CN113807858A

CN113807858A - Data processing method based on decision tree model and related equipment

Info

Publication number: CN113807858A
Application number: CN202111115082.0A
Authority: CN
Inventors: 李佳颖; 汪凌峰
Original assignee: Weikun Shanghai Technology Service Co Ltd
Current assignee: Sinosoft Co ltd
Priority date: 2021-09-23
Filing date: 2021-09-23
Publication date: 2021-12-17
Anticipated expiration: 2041-09-23
Also published as: CN113807858B

Abstract

The embodiment of the application is applied to the field of artificial intelligence and discloses a data processing method and related equipment based on a decision tree model, wherein the data processing method comprises the following steps: receiving a resource data transfer request which is sent by a target account and carries a task identifier of the target account; acquiring first account information and first historical task information of a target account, and inputting the first account information and the first historical task information into a pre-trained target decision tree model; determining a first data volume corresponding to a first task indicated by the task identifier; carrying out identity verification on the target account, and transferring resource data of a first data volume to the target account; and determining a first node corresponding to the first task and task abnormal information, determining a second data volume according to the task abnormal information, and transferring resource data of the second data volume to the target account. By adopting the embodiment of the application, the loss of the resource data of the platform can be reduced, and the security of the resource data transfer of the platform can be improved. The present application relates to blockchain technology, and the data described above may be stored in blockchains.

Description

Data processing method based on decision tree model and related equipment

Technical Field

The invention relates to the technical field of artificial intelligence, in particular to a data processing method based on a decision tree model and related equipment.

Background

With the rapid development of the internet technology, a task platform appears, which refers to a platform capable of issuing a paid task provided by a product provider instantly, wherein the paid task can be popularized and sold for a specific product, and the like, and a user can complete the task after receiving the task on the task platform, so as to obtain resource data transferred by the task platform.

Currently, in order to attract more users, a task platform generally increases the speed of resource data accounting and transferring, for example, after a user accepts and completes a task, the task platform may transfer resource data corresponding to the task to the user. However, in some scenarios, for example, in a shopping scenario, the task is to sell a specific commodity, the user sells the commodity, the resource data transferred to the task platform is obtained, and a refund occurs to the subsequent commodity. Therefore, if the task is a failure, the product provider does not transfer the resource data corresponding to the task platform, but the task platform transfers the resource data of the task to the user, so that the task platform needs to keep up with the resource data of the task to the user, which results in loss of the platform resource data.

Disclosure of Invention

The embodiment of the application provides a data processing method, a data processing device, data processing equipment and a readable medium based on a decision tree model, which are beneficial to reducing the loss of resource data of a task platform, can also reduce the transfer risk of the resource data, and can improve the security of resource transfer of the task platform.

In a first aspect, an embodiment of the present application provides a data processing method based on a decision tree model, including:

receiving a resource data transfer request sent by a target account, wherein the resource data transfer request carries a task identifier of the target account, and the task identifier is used for indicating a first task;

acquiring first account information and first historical task information of the target account, and inputting the first account information and the first historical task information into a pre-trained target decision tree model to obtain a risk evaluation score of the target account;

determining that the risk evaluation result of the target account passes and determining a first data volume corresponding to the first task under the condition that the risk evaluation score is smaller than a risk evaluation threshold;

performing identity verification on the target account, and transferring the resource data of the first data volume to the target account under the condition that the identity verification result of the target account is passed;

and under the condition that the first task does not reach the task completion node, determining a first node and task abnormal information corresponding to the first task, determining a second data volume according to the task abnormal information, and transferring the resource data of the second data volume to the target account, wherein the second data volume is a data volume smaller than zero.

Further, before the inputting the first account information and the first historical task information into a pre-trained objective decision tree model and obtaining the risk assessment score of the objective account, the method further includes:

constructing a training sample set, wherein the training sample set comprises a positive sample and a negative sample, and the positive sample and the negative sample are determined according to the risk characteristics of second account information of a reference account in a plurality of accounts and the risk characteristics of second historical task information;

determining first proportion data of the number of reference accounts corresponding to the risk characteristics of the second account information in the total account number, and determining second proportion data of the number of reference accounts corresponding to the risk characteristics of the second historical task information in the total account number;

determining an information gain of a risk characteristic of the second account information according to the first proportion data, and determining an information gain of a risk characteristic of the second historical task information according to the second proportion data;

and inputting the information gain of the risk characteristic of the second account information and the information gain of the risk characteristic of the second historical task information into a preset decision tree model for training to obtain an initial decision tree model.

Further, the constructing the training sample set includes:

acquiring second account information and second historical task information of a reference account in the plurality of accounts, and determining risk characteristics in the second account information and risk characteristics in the second historical task information according to a preset risk evaluation item, wherein the reference account is any one of the plurality of accounts;

determining the proportion of the number of second task completion nodes executed by the reference account within a preset time length to the total number of executed tasks;

and if the proportion is smaller than the preset proportion threshold, the risk features in the second account information and the risk features in the second historical task information are used as negative samples, and the positive samples and the negative samples are determined to be the training sample set.

Further, after the inputting the information gain of the risk characteristic of the second account information and the information gain of the risk characteristic of the second historical task information into a preset decision tree model for training to obtain an initial decision tree model, the method further includes:

obtaining a plurality of paths according to the paths from the root node to each leaf node in the initial decision tree model;

acquiring the number of negative samples and the total number of samples corresponding to a target path in the plurality of paths, wherein the target path is any one of the plurality of paths;

determining a risk evaluation score corresponding to the target path according to the proportion of the number of the negative samples corresponding to the target path to the total number of the samples;

and marking the risk evaluation score corresponding to the target path at a leaf node of the target path to obtain the target decision tree model.

Further, the authenticating the target account includes:

determining whether the network quality of the terminal equipment of the target account reaches an authentication condition, wherein the network quality refers to the signal intensity of the terminal equipment for data transmission;

under the condition that the network quality reaches an authentication condition, performing video authentication on the target account to obtain an authentication result of the target account;

and under the condition that the network quality does not reach the authentication condition, sending a face authentication instruction to the terminal equipment, wherein the face authentication instruction carries verification information and receives an authentication result returned by the terminal equipment, and the verification information is verification data of a face image corresponding to the target account.

