CN112184231B

CN112184231B - Credible service determination method and device

Info

Publication number: CN112184231B
Application number: CN202011204068.3A
Authority: CN
Inventors: 周璟; 宝鹏庆; 傅欣艺; 陆毅成; 肖凯; 吕乐; 王维强
Original assignee: Alipay Hangzhou Information Technology Co Ltd
Current assignee: Alipay Hangzhou Information Technology Co Ltd
Priority date: 2020-11-02
Filing date: 2020-11-02
Publication date: 2022-06-28
Anticipated expiration: 2040-11-02
Also published as: CN112184231A

Abstract

According to the method and the device for determining the trusted service, the terminal determines whether the service to be determined is the trusted service. In addition, in the credible service determining process, a credible subgraph is used as a basis for judging whether the service to be determined is credible, and the credible subgraph is obtained based on credible services in the history corresponding to the terminal and can represent the comprehensive characteristics of each historical credible service corresponding to the terminal. In addition, the trusted service determination process in this specification further incorporates information of the service object when the pending service is triggered.

Description

Credible service determination method and device

Technical Field

One or more embodiments of the present disclosure relate to the field of internet technologies, and in particular, to a method and an apparatus for determining a trusted service.

Background

The internet provides a data transmission medium for multi-terminal interaction, more and more services need to be executed by the internet, and on one hand, the internet provides convenience for service execution; on the other hand, each service executed by the internet can be executed more safely under the guarantee mechanism realized by the network.

Currently, when a service is executed, in order to ensure the safety of service execution, data of the service needs to be transmitted to a cloud, the cloud identifies whether the service has risks, and the service is processed according to an identification result.

Therefore, the risk identification needs to be performed by the cloud, and the cloud server is occupied. And the interaction process between the cloud and the terminal consumes more resources.

Disclosure of Invention

One or more embodiments of the present specification describe a mechanism for determining trusted services, which may be performed by a terminal based on information of a trusted subgraph and a service object.

According to a first aspect, there is provided a first trusted service determination method, the method being performed by a terminal, the method comprising:

after the undetermined service is detected, acquiring a credible subgraph, wherein the credible subgraph is acquired based on historical service data and comprises credible relations between the terminal and a plurality of service objects;

determining the reference credibility of the to-be-determined service according to the credibility subgraph;

and determining the comprehensive credibility of the to-be-determined service according to the reference credibility and the information of the service object triggering the to-be-determined service.

In one embodiment, the pending service includes a first interaction of the terminal with a payment account and a second interaction of the terminal with a revenue account, wherein determining a benchmark credibility of the pending service according to the credibility subgraph includes:

Determining a benchmark trustworthiness level based on at least one of: and matching degree of the first interaction and the credible subgraph, and matching degree of the second interaction and the credible subgraph.

In one embodiment, the business object includes: a user to which the terminal belongs;

determining the comprehensive credibility of the undetermined service according to the reference credibility and the information of the service object when the undetermined service is triggered, wherein the method comprises the following steps:

acquiring behavior information of the user in a preset first preset time period before the undetermined service is triggered;

determining the difference between the user behavior and the behavior of the user when the credible service is triggered historically according to the behavior information by adopting a pre-trained comprehensive credibility determination model;

and determining the comprehensive credibility according to the difference.

In one embodiment, the method further comprises:

sending the data of the pending service to the server, so that the server updates the credible subgraph according to the data of the pending service;

obtaining a credible subgraph corresponding to the terminal, including:

and receiving the updated credible subgraph corresponding to the terminal from the server.

According to a second aspect, there is provided a second trusted service determination method, the method being performed by a server, the method comprising:

Acquiring a plurality of historical service data collected by a terminal, wherein the historical service data comprises interactive data between the terminal and a service object;

generating a credible subgraph corresponding to the terminal according to the historical service data, wherein the credible subgraph comprises credible relations between the terminal and a plurality of service objects;

and sending the generated credible subgraph to the terminal, so that when the terminal detects the pending service, the terminal determines whether the pending service is credible according to the credible subgraph and the information of the service object triggering the pending service.

In one embodiment, the obtaining of the plurality of historical service data collected by the terminal includes: acquiring a plurality of historical service data acquired by a terminal in a first preset time period;

generating a credible subgraph corresponding to the terminal according to the plurality of historical service data comprises the following steps: and generating a credible sub-graph of the terminal in a second preset time period in the future according to the plurality of historical service data, wherein the second preset time period is determined based on the first preset time period.

In one embodiment, the plurality of historical traffic data includes: the first service data comprise interaction data of the terminal and a first payment account; generating a credible subgraph corresponding to the terminal according to the plurality of historical service data comprises the following steps: and determining whether the relationship between the terminal and the first payment account is a trusted relationship or not according to at least one of the quantity of the first service data, the resource corresponding to the first service data and the earliest time corresponding to the first service data, and adding the relationship between the terminal and the first payment account into the trusted subgraph if the relationship is determined to be the trusted relationship.

In one embodiment, the plurality of historical traffic data includes: second service data, wherein the second service data comprises interaction data of the terminal and a second revenue account; generating a credible subgraph of the terminal in a second preset time period in the future according to the plurality of historical service data, wherein the credible subgraph comprises the following steps:

determining whether a second relationship between the terminal and the second payment account is a trusted relationship in the second preset time period according to at least one of the frequency of the second service data and the attribute of the second revenue account; or, a preset trigger probability prediction model is adopted to determine the probability that the service corresponding to the second service data is triggered in the second preset time period; determining whether the second relation is a credible relation within the second preset time period according to the probability;

and adding the second relation into the credible subgraph under the condition that the second relation is determined to be a credible relation.

In one embodiment, determining whether the second relationship is a trusted relationship within the second preset time period according to the probability includes:

determining a preset credible attribute of the second payment account;

And determining whether the second relation is a credible relation within the second preset time period according to the probability and the credible attribute.

According to a third aspect, there is provided a first trusted service determination apparatus, the apparatus being for a terminal, the apparatus comprising:

the credible subgraph acquisition unit is configured to acquire a credible subgraph after the pending service is detected, wherein the credible subgraph is acquired based on historical service data and comprises credible relations between the terminal and a plurality of service objects;

a reference credibility determining unit configured to determine a reference credibility of the pending service according to the credibility subgraph;

and the comprehensive credibility determining unit is configured to determine the comprehensive credibility of the undetermined service according to the reference credibility and the information of the service object triggering the undetermined service.

In one embodiment, the pending service comprises a first interaction of the terminal with a payment account and a second interaction of the terminal with a revenue account;

the reference credibility determination unit is configured to: determining a benchmark trustworthiness level based on at least one of: and matching degree of the first interaction and the credible subgraph, and matching degree of the second interaction and the credible subgraph.

the comprehensive credibility determining unit is configured to: acquiring behavior information of the user in a preset first preset time period before the pending service is triggered; determining the difference between the user behavior and the behavior of the user when the credible service is triggered historically according to the behavior information by adopting a pre-trained comprehensive credibility determination model; and determining the comprehensive credibility according to the difference.

In one embodiment, the apparatus further comprises a data transmission unit configured to: sending the data of the pending service to the server, so that the server updates the credible subgraph according to the data of the pending service;

the credible subgraph acquisition unit is configured to: and receiving the updated credible subgraph corresponding to the terminal from the server.

