheme to an object.

[0091] System architecture and scenario of the present disclosure

[0092] Figure 1 It is a system architecture diagram applied by the order processing method according to the embodiments of the present disclosure. It includes terminal 140, Internet 130, gateway 120, server 110, etc.

[0093] The terminal 140 includes desktop computers, laptops, PDAs (Personal Digital Assistants), mobile phones, car-mounted terminals, home theater terminals, special-purpose terminals, and other forms. In addition, it can be a single device or a collection of multiple devices. For example, multiple devices are connected through a local area network and share a display device for collaborative work, collectively forming a terminal 140. The terminal 140 can also communicate with the Internet 130 in a wired or wireless manner to exchange data.

[0094] The server 110 refers to a computer system capable of providing certain services to the terminal 140. Compared with the ordinary terminal 140, the server 110 has high requirements in stability, security, performance, and the like. The server 110 can be a high-performance computer in a network platform, a cluster of multiple high-performance computers, a part of a high-performance computer (for example, a machine), a combination of parts of multiple high-performance computers (for example, machines), or the like.

[0095] The gateway 120 is also referred to as an inter-network connector or a protocol converter. The gateway implements network interconnection at the transport layer and is a computer system or device that acts as a conversion function. The gateway is a translator between two systems using different communication protocols, data formats or languages, or even having completely different architectures. Meanwhile, the gateway can also provide filtering and security functions. The message sent by the terminal 140 to the server 110 is sent to the corresponding server 110 through the gateway 120. The message sent by the server 110 to the terminal 140 is also sent to the corresponding terminal 140 through the gateway 120.

[0096] The order processing method of the embodiment of the present disclosure can be executed by an electronic device, which can be the terminal 140 or the server 110, that is, the order processing method can be executed by the terminal 140 or the server 110, or can be executed by the terminal 140 and the server 110 together.

[0097] The embodiment of the present disclosure can be applied in multiple scenarios, for example Figures 2A-2B the online order payment scenario shown in the figure.

[0098] As shown in the figure, Figure 2A the product detail page 210 displayed on the e-commerce platform displays the product name "product A1", displays the order payee, that is, the product provider "order payee B1", displays the product detail figure and product description, displays the required resource amount of product A1 as 100, and displays the payment confirmation control 211.

[0099] After clicking the payment confirmation control 211, the e-commerce platform displays the order payment interface 220, as shown in the figure. Figure 2B The order payment interface 220 displays the order details information such as the order payee name, the product name, and the payable resource amount, so as to be confirmed by the payment object. The order payment interface 220 also displays the payment scheme selection area 221, which displays multiple payment schemes, specifically including scheme M1, scheme M2, scheme M3, and scheme M4. The scheme M1 is determined by the order processing method of the embodiment of the present disclosure and is the payment scheme preferentially recommended to the payment object. The schemes M2, M3, and M4 are other selectable payment schemes recommended to the payment object. The payment object can click the payment control 222 to perform order payment after selecting the scheme M1 based on the recommended scheme.

[0100] After the order payment is successful, a payment result feedback interface 230 can be displayed as shown in Figure 2C . The payment result feedback interface 230 displays “payment success”, and displays the order payee name, product name, and payment resource amount, so that the payment object can confirm the payment result again.

[0101] Therefore, the order processing method provided by the embodiment of the present disclosure can accurately recommend a payment scheme to the payment object, so that the payment object can quickly obtain the most suitable payment scheme in the online order payment scenario.

[0102] Overall description of the embodiment of the present disclosure

[0103] According to an embodiment of the present disclosure, an order processing method is provided. The order processing method provided by the embodiment of the present disclosure can be applied in online order payment and other scenarios. As shown in Figure 3 , a flowchart of the order processing method provided by the present disclosure. The order processing method can include:

[0104] Step 310, receiving a payment instruction for a target order, and obtaining order environment data of multiple modalities associated with the target order.

[0105] The target order can be an order initiated by an order initiator and needs to be paid. The target order can be initiated through an online e-commerce platform, or can be generated when a product is purchased offline. The payment instruction is generated based on the payment operation for the target order.

[0106] When the target order is initiated through an online e-commerce platform, the order initiator selects a product through the online e-commerce platform, and after determining the target product, the order initiator can generate a target order through an order generation instruction. For the target order, the order initiator can choose to pay immediately or pay at a specific time. After triggering the payment operation, a payment instruction is generated for payment scheme recommendation.

[0107] When the target order is generated when the order initiator purchases a product offline, after the order initiator selects a product offline, the product provider can provide the order initiator with an order generation label, such as an order generation QR code. The order initiator can generate a target order based on the order generation label. After triggering the payment operation, a payment instruction is generated for payment scheme recommendation.

[0108] After receiving the payment instruction for the target order, order environment data of multiple modalities associated with the target order can be obtained. The order environment data can indicate the environment in which the target order is located. The modality can indicate the dimension of the environment, for example, geographical environment, time characteristics, device environment, and application environment.

[0109] The geographic environment can indicate a geographic location where the target order is generated. The geographic location can be an administrative geographic location, such as A city, B city, etc., or a geographic area divided by entities on a map, such as an XX school area, an XX mall area, etc. The administrative geographic location can affect the payment scheme that the order initiator wants to use. For example, if the order initiator is in A city, the order initiator can choose a payment scheme that has a larger discount in A city. The geographic area divided by entities can also affect the payment scheme that the order initiator wants to use. For example, if the order initiator is in the XX mall area, the order initiator can choose a payment scheme that has a larger discount in the XX mall.

[0110] The time feature can indicate the time when the target order is generated. The time can be divided into time periods, such as 9:00-11:00, 11:00-13:00, 13:00-18:00, 18:00-22:00, etc. The recommended payment scheme can be different for different time periods. For example, if the order payment time is in the 18:00-22:00 time period, although payment scheme A has a larger payment discount, the order payment time is not in the working hours of the payment platform to which payment scheme A belongs, so the order initiator can only choose other payment schemes.

[0111] The device environment can indicate the device state where the target order is generated. The device state can be device power, device network state, and device system environment, etc. The device state can also affect the selection of the order payment scheme. For example, when the device state for order payment is low, the order initiator can be preferentially recommended a payment scheme with a simpler payment process and faster payment speed to prevent the device from running out of power and being unable to normally pay. For another example, when the device system version is low, the order initiator can be preferentially selected a payment scheme that can match the current system version to prevent the payment scheme from being unable to normally execute.

[0112] The application environment can indicate an application related to the target order, such as an application for generating the target order, an interface being displayed in the application, etc. The application environment can also affect the selection of the order payment scheme. For example, when the target order is generated by application E1, the order initiator can be preferentially recommended a payment scheme supported by application E1 and having a larger discount.

[0113] The order environment data of different modalities can be obtained in different ways. For example, for the geographical environment, it can be obtained through the GPS or Beidou positioning device in the payment device or other device sensors; the time characteristics can be obtained from the system time of the payment device, which generally includes the information of year, month, day, weekday, hour, minute, and second. The time characteristics can be generated from the hour, minute, and second information. The device environment can be directly perceived from the payment device state information. The application environment can be obtained through the application information being used by the payment device.

[0114] Note that when obtaining the order environment data of multiple modalities associated with the target order, the consent of the order initiator is sought in advance. Moreover, the collection, use, and processing of these order environment data will comply with relevant laws, regulations, and standards. When seeking the consent of the order initiator, the individual permission or individual consent of the order initiator can be obtained through a pop-up window or by jumping to a confirmation page.

[0115] In an embodiment, obtaining the order environment data of multiple modalities associated with the target order includes:

[0116] The order environment data of multiple modalities associated with the target order includes the historical operation data of the order initiator.

[0117] The historical operation data of the order initiator can indicate the operation data of the order initiator in the payment device within a predetermined time period before the current time point, such as browsing records, search records, and application interaction data, etc.

[0118] The payment intention of the order initiator can be determined through the historical operation data of the order initiator, which may also affect the selection of the payment scheme. For example, if the order initiator has inquired about the installment payment method in the past, the order initiator may want to make installment payment for a large order. For another example, if the order initiator has browsed multiple restaurant information through the application in the past, the order scenario of the target order may be related to catering, and a payment scheme with a discount in catering payment can be recommended to the order initiator.

[0119] The historical operation data of the order initiator can be obtained based on device logs. The device logs record the object operations in the device within a predetermined time period, which can include browsing records, query records, and interaction records, etc.

[0120] Note that when obtaining the historical operation data of the order initiator, the consent of the order initiator is sought in advance, which will not be repeated here.

[0121] Obtaining the historical operation data of the order initiating object helps to understand the payment intention of the order initiating object, and then in the subsequent steps, the target order scene is more accurately found.

[0122] Step 320, finding the target order scene in a plurality of predetermined order scenes according to the order environment data of a plurality of modalities.

[0123] After obtaining the order environment data of a plurality of modalities, the target order scene can be found in a plurality of predetermined order scenes. The order scene can be pre-set according to a plurality of dimensions. For example, the order scene set according to the order product type can include: catering order scene, shopping order scene, transportation order scene, tourism and entertainment order scene, and fixed expenditure order scene, etc. For example, the order scene set according to the order payment resource amount can include: large payment scene, medium payment scene, and small payment scene, etc. In addition to the order scene divided according to the predetermined dimension, some order scenes in special scenarios can also be pre-set, such as multi-person payment scene or cross-region payment scene, etc.

[0124] The payment scheme recommended by different order scenes may be different. For example, in the catering order scene, the payment scheme that can provide catering payment discount can be selected; in the large payment scene, the payment scheme of multiple payments can be selected; in the cross-region payment scene, resource exchange may be required according to the cross-region resource exchange ratio before payment, so the payment scheme with more favorable resource exchange ratio can be selected. Therefore, finding the target order scene is crucial for selecting the payment scheme recommended to the order initiating object.

[0125] In an embodiment, finding the target order scene in a plurality of predetermined order scenes according to the order environment data of a plurality of modalities comprises:

[0126] Obtaining a multi-modal fusion model;

[0127] Based on the multi-modal fusion model, the order environment data of a plurality of modalities is analyzed, and the target order scene is found in a plurality of predetermined order scenes.

[0128] The multi-modal fusion model can be pre-set and trained, and can be directly called.