Further, the resource data transfer request also carries a second node corresponding to the first task; the determining the second data size according to the task exception information includes:

when the first node is a node previous to the second node, the task exception information is indication information for indicating a task execution failure;

and acquiring a first calculation method corresponding to the task exception information, wherein the first calculation method is used for indicating that the opposite number of the first data volume is determined as the second data volume.

Further, the determining the second data size according to the task exception information includes:

when the first node is a node previous to the task completion node, the task exception information is indication information for indicating that a task is not completed according to preset time;

acquiring a second calculation method corresponding to the task abnormal information, and determining the second data volume according to the second calculation method;

the determining the second data volume according to the second calculation method includes:

determining a third data amount according to the first data amount and the first node, wherein the third data amount is the data amount of the first task reaching the first node, and the third data amount is smaller than the first data amount;

the difference between the first data amount and the third data amount is determined as the second data amount.

In a second aspect, an embodiment of the present application provides a data processing apparatus based on a decision tree model, including:

a receiving unit, configured to receive a resource data transfer request sent by a target account, where the resource data transfer request carries a task identifier of the target account, and the task identifier is used to indicate a first task;

an obtaining unit, configured to obtain first account information and first historical task information of the target account, and input the first account information and the first historical task information into a pre-trained target decision tree model to obtain a risk assessment score of the target account;

a determining unit, configured to determine that a risk evaluation result of the target account passes and determine a first data volume corresponding to the first task when the risk evaluation score is smaller than a risk evaluation threshold;

a verification unit, configured to perform identity verification on the target account, and transfer the resource data of the first data size to the target account when the identity verification result of the target account passes;

the determining unit is further configured to, when it is determined that the first task does not reach the task completion node, determine a first node and task exception information corresponding to the first task, determine a second data size according to the task exception information, and transfer resource data of the second data size to the target account, where the second data size is a data size smaller than zero.

In addition, in this aspect, other alternative embodiments of the data processing apparatus based on the decision tree model can refer to the related contents of the above first aspect, and are not described in detail here.

In a third aspect, an embodiment of the present application provides a computer device, which includes a memory, a processor, and a transceiver; the processor is connected to the memory and the transceiver, respectively, where the memory stores computer program codes, and the processor and the transceiver are configured to call the program codes to execute the method provided by the first aspect and/or any possible implementation manner of the first aspect.

In a fourth aspect, embodiments provide a computer-readable storage medium storing a computer program which, when executed by a computer device, implements a decision tree model-based data processing method as disclosed in any one of the possible implementations of the first aspect.

In the embodiment of the application, the acquired account information of the target account and the historical task information are input into a pre-trained target decision tree model by receiving a resource data transfer request carrying a task identifier sent by the target account, a risk evaluation score of the target account is obtained, when the risk evaluation score is smaller than a risk evaluation threshold value, the risk evaluation result of the target account is determined to be passed, namely, the risk evaluation can be performed on the target account, when the risk is lower, a first data volume of a task executed by the target account can be determined, the identity verification is performed on the target account, when the identity verification is passed, resource data with the first data volume is transferred to the target account, and when the task is determined not to reach a task completion node, the currently reached node and task abnormal information of the task are determined, and determining a second data volume according to the task abnormal information, wherein the second data volume is a data volume smaller than zero, and transferring the resource data of the second data volume to the target account. On one hand, risk assessment is carried out on the target account through the target decision tree model, and the first resource data are transferred under the condition that the risk assessment is passed, so that the safety of transferring the task platform resource data is improved. On the other hand, the node currently reached by the task is determined, the node is divided according to the completion node of the task, and the negative resource data is determined according to the task abnormal information, so that the negative resource data is transferred to the target account, the loss of the task platform is reduced, and the transfer risk of the resource data is reduced.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.

FIG. 1 is a schematic structural diagram of a data processing system based on a decision tree model according to an embodiment of the present application;

FIG. 2 is a schematic flow chart of a data processing method based on a decision tree model according to an embodiment of the present application;

FIG. 3 is another schematic flow chart of a data processing method based on a decision tree model according to an embodiment of the present application;

FIG. 4 is a schematic structural diagram of a data processing apparatus based on a decision tree model according to an embodiment of the present application;

fig. 5 is a schematic structural diagram of a computer device according to an embodiment of the present application.

Detailed Description

The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.

A data processing method based on a decision tree model according to an embodiment of the present application is schematically described below with reference to fig. 1 to 3.

The data processing method based on the decision tree model provided by the embodiment of the application can be applied to a platform, and the task platform can be borne in a data processing system or an application program based on the decision tree model. In some embodiments, the task platform is a platform that provides paid tasks for users, and the task platform may also provide a resource data transfer function, and when a user completes a task, transfer resource data corresponding to the completed task to an account of the user. The user is the user of the task platform, and can receive the compensated task in the task platform and then obtain the resource data transferred to the account by the task platform after the task is completed. In some embodiments, the task platform may establish a communication connection with a terminal device of a user, and receive a resource data transfer request initiated by the terminal device. In some embodiments, the task platform may also establish a communication connection with a terminal device and/or a server of the task provider. In some embodiments, the manner of the Communication connection may include, but is not limited to, Wireless Communication technology (WIFI), bluetooth, Near Field Communication (NFC), and the like.

Specifically, the task platform may receive a resource data transfer request sent by a user through a target account, where the resource data transfer request carries a task identifier of the target account, and the task identifier is used to indicate a task executed by the user of the target account, that is, a task corresponding to the transferred resource data; and then the task platform acquires the account information and the historical task information of the target account, and inputs the acquired information into a pre-trained target decision tree model to obtain the risk assessment score of the target account. The method comprises the steps that risk assessment is carried out on a target account of a user, if the risk assessment is not passed, resource data are not transferred into the target account by a task platform, otherwise, when the risk assessment result is passed, namely under the condition that the risk assessment score of the target account output by a target decision tree model is smaller than a risk assessment threshold value, the risk assessment result of the target account is determined to be passed, the data volume corresponding to a task is determined, identity verification is carried out on the target account, and under the condition that the identity verification is passed, the resource data of the data volume corresponding to the task are transferred into the target account. And finally, the task platform determines the node where the task is currently located and task abnormal information under the condition that the task does not reach the task completion node, and determines another data volume according to the task abnormal information, wherein the data volume is less than zero and is used for settling the resource data of the multi-transfer target account, namely transferring the data volume to the target account for settlement.