According to a fourth aspect, there is provided a second trusted traffic determination apparatus, the apparatus being for a server, the apparatus comprising:

the historical service data acquisition unit is configured to acquire a plurality of historical service data acquired by a terminal, wherein the historical service data comprises interactive data between the terminal and a service object;

A credible subgraph generating unit configured to generate a credible subgraph corresponding to the terminal according to the plurality of historical business data, wherein the credible subgraph comprises credible relations between the terminal and a plurality of business objects;

and the credible subgraph sending unit is configured to send the generated credible subgraph to the terminal, so that when the terminal detects the pending service, whether the pending service is credible or not is determined according to the credible subgraph and the information of the service object triggering the pending service.

According to a fifth aspect, there is provided a computer readable storage medium having stored thereon a computer program which, when executed in a computer, causes the computer to perform the method of the first and/or second aspect.

According to a sixth aspect, there is provided a computing device comprising a memory having stored therein executable code, and a processor which, when executing the executable code, implements the method of the first and/or second aspect.

According to the method and the device provided by one embodiment of the specification, whether the pending service is the trusted service is determined by the terminal. The process is executed by the terminal without occupying the resources of other business objects (such as servers). And the process reduces the interaction with other business objects to a greater extent, and effectively avoids the resource consumption in the interaction process. In addition, the credible subgraph is used as a basis for judging whether the pending service is credible in the credible service determining process, and the credible subgraph is obtained based on the credible service in the history corresponding to the terminal, so that the comprehensive characteristics of each historical credible service corresponding to the terminal can be represented, and the credible determining result obtained in the credible service determining process based on the credible subgraph in the description is more fit with the actual situation of the terminal. In addition, the trusted service determination process in this specification further combines with information of the service object when the pending service is triggered, so that the obtained trusted determination result can embody a specific situation when the pending service is triggered, and the accuracy of the trusted determination result is improved.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.

FIG. 1 is a schematic diagram illustrating a risk identification process for a transaction;

FIG. 2 shows a network architecture used when some business objects interact with each other in a scenario where a terminal and a merchant perform a transaction according to an embodiment;

FIG. 3 illustrates a schematic diagram of a graph and subgraph of one embodiment;

FIG. 4 illustrates a flow diagram for generating a trusted subgraph, according to one embodiment;

FIG. 5 shows a schematic diagram of a first interaction and a second interaction;

FIG. 6 illustrates a schematic diagram of determining whether a first interaction and a second interaction are trustworthy;

FIG. 7 illustrates a process diagram for determining whether the second interaction is trustworthy based on the prediction data, according to one embodiment;

FIG. 8 illustrates a flow diagram for determining trusted traffic, according to one embodiment;

FIG. 9 illustrates a schematic diagram for determining a comprehensive trustworthiness rating in one embodiment;

FIG. 10 illustrates the structure of the synthetic confidence level determination model in one embodiment;

FIG. 11 illustrates a process diagram for trusted and untrusted traffic, respectively, according to one embodiment;

FIG. 12 shows a schematic block diagram of a trusted traffic determination means according to one embodiment;

FIG. 13 shows a schematic block diagram of a trusted traffic determination means according to another embodiment;

FIG. 14 shows a schematic diagram of an electronic device according to one embodiment.

Detailed Description

The present specification will be described in further detail with reference to the accompanying drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the relevant invention and not restrictive of the invention. The described embodiments are only a subset of the embodiments described herein and not all embodiments described herein. All other embodiments obtained by a person skilled in the art based on the embodiments in the present specification without any inventive step are within the scope of the present application.

It should be noted that, for convenience of description, only the portions related to the related invention are shown in the drawings. The embodiments and features of the embodiments in the present description may be combined with each other without conflict.

As described above, the internet provides convenience and security for services performed through multi-terminal interaction.

Fig. 1 shows a schematic diagram of a risk identification process for a transaction to be processed. In a service execution process, as shown in fig. 1, a terminal detects a service to be processed, and sends a processing request of the service to be processed to a server in the cloud. And the server identifies the risk of the service to be processed according to the received processing request and sends an identification result to the terminal. This process may result in a busy server, a heavy consumption of resources (e.g., network resources, computing resources of the server, time resources consumed during data transmission, etc.), and so on. Moreover, the risk identification performed by the server is completely relied on, so that the requirement of real-time performance of the business risk identification can be met under the conditions that the computing resources of the server are sufficient and the network between the server and the terminal of the user is unobstructed, and the actual business execution scene does not necessarily meet the condition.

Furthermore, factors affecting the security of the service are various, taking the service as a transaction as an example, factors such as gambling behavior, behavior of a lawbreaker stealing a user account and the like all negatively affect the security of the transaction, and if each factor which can bring risks to the execution of the service is checked one by one, a large amount of resources are consumed, and the execution efficiency of the service is affected.

In addition, the number of users corresponding to the service provided by the server is large, and the server often judges the business risk based on the universal business rule, so that it is difficult to use the personal condition of the user as the basis of risk judgment, and the risk judgment result of the server for the business cannot reflect the individual difference of the user, which causes the phenomenon that the judgment result is not suitable for some users.

In view of the above, the present specification provides a trusted service determination process, which involves the following concepts in part:

in this specification, a service may be a transaction to be processed by a terminal. The specific meaning of the service can be determined according to the actual scenario.

Fig. 2 illustrates a network architecture adopted when a part of business objects interact with each other in a scenario in which a terminal and a merchant perform a transaction according to an embodiment.

For example, in a scenario where a user conducts a transaction with a merchant through a terminal (the related network architecture is shown in fig. 2), a transaction or a link experienced during the transaction may be regarded as a service. In this case, the terminal may be a terminal device of the user, for example, a mobile phone, a tablet computer, a Personal computer, a notebook computer, a Personal Digital Assistant (PDA), a wearable device (such as smart glasses and a smart watch), and the like.

A server refers to a computer system in a network that can provide services to other devices. The terminal is the object served by the server. The server and the terminal can be in communication connection in a wired or wireless mode. The implementation manner of the server is various, and may be a single computer device, or may be a combination of multiple computer devices (e.g., a cluster server, a cloud server, etc.). The server may also be referred to as a server, a cloud, etc. in some application scenarios.

Risk refers to a factor that negatively impacts at least one of efficiency, security, and effectiveness of business execution. In the scenario where a user conducts a transaction with a merchant through a terminal, the risks may include a fraud risk (e.g., fraudulent use of the terminal, fraudulent use of a payment account), a fraud risk, a gambling risk, a false transaction risk, etc.

Confidence refers to a pattern in which the probability of risk occurrence is much lower than the average risk probability. For a certain service, it can be determined that the service is authentic in the case that the combined risk obtained from various risks of the service is low.

Each service, whether it is a trusted service or not, will be referred to as a pending service in this specification. As can be seen, when the trusted status is taken as the basis for the division, the services in this specification may include: trusted traffic for which the trusted state is known, untrusted traffic, and pending traffic for which the trusted state is unknown.

The business objects in this specification may be respective objects involved in the business execution process. The business objects may include: at least one of an entity business object, a process business object, and an event business object. The meaning of the business object may be determined according to the actual scenario.