[0129] The multi-modal fusion model can analyze data for each modal order environment data, and can also analyze fused data of multiple modal order environment data. The multi-modal fusion model can first process data of different modalities through a special network, for example, using Bidirectional Encoder Representations from Transformers (BERT) to process text data, using Convolutional Neural Network (CNN) to process image data, using Transformer to process time series data, etc. After data processing, the data of different modalities can be mapped to a unified semantic space to realize cross-modal semantic alignment, thereby establishing a correlation between data of different modalities, for example, correlating address environment data and device environment data.

[0130] After obtaining the multi-modal fusion model, the order environment data of multiple modalities can be input into the multi-modal fusion model for data analysis, so as to find the target order scenario in multiple predetermined order scenarios.

[0131] In an embodiment, based on the multi-modal fusion model, the order environment data of multiple modalities is analyzed to find the target order scenario in multiple predetermined order scenarios, including:

[0132] Obtaining order category information of the target order, and based on the order category information, finding multiple target modalities for data analysis;

[0133] Based on the multi-modal fusion model, the order environment data of multiple target modalities is analyzed to find the target order scenario in multiple predetermined order scenarios.

[0134] The order category information of the target order can indicate the type of the target order, and can specifically include offline payment orders and e-commerce platform orders, etc. For different types of orders, the key information for determining the target order scenario can be different. For example, for offline payment orders, the generation of the order mainly relies on offline product selection, and after selecting the product, the payment is made through an online payment scheme, therefore, the application environment has less influence on the selection of the payment scheme, and the geographic environment, time characteristics and environmental devices can be used as target modalities. For example, for e-commerce platform orders, the order is generated for payment after selecting the product on the e-commerce platform, therefore, the geographic environment has less influence on the selection of the payment scheme, and the time characteristics, device environment and application environment can be used as target modalities.

[0135] After finding the target modality, the order environment data corresponding to the target modality can be input into the multi-modal fusion model for data analysis, and then the target order scene can be found in the plurality of predetermined order scenes.

[0136] The multi-modal fusion model is used to analyze the order environment data of the plurality of target modalities, which is beneficial to filter out the key data for finding the target order scene from the order environment data of the plurality of modalities, and to eliminate irrelevant data, thereby improving the efficiency and accuracy of finding the target order scene.

[0137] After the multi-modal fusion model is used to analyze the order environment data of the plurality of modalities, the target order scene is found, which is beneficial to fully fuse the order environment data of the plurality of modalities, and then find the target order scene that best matches the order environment from the plurality of order scenes, thereby improving the accuracy of finding the target order scene.

[0138] In an embodiment, finding the target order scene in the plurality of predetermined order scenes according to the order environment data of the plurality of modalities comprises:

[0139] Obtaining a plurality of historical order environment data in a predetermined time period;

[0140] Finding the target order scene in the plurality of predetermined order scenes according to the order environment data of the plurality of modalities and historical order scenes corresponding to the plurality of historical order environment data.

[0141] The historical order environment data indicates the environment data of a plurality of modalities of a historical order in which the order initiating object makes payment through the same payment device in a predetermined time period before the current time point. The predetermined time period can be 10 minutes, 30 minutes or one hour before the current time point, etc. The historical order environment data can be obtained from the order record in the payment device. The payment data in the predetermined time period is stored in the order record. It should be noted that when obtaining the historical order environment data of the order initiating object, the consent of the order initiating object should be obtained in advance, which will not be described here.

[0142] Since the order initiating object may make payment for the same order scene in a period of time, the historical order environment data can be similar to the order scene in which the target order is located, and obtaining the historical order environment data can provide a reference for finding the target order scene, thereby improving the accuracy and efficiency of finding the target order scene.

[0143] After obtaining the historical order environment data, the target order scene can be found based on the order environment data and the historical order scene corresponding to the historical order environment data.

[0144] In an embodiment, the target order scenario is found in the plurality of predetermined order scenarios according to the order environment data of the plurality of modalities and historical order scenarios corresponding to the plurality of historical order environment data, comprising:

[0145] The reference order environment data is queried in the plurality of historical order environment data, which meets a predetermined condition with the order environment data of the plurality of modalities;

[0146] The target order scenario is found in the plurality of predetermined order scenarios based on the reference order environment data.

[0147] The matching degree between the order environment data of the plurality of modalities and the historical order environment data meeting the predetermined condition can be that the consistency between the historical order environment data corresponding to a historical order and the order environment data reaches a preset value. For example, the plurality of modalities includes geographical environment, time characteristics, device environment and application environment, and the predetermined condition is that the data of at least three modalities is consistent, that is, when the historical order environment data corresponding to a historical order and the order environment data exist in three or more of geographical environment, time characteristics, device environment and application environment, the historical order environment data can be used as the reference order environment data meeting the predetermined condition.

[0148] After the reference order environment data is found in the plurality of historical order environment data, the target order scenario can be found in the plurality of predetermined order scenarios based on the reference order environment data.

[0149] In an embodiment, the order scenario of the historical order corresponding to the reference order environment data can be directly used as the target order scenario.

[0150] The reference order environment data is found based on the matching degree between the plurality of historical order environment data and the order environment data of the plurality of modalities, and the target order scenario is found based on the reference order environment data, which can quickly find the order scenario and improve the efficiency of the payment scheme recommendation.

[0151] In an embodiment of step 310, in addition to the order environment data of the plurality of modalities, historical operation data of the order initiator is also obtained. Based on this, in an embodiment, the target order scenario is found in the plurality of predetermined order scenarios according to the order environment data of the plurality of modalities, comprising:

[0152] The target order scenario is found in the plurality of predetermined order scenarios according to the order environment data of the plurality of modalities and the historical operation data.

[0153] The historical operation data of the order initiating object based on the order initiating object can understand the order payment intention of the order initiating object, and thus the environmental information of the target order and the order payment intention can be fused and understood to improve the accuracy of finding the target order scene. For example, through the order environmental data of multiple modalities of the target order, it is found that the current geographic environment of the order initiating object is in a mall, and the time period at the current time point is 11:00-13:00. In this way, only through the order environmental data, the target order scene can be identified as a daily consumption scene. However, through the historical operation data, it is found that the order initiating object has browsed multiple restaurant information through the application in the past period of time, so it can be judged that the order initiating object may have a meal in the restaurant of the mall, and the target order scene is a catering payment scene.

[0154] In an implementation manner, the target order scene is found according to the order environmental data of multiple modalities and the historical operation data in multiple predetermined order scenes, comprising:

[0155] The target operation data related to the order environmental data is queried in the historical operation data;

[0156] The target order scene is found in multiple predetermined order scenes based on the environmental analysis of the order environmental data of multiple modalities and the target operation data by the first neural network.

[0157] The target operation data can be historical operation data related to the order environmental data in the historical order. The historical operation data includes browsing records, query records and interaction records of the order initiating object, but not all historical operation data is helpful for finding the target order scene. For example, the historical operation data shows that the order initiating object browses multiple restaurant information in the mall, and in this process, the order initiating object switches to the instant messaging application to send a communication message. In these historical operation data, browsing multiple restaurant information in the mall is helpful for finding the target order scene and can be used as target operation data, while sending a communication message by using the instant messaging application is less helpful for finding the target order scene.

[0158] The target operation data can be found in the historical operation data based on the order environmental data. For example, the multiple modalities include geographic environment, time characteristics, device environment and application environment, and when finding the target operation data, the geographic environment, time characteristics, device environment and application environment related to the target order can be found in the historical operation data.

[0159] After the target operation data is found, the order environment data and the target operation data can be input into a first neural network for environment analysis, and then the target order scenario is found. The first neural network can be pre-designed and trained, and the order environment data and the target operation data can be fused and analyzed to realize cross-modal semantic alignment, thereby establishing a correlation between different data. In an embodiment, the first neural network can be a deep learning network constructed based on an encoder.

[0160] Querying the target operation data related to the order environment data in the historical operation data can eliminate the historical operation data that is less helpful for finding the target order scenario, reduce the data amount, and improve the efficiency of finding the target order scenario. Based on the first neural network, the order environment data and the target operation data are analyzed, which helps to fully understand and fuse the order environment data and the target operation data, and helps to improve the efficiency of finding the target order scenario.

[0161] Step 330, based on the target order scenario, calling a plurality of order agents for order analysis to obtain a plurality of order analysis results.

[0162] An order agent can be an artificial intelligence module capable of perceiving an environment and making corresponding decisions. Different order agents can have different functions, and each agent obtains corresponding data according to its function and uses a corresponding artificial intelligence model to analyze the data, thereby obtaining an order analysis result. For example, order agent X1 can be responsible for analyzing the income and expenditure of the order initiator. Based on this, order agent X1 can obtain data related to the income and expenditure of the order initiator for analysis to obtain a corresponding order analysis result. Order agent X2 can be responsible for analyzing the product value of the product corresponding to the target order. For this, order agent X2 can obtain product-related data for analysis to obtain a corresponding order analysis result. Order agent X3 can be responsible for evaluating whether the payment environment of the target order is safe. For this, order agent X3 can obtain data related to the payment environment for analysis to obtain a corresponding order analysis result.

[0163] An order agent can be a pre-trained artificial intelligence model. In an embodiment, the training process of an order agent includes:

[0164] Obtaining an initialized order agent and a historical order data training set in a predetermined time period;

[0165] Updating the initialized order agent based on the historical order scenario information corresponding to each historical order data and the corresponding historical payment scheme.

[0166] The predetermined time period can be a predetermined time before the current time point. The training process of the order agent can be performed according to a predetermined period, for example, once a day, once a week, or once a month. The predetermined time period can be a corresponding time period in a period. The initialized order agent can be an order agent before training.

[0167] The historical order data training set can include historical order data of the order initiator in the predetermined time period, and the historical order scenario information of the historical order data is used as input, and the historical payment scheme selected by the order initiator for the historical order scenario information is used as a label. The historical order scenario information is input into the order agent to obtain a predicted payment scheme, and the parameters of the initialized order agent are updated based on the predicted payment scheme and the historical payment scheme.