The data processing method based on the decision tree model provided by the embodiment of the application relates to the technologies of artificial intelligence, machine learning and the like, wherein: artificial Intelligence (AI) is a theory, method, technique and application system that uses a digital computer or a machine controlled by a digital computer to simulate, extend and expand human Intelligence, perceive the environment, acquire knowledge and use the knowledge to obtain the best results. The artificial intelligence infrastructure generally includes technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technologies, operation/interaction systems, mechatronics, and the like. The artificial intelligence software technology mainly comprises a computer vision technology, a robot technology, a biological recognition technology, a voice processing technology, a natural language processing technology, machine learning/deep learning and the like. Machine learning is a multi-disciplinary cross specialty, covers probability theory knowledge, statistical knowledge, approximate theoretical knowledge and complex algorithm knowledge, uses a computer as a tool and is dedicated to a real-time simulation human learning mode, and knowledge structure division is carried out on the existing content to effectively improve learning efficiency.

Based on the above description, a data processing system based on a decision tree model provided by the embodiment of the present application is schematically illustrated in conjunction with fig. 1.

Referring to fig. 1, fig. 1 is a schematic structural diagram of a data processing system based on a decision tree model according to an embodiment of the present disclosure. The data processing system based on the decision tree model comprises: a task platform 101, one or more terminal devices (terminal device 102 and terminal device 103), and one or more task providing devices (task providing device 104, task providing device 105, and task providing device 106). One or more terminal devices and the task platform 101 may be connected directly or indirectly through a wired or wireless manner. One or more task providing devices may also be directly or indirectly connected to the task platform 101 via wires or wirelessly. Optionally, the data processing system based on the decision tree model may further include a training device 107, where the training device may be configured to obtain data in the task platform 101, and train the decision tree model according to the obtained data to obtain the target decision tree model. The task platform 101 and the training device 107 may be the same device or different devices, and in the case that the task platform 101 and the training device 107 are not the same device, the task platform 101 and the training device 107 may be directly or indirectly connected in a wired or wireless manner.

It should be noted that the number and the form of the devices shown in fig. 1 are used for example, and do not constitute a limitation to the embodiments of the present application, and in practical applications, the data processing system based on the decision tree model may further include three or more terminal devices, or three or more task providing devices, that is, the task platform may be directly or indirectly connected with three or more terminal devices in a wired or wireless manner, and the task platform may also be directly or indirectly connected with three or more task providing devices in a wired or wireless manner. In the embodiment of the present application, two terminal devices (terminal device 102 and terminal device 103), one task platform 101, and three task providing devices (task providing device 104, task providing device 105, and task providing device 106) are taken as examples, and the task platform 101 and the training device 107 are the same device.

The two terminal devices (terminal device 102 and terminal device 103) can be terminal devices of different users, the task platform 101 can provide a paid task for each user, the two different users can check the paid tasks provided by the task platform 101 through the terminal devices respectively and interact with the task platform 101, so that a plurality of paid tasks provided by the task platform 101 can be checked, the paid tasks provided by the task platform 101 can be received through the terminal devices, and after the users complete the tasks, the tasks can return to the task platform 101 through the terminal devices, so that resource data transferred into the task platform 101 can be obtained. The user registers an account in the task platform 101 through the terminal device, and the task platform 101 counts the task execution condition of the user and transfers the resource data to the user through the account of the user.

Specifically, after a user completes a task, a resource data transfer request can be sent to the task platform 101 through the terminal device, the account of the user is a target account, the resource data transfer request carries a task identifier of the target account, the task identifier is used for indicating a task executed by the user, further, the task platform 101 obtains account information and historical task information of the target account, and the account information and the historical task information are input into a pre-trained target decision tree model to obtain a risk assessment score of the target account of the user; determining that the risk evaluation result of the target account passes under the condition that the risk evaluation score is smaller than a risk evaluation threshold value, and determining a first data volume corresponding to the task, namely, under the condition that the risk evaluation result passes, transferring resource data to the target account, performing identity verification on the target account before transferring the resource data, and under the condition that the identity verification result of the target account passes, transferring the resource data of the first data volume to the target account; if the task platform 101 determines that the task executed by the user does not reach the task completion node, the node where the task is currently located and the task exception information are determined, a second data volume is determined according to the task exception information, the second data volume is a negative data volume, and the resource data of the second data volume are transferred to the target account, so that the loss of the task platform can be avoided.

The task platform 101, the terminal devices (terminal device 102 and terminal device 103), the task providing platform (task providing device 104, task providing device 105, and task providing device 106), and the training device 107 may be a smart phone, a tablet computer, a notebook computer, a desktop computer, a smart speaker, a smart watch, or the like; the device 101 of the application platform and the device 102 of the authorization platform may also be servers, for example, independent physical servers, a server cluster or a distributed system formed by a plurality of physical servers, or cloud servers providing basic cloud computing services such as cloud service, a cloud database, cloud computing, a cloud function, cloud storage, Network service, cloud communication, middleware service, domain name service, security service, Content Delivery Network (CDN), big data, and an artificial intelligence platform, which is not limited in this application.

Referring to fig. 2, please refer to fig. 2 for further detailed description of a data processing method based on a decision tree model provided in an embodiment of the present application, and fig. 2 is a schematic flow diagram of the data processing method based on the decision tree model provided in the embodiment of the present application. As shown in fig. 2, the data processing method based on the decision tree model may include:

201. and receiving a resource data transfer request sent by the target account.

In the embodiment of the application, in order to better enable the user to receive the compensated task of the platform and complete the task, the task platform may receive a resource data transfer request initiated by the user after the user receives and executes the task, and further transfer the resource data corresponding to the task received by the user to an account of the user. The target account is an account used by the user, and the user is a user of the task platform, receives the task of the task platform and completes the task. The resource data transfer request may be initiated after a user receives a task and performs a task, that is, when a target account reaches an execution node behind a task receiving node, the user may initiate a resource data transfer request to a task platform through the target account, where the resource data transfer request carries a task identifier of the target account, the task is used to indicate a first task, and the first task initiates a task corresponding to the transferred resource data for the target account and is a task that has been received and is being executed by the target account.

202. Acquiring first account information and first historical task information of the target account, and inputting the first account information and the first historical task information into a pre-trained target decision tree model to obtain a risk evaluation score of the target account.