In a scene that a user conducts transaction with a merchant through a terminal, the user needs to transfer the amount of money in a payment account of the user to a profit account of the merchant through the terminal, and all interaction media adopted in the process can be used as business objects. At this time, the business objects include, but are not limited to: a user's terminal, payment and revenue accounts, a network, a card, a merchant's POS, etc. The relationships between the business objects involved in the process include, but are not limited to: the relationship between the terminal and the payment account, the relationship between the terminal and the revenue account, the relationship between the terminal of the user and the network, and the like, which are not described herein in detail.

When various different business objects are used as different types of nodes, and the relationship between the business objects in the business generation process is used as an edge, a graph structure can be formed.

FIG. 3 illustrates a schematic diagram of a graph and a sub-graph of an embodiment. A graph is a topology consisting of nodes and edges. The subgraph consists of part of nodes and part of edges in the graph. In an alternative embodiment of the present description, the graph representation of the graph and the subgraph may be as shown in fig. 3, taking as an example that the graph has a quaternary structure (i.e., the nodes include four types of nodes: a first node, a second node, a third node, and a fourth node). If each node in fig. 3 and the edges between the nodes are taken as a graph, a part of the graph (as shown by the dashed box in fig. 3) can be taken as a sub-graph of the graph. The subgraph includes 1 first node, 2 second nodes and 1 third node, and does not include a fourth node. In addition, the graph may further include other multiple subgraphs, and there may be overlaps between subgraphs, which are not listed here.

It should be noted that the present specification does not limit the specific representation manner of the subgraph. For example, when a node is a business object, a subgraph having a ternary structure can also be represented by the following structural formula: { Business object A, [ Business object B₁Business object B₂Business object B₃……][ business object C₁Business object C₂Business object C₃……]}. In the scene that the user carries out transaction with the merchant through the terminal, the business object A can be the terminal of the user; business object B_nMay be the nth payment account of the user; business object C_mMay be the mth revenue account with which the transaction occurred with the terminal. The relationship between the business objects may be as shown in FIG. 2.

The trusted service determination process provided by the present specification is performed based on trusted subgraphs, where one trusted subgraph corresponds to one terminal and indicates the trusted relationship between the terminal and other service objects, and is used to characterize the synthesis of each trusted service executed by using the terminal. The process of generating the child trust map is first described below.

Firstly, generating a credible subgraph.

In different embodiments, the trusted subgraph can be generated by the terminal itself or by the server. In a typical case, a server generates a corresponding credible subgraph for each terminal. The following description is mainly made for this case.

FIG. 4 illustrates a flow diagram for generating a trusted subgraph, which can be performed by a server, according to one embodiment. As shown in fig. 4, in an alternative embodiment of the present specification, the process of generating a trusted subgraph may include one or more of the following steps:

s400: acquiring a plurality of historical service data collected by a terminal, wherein the historical service data comprises interactive data between the terminal and a service object.

The present specification does not specifically limit the time range related to the historical service data. The historical traffic data is at least capable of reflecting information associated with the traffic over a longer period of time. For example, the history data may be data of each service transmitted within 30 days from the current time in the history.

The historical traffic data may be data characterizing the execution status and/or execution results of the historical traffic. In an alternative embodiment of the present specification, the historical service data may include: the historical service relates to at least one of individual (historical) service objects, relationships between service objects, resources consumed by executing the service, and a time at which the service is triggered.

In the foregoing scenario in which the user performs a transaction with the merchant through the terminal, the resources consumed for executing the transaction service include but are not limited to: the amount of the transaction. The time at which the service is triggered includes, but is not limited to: the time when the user performs the payment operation in the terminal (for example, the time when the user scans the two-dimensional payment code of the merchant, the time when the user clicks "confirm payment", etc.).

The present specification does not specifically limit the timing of obtaining the historical service data by the server, and the server may obtain the historical service data when the user performs the operation of triggering the historical service; the terminal may store the data of the historical service when the user performs an operation of triggering the historical service, and then periodically transmit the stored data to the server.

S402: and generating a credible subgraph corresponding to the terminal according to the plurality of historical business data, wherein the credible subgraph comprises the credible relation between the terminal and a plurality of business objects.

As can be seen from the foregoing, the trusted subgraph includes the trusted relationship between the terminal and several service objects, thereby depicting the integration of each trusted service executed by the user using the terminal. Therefore, the service object related to the terminal and the relationship between the service objects are necessary conditions for obtaining the trusted subgraph, and then the data of each historical service processed by the terminal can be processed to obtain each service object (which can be used as a node of the trusted subgraph) for generating the trusted subgraph and the relationship between the service objects (which can be used as an edge of the trusted subgraph).

In an optional embodiment of the present specification, the service object may be determined according to a preset service meaning; and determining the relation between the business objects according to the historical business data.

Further, in one embodiment, the plurality of historical traffic data are historical traffic data in a past certain period (hereinafter referred to as a first preset period), and the generated credibility sub-graph is used for representing credibility relation in a future second preset period for a period (hereinafter referred to as a second preset period). Generally, the second preset time period is determined based on the first preset time period, and the specific duration of the second preset time period can be determined according to the actual scene. For example, the second predetermined period may be one day, 8 hours, etc., or may be a longer time frame, such as one month in the future.

As can be seen from the content recorded in step S400, when the server acquires the historical service data, it is not necessary to determine whether the historical service is a trusted service, and there is a possibility that the acquired historical service data is not data of a trusted service in a second preset time period in the future, and the historical service data is not suitable for generating a trusted subgraph. It is necessary to determine, according to the data of each historical service, a trusted relationship that can be used for generating a corresponding trusted subgraph of the terminal in a second preset time period from the relationships corresponding to each historical service.

Specifically, the relationship determined according to the data of the historical service in the present specification may include a trusted relationship and an untrusted relationship. The process of determining the trusted relationship among the relationships determined according to the data of the historical service may be: and determining whether the relation meets a preset credible relation determination condition or not according to each relation determined according to the data of each historical service. If yes, the relationship is a trusted relationship; if not, the relationship is an untrusted relationship.

In an optional embodiment of the present specification, a service object corresponding to each trusted relationship may be determined as a node; and determining the edges of the connecting nodes according to the credibility relations to obtain the credibility subgraph corresponding to the terminal.

S406: and sending the generated credible subgraph to the terminal, so that when the terminal detects the pending service, the terminal determines whether the pending service is credible according to the credible subgraph and the information of the service object triggering the pending service.

After the credible subgraph is obtained, the credible subgraph can be sent to a terminal corresponding to the credible subgraph, so that the terminal determines whether the service to be determined is credible or not according to the credible subgraph.

From the foregoing, when generating a trusted subgraph, how to determine a trusted relationship among relationships is described based on the business objects and the trusted relationships between the business objects.

Still taking the aforementioned scenario that the user conducts transaction with the merchant through the terminal as an example, the first node represents the terminal, the second node represents the payment account corresponding to the historical service, and the third node represents the revenue account corresponding to the historical service. And, characterizing a relationship between the first node and a second node with a first interaction, and characterizing a relationship between the first node and a third node with a second interaction.

How to determine whether the first interaction is trusted and how to determine whether the second interaction is trusted is described below in conjunction with fig. 5 and 6, where fig. 5 shows a schematic diagram of the first interaction and the second interaction and fig. 6 shows a schematic diagram of determining whether the first interaction and the second interaction are trusted.