[0168] Training the order agent using the historical order data training set in the predetermined time period can enable the order agent to continuously learn the payment preferences of the order initiator, so that the order analysis of the order agent is more in line with the decision-making mode of the order initiator, and the order agent can achieve personalized analysis and improve the accuracy of payment scheme recommendation to the order initiator.

[0169] In an embodiment, a plurality of order agents are called based on a target order scenario to perform order analysis, and a plurality of order analysis results are obtained, including:

[0170] The financial analysis agent, the product analysis agent, the security analysis agent, and the payment method analysis agent are called based on the target order scenario to perform order analysis, and a financial analysis result, a product analysis result, a security analysis result, and a payment method analysis result are obtained.

[0171] The financial analysis agent can be used to analyze the financial status of the order initiator. Specifically, the budget of the order initiator and whether the cash flow in a period is healthy can be analyzed. When the order initiator wants to pay for a target order, the financial analysis agent can obtain the received resource amount data and the expenditure data of the order initiator in a predetermined period, and the overall resource storage data of the order initiator and other financial status related data, and then can perform order analysis on the financial status related data. The predetermined period can be a week, half a month, or a month, etc.

[0172] The order analysis based on the financial status can include multiple dimensions, such as a payment budget of the order initiator, a short-term financial health, and a long-term financial health. The payment budget of the order initiator can be determined based on the financial status related data. For example, when the order initiator has only 100 resources available in the current month, 100 can be taken as the payment budget. The short-term financial health of the order initiator can be determined based on the resource income and expenditure in a predetermined period. For example, in the current month, the order initiator has already spent much more resources than received, which can indicate that the short-term financial status of the order initiator is not healthy. The long-term financial health of the order initiator can be further analyzed based on the long-term resource storage and other information on the basis of the short-term income and expenditure. For example, in the current month, although the order initiator has already spent much more resources than received, the long-term resource storage is relatively large and can maintain a positive income, which can indicate that the long-term financial status of the order initiator is relatively healthy.

[0173] When the financial analysis agent analyzes the order based on the financial status, the financial analysis agent can determine whether each order payment scheme is worth recommending based on the multiple dimensions of the above analysis. The financial status of the order initiator can affect the selection of the payment scheme. For example, when the financial analysis agent determines that the payment budget of the order initiator is low, the financial analysis agent can recommend a payment scheme of installment payment to the order initiator. For another example, when the financial analysis agent determines that the short-term financial status of the order initiator is not healthy, the financial analysis agent can recommend a payment scheme with a high credit limit to the order initiator. Therefore, the financial analysis result obtained by the financial analysis agent in the order analysis includes the order payment scheme recommendation result obtained based on the financial status analysis.

[0174] Note that the consent of the order initiator is required in advance when the financial status related data is obtained.

[0175] The product analysis agent can be used to analyze the product value and the product cost performance of the target product in the target order. Specifically, the product analysis agent can analyze whether the value of the target product is reasonable and the cost performance of the target product. The value of the target product can be directly represented by the required resource amount of the target product. The product cost performance can be analyzed by analyzing the matching degree of the required resource amount of the target product and the product quality. When the product quality is low but the required resource amount is large, the product cost performance is low. When the product quality is high but the required resource amount is low, the product cost performance is high. When the order initiator wants to pay for the target order, the product analysis agent can obtain product related data such as the required resource amount of the product, the average required resource amount of the product in the market where the target product is located, the product use evaluation, and the product information and resource amount information of the same type of product of the target product, and then can analyze the product value and the product cost performance of the target product.

[0176] The product value analysis mainly analyzes the amount of resources required by the product. The product cost performance order analysis mainly analyzes whether the product quality matches the amount of resources required. For example, when the target product in the market has other products of the same quality, and the amount of resources required by the other products is much smaller than that of the target product, it can be considered that the product cost performance of the target product is low.

[0177] After the product analysis agent analyzes the target product, it can determine whether each order payment scheme is worth recommending based on the analysis result. The product value may affect the selection of the payment scheme. For example, when the product analysis agent considers that the product value of the target product is high, it can recommend a payment scheme with a large payment guarantee to the order initiator. The product cost performance may also affect the selection of the payment scheme. For example, when the product analysis agent considers that the product cost performance of the target product is low, it can recommend a payment scheme with a price protection function to the order initiator, that is, when the subsequent product price decreases, it can also provide protection for the order initiator. Therefore, the product analysis result obtained by the product analysis agent includes the order payment scheme recommendation result obtained based on the target product value and the target product cost performance analysis.

[0178] The security analysis agent can be used to analyze the credit level of the product provider and the security of the order payment environment. When the order initiator wants to pay for the target order, the security analysis agent can obtain the product providing record of the product provider and the security certification information provided by the official organization. Through the product providing record, the security analysis agent can analyze whether the order initiator has a bad operation record; through the security certification information, the credit level of the order initiator can be analyzed, and whether there has been a bad event around the order initiator. In addition, the security analysis agent can also obtain payment device and network information. Through the payment device and network information, the security analysis agent can analyze whether the payment device has the risk of being invaded, whether the payment link is safe, etc. After analyzing the security of the product provider and the order payment environment, the order payment scheme can be recommended to the order initiator based on the security analysis result.

[0179] The security analysis of the product provider and the order payment environment can affect the recommended result of the order payment scheme. For example, when the security analysis agent determines that the product provider has a low credit level and there is a risk of payment, the order initiator can be recommended an order payment scheme with higher payment security. For another example, when the security analysis agent determines that the current network environment is not secure, the order initiator can be recommended an order payment scheme with additional security authentication. Therefore, the security analysis result obtained by the security analysis agent in the order analysis includes the order payment scheme recommendation result based on the security level analysis of the product provider and the order payment environment.

[0180] The payment method analysis agent can be used to directly analyze the availability of multiple payment schemes. When the order initiator wants to pay for the target order, the payment method analysis agent can obtain payment scheme related data of each payment scheme that can be supported, such as discount information corresponding to each payment scheme, available amount of the order initiator in each payment scheme, and additional resource expenditure required by each payment scheme, and then perform order analysis by using the payment scheme related data.

[0181] The payment method analysis agent determines whether to make a recommendation by comprehensively analyzing the comprehensive situation of each payment scheme. For example, a payment scheme has a large discount for the target order scenario, requires a low additional resource expenditure, and has a high available amount of the order initiator in the payment scheme, so the payment scheme is worth recommending. Therefore, the payment method analysis result obtained by the payment method analysis agent in the order analysis includes the recommended result after direct analysis of multiple order payment schemes.

[0182] After finding the target order scenario, the financial analysis agent, the product analysis agent, the security analysis agent, and the payment method analysis agent can be called simultaneously to perform parallel order analysis, as shown in FIG. 4. Figure 4

[0183] By using the financial analysis agent, the product analysis agent, the security analysis agent, and the payment method analysis agent to perform order analysis, the target order can be comprehensively evaluated from multiple aspects, such as the financial condition of the order initiator, the product value of the target product, the security of the payment environment, and the applicability of the payment method, so as to recommend a payment scheme to the order initiator from multiple dimensions, thereby improving the accuracy of generating the payment scheme.

[0184] In one embodiment, all available order agents can be called for the target order scenario. That is, each available order agent performs independent order analysis when performing order analysis.

[0185] In another embodiment, multiple order agents are called based on the target order scenario to perform order analysis and obtain multiple order analysis results, including:​

[0186] Based on the target order scenario, find multiple target order agents from multiple order agents;

[0187] Multiple target order intelligent agents are used to analyze target order scenarios and obtain multiple order analysis results.

[0188] The order analysis elements that need to be focused on differ depending on the target order scenario. For example, multiple order intelligence agents may include financial analysis agents, product analysis agents, security analysis agents, and payment method analysis agents. When the target order scenario is a daily fixed expense scenario, the analysis requirements for the product's intrinsic value and cost-effectiveness are relatively low. This is because daily fixed expenses, such as rent or electricity costs, are mandatory payments, and the product's value and cost-effectiveness have a smaller impact on payment option recommendations. Therefore, the financial analysis agent, security analysis agent, and payment method analysis agent can be used as the target order intelligence agents.

[0189] In one implementation, the process of finding multiple target order agents from multiple order agents based on a target order scenario can be pre-defined. Since the order scenarios are pre-defined, the target order agents corresponding to each order scenario can be set simultaneously. For example, the target order agents corresponding to a restaurant order scenario may include a financial analysis agent, a security analysis agent, and a payment method analysis agent; the target order agents corresponding to a transportation order scenario may include a security analysis agent and a payment method analysis agent; and the target order agents corresponding to a large-amount payment scenario may include a financial analysis agent, a product analysis agent, a security analysis agent, and a payment method analysis agent. Pre-defining the target order agents corresponding to the target order scenarios can improve the efficiency of finding target order agents.

[0190] In another implementation, finding multiple target order agents from multiple order agents based on a target order scenario can be performed using a pre-defined neural network model. Specifically, the scenario data of the target order scenario can be input into the pre-defined neural network model for scenario analysis, thereby finding multiple target order agents most suitable for the target order scenario. Using a pre-defined neural network model to find target order agents improves the accuracy of the search.

[0191] After identifying the target order agent, multiple target order agents can be used to analyze the target order scenario, yielding multiple order analysis results. Each target order agent acquires corresponding data based on the item it analyzes and uses a corresponding artificial intelligence model for data analysis.

[0192] According to the target order scene selection target order agent, the target order scene can be analyzed in a targeted manner, which is beneficial to improve the accuracy of order analysis.

[0193] In step 340, the historical order payment data of the order initiating object of the target order is obtained, and the target order payment scheme is generated according to the historical order payment data and the plurality of order analysis results.

[0194] The historical order payment data of the order initiating object can include historical order payment scenes corresponding to historical orders of the order initiating object and historical order payment schemes selected by the order initiating object. The historical order payment data of the order initiating object can be obtained based on the payment records recorded in the order payment device.

[0195] According to the historical order payment data, the order payment preference of the order initiating object can be determined. In an embodiment, the payment scheme preferred by the order initiating object can be determined based on the frequency of use of the payment scheme within a predetermined time period. For example, among the 100 order payments of the order initiating object in a week, 60 times use order payment scheme Y1, and 30 times use order payment scheme Y2, which can indicate that the order initiating object is used to order payment scheme Y1, followed by order payment scheme Y2. Then, when recommending the order payment scheme to the order initiating object, the order payment scheme commonly used by the order initiating object can be recommended.