In one possible implementation manner, the account information is basic information of an account, and the first account information is account information of the target account, where the account information may include a name of the user, a company of the user, a position of the user, a work place of the user, a type of a task performed by the user, a recommender of the user, and the like. Wherein the first account information may be derived based on information entered by the user. The first historical task information is information of tasks executed by the target account and task conditions, and may specifically include information of each task executed by the account. Specifically, the time of the task received by each task, the time of initiating the resource data transfer request, the time length for executing the task, and the like may be included; the task completion condition can also be included, and the task completion condition specifically includes whether the task executed by the target account reaches the task completion node, the proportion of the number of the tasks of the task completion node to the total number of the tasks, the time for the task completion node to receive the task and the time for the task completion node to reach; the method can further include task information of the nodes which do not reach the task, and specifically can include task nodes where the tasks which do not reach the task completion nodes are located, resource data transfer time corresponding to the tasks which do not reach the task completion nodes, execution time corresponding to the tasks which do not reach the task completion nodes, and the like.

In a possible implementation manner, the first account information and the first historical task information of the target account are input into a pre-trained target decision tree model, so as to obtain a risk assessment score of the target account. The decision tree is a classifier for training and predicting samples, and is a prediction model; he represents a mapping between object properties and object values. The decision tree comprises a plurality of nodes, the nodes are divided into root nodes and leaf nodes, the root nodes can correspond to at least two bifurcation paths, each path corresponds to one leaf node, and then the bifurcation can be continued until the bifurcation can not be performed, so that a decision tree model is obtained. Each node in the decision tree represents an object and each divergent path represents a possible attribute value, and each leaf node corresponds to the value of the object represented by the path traversed from the root node to the leaf node. The target decision tree model is pre-trained, and when the risk assessment score of the target account is actually required to be determined, the risk assessment score of the target account is determined only by directly using the target decision tree. Specifically, the first account information and the first historical task information of the target account are input into the pre-trained target decision tree model, so that the target decision tree model can determine a path from a root node to the leaf node in the target decision tree model according to the first account information and the first historical task information of the target account, and further determine a risk assessment score of the target account according to a score corresponding to the path determined in the target decision tree model.

Further, after the risk assessment score of the target account is obtained, the risk assessment score is judged, if the risk assessment score is larger than or equal to a risk assessment threshold value, it is determined that the risk of the target account is large, and it is determined that the risk assessment result of the target account does not pass. Otherwise, if the risk assessment score is smaller than the risk assessment threshold, determining that the risk of the target account is smaller, and determining that the risk assessment result of the target account passes.

203. And determining that the risk evaluation result of the target account passes and determining a first data volume corresponding to the first task when the risk evaluation score is smaller than a risk evaluation threshold.

In a possible implementation manner, the mission platform stores a risk assessment threshold, compares the risk assessment threshold with the risk assessment score of the target account, and if the mission platform determines that the risk assessment score is smaller than the risk assessment threshold, the mission platform determines that the risk of the target account is smaller, that is, the risk assessment result is passed. Furthermore, the task platform may transfer the resource data to the target account, that is, determine a first data volume corresponding to the task identifier carried in the resource data transfer request initiated by the target account, where the first data volume is a data volume of the resource data of the first task. When the task provider provides the tasks to the task platform through the task providing device, the task platform determines the data volume according to the data volume provided by the task provider. Therefore, the task platform only needs to acquire the first data volume corresponding to the first task according to the task identifier.

In a possible implementation manner, if the mission platform determines that the risk assessment score is greater than or equal to the risk assessment threshold, the mission platform determines that the risk of the target account is greater, that is, the risk assessment result is failed. And then the task platform does not need to acquire the first data volume corresponding to the first task and does not transfer the resource data to the target account. Optionally, the task platform may also send a prompt message that the risk assessment result does not pass to the target account, and prompt the user to complete the task, that is, increase the probability that the task reaches the task completion node to improve the risk assessment score, so as to strive to improve the risk assessment result of the target account in the subsequent risk assessment.

204. And performing identity verification on the target account, and transferring the resource data of the first data volume to the target account under the condition that the identity verification result of the target account is passed.

In a possible implementation manner, before transferring the resource data of the first data volume to the target account, the identity of the target account needs to be verified, and when the identity verification result of the target account is passed, the resource data of the first data volume is transferred to the target account. The task platform can obtain the network quality of the terminal device of the current target account, wherein the network quality refers to the signal strength of the terminal device for data transmission, after the task platform initiates an authentication request for the target account, the terminal device bearing the target account can feed back the current network quality of the terminal device to the task platform, and the network quality does not meet the authentication condition and can include that the network transmission rate is lower than a rate threshold, the network bandwidth is lower than a bandwidth threshold, the network packet loss rate is greater than a packet loss rate threshold, the network response duration is greater than a duration threshold, and the like.

In a possible implementation manner, the terminal device may send, to the task platform, indication information of network quality, where the indication information is used to indicate whether the terminal device may perform video authentication conditions, and the indication information may also indicate parameters including a current network transmission rate, a network packet loss rate, a network response duration, and the like of the terminal device, and the task platform determines whether the current terminal device reaches the video authentication conditions, and if the task platform determines that the current terminal device reaches the video authentication conditions, sends a request of video authentication to the terminal device, so as to perform video authentication on a face of a user in a video.

In another possible implementation manner, when the task platform determines that the current terminal device does not meet the video authentication condition, a face authentication instruction may be sent to the terminal device, where the face authentication instruction carries verification information, and the verification information is verification data of a face image corresponding to the target account. Specifically, the verification information may be verification data converted by the platform according to the facial image data of the user corresponding to the target account during registration. For example, after acquiring one or more images of a user corresponding to a target account, a platform converts the one or more images into verification data, where the verification data may be positions of feature points identified in a face image, such as positions of eyes, a nose, a mouth, and a chin, and calculates euclidean distances, curvatures, angles, and the like between the feature points, and may determine the positions of the feature points and associated information between the feature points as the verification data corresponding to the face image.

Specifically, the terminal device receives and collects a face image of a target user, converts the face image collected during authentication into feature data in the same manner, where the feature data may also include multiple feature points of the face of the user and associated information between the multiple feature points, and the terminal device may further match the verification data with the feature data, and if the matching degree is greater than a matching threshold, it is determined that the authentication result of the user passes, or it is determined that the authentication result of the user does not pass. The matching degree can be the similarity between the verification data and the feature data, and if the similarity is greater than a similarity threshold, the feature data of the face image is determined to be matched with the verification data, and the identity authentication of the user is determined to be passed. Otherwise, if the similarity between the verification data and the feature data is smaller than the similarity threshold, the feature data of the face image is determined not to be matched with the verification data, and the identity verification result of the user is determined not to be passed.