1. It is determined whether the first interaction is authentic.

As can be seen from the foregoing, whether the first interaction is trusted may have temporal attributes. A first interaction in this specification may be used to characterize that a first node and a second node corresponding to the first interaction were present in the same historical traffic. However, the presence of the first interaction in the same historical service does not indicate that the first interaction is authentic within the second predetermined period of time.

For example, a first interaction determined according to the ith historical service, which is represented in a graph, can be as shown in fig. 5. As shown in fig. 5, in the process of determining whether a first interaction is trusted, the present specification determines, for each historical service, a first interaction corresponding to the historical service, where, taking an ith historical service in fig. 5 as an example, i may be an integer greater than 0. Then, as shown in fig. 6, for each first interaction, at least one of the number of the historical services including the first interaction in each historical service, the resource corresponding to each historical service including the first interaction, and the time for processing the historical service including the first interaction with the terminal for the first time is determined as the first data. If the first data meet a preset first condition, determining that the first interaction is in a trusted relationship within a second preset time period; and if not, determining that the first interaction is in the untrusted relationship within a second preset time period. It can be seen that the first condition in this specification is to determine whether the first interaction to which the first data belongs is trusted based on the first data.

In an optional embodiment of the present specification, the process of determining whether the first data meets the preset first condition may be: and searching the historical service containing the first interaction in each historical service, judging whether the quantity of the historical service containing the first interaction is larger than a preset quantity threshold value, and if so, determining that the first interaction is credible. Still taking the foregoing scenario in which the user performs a transaction with the merchant through the terminal as an example, if the determination result of the first interaction with respect to the transaction is yes, it indicates that the transaction performed through the payment account by using the terminal in history is relatively frequent, the probability that the payment account and the terminal belong to the same user is relatively high (that is, the relationship between the payment account and the terminal is relatively stable), the probability that the payment account is in a stolen state is relatively low, and it may be determined that the first interaction with respect to the transaction is authentic.

Or, the process of determining whether the first data meets the preset first condition may be: and searching the historical service containing the first interaction in each historical service, judging whether the accumulated transaction amount of the historical service containing the first interaction is larger than a preset amount threshold value, and if so, determining that the first interaction is credible. Still taking the aforementioned scenario that the user performs the transaction with the merchant through the terminal as an example, if the determination result is yes, it indicates that the possibility that the transaction performed through the payment account by using the terminal in history is real and effective is high, and it may be determined that the first interaction is authentic.

Or, determining that the first interaction is credible under the condition that the quantity of the historical service containing the first interaction is larger than a preset quantity threshold value and the accumulated transaction amount of the historical service containing the first interaction is larger than a preset amount threshold value.

In the foregoing scenario where the user performs the transaction with the merchant through the terminal, the transaction performed by the user Zhang san is taken as an example of the historical service. If the number of historical transactions conducted by the payment account a through the terminal D is large (for example, more than 100 transactions), the result indicates that the first interaction between the terminal D and the payment account a is stable, and the probability that the first interaction is a trusted relationship is high. If Zhang III only conducts 1 transaction through the payment account B by using the terminal D, it indicates that the first interaction between the terminal D and the payment account B is unstable.

If the accumulated amount of historical transactions conducted by the terminal D through the payment account A is more, the fact that the first interaction between the terminal D and the payment account A is stable is shown, and the probability that the first interaction is a credible relation is higher.

In addition, the present specification may further determine whether the first interaction is trusted according to a time when the first interaction is first triggered in the history as the first data. The earlier the first interaction is triggered, the greater the likelihood that the first interaction is a trusted relationship.

In an optional embodiment of the present specification, a preset first interaction judgment model may be adopted to determine whether the first interaction is trusted. Specifically, the process of determining whether the first interaction is trusted by using the preset first interaction judgment model may be: first data of the first interaction is determined, the content of the first data may be as described above. And inputting the first data of the first interaction into a preset first interaction judgment model to obtain a first credible score of the first interaction output by the first interaction judgment model. Comparing the first credibility score with a preset first credibility score threshold, and if the first credibility score is larger than the preset first credibility score threshold, determining that the first interaction is credible within a second preset time period; and if not, determining that the first interaction is not credible within a second preset time period.

Wherein the first confidence score is used to characterize a likelihood that the first interaction is triggered within a second predetermined period of time, e.g., the likelihood may be characterized by a probability. The likelihood that the first interaction is triggered within the second preset time period may be different from the likelihood that the first interaction is triggered within a time period other than the second preset time period. It can be seen that the same first interaction may result in different first confidence scores over different time periods.

The first interaction judgment model learns how to generalize a rule of triggering the first interaction in the historical service according to the first data in a pre-training process (for example, the rule may be a rule of triggering the first interaction over time), and obtains a probability that the first interaction reoccurs within a second preset time period (for example, the second preset time period may be within a day in the future and within 8 hours in the future) according to the rule. The higher the likelihood, the higher the confidence level of the corresponding first interaction within the second preset time period.

Optionally, the first credible score is positively correlated with at least one of the number of the historical services including the first interaction, the resources corresponding to the historical services including the first interaction, and the time length from the time of the historical service including the first interaction to the current time when the terminal is adopted to process the historical service including the first interaction for the first time. For example, the earlier the terminal is first used to process historical traffic containing the first interaction, the higher the first confidence score.

In an alternative embodiment of the present specification, at least one of the amount of the historical service including the first interaction and the accumulated transaction amount of the historical service including the first interaction may be obtained from data of the historical service generated within a preset first historical period (for example, the past 30 days) before the current time.

As can be seen, in the foregoing scenario where the user performs a transaction with the merchant through the terminal, the description can characterize the risk degree that the payment account may be stolen in the service execution process through at least the first interaction between the terminal and the payment account.

2. It is determined whether the second interaction is authentic.

Similarly, whether a second interaction in this specification is trusted may have temporal attributes. The second interaction in this specification may be used to characterize that the first node and the third node were present in the same historical traffic. For example, a second interaction determined according to the ith historical service, which is represented in a graph, may be as shown in fig. 5.

For whether the second interaction is trusted within the second preset time period, the following two determination methods are exemplarily provided in the present specification. Then, in generating the trusted subgraph, the trusted second interaction may be determined in each second interaction by using any one or a combination of the following two methods.

(1) It is determined whether the second interaction is authentic based on the historical data.

As shown in fig. 5 and 6, for each historical service, a second interaction corresponding to the historical service is determined. Then, taking the ith historical service as an example, for each second interaction, determining at least one of the triggered frequency of the historical service of the second interaction and a preset credible attribute of the third node in each historical service as second data. If the second data meets a preset second condition, determining that the second interaction is in a trusted relationship within a second preset time period; and if the second data does not meet a preset second condition, determining that the second interaction is in an untrusted relationship within a second preset time period. It can be seen that the second condition in this specification is to determine whether the second interaction to which the second data belongs is trusted based on the second data.