[0196] According to the historical order payment data and the plurality of order analysis results, the target order payment scheme is generated, that is, the target order payment scheme is generated based on the order payment preference of the order initiating object and the plurality of order analysis results.

[0197] In the foregoing embodiment, the financial analysis agent, the product analysis agent, the security analysis agent and the payment method agent are called based on the target order scene to perform order analysis, and the financial analysis result, the product analysis result, the security analysis result and the payment method analysis result are obtained. Based on this, in an embodiment, the historical order payment data of the order initiating object of the target order is obtained, and the target order payment scheme is generated according to the historical order payment data and the plurality of order analysis results, including:

[0198] The historical order payment data of the order initiating object of the target order is obtained, and the target order payment scheme is generated according to the historical order payment data, the financial analysis result, the product analysis result, the security analysis result and the payment method analysis result.

[0199] In an embodiment, generating the target order payment scheme according to the historical order payment data, the financial analysis result, the product analysis result, the security analysis result, and the payment method analysis result can utilize a multi-objective optimization algorithm. Through the multi-objective optimization algorithm, an optimal target order payment scheme that can balance the historical payment preferences of the order initiator, the financial analysis result, the product analysis result, the security analysis result, and the payment method analysis result can be found. The multi-objective optimization algorithm can weigh the following factors when making decisions: short-term financial gains, long-term financial gains, payment security, and historical payment preferences, as shown in the following formula: Figure 5 By comprehensively balancing the above factors, the target order payment scheme is obtained.

[0200] In an embodiment, the financial analysis result includes a first recommendable score corresponding to each of the order payment schemes, the product analysis result includes a second recommendable score corresponding to each of the order payment schemes, the security analysis result includes a third recommendable score corresponding to each of the order payment schemes, and the payment method analysis result includes a fourth recommendable score corresponding to each of the order payment schemes.

[0201] The first recommendable score refers to a score that is worth recommending based on the financial analysis, and the higher the first recommendable score corresponding to an order payment scheme is, the more worth recommending it is from the perspective of financial analysis. The second recommendable score refers to a score that is worth recommending based on the product analysis, and the higher the second recommendable score corresponding to an order payment scheme is, the more worth recommending it is from the perspective of product analysis. The third recommendable score refers to a score that is worth recommending based on the security analysis, and the higher the third recommendable score corresponding to an order payment scheme is, the more worth recommending it is from the perspective of security analysis. The fourth recommendable score refers to a score that is worth recommending based on the payment method analysis, and the higher the fourth recommendable score corresponding to an order payment scheme is, the more worth recommending it is from the perspective of payment method analysis.

[0202] Based on this, generating the target order payment scheme according to the historical order payment data, the financial analysis result, the product analysis result, the security analysis result, and the payment method analysis result includes:

[0203] Obtaining usage scores corresponding to the order payment schemes based on the historical order payment data;

[0204] Generating the target order payment scheme according to the usage scores corresponding to the order payment schemes, the first recommendable scores, the second recommendable scores, the third recommendable scores, and the fourth recommendable scores.

[0205] The use score corresponding to the order payment scheme can represent the preference degree of the order payment scheme for the order initiating object. The higher the use score is, the more the order initiating object likes to use the order payment scheme.

[0206] In an embodiment, the use score corresponding to the order payment scheme can be preset based on the use frequency of each order payment scheme in the historical order payment data. For example, for a predetermined time period before the current time point, the order payment scheme with the highest use frequency has a use score of 10, the order payment scheme with the second highest use frequency has a use score of 8, the order payment scheme with the third highest use frequency has a use score of 6, and so on.

[0207] In another embodiment, obtaining the use score corresponding to each order payment scheme based on the historical order payment data can be based on a preset neural network model for analysis and calculation. That is, the historical order payment data is input into the preset neural network model, and the use score corresponding to each order payment scheme is output. Using the neural network model to calculate the use score of each order payment scheme is conducive to improving the accuracy of calculating the use score.

[0208] After obtaining the use score corresponding to each order payment scheme, the target order payment scheme can be generated according to the use score corresponding to each order payment scheme, the first recommendable score, the second recommendable score, the third recommendable score, and the fourth recommendable score.

[0209] In an embodiment, the target order payment scheme can be generated by calculating the average of the use score, the first recommendable score, the second recommendable score, the third recommendable score and the fourth recommendable score corresponding to each order payment scheme. For example, the use score corresponding to the order payment scheme Y1 is 8, the first recommendable score is 7.5, the second recommendable score is 6, the third recommendable score is 8.5 and the fourth recommendable score is 5, then the average score corresponding to the order payment scheme Y1 is (8+7.5+6+8.5+5) / 5=7. The use score corresponding to the order payment scheme Y2 is 6, the first recommendable score is 8.5, the second recommendable score is 8, the third recommendable score is 9 and the fourth recommendable score is 7, then the average score corresponding to the order payment scheme Y2 is (6+8.5+8+9+7) / 5=7.7. The use score corresponding to the order payment scheme Y3 is 10, the first recommendable score is 5, the second recommendable score is 9, the third recommendable score is 7 and the fourth recommendable score is 6, then the average score corresponding to the order payment scheme Y3 is (10+5+9+7+6) / 5=7.4. The order payment schemes are sorted in descending order according to the average score, and the result is the order payment scheme Y2, the order payment scheme Y3 and the order payment scheme Y1. Therefore, the order payment scheme Y2 can be used as the target order payment scheme. The target order payment scheme is generated by using the average score, which can ensure that each factor has the same influence on the generation of the target order payment scheme, and improve the fairness of generating the target order payment scheme.

[0210] In another embodiment, the target order payment scheme is generated according to the use score, the first recommendable score, the second recommendable score, the third recommendable score and the fourth recommendable score corresponding to the plurality of order payment schemes, comprising:

[0211] obtaining the first weight corresponding to the historical order payment data, the second weight corresponding to the financial analysis agent, the third weight corresponding to the product analysis agent, the fourth weight corresponding to the security analysis agent and the fifth weight corresponding to the payment method analysis agent;

[0212] based on the first weight, the second weight, the third weight, the fourth weight and the fifth weight, the use score, the first recommendable score, the second recommendable score, the third recommendable score and the fourth recommendable score are weighted and calculated, and the target order payment scheme is generated from the plurality of order payment schemes.

[0213] The first weight can indicate an influence degree of the historical payment preference on generation of the target order payment scheme, the second weight can indicate an influence degree of the financial analysis result on generation of the target order payment scheme, the third weight can indicate an influence degree of the product analysis result on generation of the target order payment scheme, the fourth weight can indicate an influence degree of the security analysis result on generation of the target order payment scheme, and the fifth weight can indicate an influence degree of the payment method analysis result on generation of the target order payment scheme.

[0214] In an embodiment, the first weight, the second weight, the third weight, the fourth weight, and the fifth weight can be preset. When generating the target order payment scheme, the influence degrees of the historical payment preference, the financial analysis result, the product analysis result, the security analysis result, and the payment method analysis result on the target payment scheme can be determined based on needs of actual application. For example, when generating the target order payment scheme, if more attention is paid to the financial analysis on the payment budget, the second weight can be increased; if more attention is paid to the payment environment security, the fourth weight can be increased.

[0215] In another embodiment, the first weight corresponding to the historical order payment data, the second weight corresponding to the financial analysis agent, the third weight corresponding to the product analysis agent, the fourth weight corresponding to the security analysis agent, and the fifth weight corresponding to the payment method analysis agent are obtained, including:

[0216] The second neural network is obtained.

[0217] The order environment data of the multiple modalities is subjected to environment analysis based on the second neural network, to generate the first weight corresponding to the historical order payment data, the second weight corresponding to the financial analysis agent, the third weight corresponding to the product analysis agent, the fourth weight corresponding to the security analysis agent, and the fifth weight corresponding to the payment method analysis agent.

[0218] The second neural network can be a pre-trained neural network model. The order environment data of the multiple modalities is input into the second neural network, and the second neural network can output the first weight corresponding to the historical order payment data, the second weight corresponding to the financial analysis agent, the third weight corresponding to the product analysis agent, the fourth weight corresponding to the security analysis agent, and the fifth weight corresponding to the payment method analysis agent.

[0219] The second neural network is used to analyze the order environment data of multiple modalities, thereby generating the first weight, the second weight, the third weight, the fourth weight, and the fifth weight. The weights can be automatically set according to the order payment scene, so that the influence of the historical payment preference, the financial analysis result, the product analysis result, the security analysis result, and the payment method analysis result on the target payment scheme can be flexibly adjusted based on the order environment data, so as to ensure that the target order payment scheme is more suitable for the order payment scene, and the accuracy of generating the target order payment scheme is improved.

[0220] After the first weight, the second weight, the third weight, the fourth weight, and the fifth weight are generated, the usage score, the first recommendable score, the second recommendable score, the third recommendable score, and the fourth recommendable score can be weighted and calculated. For example, the first weight is 0.2, the second weight is 0.2, the third weight is 0.1, the fourth weight is 0.2, and the fifth weight is 0.3. The usage score corresponding to the order payment scheme Y1 is 8, the first recommendable score is 7.5, the second recommendable score is 6, the third recommendable score is 8.5, and the fourth recommendable score is 5. Then, the weighted average score of the order payment scheme Y1 is 8*0.2+7.5*0.2+6*0.1+8.5*0.2+5*0.3=6.9. The usage score corresponding to the order payment scheme Y2 is 6, the first recommendable score is 8.5, the second recommendable score is 8, the third recommendable score is 9, and the fourth recommendable score is 7. Then, the average score of the order payment scheme Y2 is 6*0.2+8.5*0.2+8*0.1+9*0.2+7*0.3=7.6. The usage score corresponding to the order payment scheme Y3 is 10, the first recommendable score is 5, the second recommendable score is 9, the third recommendable score is 7, and the fourth recommendable score is 6. Then, the average score of the order payment scheme Y3 is 10*0.2+5*0.2+9*0.1+7*0.2+6*0.3=7.1. The three order payment schemes are sorted in descending order according to the average score, and the result is the order payment scheme Y2, the order payment scheme Y3, and the order payment scheme Y1. Therefore, the order payment scheme Y2 can be used as the target order payment scheme.