Further, after obtaining the authentication result of the target account, the terminal device may return the authentication result to the task platform. The task platform can transfer the resource data of the first data volume to the target account under the condition that the identity verification result of the target account is passed.

Optionally, the task platform does not transfer the resource data of the first data size to the target account under the condition that the identity authentication result of the target account is determined to be invalid, and may send a prompt message indicating that the identity authentication is invalid to the terminal device corresponding to the target account, where the prompt message may also be used to prompt the user to perform the second identity authentication, and the prompt message may also be used to prompt the user to keep a camera of the terminal device clean and clear, and so on.

205. And under the condition that the first task does not reach the task completion node, determining the first node and the task abnormal information corresponding to the first task, determining a second data volume according to the task abnormal information, and transferring the resource data of the second data volume to the target account.

At present, after the resource data is transferred to the target account, if the first task is not completed, the task platform may determine a first node and task exception information corresponding to the first task, and further determine a second data volume corresponding to the task exception information, where the second data volume is a negative data volume, and transfer the resource data of the second data volume to the target account. The first node is a node where the current task is located, and the task abnormal information is the reason why the current task is executed and finished but the task completion node is not reached. The second amount of data is determined based on the task exception information. And transferring the resource data corresponding to the second data volume smaller than zero to the target account so as to reduce the condition that the user does not complete the first task because the user obtains the resource data of the first data volume of the first task.

For example, taking a shopping scenario as an example, a user may market a certain product to a consumer, the consumer purchases the product, but since the product can be returned within a certain time threshold, the consumer does not confirm the receipt of the product, and initiates a request for returning money and returning goods, the task provider does not obtain a profit based on the returned goods, and therefore the resource data of the task is not transferred to the task platform, but the task platform transfers the resource data of the first task to the target account, which results in a loss of the resource data of the task platform. Therefore, the task platform can determine a second data volume based on the task and transfer the resource data of the second data volume to the target account, thereby ensuring the security of the resource data of the task platform.

The task can be divided into a plurality of processing nodes, and the specific task node can be determined according to the task type provided by the task provider and the role type of the task provider. For example, in a scenario of selling a certain product, a task provider may be a brand party of a certain product, i.e., a role type may be a brand party, and a task is to sell a product provided by the brand party. The nodes of the task may include: the method comprises the following steps of releasing a task, receiving the task by a user, selling a product, receiving goods by a consumer and confirming completion of the task by a brand party. As another example, in the scenario of a loan product, the task provider may be some loan platform, i.e., the role type may be the loan platform and the task is to offer loans. The nodes of the task may include: the method comprises the following steps of releasing a task, receiving the task by a user, determining a client, paying by a product provider, repaying by the client and confirming the completion of the task by the product provider.

In a possible implementation manner, the task platform determines a current node of the first task, that is, the first node, in a case where it is determined that the task has completed the node that is not reached by the first task. When the target account initiates a resource data transfer request to the task platform, the resource data transfer request may carry a second node, where the second node may be a previous node of the first node, and the task exception information may be used to indicate that the first task fails to be executed, and then the task platform may obtain a first calculation method corresponding to the task exception information, and determine the second data size by using the first calculation method. The first calculation method is used for indicating that the opposite number of the first data volume is determined as the second data volume, namely the task fails to execute, and returning all the resource data previously transferred into the task platform.

Illustratively, in a scene of live delivery, a user corresponding to the target account is an anchor, the first task is to sell a certain product, the anchor receives a task, the first task reaches a sold node (a second node) in the live delivery process, at this time, the anchor can initiate a resource data transfer request, and the resource data transfer request can carry the sold node. And after the risk assessment is passed, the task platform carries out identity verification on the target account, and after the identity verification is passed, the resource data corresponding to the first task, namely the resource data with the first data volume, is transferred to the target account. But the subsequent consumer initiates the goods return and refund, which is equivalent to that the task returns to the last node, namely the first node, the node of the task accepted by the user, and the task exception information is used for indicating that the task executed by the user fails. The task platform may obtain a first calculation method corresponding to the task exception information when the first node is a previous node of the second node and the task exception information is used to indicate that the task execution fails, where the first calculation method is used to indicate that the inverse of the first data size is determined as the second data size. The task exception information may include a task type, an execution condition, and a failure reason of the task. The second data volume is a negative data volume, which is equivalent to the failure of the first task execution of the target account, and the full amount is returned to the resource data.

In another possible implementation manner, the first node may be a previous node of the task completion node, and the task exception information may be used to indicate that the task is not completed according to the preset time, and the task platform may obtain a second calculation method corresponding to the task exception information and determine the second data size according to the second calculation method. The specific process of determining the second data amount according to the second calculation method may be: and the task platform determines a third data volume according to the first data volume and the first node, and determines the third data volume as the second data volume according to the first data volume and the third data volume. The third data volume is a corresponding data volume reaching the first task node, the third data volume is smaller than the first data volume, and the second calculation method is to determine a difference between the first data volume and the third data volume as the second data volume.

For example, in the above scenario of the loan product, the user sells the loan product to the customer, the customer makes a payment in advance in a payment period, that is, the customer does not make a payment according to the preset time, and if the payment is not made according to the preset time, the task is not completed according to the preset time, the current task node (that is, the first node) is a node previous to the task completion node, that is, the customer payment node, and the task exception information may be used to indicate that the task is not completed according to the preset time. The task platform may determine a third data volume according to the first data volume (i.e., the data volume obtained by the user who completes the task) and the first node (i.e., the customer payment node), where the third data volume is a data volume determined by the task node according to a certain proportion in the third data volume, i.e., the data volume obtained by the user executing the task, and therefore, the third resource data is smaller than the first resource data.

Further, the task platform may determine a difference between the third amount of data and the first amount of data as the second amount of data. The third data volume, the first data volume, yields the second data volume. Since the third data amount is smaller than the first data amount, the second data amount is a number smaller than zero. And transferring the second amount of data to the target account.

Optionally, if the resource data in the balance of the target account is not presented, the remaining resource data and the second data amount are settled to obtain final remaining resource data. If the target account executes other tasks and can obtain resource data of other tasks, the other resource data can be used for offsetting the resource data of the second data volume, and when the balance of the target account is positive, the target account can initiate a discovery request.