In an alternative embodiment of the present specification, the process of determining whether the second data satisfies the preset second condition may be: and searching the historical service containing the second interaction in each historical service triggered within a preset second historical time period before the current time, judging whether the triggered frequency of the historical service containing the second interaction is greater than a preset frequency threshold value, and if so, determining that the second interaction is credible within the second preset time period. Still taking the foregoing scenario in which the user performs the transaction with the merchant through the terminal as an example, if the determination result of the second interaction with respect to the transaction is yes, it indicates that the transaction performed between the terminal and the revenue account in history is relatively frequent, the transaction between the user corresponding to the terminal and the merchant is relatively credible, and it may be determined that the second interaction is credible within a second preset time period.

Alternatively, the second historical time period may include a plurality of time ranges, and then the frequency with which the second interaction is triggered in the time range may be determined for each time range, and whether the second interaction is trusted may be determined for the obtained respective frequencies and the relationship between the respective frequencies and the frequency threshold. For example, the second historical period of time may include a time range of the past 90 days, the past 60 days, and the past 30 days.

Or, the process of determining whether the second data meets the preset second condition may be: and determining the credible attribute of the third node according to the state of the third node corresponding to the second condition in the history. And if the credible attribute is matched with the preset standard attribute, determining that the second interaction is credible.

In the foregoing scenario that the user performs a transaction with the merchant through the terminal, the trusted attribute of the third node may be determined according to the historical behavior of the merchant corresponding to the third node. And judging whether the credible attribute of the third node is matched with the preset standard attribute. If the result of the judgment is yes, it indicates that the merchant corresponding to the third node performs well in the history, and there is no or few illegal behaviors, so that it can be determined that the second interaction is credible within a second preset time period.

Optionally, the credibility attribute can be used for characterizing whether the merchant corresponding to the third node has illegal behaviors such as fraud, false transactions, gambling and the like in the history. Further, the trusted attributes of the third node may include: any one of the third node being trusted and the third node being untrusted. If the credible attribute of the third node is credible for the third node, matching the credible attribute of the third node with a preset standard attribute; and if the credible attribute of the third node is that the third node is not credible, the credible attribute of the third node is not matched with the preset standard attribute.

Still alternatively, the process of determining whether the second data meets the preset second condition may be: and under the condition that the credible attribute of the third node is matched with the preset standard attribute, if the triggered frequency of the historical service containing the second interaction is greater than a preset frequency threshold, determining that the second interaction is credible within a second preset time period.

In addition, in an optional embodiment of the present specification, a preset second interaction determination model may be used to determine whether the second interaction is trusted within a second preset time period. Specifically, the process of determining whether the second interaction is trusted by using the preset second interaction judgment model may be: and determining second data of the second interaction, and inputting the second data of the second interaction into a preset second interaction judgment model to obtain a second credibility score of the second interaction output by the second interaction judgment model.

Then, comparing the second credibility score with a preset second credibility score threshold, and if the second credibility score is larger than the preset second credibility score threshold, determining the second interaction credibility; and if the second credibility score is not larger than a preset second credibility score threshold, determining that the second interaction is not credible in a second preset time period. And if the second data show that the credible attribute of the third node is not matched with the preset standard attribute, determining that the second interaction is not credible in a second preset time period.

Optionally, the second credible score is positively correlated with the frequency of triggering the historical service containing the second interaction in each historical service; and/or the second credibility score is positively correlated with the matching degree of the credibility attribute of the third node and the preset standard attribute.

(2) Determining whether the second interaction is authentic based on the prediction data.

As shown in FIG. 7, second data of the second interaction is determined in each historical service according to the second interaction. The second data includes at least one of a frequency of triggering the second interaction in a second historical time period (the second historical time period may include several time ranges), a number of times the second interaction is triggered in the history, and accumulated resources consumed by executing each historical service corresponding to the second interaction. And then, inputting the second data into a preset trigger probability prediction model to obtain the probability that the service which is output by the trigger probability prediction model and contains the second interaction is triggered by the user in a second preset time period in the future. And then, according to the probability, determining whether the second interaction is in a credible relationship within a second preset time period. The higher the determined probability, the greater the likelihood that the second interaction is a trusted relationship within a second predetermined period of time.

Optionally, the probability is positively correlated with at least one of a frequency of triggering the second interaction in a second historical time period, a number of times of triggering the second interaction in the history, and an accumulated resource consumed by executing each historical service corresponding to the second interaction.

The trigger probability prediction model is obtained by training according to data of historical services. In the process of model training, the trigger probability prediction model can learn, according to the data of the historical services, a rule that the second interaction of each historical service is triggered (for example, the rule may be a rule that the second interaction is triggered over time), and obtain, according to the rule, a possibility that the second interaction occurs again in a future certain period (for example, the future certain period may be a second preset period). The higher the likelihood, the more trustworthy the second interaction that the likelihood corresponds to.

Optionally, the process of determining whether the second interaction is a trusted relationship according to the probability may be: and judging whether the probability is greater than a preset probability threshold value. If the second interaction is determined to be credible in the second preset time period, the probability that the service corresponding to the second interaction is triggered in the future second preset time period is higher, the service corresponding to the second interaction is triggered in history according to a certain rule without causing risks, and the service corresponding to the second interaction is shown to be credible in the second preset time period.

It will be appreciated that the above trigger probabilities and confidence decisions are related to future time periods. Take user Zhang III buying breakfast at business A as an example. Through the above process, it can be determined, for example, that a second interaction between the terminal D and the revenue account a of the merchant a is more likely to be a trusted relationship on weekdays (i.e., a second preset period), but that the second interaction is less likely to be a trusted relationship on weekends (saturday and sunday). And, zhang san uses terminal D to conduct a transaction with revenue account B of restaurant B on weekends (e.g., the transaction may be a transaction to purchase takeoffs at restaurant B), it is determined through the processes in this specification that there is a high likelihood that the second interaction between terminal D and revenue account B is a trusted relationship on weekends (i.e., the second predetermined period of time).

Furthermore, in an alternative embodiment of the present description, the process of determining whether the second interaction is trusted may also incorporate trusted attributes of the third node. The process of determining whether the second interaction is a trusted relationship according to the probability and the trusted attribute of the third node may be: determining a preset credible attribute of the third node; and if the credible attribute of the third node is matched with the preset standard attribute and the probability is greater than a preset probability threshold, determining that the second interaction is credible within a second preset time period.

It can be seen that the present description is capable of characterizing, at least through a second interaction between the terminal and the revenue account, whether there is a risk of fraudulent transactions, fraud, gambling, etc. during the execution of the service.

In one embodiment, the operation of generating the trusted subgraph for the second preset period may be performed periodically. The period may be determined according to actual requirements. Optionally, the duration of the period is the same as the duration of the second preset period.

It should be noted that the trusted service determination process in this specification is not only applicable to the aforementioned transaction scenario (e.g., an offline transaction scenario), but also applicable to a transfer scenario, a water and electricity fee payment scenario, and the like, which is not described herein again. In the process of determining the credible subgraph, the execution sequence of determining whether the first interaction is credible and determining whether the second interaction is credible are not in sequence.

After the trusted first interaction and the trusted second interaction are determined to be the trusted relationships by adopting the steps, a trusted subgraph suitable for a second preset time period can be formed according to the trusted relationships, nodes in the subgraph represent service objects corresponding to the trusted relationships, and edges represent the trusted relationships. The credible subgraph can be used for determining whether the service to be determined is the credible service in the second preset time period in the subsequent steps.