[0221] The weighted calculation based on the first weight, the second weight, the third weight, the fourth weight, and the fifth weight can adjust the influence of each factor on the generation of the target order payment scheme according to the actual application needs, which is beneficial to improve the flexibility of the generation of the target order payment scheme.

[0222] The target order payment plan is generated based on the use of scores, the first recommended score, the second recommended score, the third recommended score, and the fourth recommended score. Each order payment plan can be evaluated in the form of a quantifiable score, ensuring that the target order payment plan comprehensively considers historical preferences, financial analysis results, product analysis results, security analysis results, and payment method analysis results, thereby improving the accuracy of generating target order payment plans.

[0223] Based on historical order payment data, financial analysis results, product analysis results, security analysis results, and payment method analysis results, a target order payment plan is generated. This plan comprehensively weighs the short-term and long-term financial benefits, payment security, and historical payment preferences of the order initiator, thus ensuring the accuracy of the generated target order payment plan.

[0224] Step 350: Recommend the target order payment plan to the order initiator.

[0225] After generating the target order payment plan, it can be recommended to the order initiator. Specifically, the target order payment plan can be rendered and displayed on the order initiator's payment device.

[0226] In one implementation, after recommending the target order payment scheme to the order initiator, the method further includes:

[0227] In response to the target order payment scheme recommended to the order initiator, the order payment interface is displayed, and a payment method selection area is shown in the order payment interface;

[0228] In response to a triggered action on the selected payment method in the payment method selection area, the order payment result is displayed.

[0229] After recommending the target order payment method to the order initiator, the order payment interface is rendered in response to the recommendation. The order payment interface includes a payment method selection area, which includes the target order payment method. Besides the target order payment method, the payment method selection area may also include other order payment methods for the order initiator to choose from. Once the order initiator selects a payment method in the payment method selection area, the order payment result can be displayed. Simultaneously, the selected payment method is recorded. Based on one of the aforementioned implementation methods, the order agent can also be updated based on the selected payment method, such as... Figure 6 As shown.

[0230] Order payment interface as follows Figure 7As shown, the order payment interface 710 displays product provider information and payable resource quantity information so that the order initiator can confirm the accuracy of the payment information. It also displays a payment method selection area 711. The payment method selection area 711 displays the target order payment plan. Besides the target order payment plan, other order payment plans are also available. The order initiator can choose the recommended target order payment plan or select other order payment plans independently. Figure 7 In the process, after selecting the target order payment plan, click payment control 712 to confirm payment.

[0231] In one implementation, in response to a target order payment scheme recommended to the order initiator, an order payment interface is displayed, and a payment method selection area is shown within the order payment interface, including:

[0232] In response to the target order payment scheme recommended to the order initiator, the order payment interface is displayed, and the payment method selection area and payment scenario analysis area are displayed in the order payment interface;

[0233] In response to a triggering action on the payment adjustment control, a payment adjustment pop-up window is displayed.

[0234] The payment scenario analysis area displays order analysis information for the target order, which may include budget analysis, product analysis, and security information. For example, the payment scenario analysis area may prompt the order initiator with "The payment environment is secure, please pay with confidence," and "The resource requirements of this product exceed the market average, please confirm before payment." Scenario analysis information can be generated based on the order analysis results of multiple order agents.

[0235] The payment scenario analysis area includes payment adjustment controls. Since the payment scenario analysis displays order analysis information from multiple dimensions, the order initiator may want to modify or cancel the target order after reviewing the information. Therefore, the order initiator can adjust the target order using the payment adjustment controls. For example, if the payment scenario analysis area prompts the order initiator, "This order exceeds 80% of the budget; it is recommended to control spending," and the order initiator decides to accept the suggestion and control spending, then cancels the target order by triggering the payment adjustment controls.

[0236] For example, the order payment interface is like Figure 8A As shown, the order payment interface 710 displays a payment method selection area 711 and a payment scenario analysis area 811. The payment scenario analysis area 811 displays messages such as "This order exceeds 80% of the budget; it is recommended to control spending," "The payment environment is secure; please pay with confidence," and a payment adjustment control 812. After triggering the payment adjustment control, as shown... Figure 8BAs shown, a payment adjustment pop-up window 821 can be displayed, through which "cancel order", "modify order" or "return payment" can be selected. After "cancel order" is selected, an adjustment confirmation control 822 can be triggered to confirm cancellation of the order.

[0237] Displaying the payment scenario analysis area in the order payment interface can enable the order initiating object to view the order analysis conclusion, so that the order initiating object can adjust the target order, thereby improving the payment experience of the order initiating object.

[0238] In an embodiment, in response to a triggering operation on a selected payment scheme in the payment method selection area, an order payment result is displayed, including:

[0239] In response to a triggering operation on a selected payment scheme in the payment method selection area, a payment result display interface is displayed.

[0240] After confirming the order payment, a payment result display interface can be displayed. The payment result display interface can include the order payment result and the order payment analysis conclusion. The order payment analysis conclusion can include an analysis of this payment operation, for example, "this payment uses the target order payment scheme, saving 15 resources", and can also include future payment suggestions, for example, "based on historical order record analysis, you have purchased the product multiple times, it is recommended to open a preferential card". As Figure 9 As shown, the order payment result and the order payment analysis conclusion are displayed in the payment result display interface 910.

[0241] Displaying the order payment result and the order payment analysis result in the payment result display interface enables the order initiating object to view the analysis conclusion of the payment operation, which is conducive to improving the payment experience of the order initiating object; payment suggestions for the order initiating object can also help the order initiating object form a healthier financial habit.

[0242] Displaying the target order payment scheme in the payment method selection area for the order initiating object to make a payment can enable the order initiating object to quickly find the target order payment scheme for payment, which is conducive to improving the efficiency of the target order payment.

[0243] In summary, in an embodiment, the system architecture of the order processing method in the embodiments of the present disclosure is as follows: Figure 10As shown, it includes a display layer 1010, a perception layer 1020, an agent coordination layer 1030, a decision layer 1040, and a data layer 1050. The display layer 1010 mainly includes a terminal interface 1011, which is responsible for interacting with the order initiating object, receiving information input by the order initiating object, and presenting system information to the order initiating object. The perception layer 1020 mainly includes a scene perceiver 1021, which is responsible for perceiving the target order scene corresponding to the target order. The agent coordination layer 1030 first includes a coordination scheduler 1031, which activates the corresponding order agent based on the target order scene. The order agent includes a financial analysis agent 1032, a product analysis agent 1033, a security analysis agent 1034, and a payment method analysis agent 1035. After order analysis by multiple order agents, the decision engine in the decision layer 1040 generates an optimal target order payment scheme using a multi-objective optimization algorithm. The data layer 1050 provides various types of data required for the order processing process, such as payment data of the order initiating object, historical order data, and financial related data, etc.

[0244] Therefore, after receiving the payment instruction of the target order, the order processing method of the embodiment of the disclosure can find the target order scene based on the order environment data of the target order in multiple modalities. The difference of the target order scene will affect the selection of the order payment scheme. After order analysis by multiple order agents corresponding to the target order scene, the target order payment scheme is generated based on the historical order payment data of the order initiating object and the order analysis result. That is, the target order payment scheme recommended to the order initiating object in the embodiment of the disclosure is determined based on the scene information of the target order and the historical payment preference of the object, so that the target order payment scheme can better meet the needs of the order payment scene, and the accuracy of recommending the order payment scheme to the object is improved.

[0245] The embodiment of the disclosure is described in detail in combination with a specific application scenario

[0246] As Figure 11 shown, the specific process diagram of the order processing method provided by the disclosure applied to the online order payment scene is shown, in combination with Figure 12 shown in the multi-terminal interaction diagram, the order processing method includes:

[0247] Step 1101, the order initiating object generates a target order, the perception layer 1020 receives a payment instruction of the target order, and obtains order environment data of multiple modalities associated with the target order.

[0248] The target order is an order initiated by the order initiating object through an online e-commerce platform, which needs to be paid.

[0249] After receiving the payment instruction of the target order, the perception layer 1020 can obtain the order environment data of multiple modalities associated with the target order. The order environment data can indicate the environment in which the target order is located. The modalities can indicate the dimensions of the environment, such as the geographic environment, the time feature, the device environment, and the application environment, etc.

[0250] The perception layer 1020 can obtain the order environment data of different modalities in different ways. For example, the geographic environment can be obtained through GPS or Beidou positioning device and other device sensors; the time feature can be obtained from the system time of the payment device. The device environment can be directly perceived from the payment device state information. The application environment can be obtained through the application information that the payment device is using.

[0251] Step 1102, the perception layer 1020 obtains the historical operation data of the order initiating object;

[0252] The historical operation data can indicate the operation data of the order initiating object in the order payment device within a predetermined time period before the current time point, such as browsing records, search records, and application interaction data, etc. The payment intention of the order initiating object can be determined through the historical operation data of the order initiating object, which may also affect the selection of the payment scheme.

[0253] The historical operation data of the order initiating object can be obtained based on the device log. The device log records the object operations occurring in the device within a predetermined time period, which can include browsing records, query records, and interaction records, etc.

[0254] Step 1103, the perception layer 1020 queries the target operation data related to the order environment data in the historical operation data;

[0255] The target operation data can be the historical operation data related to the order environment data in the historical order. The historical operation data includes browsing records, query records, and interaction records of the order initiating object, etc., but not all historical operation data is helpful for the search of the target order scenario. For example, the historical operation data shows that the order initiating object browses multiple restaurant information in the shopping mall, and in this process, the order initiating object switches to the instant messaging application to send a communication message. Among these historical operation data, the browsing of multiple restaurant information in the shopping mall is helpful for the search of the target order scenario and can be used as the target operation data, while the sending of the communication message through the instant messaging application is less helpful for the search of the target order scenario.

[0256] The target operation data can be found in the historical operation data based on the order environment data. For example, the plurality of modalities include a geographical environment, a time feature, a device environment, and an application environment, and when the target operation data is found, the geographical environment, the time feature, the device environment, and the application environment corresponding to the target order can be found from the historical operation data.