Referring to fig. 3, fig. 3 is another schematic flow chart of a data processing method based on a decision tree model according to an embodiment of the present disclosure. It should be noted that, in the present application, the same or similar parts between the various embodiments may be mutually referred to. In the embodiments and the implementation methods/implementation methods in the embodiments in the present application, unless otherwise specified or conflicting in logic, terms and/or descriptions between different embodiments and between various implementation methods/implementation methods in various embodiments have consistency and can be mutually cited, and technical features in different embodiments and various implementation methods/implementation methods in various embodiments can be combined to form new embodiments, implementation methods, or implementation methods according to the inherent logic relationships thereof. The above-described embodiments of the present application do not limit the scope of the present application. As shown in fig. 3, the data processing method based on the decision tree model may include:

301. and constructing a training sample set, wherein the training sample set comprises positive samples and negative samples.

In a possible implementation manner, before applying the target decision tree, the task platform may construct a training sample set pair to construct a target decision tree model, where the training sample set for constructing the target decision tree model may be a plurality of accounts on the task platform, and specifically, the task platform may obtain second account information and second historical task information in reference accounts of the plurality of accounts, where the reference account is any one of the plurality of accounts. The second account information and the second historical task information may include a feature characterizing the reference account, where the feature may be a user feature corresponding to the reference account or a behavior feature of the reference account. For the reference account, the second account information and the second historical task information of the reference account are used as data sets, the data sets are classified, and each reference data is divided into a positive sample or a negative sample to obtain a training sample set.

Optionally, after the second account information and the second historical task information of the reference account are obtained, the second account information and the second historical task information of the reference account may be preprocessed, where the preprocessing may include removing data that does not have any help on data used for building the target decision tree model. Illustratively, the second account information includes a name of the user corresponding to the reference account, a company of the user corresponding to the reference account, a position of the user corresponding to the reference account, a work place of the user corresponding to the reference account, a type of the task executed by the user corresponding to the reference account, a recommender of the user corresponding to the reference account, and the like. During the preprocessing, the name of the user corresponding to the reference account, the company of the user corresponding to the reference account, the work place of the user corresponding to the reference account, and the recommender of the user corresponding to the reference account may be deleted.

Specifically, the risk characteristics in the second account information and the risk characteristics in the second historical task information may be determined according to preset risk assessment items. The preset risk items are categories used for screening out risk characteristics from the second account information and the second historical information, and the preset risk items can be set according to specific application scenarios. Optionally, if a certain risk feature in the second account information or the second historical information is not a discrete feature, that is, cannot be used to construct the objective decision tree, the preset risk item may further include a partition threshold for the risk feature, so that the risk feature is classified, and the objective decision tree model is better constructed. For example, when a risk feature in the second historical task is a task execution duration, and the duration is a continuous value, the preset risk item may further include a duration threshold corresponding to the risk feature, where the duration threshold is used to divide the risk feature, samples greater than or equal to the duration threshold are taken as one class, and samples less than the duration threshold are taken as another class. Further, the time length threshold may be two thresholds, and the time length for executing the task is divided into three categories, which is not limited in the present application.

Further, the task platform may determine that the number of tasks of a second task, executed by the reference account within a preset time period, reaching the task completion node is a proportion of the total number of executed tasks, if the proportion is greater than a preset proportion threshold, the risk features included in the second account information and the second historical task information of the reference account are used as positive samples, otherwise, if the proportion is less than or equal to the preset proportion threshold, the risk features included in the second account information and the second historical task information of the reference account are used as negative samples, and a training sample set is obtained.

302. And determining first proportion data of the number of the reference accounts corresponding to the risk characteristics of the second account information in the total account number, and determining second proportion data of the number of the reference accounts corresponding to the risk characteristics of the second historical task information in the total account number.

In one possible implementation, the task platform may construct a target decision tree model based on the positive and negative examples described above. Specifically, the task platform may determine the risk characteristics in the second account information and the proportion data of the risk characteristics in the second historical task information to obtain first proportion data and second proportion data, respectively. For example, one risk feature included in the second historical task information is a task execution time length, the task execution time length may be divided into two types, one type is greater than or equal to the time length threshold, and the other type is smaller than the time length threshold, then ratio data of the number of reference accounts corresponding to the two types in the risk feature in the total number of reference accounts is respectively counted, and then the second ratio data of the risk feature includes ratio data corresponding to each type in the risk feature of the task execution time length.

303. And determining an information gain of the risk characteristic of the second account information according to the first proportion data, and determining an information gain of the risk characteristic of the second historical task information according to the second proportion data.

After the proportion data of all risk features are obtained, the information entropy of the risk features of the second account information can be determined according to the first proportion data, the information gain of the risk features of the second account information is further determined according to the information entropy, similarly, the information gain of the risk features of the second historical task information is determined according to the second proportion data, and then the decision tree model is trained according to the information gain. Specifically, the information entropy can be calculated according to the ratio data as shown in formula 1:

where, ent (D) represents information entropy, D represents a training sample set, and K represents a total number of classes, for example, if the execution duration includes two classes, K is 2, pk is a proportion occupied by a current class sample, that is, in the first proportion data and the second proportion data, K represents that there may be K values of the risk feature, and a proportion in the kth class.

After the information entropy of each risk feature is obtained, the information gain of each risk feature is determined, wherein the information gain can be used for representing the influence of a certain risk feature on the result of classifying the branch node of the risk feature obtained after the training sample set D is divided. Specifically, calculating the information gain according to the information entropy of a certain risk characteristic can be shown as formula 2:

wherein Ent (D) represents information entropy, and D represents training sampleIn this set, K indicates that the risk feature a may have K categories, D_kRepresenting that the sample is concentrated on the risk characteristic a and takes the value of a_kThe number of samples of (1).

And further, learning and training the decision tree model according to the information gain of each risk characteristic to construct a target decision tree model.

304. And inputting the information gain of the risk characteristic of the second account information and the information gain of the risk characteristic of the second historical task information into a preset decision tree model for training to obtain an initial decision tree model.

The larger the value of the information gain is, the larger the influence on the subsequent classification is, so that the calculated information gains are sequenced from large to small, and the risk characteristic with the largest information gain is used as the root node of the target decision tree model.