And secondly, determining the credible service.

The credible subgraph generated based on the mode can be used for judging whether the to-be-determined business is the credible business or not. The trusted service determination process provided by the present specification may be performed by a terminal. In particular, FIG. 8 illustrates a flow diagram for determining trusted services, according to one embodiment. As shown in fig. 8, the process may include one or more of the following steps:

s800: and after the pending service is detected, acquiring a credible subgraph, wherein the credible subgraph is acquired based on historical service data and comprises credible relations between the terminal and a plurality of service objects.

In one embodiment, the trusted subgraph is generated periodically by the terminal itself, and the terminal can acquire the newly generated trusted subgraph. In another embodiment, the trusted subgraph is generated by a server, as described above in connection with FIG. 4. In such a case, the terminal may acquire the trusted subgraph generated in the above manner from the server.

As can be seen from the foregoing, the credibility in this specification is a concept having a time attribute, so that the obtained credible subgraph also has a time attribute. The reliability of the credible subgraph for judging whether the service to be determined is credible is influenced by the time attribute.

In order to match with the time attribute of the credible subgraph, the credible subgraph can show the change of the credible relationship, and further the terminal can determine whether the pending service is credible based on the credible subgraph suitable for the current condition. And then, the server sends the updated credible subgraph to the terminal, so that the terminal determines the credible service according to the updated credible subgraph.

In another embodiment, when the time length from the time when the terminal receives the credible subgraph from the server last time to the current time does not exceed a certain threshold, the terminal may obtain the previously received and stored credible subgraph for the judgment of the pending service.

S802: and determining the benchmark credibility of the undetermined service according to the credibility subgraph.

In an embodiment, the reference credibility may represent a similarity between the currently pending service and the trusted service indicated in the trusted subgraph. Correspondingly, if the benchmark credibility of the pending service is higher, the probability that the pending service is the credible service is higher.

According to the description of the generation process of the credible subgraph, the credible subgraph in the description can comprehensively characterize the credible business triggered in the history from at least two aspects of the first interaction and the second interaction. And when the reference credibility of the service to be determined in the second preset time period is determined on the line according to the credibility subgraph, the two aspects of the first interaction and the second interaction can be started respectively.

In an optional embodiment of this specification, the process of determining the reference credibility of the pending service according to the first interaction and the second interaction of the pending service may be: and determining the relation between the payment account corresponding to the pending service and the terminal as a first interaction. And determining the relationship between the terminal and a revenue account corresponding to the pending service as a second interaction. And determining the comprehensive matching degree according to at least one of the first matching degree of the first interaction and the credible subgraph and the second matching degree of the second interaction and the credible subgraph. And determining a reference credibility according to the comprehensive matching degree, wherein the reference credibility is positively correlated with the comprehensive matching degree.

And under the condition that the comprehensive matching degree is obtained according to the first matching degree, positively correlating the comprehensive matching degree with the first matching degree. And under the condition that the comprehensive matching degree is obtained according to the second matching degree, positively correlating the comprehensive matching degree with the second matching degree. And under the condition that the comprehensive matching degree is obtained according to the first matching degree and the second matching degree, positively correlating the comprehensive matching degree with both the first matching degree and the second matching degree.

Optionally, under the condition that the comprehensive matching degree is obtained according to the first matching degree and the second matching degree, the process of determining the first matching degree, the process of determining the second matching degree, and the process of determining the comprehensive matching degree may be determined as processes of solving the classification problem.

Specifically, the process may be: and matching the pending service with the credible subgraph as the benchmark credibility of the pending service under the condition that the first interaction corresponding to the pending service is matched with the first interaction in the credible subgraph and the second interaction of the pending service is matched with the second interaction in the credible subgraph. And if the first interaction corresponding to the to-be-determined service is not matched with the first interaction in the credible subgraph and/or the second interaction corresponding to the to-be-determined service is not matched with the second interaction in the credible subgraph, the to-be-determined service is not matched with the credible subgraph and is used as the benchmark credibility of the to-be-determined service.

Taking the first interaction as an example, the process of determining whether the first interaction corresponding to the pending service matches the first interaction in the trusted subgraph may be: and searching whether a first interaction corresponding to the pending service exists in each first interaction of the credible subgraph. If yes, determining that a first interaction corresponding to the service to be determined is matched with a first interaction in the credible subgraph; if not, determining that the first interaction corresponding to the service to be determined is not matched with the first interaction in the credible subgraph.

In other examples, the benchmark trustworthiness may also be a score reflecting how similar the first/second interaction is to the trustworthy subgraph.

Therefore, the reference credibility of the undetermined service is determined through the method.

Then, in step S804, the comprehensive credibility of the pending service is determined according to the benchmark credibility and the information of the service object for triggering the pending service.

As mentioned above, various business objects may be involved in the execution of a business, and the state of the business object may be constantly changing. In this embodiment, information of a service object that triggers an undetermined service when the service object is triggered is considered, so that the reliability of the undetermined service can be influenced to a certain extent.

The service object triggering the pending service is not specifically limited in this specification. In one embodiment, the pending service is triggered by a user. In this case, the information of the business object is information of the user. The information of the user may include: at least one of behavior information of the user, information of the geographical position where the user is located and transaction record information of the user.

Wherein, the specific meaning of the behavior information of the user can be determined according to the actual scene. Optionally, the behavior information of the user may include an operation (e.g., a click operation, a screen capture operation, etc.) performed by the user with respect to the terminal. The behavior may be a behavior of the user at a time of triggering the pending service; the action in a certain time period including the time when the user triggers the pending service may also be, for example, the action information of the user in a certain time period before the pending service is triggered, where the certain time period may be determined according to the service requirement. The behavior information may be obtained by an RPC (Remote Procedure Call), and/or the behavior information may be obtained by a point burying method using an SPM (Super Position Model).

The information of the geographical location where the user is located may include: the geographical position of the user at the moment of triggering the pending service and/or the movement track formed by the user within a certain time period.

The transaction record information of the user may include: the user forms a record of each transaction over a certain period of time, such as the first preset period of time.

In other examples, the business object that triggers the pending business may also be other business objects. The information of the service object can be correspondingly acquired.

FIG. 9 illustrates a schematic diagram that illustrates determining a comprehensive trustworthiness rating in one embodiment. As shown in fig. 9, information of the service object triggering the pending service is acquired. Inputting the information of the business object into a pre-trained comprehensive credibility determination model to obtain the difference between the information of the business object output by the comprehensive credibility determination model and the information of the business object historically triggering credible business. And determining the comprehensive credibility according to the difference and the benchmark credibility. The aggregate confidence level is inversely related to the difference, and the aggregate confidence level is positively related to the baseline confidence level.

In the aforementioned scenario that the user performs a transaction with the merchant through the terminal, the behavior information of the user in a preset time period before the pending service is triggered may be first acquired as the information of the aforementioned service object. The difference obtained is the difference between the user's behavior when triggering the pending service and the behavior when historically triggering a trusted service.

In the foregoing scenario where the user performs a transaction with the merchant through the terminal, the behavior information includes: and the user checks a personal information page, checks a bill, deletes a service record, sets a service execution mode (such as a payment mode), and sets user information (such as a new user fingerprint is added) screenshot behaviors within a certain preset time period.