[0257] In step 1104, the perception layer 1020 finds a plurality of target modalities for data analysis from the plurality of modalities, and acquires a multi-modality fusion model.

[0258] The target modality can indicate a modality that has a greater impact on the search for the target order scenario. The target modality can be determined by order category information of the target order.

[0259] The order category information of the target order can indicate the type of the target order, and can specifically include an offline payment order and an e-commerce platform order. For different types of orders, the key information for determining the target order scenario can be different. For example, for an offline payment order, the generation of the order mainly relies on the selection of products offline, and after the selection of products, the payment is made by using an online payment scheme, therefore, the application environment has a smaller impact on the selection of the payment scheme, and the geographical environment, the time feature, and the device environment can be used as the target modality. For example, for an e-commerce platform order, the order is generated for payment after selecting products on an e-commerce platform, therefore, the geographical environment has a smaller impact on the selection of the payment scheme, and the time feature, the device environment, and the application environment can be used as the target modality.

[0260] The multi-modality fusion model can analyze the order environment data of each modality, and then analyze the data of the plurality of modalities. The multi-modality fusion model can first process the data of different modalities by using a special network, for example, using BERT to process text data, using CNN to process image data, and using Transformer to process time series data; after data processing, the data of different modalities can be mapped to a unified semantic space to realize cross-modality semantic alignment, thereby establishing a correlation between the data of different modalities, for example, correlating the address environment data and the device environment data.

[0261] In step 1105, the perception layer 1020 performs data analysis on the order environment data of the plurality of target modalities and the target operation data based on the multi-modality fusion model, and finds the target order scenario in a plurality of predetermined order scenarios.

[0262] After the perception layer 1020 acquires the order environment data of the plurality of target modalities and the target operation data, the perception layer 1020 can find the target order scenario in a plurality of predetermined order scenarios.

[0263] The target order scene can be found from the multiple predetermined order scenes according to the order environment data of the multiple target modalities and the target operation data.

[0264] The multi-modal fusion model can analyze the data of each target modality and the target operation data, and then analyze the order environment data and the target operation data of the multiple target modalities. In this way, the target order scene that matches the order data and the target operation data of the multiple target modalities is found from the multiple predetermined order scenes through comprehensive analysis of the order environment data and the target operation data of the multiple target modalities.

[0265] After finding the target order scene, the perception layer 1020 sends the target order scene to the agent coordination layer 1030.

[0266] In step 1106, the agent coordination layer 1030 finds the target order agent from the multiple order agents based on the target order scene sent by the perception layer 1020.

[0267] The order agent can be an artificial intelligence module that can perceive the environment and make corresponding decisions. Different order agents can have different functions, and each agent obtains corresponding data for its function and analyzes the data using a corresponding artificial intelligence model to obtain an order analysis result.

[0268] For different target order scenes, the order analysis elements that need to be focused on are different. For example, the multiple order agents include a financial analysis agent, a product analysis agent, a security analysis agent, and a payment method analysis agent. When the target order scene is a daily fixed expenditure scene, the analysis of the value and cost performance of the product is not required. This is because daily fixed expenditure, such as rent or electricity expenditure, is a scenario that must be paid, and the product value and product cost performance have little effect on the payment scheme recommendation. Therefore, the financial analysis agent, the security analysis agent, and the payment method analysis agent can be used as the target order agent.

[0269] The target order scene target order agent can be pre-set or found by a pre-set neural network model from the scene data of the target order scene.

[0270] In step 1107, the agent coordination layer 1030 calls multiple target order agents for order analysis to obtain multiple order analysis results.

[0271] The agent coordination layer 1030 can call multiple target order agents. Each target order agent can obtain data based on the corresponding function and perform analysis to obtain a corresponding order analysis result.

[0272] The order agent includes a financial analysis agent, a product analysis agent, a security analysis agent, and a payment method analysis agent. The financial analysis agent can be used to analyze the financial status of the order initiator, and the financial analysis result obtained by the financial analysis agent includes an order payment scheme recommendation result obtained based on the financial status analysis.

[0273] The product analysis agent can be used to analyze the product value and product cost performance of the target product in the target order, and the product analysis result obtained by the product analysis agent includes an order payment scheme recommendation result obtained based on the target product value and target product cost performance analysis.

[0274] The security analysis agent can be used to analyze the credit level of the product provider and the security of the order payment environment. The security analysis result obtained by the security analysis agent includes an order payment scheme recommendation result obtained based on the product provider and the order payment environment security level analysis.

[0275] The payment method analysis agent can be used to directly analyze the availability of multiple payment schemes. The payment method analysis result obtained by the payment method analysis agent includes a recommendation result obtained by directly analyzing multiple order payment schemes.

[0276] Step 1108, the decision layer 1040 obtains historical order payment data of the order initiator, and obtains usage scores corresponding to multiple order payment schemes based on the historical order payment data.

[0277] The historical order payment data of the order initiator can include historical order payment scenarios corresponding to historical orders of the order initiator and historical order payment schemes selected by the order initiator. The historical order payment data of the order initiator can be obtained based on payment records recorded in the order payment device. According to the historical order payment data, the order payment preference of the order initiator can be determined.

[0278] The usage score corresponding to the order payment scheme can represent the preference degree of the order payment scheme for the order initiator. The higher the usage score, the more the order initiator prefers to use the order payment scheme.

[0279] The usage score corresponding to the order payment scheme can be obtained by data analysis of the historical order payment data through a preset neural network model. The historical order payment data is input into the preset neural network model, and the usage score corresponding to each order payment scheme is output.

[0280] Step 1109, the decision layer 1040 obtains weights corresponding to multiple target order agents and historical order payment data.

[0281] The weights corresponding to the plurality of target order agents and the historical order payment data can indicate the influence degree of the plurality of target order agents and the historical order payment data on the generation of the target order payment scheme.

[0282] The weights can include a first weight corresponding to the historical order payment data, a second weight corresponding to the financial analysis agent, a third weight corresponding to the product analysis agent, a fourth weight corresponding to the security analysis agent, and a fifth weight corresponding to the payment method analysis agent. The first weight can indicate the influence degree of the historical payment preference on the generation of the target order payment scheme, the second weight can indicate the influence degree of the financial analysis result on the generation of the target order payment scheme, the third weight can indicate the influence degree of the product analysis result on the generation of the target order payment scheme, the fourth weight can indicate the influence degree of the security analysis result on the generation of the target order payment scheme, and the fifth weight can indicate the influence degree of the payment method analysis result on the generation of the target order payment scheme.

[0283] The first weight, the second weight, the third weight, the fourth weight, and the fifth weight can be obtained by the second neural network performing data analysis on the plurality of target modal order environment data.

[0284] In step 1110, the decision layer 1040 generates the target order payment scheme based on the weights corresponding to the plurality of target order agents and the historical order payment data, the order analysis results sent by the agent cooperation layer 1030, and the historical order payment data.

[0285] The order analysis result corresponding to each target order agent includes a recommendable score corresponding to each order payment scheme. Therefore, the order analysis result sent by the agent cooperation layer 1030 to the decision layer 1040 includes: the first recommendable score corresponding to the plurality of order payment schemes in the financial analysis result; the second recommendable score corresponding to the plurality of order payment schemes in the product analysis result; the third recommendable score corresponding to the plurality of order payment schemes in the security analysis result; and the fourth recommendable score corresponding to the plurality of order payment schemes in the payment method analysis result.

[0286] Generating the target order payment scheme according to the historical order payment data and the plurality of order analysis results can utilize the historical order payment data and the weights corresponding to the plurality of target order agents, and perform weighted calculation on the use score of each order payment scheme, the recommendable score corresponding to each order payment scheme in the order analysis result corresponding to the plurality of target order agents, and generate the target order payment scheme.

[0287] Further, based on the first weight, the second weight, the third weight, the fourth weight and the fifth weight, the usage score, the first recommendable score, the second recommendable score, the third recommendable score and the fourth recommendable score corresponding to each order payment scheme are weighted and calculated to obtain a recommendation score of each order payment scheme, and the target order payment scheme is generated based on the recommendation scores corresponding to the plurality of order payment schemes.

[0288] In step 1111, the display layer 1010 displays the target order payment scheme sent by the decision layer 1040 to the order initiator.

[0289] The display layer 1010 can display an order payment interface. The order payment interface includes a payment method selection area, and the payment method selection area includes the target order payment scheme and other selectable order payment schemes. The order initiator can select the target order payment scheme according to the actual scene needs, or can select an order payment scheme that is not recommended.

[0290] The order payment interface can also display a payment scene analysis area. The payment scene analysis area displays order analysis information of the target order, which can include budget analysis, product analysis and security information of the target order, etc.

[0291] In step 1112, the decision layer 1040 records the selected order payment scheme and updates the historical preference model based on the selected order payment scheme.

[0292] The payment method selected by the order initiator can be used for subsequent updating of the historical preference model, so that the decision layer can further understand the payment preferences of the order initiator.

[0293] The historical preference model can be updated and trained according to a predetermined period. In each training period, the order data stored in the period is used for model updating to improve the prediction accuracy of the historical preference model.

[0294] In step 1113, the display layer 1010 displays the order payment result and the order payment analysis conclusion to the order initiator.

[0295] After confirming the order payment, the display layer 1010 can display a payment result display interface. The payment result display interface can include the order payment result and the order payment analysis conclusion. The order payment analysis conclusion can include an analysis of this payment operation.

[0296] The order payment result and the order payment analysis result are displayed in the payment result display interface, so that the order initiator can view the analysis conclusion of the payment operation, which is conducive to improving the payment experience of the order initiator. The payment suggestions provided for the order initiator can also help the order initiator form a healthier financial habit.

[0297] Apparatuses and devices of embodiments of the present disclosure

[0298] It can be understood that, although each step in each of the above flowcharts is displayed in sequence represented by an arrow, these steps are not necessarily executed in the order represented by the arrow. Unless otherwise specified in the embodiments, the execution of these steps is not strictly limited in sequence, and these steps can be executed in other orders. Moreover, at least some of the steps in the above flowcharts can include multiple steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily sequential, but can be executed alternately or alternately with at least part of other steps or steps or stages in other steps.