Further, after the root node is selected, each branch of the risk characteristics of the root node continues to be divided. Similarly, the first branch (i.e. the first category of the risk feature) under the risk feature is divided according to another risk feature, the information gain of each risk feature under the first branch can be calculated, the risk feature with the largest information gain is selected as the leaf node of the first branch, and so on, until the division can not be performed, the learning and training of the decision tree are performed, and the initial decision tree model is obtained.

Further, according to the path from the root node to each leaf node in the initial decision tree model, obtaining a plurality of paths in the initial decision tree model, obtaining the number of negative samples of each path, determining a risk assessment score corresponding to the path according to the number of the negative samples and the total number of the samples, and marking the risk assessment score of the path at the position of the leaf node at the lowest layer of the path to obtain the target decision tree model. The negative sample is classified based on the fact that the proportion of the task executed by the reference account reaching the task completion node is smaller than or equal to a preset proportion threshold, and therefore the risk assessment score of the target decision tree model can be used for representing the risk that the task executed by the user corresponding to the path cannot reach the task completion node, and therefore the risk assessment score is used for carrying out risk assessment on the target account.

Referring to fig. 4, fig. 4 is a schematic structural diagram of a data processing apparatus based on a decision tree model according to an embodiment of the present disclosure. The decision tree model-based data processing apparatus 400 includes:

a receiving unit 401, configured to receive a resource data transfer request sent by a target account, where the resource data transfer request carries a task identifier of the target account, and the task identifier is used to indicate a first task;

an obtaining unit 402, configured to obtain first account information and first historical task information of the target account, and input the first account information and the first historical task information into a pre-trained target decision tree model to obtain a risk assessment score of the target account;

a determining unit 403, configured to determine that a risk evaluation result of the target account passes and determine a first data volume corresponding to the first task when the risk evaluation score is smaller than a risk evaluation threshold;

a verification unit 404, configured to perform identity verification on the target account, and transfer the resource data of the first data amount to the target account if the identity verification result of the target account is passed;

the determining unit 403 is configured to, if it is determined that the first task does not reach the task completion node, determine a first node and task exception information corresponding to the first task, determine a second data size according to the task exception information, and transfer resource data of the second data size to the target account, where the second data size is a data size smaller than zero.

Further, the decision tree model-based data processing apparatus 400 further includes:

the input unit 405, configured to input the first account information and the first historical task information into a pre-trained objective decision tree model, before obtaining the risk assessment score of the objective account, further includes:

a constructing unit 406, configured to construct a training sample set, where the training sample set includes a positive sample and a negative sample, and the positive sample and the negative sample are determined according to a risk feature of second account information of a reference account in the multiple accounts and a risk feature of second historical task information;

the determining unit 403 is further configured to determine first proportion data of the number of reference accounts corresponding to the risk characteristic of the second account information to the total number of accounts, and determine second proportion data of the number of reference accounts corresponding to the risk characteristic of the second historical task information to the total number of accounts;

the determining unit 403 is further configured to determine an information gain of the risk characteristic of the second account information according to the first proportion data, and determine an information gain of the risk characteristic of the second historical task information according to the second proportion data;

a training unit 407, configured to input the information gain of the risk feature of the second account information and the information gain of the risk feature of the second historical task information into a preset decision tree model for training, so as to obtain an initial decision tree model.

Further, the constructing unit 406 is specifically configured to:

the obtaining unit 402 is configured to obtain a plurality of paths according to paths from a root node to each leaf node in the initial decision tree model; acquiring the number of negative samples and the total number of samples corresponding to a target path in the plurality of paths, wherein the target path is any one of the plurality of paths;

the determining unit 403 is further configured to determine a risk assessment score corresponding to the target path according to a ratio of the number of negative samples corresponding to the target path to the total number of samples;

the labeling unit 408 is configured to label the risk assessment score corresponding to the target path at a leaf node of the target path, so as to obtain the target decision tree model.

Further, the verification unit 404 is specifically configured to:

Further, the resource data transfer request also carries a second node corresponding to the first task; the determining unit 403 is specifically configured to:

Further, the determining unit 403 is specifically configured to:

The detailed descriptions of the receiving unit 401, the obtaining unit 402, the determining unit 403, the verifying unit 404, the inputting unit 405, the constructing unit 406, the training unit 407, and the labeling unit 408 can be directly obtained by directly referring to the related descriptions in the method embodiments shown in fig. 2 to fig. 3, which are not repeated herein.

Referring to fig. 5, fig. 5 is a schematic structural diagram of a computer device according to an embodiment of the present disclosure, and as shown in fig. 5, a computer device 500 according to an embodiment of the present disclosure may include:

the processor 501, the transceiver 502, and the memory 505, the computer device 500 may further include: a user interface 504, and at least one communication bus 503. Wherein a communication bus 503 is used to enable connection communication between these components. The user interface 504 may include a Display (Display) and a Keyboard (Keyboard), and the memory 505 may be a high-speed RAM memory or a non-volatile memory (non-volatile memory), such as at least one disk memory. The memory 505 may alternatively be at least one memory device located remotely from the processor 501 and the transceiver 502. As shown in fig. 5, the memory 505, which is a type of computer storage medium, may include therein an operating system, a network communication module, a user interface module, and a device control application program.

In the computer device 500 shown in fig. 5, the transceiver 502 may provide network communication functions to enable communication between servers; while user interface 504 is primarily an interface for providing input to a user; and processor 501 may be configured to invoke a device control application stored in memory 505 to perform the following operations:

the transceiver 502 is configured to receive a resource data transfer request sent by a target account, where the resource data transfer request carries a task identifier of the target account, and the task identifier is used to indicate a first task;

the processor 501 is configured to obtain first account information and first historical task information of the target account, and input the first account information and the first historical task information into a pre-trained target decision tree model to obtain a risk assessment score of the target account;

the processor 501 is configured to determine that a risk evaluation result of the target account passes and determine a first data volume corresponding to the first task when the risk evaluation score is smaller than a risk evaluation threshold;

the processor 501 is configured to perform authentication on the target account, and transfer the resource data of the first data size to the target account if the authentication result of the target account passes;

the processor 501 is configured to, when it is determined that the first task does not reach the task completion node, determine a first node and task exception information corresponding to the first task, determine a second data size according to the task exception information, and transfer resource data of the second data size to the target account, where the second data size is a data size smaller than zero.