Under normal conditions, the user can not adopt the terminal to execute the action irrelevant to the transaction with a larger probability when the transaction is carried out. Under the condition that the terminal and/or the payment account are/is stolen by a lawbreaker (the risk of theft exists), the lawbreaker usually executes operations such as deleting business records and the like after adopting the stolen terminal and/or payment account to carry out transaction; or in order to reach a transaction, a lawbreaker usually views a personal information page, sets a service execution mode, sets user information, and the like. In the case of fraud and/or billing (fraud risk), a lawbreaker often conducts multiple transactions continuously, and captures a transaction page to obtain corresponding rewards according to the captured image. These extraneous behaviors all constitute a difference from the information of the business object that historically triggered the trusted business.

And based on the difference and the reference credibility, obtaining the comprehensive credibility of the undetermined service.

In another embodiment, the information of the service object when the service to be determined is triggered and the reference credibility can be input into another comprehensive credibility determination model together, and the comprehensive credibility determination model directly outputs the comprehensive credibility.

FIG. 10 illustrates the structure of the integrated confidence level determination model in one embodiment.

Specifically, the comprehensive credibility determination model may include an input layer, a first decision layer, a second decision layer and an output layer, which are sequentially arranged as shown in fig. 10, and then the process of determining the comprehensive credibility may be: inputting the reference credibility and the information of the service object when the pending service is triggered into an input layer of the comprehensive credibility determination model, inputting the reference credibility into a first decision layer by the input layer, and inputting the information of the service object when the pending service is triggered into a second decision layer by the input layer. The first decision layer inputs a first decision result obtained according to the benchmark credibility to the second decision layer. And the second decision layer obtains a second decision result of the second decision layer according to the first decision result output by the first decision layer and the information of the service object when the service to be determined is triggered, and outputs the second decision result to the output layer. The output layer outputs the second decision result output by the second decision layer as the comprehensive credibility.

Optionally, the first decision layer is configured to determine whether the pending service is an untrusted service according to the reference trustworthiness. In particular, the first decision result may comprise a first result and a second result. If the benchmark credibility of the pending service is as follows: if the pending service is matched with the credible subgraph, the first decision layer outputs the first result, wherein the first result is used for indicating the second decision layer to make further decision according to the information of the service object; if the benchmark credibility of the pending service is: and if the pending service is not matched with the credible subgraph, the first decision layer outputs the second result, and the second result is used for indicating the second decision layer to output the comprehensive credibility degree which does not meet the preset credible service determination condition.

And the second decision layer is used for determining whether the comprehensive credibility of the undetermined service meets a preset credible service determination condition or not according to the output of the first decision layer and the information of the service object when the undetermined service is triggered. Specifically, if the output of the first decision layer is the first result, the second decision layer determines that various risks may exist in the execution of the pending service according to the information of the service object when the pending service is triggered, and if it is determined that the risk of executing the pending service is low after the various risks are integrated, the second decision layer outputs the integrated credibility degree meeting the preset credible service determination condition; and if the risk of executing the undetermined service is determined to be higher after various risks are synthesized, the second decision-making layer outputs the comprehensive credibility which does not meet the preset credible service determination condition.

The sample used for training the comprehensive credibility determination model may include, but is not limited to, information of a service object historically triggering a pending service, which is collected by the terminal. That is, the comprehensive credibility determination model in the present specification may have universality for each business object (e.g., the user to which the terminal belongs).

For example, the information of the business object of the pending service with a low trigger risk may be used as a positive example, the information of the business object of the pending service with a high trigger risk may be used as a negative example, and the training samples are respectively determined according to the positive example and the negative example.

It can be known from the foregoing that, the first interaction in the trusted sub-graph is obtained according to the historical data, and still taking a scenario in which the user performs a transaction with the merchant through the terminal as an example, if the first interaction between the terminal D of zhang san and the payment account a is trusted within the second preset time period, and the second interaction between the terminal D and the income account B is trusted within the second preset time period, zhang san loses the terminal D, after a lawless person finds the terminal D, the terminal D is adopted to perform a transaction through the account a and the income account B within the second preset time period, and the difficulty of the theft risk determined according to the trusted sub-graph is large. If the trusted subgraph is used alone for determining the trusted service, the phenomenon of misjudgment is possible to occur.

By the comprehensive credibility determination model in the specification, the relationship between the undetermined service represented by the reference credibility and each credible service triggered in a longer (macroscopic) time range and the state of the service object represented by the information of the service object in a shorter (microscopic) time range are comprehensively considered, so that the comprehensive credibility with stronger comprehensiveness and more comprehensive characterization dimensionality is obtained, whether the undetermined service is credible in a second preset time period is measured according to the comprehensive credibility, the determination accuracy and the determination precision of the credible service can be improved to a greater extent, and the possibility of misjudgment is reduced.

In an optional embodiment of the present specification, on the basis of obtaining the comprehensive credibility of the pending service, it may further determine whether the pending service is credible according to the comprehensive credibility.

If the comprehensive credibility of the pending service is higher, the probability that the pending service is the credible service is higher. In an optional embodiment of this specification, the process of determining whether the pending service is trusted according to the comprehensive credibility may be a process of solving a two-classification problem, and the credibility determination result for the pending service may be any one of "trusted service" and "untrusted service". Specifically, if the comprehensive credibility of the pending service meets a preset threshold, the pending service is a credible service in a second preset period; if not, the pending service is an untrustworthy service.

And thirdly, processing the service to be determined according to the credibility determination result.

After the trusted determination result for the pending service is obtained through the trusted service determination process, the processing mode of the pending service can be determined according to the trusted determination result, and the processing is performed according to the determined processing mode.

Specifically, the process may be: if the service is a trusted service in a second preset time period, processing the trusted service by adopting a first processing mode; and if the service is the untrusted service in the second preset time period, processing the untrusted service by adopting a second processing mode. The security level of the second processing mode is higher than that of the first processing mode.

Optionally, the process of processing the trusted service by using the first processing mode may be: and directly processing the credible service and generating a processing result.

Optionally, the process of processing the untrusted service by using the second processing manner may be: and further identifying the risk of the untrusted service, and processing the untrusted service according to the result of the further risk identification.

Wherein the above-mentioned further risk identification may be performed by other entities having higher identification capability than the terminal, such as a server. A processing request for the untrusted service may be generated by the terminal as shown in fig. 11. Then, the terminal sends the processing request to the server, and the server processes the untrusted service after receiving the service request.

Therefore, through the process in the present specification, after it is determined that the service to be determined is an untrusted service, in order to avoid that the execution of the service is affected by a risk that the untrusted service may face, a server is adopted to perform further risk identification on the untrusted service, and the process of the risk identification is different from the process of determining whether the service is trusted or not.

If the server identifies the risk of the untrusted service as follows: although the untrusted service is risky, the risk of the untrusted service is not enough to start a subsequent verification step (that is, the untrusted service does not meet the verification start condition), a service processing instruction is generated and sent to the terminal, and the terminal continues to process the untrusted service; and if the risk identification result of the server to the untrusted service is that the untrusted service meets the verification starting condition, generating a verification instruction to be sent to the terminal, and verifying the untrusted service by the terminal.