[0299] It should be noted that, in each specific embodiment of the present application, when it is necessary to perform relevant processing according to data related to the characteristics of the target content, such as target content attribute information or attribute information set, the permission or consent of the target content will be obtained first, and the collection, use and processing of these data will comply with relevant laws, regulations and standards. In addition, when the present application needs to obtain target content attribute information, it will obtain the separate permission or separate consent of the target content through a pop-up window or by jumping to a confirmation page, and after obtaining the separate permission or separate consent of the target content, the necessary target content related data for the normal operation of the present application will be obtained.

[0300] Figure 13 A structural schematic diagram of an order processing apparatus 1300 provided by an embodiment of the present disclosure is provided. The order processing apparatus 1300 comprises:

[0301] An acquisition unit 1310, configured to receive a payment instruction for a target order, and acquire order environment data of multiple modalities associated with the target order;

[0302] A search unit 1320, configured to search for a target order scene from a plurality of predetermined order scenes according to the order environment data of multiple modalities;

[0303] An analysis unit 1330, configured to call a plurality of order agents for order analysis based on the target order scene, and obtain a plurality of order analysis results;

[0304] A generation unit 1340, configured to acquire historical order payment data of an order initiating object of the target order, and generate a target order payment scheme according to the historical order payment data and the plurality of order analysis results;

[0305] A recommendation unit 1350, configured to recommend the target order payment scheme to the order initiating object.

[0306] Optionally, in an implementation, the finding unit 1320 is specifically configured to:

[0307] obtain the multi-modal fusion model;

[0308] perform data analysis on the order environment data of the plurality of modalities based on the multi-modal fusion model, and find the target order scenario from the plurality of predetermined order scenarios.

[0309] Optionally, in an implementation, the finding unit 1320 is specifically configured to:

[0310] obtain order category information of the target order, and find a plurality of target modalities for performing data analysis based on the order category information;

[0311] perform data analysis on the order environment data of the plurality of target modalities based on the multi-modal fusion model, and find the target order scenario from the plurality of predetermined order scenarios.

[0312] Optionally, in an implementation, the finding unit 1320 is specifically configured to:

[0313] obtain a plurality of historical order environment data in a predetermined time period;

[0314] find the target order scenario from the plurality of predetermined order scenarios according to the order environment data of the plurality of modalities and historical order scenarios corresponding to the plurality of historical order environment data.

[0315] Optionally, in an implementation, the finding unit 1320 is specifically configured to:

[0316] query, from the plurality of historical order environment data, reference order environment data that has a matching degree reaching a predetermined condition with the order environment data of the plurality of modalities;

[0317] find the target order scenario from the plurality of predetermined order scenarios based on the reference order environment data.

[0318] Optionally, in an implementation, the obtaining unit 1310 is specifically configured to:

[0319] obtain historical operation data of an order initiating object and order environment data of a plurality of modalities associated with a target order;

[0320] the finding unit 1320 is specifically configured to:

[0321] find the target order scenario from the plurality of predetermined order scenarios according to the order environment data of the plurality of modalities and the historical operation data.

[0322] Optionally, in an implementation, the finding unit 1320 is specifically configured to:

[0323] query target operation data related to the order environment data in historical operation data;

[0324] perform environment analysis on the order environment data and the target operation data of multiple modalities based on the first neural network, and find a target order scenario in multiple predetermined order scenarios.

[0325] Optionally, in an embodiment, the analysis unit 1330 is specifically configured to:

[0326] invoke the financial analysis agent, the product analysis agent, the security analysis agent, and the payment method analysis agent based on the target order scenario to perform order analysis, and obtain a financial analysis result, a product analysis result, a security analysis result, and a payment method analysis result;

[0327] The generation unit 1340 is specifically configured to:

[0328] obtain historical order payment data of an order initiator of the target order, and generate a target order payment scheme based on the historical order payment data, the financial analysis result, the product analysis result, the security analysis result, and the payment method analysis result.

[0329] Optionally, in an embodiment, the financial analysis result includes a first recommendable score corresponding to multiple order payment schemes, the product analysis result includes a second recommendable score corresponding to the multiple order payment schemes, the security analysis result includes a third recommendable score corresponding to the multiple order payment schemes, and the payment method analysis result includes a fourth recommendable score corresponding to the multiple order payment schemes;

[0330] The generation unit 1340 is specifically configured to:

[0331] obtain a use score corresponding to the multiple order payment schemes based on the historical order payment data;

[0332] generate the target order payment scheme based on the use score, the first recommendable score, the second recommendable score, the third recommendable score, and the fourth recommendable score corresponding to the multiple order payment schemes.

[0333] Optionally, in an embodiment, the generation unit 1340 is specifically configured to:

[0334] obtain a first weight corresponding to the historical order payment data, a second weight corresponding to the financial analysis agent, a third weight corresponding to the product analysis agent, a fourth weight corresponding to the security analysis agent, and a fifth weight corresponding to the payment method analysis agent;

[0335] The first weight, the second weight, the third weight, the fourth weight and the fifth weight are used to weight the use score, the first recommendable score, the second recommendable score, the third recommendable score and the fourth recommendable score, and a target order payment scheme is generated from the plurality of order payment schemes.

[0336] Optionally, in an embodiment, the generating unit 1340 is specifically configured to:

[0337] obtain a second neural network;

[0338] perform environment analysis on the order environment data of the plurality of modalities based on the second neural network, and generate the first weight corresponding to the historical order payment data, the second weight corresponding to the financial analysis agent, the third weight corresponding to the product analysis agent, the fourth weight corresponding to the security analysis agent and the fifth weight corresponding to the payment method analysis agent.

[0339] Optionally, in an embodiment, the analyzing unit 1330 is specifically configured to:

[0340] find a plurality of target order agents from the plurality of order agents based on the target order scenario;

[0341] perform order analysis on the target order scenario by using the plurality of target order agents, and obtain a plurality of order analysis results.

[0342] Optionally, in an embodiment, the order processing apparatus 1300 further comprises:

[0343] a first display unit (not shown) configured to display an order payment interface in response to the target order payment scheme recommended to the order initiating object, and display a payment method selection area in the order payment interface, the payment method selection area including the target order payment scheme;

[0344] a second display unit (not shown) configured to display an order payment result in response to a triggering operation on the selected payment scheme in the payment method selection area.

[0345] Optionally, in an embodiment, the first display unit (not shown) is specifically configured to:

[0346] display an order payment interface in response to the target order payment scheme recommended to the order initiating object, and display a payment method selection area and a payment scenario analysis area in the order payment interface, the payment scenario analysis area including a payment adjustment control;

[0347] in response to a triggering operation on the payment adjustment control, jump to a payment adjustment pop-up window.

[0348] Optionally, in an embodiment, the second display unit (not shown) is specifically configured to:

[0349] In response to the triggering operation on the selected payment scheme in the payment method selection area, a payment result display interface is displayed, and the payment interface display interface includes an order payment result and an order payment analysis conclusion.

[0350] Reference Figure 14 , Figure 14 To realize the structure block diagram of part of the terminal 140 of the embodiment of the present disclosure, the terminal includes: radio frequency (RF) circuit 1410, memory 1415, input unit 1430, display unit 1440, sensor 1450, audio circuit 1460, wireless fidelity (WiFi) module 1470, processor 1480, and power supply 1490, and the like. Those skilled in the art can understand that the terminal 140 structure shown does not constitute a limitation on mobile phones or computers, and can include more or fewer components than shown, or combine certain components, or different component arrangements. Figure 14 The terminal 140 structure shown does not constitute a limitation on mobile phones or computers, and can include more or fewer components than shown, or combine certain components, or different component arrangements.

[0351] The RF circuit 1410 can be used for receiving and sending signals in the process of receiving or calling, and in particular, receiving the downlink information of the base station and processing it by the processor 1480; in addition, the uplink data is sent to the base station.

[0352] The memory 1415 can be used to store software programs and modules, and the processor 1480 executes various functions of the content terminal and text summary generation by running the software programs and modules stored in the memory 1415.

[0353] The input unit 1430 can be used to receive input digital or character information, and generate key signal input related to the setting and function control of the content terminal. Specifically, the input unit 1430 can include a touch panel 1431 and other input devices 1432.

[0354] The display unit 1440 can be used to display input information or provided information and various menus of the content terminal. The display unit 1440 can include a display panel 1441.

[0355] The audio circuit 1460, speaker 1461, and microphone 1462 can provide an audio interface.

[0356] In the present embodiment, the processor 1480 included in the object terminal 140 can execute the order processing method of the preceding embodiments.

[0357] The object terminal 140 of the embodiments of the present disclosure includes, but is not limited to, a mobile phone, a computer, a smart voice exchange device, a smart home appliance, a vehicle-mounted terminal, an aircraft, and the like. The embodiments of the present disclosure can be applied to various scenarios, including but not limited to e-commerce, online payment, and the like.

[0358] Figure 15 A structural block diagram of a part of the server 110 of the embodiments of the present disclosure is shown. The server 110 can vary greatly due to different configurations or performances, and can include one or more central processing units (CPUs) 1522 (for example, one or more processors) and a memory 1532, one or more storage media 1530 (for example, one or more mass storage devices) storing application programs 1542 or data 1544. Among them, the memory 1532 and the storage media 1530 can be temporary storage or persistent storage. The programs stored in the storage media 1530 can include one or more modules (not shown in the figure), each of which can include a series of instruction operations in the server. Further, the central processing unit 1522 can be configured to communicate with the storage medium 1530 and execute a series of instruction operations in the storage medium 1530 on the server.

[0359] The server 110 can also include one or more power supplies 1526, one or more wired or wireless network interfaces 1550, one or more input / output interfaces 1558, and / or one or more operating systems 1541, such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM, and the like.

[0360] The central processing unit 1522 in the server 110 can be used to execute the order processing method of the embodiments of the present disclosure.

[0361] The embodiments of the present disclosure also provide a computer-readable storage medium for storing program codes, the program codes being used to execute the order processing method of the above-mentioned various embodiments.

[0362] The embodiments of the present disclosure also provide a computer program product including a computer program. The processor of the electronic device reads the computer program and executes, so that the electronic device executes the order processing method as described above.