In a possible implementation manner, before the first account information and the first historical task information are input into a pre-trained objective decision tree model and a risk assessment score of the objective account is obtained, the processor 501 is further configured to:

In a possible implementation manner, the processor 501 constructs a training sample set, specifically for:

In a possible implementation manner, after the information gain of the risk feature of the second account information and the information gain of the risk feature of the second historical task information are input into a preset decision tree model for training to obtain an initial decision tree model, the processor 501 is further configured to:

In a possible implementation manner, the processor 501 is configured to perform authentication on the target account, and specifically, to:

In a possible implementation manner, the resource data transfer request further carries a second node corresponding to the first task; the processor 501 is configured to determine a second data size according to the task exception information, and specifically configured to:

In a possible implementation manner, the processor 501 is configured to determine a second data size according to the task exception information, and specifically configured to:

It should be understood that, in some possible embodiments, the processor 501 may be a Central Processing Unit (CPU), and the processor 501 may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), field-programmable gate arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, and the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

The memory 505 may include both read-only memory and random access memory and provides instructions and data to the processor. A portion of memory 505 may also include non-volatile random access memory.

In a specific implementation, the computer device 500 may execute the implementation manners provided in the steps in fig. 2 and fig. 3 through the built-in functional modules, which may specifically refer to the implementation manners provided in the steps, and are not described herein again.

Further, here, it is to be noted that: an embodiment of the present application further provides a computer-readable storage medium, where the computer-readable storage medium stores a computer program executed by the aforementioned computer device, and the computer program includes program instructions, and when the processor executes the program instructions, the processor can perform the description of any one of the methods in the embodiment corresponding to any one of fig. 2 or fig. 3, and therefore, the description of any one of the methods will not be repeated here. In addition, the beneficial effects of the same method are not described in detail. For technical details not disclosed in embodiments of the computer-readable storage medium referred to in the present application, reference is made to the description of embodiments of the method of the present application.

It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware associated with computer program instructions, and the above programs can be stored in a computer readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM), a Random Access Memory (RAM), or the like.

It is emphasized that the data may also be stored in a node of a blockchain in order to further ensure the privacy and security of the data. The block chain is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, a consensus mechanism and an encryption algorithm. A block chain (Blockchain), which is essentially a decentralized database, is a series of data blocks associated by using a cryptographic method, and each data block contains information of a batch of network transactions, so as to verify the validity (anti-counterfeiting) of the information and generate a next block. The blockchain may include a blockchain underlying platform, a platform product service layer, an application service layer, and the like.

The above disclosure is only for the purpose of illustrating the preferred embodiments of the present application and is not to be construed as limiting the scope of the present application, so that the present application is not limited thereto, and all equivalent variations and modifications can be made to the present application.

Claims

1. A data processing method based on a decision tree model is characterized by comprising the following steps:

determining that the risk evaluation result of the target account passes and determining a first data volume corresponding to the first task when the risk evaluation score is smaller than a risk evaluation threshold;

and under the condition that the first task does not reach the task completion node, determining a first node and task abnormal information corresponding to the first task, determining a second data volume according to the task abnormal information, and transferring resource data of the second data volume to the target account, wherein the second data volume is less than zero.

2. The method of claim 1, wherein before inputting the first account information and the first historical task information into a pre-trained goal decision tree model and obtaining the risk assessment score of the goal account, the method further comprises:

determining the information gain of the risk characteristic of the second account information according to the first proportion data, and determining the information gain of the risk characteristic of the second historical task information according to the second proportion data;

3. The method of claim 2, wherein constructing the training sample set comprises:

acquiring second account information and second historical task information of a reference account in the plurality of accounts, and determining risk characteristics in the second account information and risk characteristics in the second historical task information according to a preset risk assessment item, wherein the reference account is any one of the plurality of accounts;

if the proportion is larger than a preset proportion threshold value, the risk features in the second account information and the risk features in the second historical task information are used as positive samples, and if the proportion is smaller than the preset proportion threshold value, the risk features in the second account information and the risk features in the second historical task information are used as negative samples, and the positive samples and the negative samples are determined to be the training sample set.

4. The method according to claim 2, wherein the training of inputting the information gain of the risk feature of the second account information and the information gain of the risk feature of the second historical task information into a preset decision tree model to obtain an initial decision tree model further comprises:

5. The method of claim 1, wherein the authenticating the target account comprises:

determining whether the network quality of the terminal equipment of the target account reaches an identity verification condition, wherein the network quality refers to the signal intensity of the terminal equipment for data transmission;

and under the condition that the network quality does not reach the authentication condition, sending a face authentication instruction to the terminal equipment, wherein the face authentication instruction carries verification information, and receiving an authentication result returned by the terminal equipment, wherein the verification information is verification data of a face image corresponding to the target account.

6. The method of claim 1, wherein the resource data transfer request further carries a second node corresponding to the first task; the determining a second data size according to the task exception information includes:

when the first node is the previous node of the second node, the task exception information is indication information for indicating that task execution fails;

7. The method of claim 1, wherein determining a second amount of data based on the task exception information comprises:

when the first node is the previous node of the task completion nodes, the task abnormal information is indicating information for indicating that the task is not completed according to preset time;

said determining said second amount of data according to said second calculation method comprises:

determining a third data volume according to the first data volume and the first node, wherein the third data volume is the data volume of the first task reaching the first node, and the third data volume is smaller than the first data volume;

determining a difference between the first amount of data and the third amount of data as the second amount of data.

8. A decision tree model-based data processing apparatus, comprising:

the system comprises a receiving unit, a processing unit and a processing unit, wherein the receiving unit is used for receiving a resource data transfer request sent by a target account, the resource data transfer request carries a task identifier of the target account, and the task identifier is used for indicating a first task;

the acquiring unit is used for acquiring first account information and first historical task information of the target account, and inputting the first account information and the first historical task information into a pre-trained target decision tree model to obtain a risk assessment score of the target account;

the determining unit is used for determining that the risk evaluation result of the target account passes and determining a first data volume corresponding to the first task when the risk evaluation score is smaller than a risk evaluation threshold;

the verification unit is used for performing identity verification on the target account, and transferring the resource data of the first data volume to the target account under the condition that the identity verification result of the target account passes;

the determining unit is further configured to determine a first node and task exception information corresponding to the first task, determine a second data volume according to the task exception information, and transfer resource data of the second data volume to the target account, where the second data volume is a data volume smaller than zero, when it is determined that the first task does not reach a task completion node.

9. A computer device comprising a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform the method of any one of claims 1-7.

10. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program which is executed by a processor to implement the method of any one of claims 1-7.