[0363] The terms "first", "second", "third", "fourth" and the like in the description of the present disclosure and the above drawings, if any, are used to distinguish similar objects, and do not necessarily indicate a particular order or sequence. It should be understood that the data thus used can be interchanged, where appropriate, so that the embodiments of the present disclosure described herein can be implemented in other sequences than those illustrated or described herein. In addition, the terms "comprise" and "include" and any variations thereof, are intended to cover non-exclusive inclusion, for example, a process, method, system, product or apparatus that includes a list of steps or units as an example is not necessarily limited to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to such processes, methods, products or apparatuses.

[0364] It should be understood that in the present disclosure, "at least one" means one or more, and "multiple" means two or more. "And / or" is used to describe the relationship between the associated contents, which means that there can be three relationships, for example, "A and / or B" can represent three cases: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally represents an "or" relationship between the associated contents. "At least one of the following" or similar expressions means any combination of these items, including any combination of single or multiple items. For example, at least one of a, b or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0365] It should be understood that in the description of the embodiments of the present disclosure, the meaning of multiple (or multiple items) is two or more, greater than, less than, more than, etc. are not included in the number, above, below, etc. are understood to include the number.

[0366] In the embodiments of the present disclosure, the term "module" or "unit" refers to a computer program or a part of a computer program with a predetermined function, and works together with other related parts to achieve a predetermined target, and can be implemented entirely or partially by using software, hardware (such as a processing circuit or a memory) or a combination thereof. Similarly, one processor (or multiple processors or memories) can be used to implement one or more modules or units. In addition, each module or unit can be a part of an overall module or unit that includes the functions of the module or unit.

[0367] In several embodiments provided in the present disclosure, it should be understood that the disclosed system, device and method can be implemented in other manners. For example, the described device embodiments are merely schematic. For example, the division of the units is merely logical function division. There can be another division manner for the actual implementation. For example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections can be indirect couplings or communication connections through some interfaces, devices or units, and can be in electrical, mechanical or other forms.

[0368] The units described as separated components can or can not be physically separated, and the components displayed as units can or can not be physical units, i.e., can be located in one place, or can be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purposes of the embodiments.

[0369] In addition, each functional unit in the various embodiments of the present disclosure can be integrated in one processing unit, or each unit can exist physically as a separate unit, or two or more units can be integrated in one unit. The integrated unit can be implemented in the form of hardware, or in the form of software functional units.

[0370] If the integrated unit is implemented in the form of software functional units and sold or used as an independent product, it can be stored in a computer readable storage medium. Based on such an understanding, the technical solutions of the present disclosure essentially or the part that makes a contribution to the prior art, or all or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods in the various embodiments of the present disclosure. The foregoing storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), magnetic disk or optical disk, and various media that can store program codes.

[0371] It should also be understood that the various embodiments provided by the present disclosure can be combined in any manner to achieve different technical effects.

[0372] The above is a specific explanation of the embodiments of the present disclosure, but the present disclosure is not limited to the above-described embodiments, and those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present disclosure, and these equivalent modifications or substitutions are included in the scope defined by the claims of the present disclosure.

Claims

1. An order processing method, characterized by, The method comprises the following steps: receiving a payment instruction for a target order, obtaining order environment data of multiple modalities associated with the target order; finding a target order scenario from a plurality of predetermined order scenarios according to the order environment data of the multiple modalities; calling a plurality of order agents for order analysis based on the target order scenario, and obtaining a plurality of order analysis results; obtaining historical order payment data of an order initiator of the target order, and generating a target order payment scheme according to the historical order payment data and the plurality of order analysis results; recommending the target order payment scheme to the order initiator.

2. The method of claim 1, wherein, The method comprises the following steps: obtaining a multi-modal fusion model; based on the multi-modal fusion model, performing data analysis on the order environment data of the multiple modalities to find a target order scenario from a plurality of predetermined order scenarios.

3. The method of claim 2, wherein, The method comprises the following steps: obtaining order category information of the target order, and finding a plurality of target modalities for data analysis based on the order category information; based on the multi-modal fusion model, performing data analysis on the order environment data of the plurality of target modalities to find a target order scenario from a plurality of predetermined order scenarios.

4. The method of claim 1, wherein, The method comprises the following steps: obtaining a plurality of historical order environment data within a predetermined time period; based on the order environment data of the multiple modalities and the historical order scenarios corresponding to the plurality of historical order environment data, finding a target order scenario from a plurality of predetermined order scenarios.

5. The method of claim 4, wherein, The method comprises the following steps: querying the plurality of historical order environment data to find reference order environment data that meets a predetermined condition with the order environment data of the multiple modalities; based on the reference order environment data, finding a target order scenario from a plurality of predetermined order scenarios.

6. The method of claim 1, wherein, The method comprises the following steps: obtaining order environment data of multiple modalities associated with the target order and historical operation data of an order initiator; The method comprises the following steps: based on the order environment data of the multiple modalities and the historical operation data, finding a target order scenario from a plurality of predetermined order scenarios.

7. The method of claim 6, wherein, The method comprises the following steps: querying the historical operation data to find target operation data related to the order environment data; The method comprises the following steps: querying the historical operation data to find target operation data related to the order environment data; The first neural network is used to analyze the order environment data and the target operation data of the plurality of modalities, and a target order scenario is found in a plurality of predetermined order scenarios.

8. The method of claim 1, wherein, The target order scenario is used to call a plurality of order agents for order analysis, and a plurality of order analysis results are obtained, including: The target order scenario is used to call a plurality of order agents for order analysis, and a plurality of order analysis results are obtained, including: The historical order payment data of the order initiator of the target order is obtained, and a target order payment scheme is generated based on the historical order payment data and the plurality of order analysis results. The historical order payment data of the order initiator of the target order is obtained, and a target order payment scheme is generated based on the historical order payment data, the financial analysis result, the product analysis result, the safety analysis result, and the payment method analysis result.

9. The method of claim 8, wherein, The financial analysis result includes a first recommendable score corresponding to a plurality of order payment schemes, the product analysis result includes a second recommendable score corresponding to the plurality of order payment schemes, the safety analysis result includes a third recommendable score corresponding to the plurality of order payment schemes, and the payment method analysis result includes a fourth recommendable score corresponding to the plurality of order payment schemes. The target order payment scheme is generated based on the historical order payment data, the financial analysis result, the product analysis result, the safety analysis result, and the payment method analysis result, including: The usage score corresponding to the plurality of order payment schemes is obtained based on the historical order payment data. The target order payment scheme is generated based on the usage score, the first recommendable score, the second recommendable score, the third recommendable score, and the fourth recommendable score corresponding to the plurality of order payment schemes.

10. The method of claim 9, wherein, The target order payment scheme is generated based on the usage score, the first recommendable score, the second recommendable score, the third recommendable score, and the fourth recommendable score corresponding to the plurality of order payment schemes, including: The first weight corresponding to the historical order payment data, the second weight corresponding to the financial analysis agent, the third weight corresponding to the product analysis agent, the fourth weight corresponding to the safety analysis agent, and the fifth weight corresponding to the payment method analysis agent are obtained. The usage score, the first recommendable score, the second recommendable score, the third recommendable score, and the fourth recommendable score are weighted and calculated based on the first weight, the second weight, the third weight, the fourth weight, and the fifth weight, and a target order payment scheme is generated from the plurality of order payment schemes.

11. The method of claim 10, wherein, The acquiring the first weight corresponding to the historical order payment data, the second weight corresponding to the financial analysis agent, the third weight corresponding to the product analysis agent, the fourth weight corresponding to the security analysis agent and the fifth weight corresponding to the payment method analysis agent comprises: acquiring a second neural network; based on the second neural network, the order environment data of the plurality of modalities is analyzed, and the first weight corresponding to the historical order payment data, the second weight corresponding to the financial analysis agent, the third weight corresponding to the product analysis agent, the fourth weight corresponding to the security analysis agent and the fifth weight corresponding to the payment method analysis agent are generated.

12. The method of claim 1, wherein, The order analysis based on the target order scene is called by a plurality of order agents, and a plurality of order analysis results are obtained, comprising: Based on the target order scene, a plurality of target order agents are found from a plurality of order agents; Using the plurality of target order agents, the target order scene is analyzed to obtain a plurality of order analysis results.

13. The method of claim 1, wherein, After the target order payment scheme is recommended to the order initiating object, it further comprises: In response to the target order payment scheme recommended to the order initiating object, an order payment interface is displayed, and a payment method selection area is displayed in the order payment interface, and the payment method selection area includes the target order payment scheme; In response to the triggering operation of the selected payment scheme in the payment method selection area, the order payment result is displayed.

14. The method of claim 13, wherein, In response to the target order payment scheme recommended to the order initiating object, an order payment interface is displayed, and a payment method selection area and a payment scene analysis area are displayed in the order payment interface, and the payment scene analysis area contains a payment adjustment control; In response to the triggering operation of the payment adjustment control, the payment adjustment pop-up window is jumped to. In response to the triggering operation of the selected payment scheme in the payment method selection area, the order payment result is displayed, comprising:

15. The method of claim 13, wherein, In response to the triggering operation of the selected payment scheme in the payment method selection area, a payment result display interface is displayed, and the payment interface display interface includes order payment result and order payment analysis conclusion. Comprising:

16. An order processing apparatus, characterized by comprising: The acquisition unit is used for receiving the payment instruction of the target order, and acquiring the order environment data of a plurality of modalities associated with the target order; The finding unit is used for finding the target order scene in a plurality of predetermined order scenes according to the order environment data of the plurality of modalities; The analysis unit is used for calling a plurality of order agents based on the target order scene to perform order analysis and obtain a plurality of order analysis results; The generation unit is used for acquiring the historical order payment data of the order initiating object of the target order, and generating a target order payment scheme according to the historical order payment data and the plurality of order analysis results; The recommendation unit is used for recommending the target order payment scheme to the order initiating object. ​ 17. An electronic device, comprising a memory and a processor, the memory storing a computer program, characterized in that, The computer program is executed by the processor to implement the order processing method according to any one of claims 1-15.

18. A storage medium storing a computer program, characterized in that, The computer program is executed by the processor to implement the order processing method according to any one of claims 1-15.

19. A computer program product comprising a computer program which is read and executed by a processor of an electronic device, such that the electronic device implements the order processing method according to any one of claims 1-15.

Images