And after receiving the verification instruction, the terminal verifies the non-trusted service. If the verification is passed, processing the untrusted service; and if the verification fails, displaying the processing result of the failure in the verification. Alternatively, the presentation may be by sound, light, vibration, graphical information, text, etc. information.

In the foregoing scenario where the user performs a transaction with the merchant through the terminal, the pending service may be an order to be paid. If the order to be paid is the credible service, determining that the payment mode of the order to be paid is a first payment mode; and if the order to be paid is the untrusted service, determining that the payment mode of the order to be paid is a second payment mode, wherein the security level of the second payment mode is higher than that of the first payment mode.

Therefore, when the process in the specification determines that the service is not credible, the service is further risk-identified in other modes, so that the accuracy of a risk identification result is improved, and the safety of service execution is improved.

Reviewing the whole process, the description represents the trusted service corresponding to the terminal in the form of a "trusted subgraph", and further, the service to be determined can be evaluated based on the trusted subgraph. The credible subgraph can be designed according to actual service requirements (for example, the number of the adjustment nodes and the number of the adjustment edges), so that the credible subgraph has high expandability, the flexibility of credible service judgment can be improved, and the process in the description can be suitable for different scenes. Further, since the information represented by the credible subgraph does not only include the nodes included in the subgraph, but also includes the relationships between the nodes, the description uses the credible subgraph to measure whether the business is credible or not at least from the business objects corresponding to the business and the dimensions of the relationships between the business objects, and the accuracy of the obtained credible determination result is improved.

According to an embodiment of another aspect, further referring to fig. 12, as an implementation of the methods shown in some of the above figures, the present specification provides an embodiment of a first trusted service determining apparatus, which corresponds to the method embodiment shown in fig. 8, and which can be applied to a terminal as shown in fig. 2.

As shown in fig. 12, the trusted service determining apparatus of this embodiment may include:

a credibility subgraph obtaining unit 1200 configured to obtain a credibility subgraph after the pending service is detected, wherein the credibility subgraph is obtained based on historical service data and includes credibility relationships between the terminal and a plurality of service objects;

a reference credibility determining unit 1202, configured to determine a reference credibility of the pending service according to the credible subgraph;

a comprehensive credibility determining unit 1204, configured to determine a comprehensive credibility of the pending service according to the reference credibility and the information of the service object triggering the pending service.

the reference credibility determination unit 1202 is configured to: determining a benchmark trustworthiness level based on at least one of: and matching degree of the first interaction and the credible subgraph, and matching degree of the second interaction and the credible subgraph.

the comprehensive credibility determining unit 1204 is configured to: acquiring behavior information of the user in a preset time period before the undetermined service is triggered; determining the difference between the user behavior and the behavior of the user when the credible service is triggered historically according to the behavior information by adopting a pre-trained comprehensive credibility determination model; and determining the comprehensive credibility according to the difference.

In one embodiment, the apparatus further comprises a data transmitting unit 1206, the data transmitting unit 1206 configured to: sending the data of the pending service to the server, so that the server updates the credible subgraph according to the data of the pending service;

the credible sub-graph obtaining unit 1200 is configured to: and receiving the updated credible subgraph corresponding to the terminal from the server.

With further reference to fig. 13, as an implementation of the methods shown in some of the above figures, the present specification provides an embodiment of a second trusted service determination apparatus, which corresponds to the method embodiment shown in fig. 4, and which may be applied to a server as shown in fig. 2.

As shown in fig. 13, the trusted service determining apparatus of this embodiment may include:

a historical service data obtaining unit 1300 configured to obtain a plurality of historical service data collected by a terminal, where the historical service data includes interaction data between the terminal and a service object;

a credible subgraph generating unit 1302, configured to generate a credible subgraph corresponding to the terminal according to the plurality of historical business data, where the credible subgraph includes credible relationships between the terminal and a plurality of business objects;

and a trusted subgraph sending unit 1304 configured to send the generated trusted subgraph to the terminal, so that when the terminal detects a pending service, the terminal determines whether the pending service is trusted according to the trusted subgraph and information of a service object triggering the pending service.

In one embodiment, the historical service data obtaining unit 1300 is configured to: acquiring a plurality of historical service data acquired by a terminal in a first preset time period;

the trusted graph generation unit 1302 is configured to: and generating a credible sub-graph of the terminal in a second preset time period in the future according to the plurality of historical service data, wherein the second preset time period is determined based on the first preset time period.

In one embodiment, the plurality of historical traffic data includes: the first service data comprise interaction data of the terminal and a first payment account; wherein the credible subgraph generating unit 1302 is configured to: and determining whether the relationship between the terminal and the first payment account is a trusted relationship or not according to at least one of the quantity of the first service data, the resource corresponding to the first service data and the earliest time corresponding to the first service data, and adding the relationship between the terminal and the first payment account into the trusted subgraph if the relationship is determined to be the trusted relationship.

In one embodiment, the plurality of historical traffic data includes: second service data, wherein the second service data comprises interaction data of the terminal and a second revenue account; the credible sub-graph generating unit 1302 is configured to:

In one embodiment, the trusted subgraph generating unit 1302 is configured to: determining a preset credible attribute of the second income account; and determining whether the second interaction is in a credible relation within a future preset second preset time period according to the probability and the credible attribute.

Embodiments of the present specification further provide a first computer-readable storage medium, where the storage medium stores a computer program, and the computer program is operable to execute the trusted service determination process provided in fig. 8.

A second computer-readable storage medium is provided in an embodiment of the present specification, and the storage medium stores a computer program, where the computer program is operable to execute the trusted service determination process provided in fig. 4.

The embodiment of the present specification also proposes a schematic structural diagram of the electronic device shown in fig. 14. As shown in fig. 14, at the hardware level, the electronic device may include a processor, an internal bus, a network interface, a memory, and a non-volatile memory, and may also include hardware required for other services. The processor reads the corresponding computer program from the nonvolatile memory into the memory and then runs the computer program to realize any one of the above trusted service determination processes.

Of course, besides the software implementation, the present specification does not exclude other implementations, such as a combination of logic devices or software and hardware, and the like, that is, the execution subject of the following processing flow is not limited to each logic unit, and may also be hardware or a logic device.

The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the system embodiment, since it is substantially similar to the method embodiment, the description is simple, and for the relevant points, reference may be made to the partial description of the method embodiment.

The foregoing description of specific embodiments has been presented for purposes of illustration and description. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.

It will be further appreciated by those of ordinary skill in the art that the elements and algorithm steps of the examples described in connection with the embodiments disclosed herein may be embodied in electronic hardware, computer software, or combinations of both, and that the components and steps of the examples have been described in a functional general in the foregoing description for the purpose of illustrating clearly the interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application. The software modules may reside in Random Access Memory (RAM), memory, Read Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the present invention in further detail, and it should be understood that the above-mentioned embodiments are merely exemplary embodiments of the present invention, and are not intended to limit the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.

Claims

1. A trusted service determination method, the method being performed by a terminal, the method comprising:

after the undetermined service is detected, acquiring a credible subgraph, wherein the credible subgraph is acquired based on historical service data and comprises service objects corresponding to credible relations in the historical service data as nodes, and edges connecting the nodes are determined according to the credible relations; the pending service comprises a first interaction and a second interaction;

determining a benchmark credibility of the service to be determined according to at least one of the matching degree of the first interaction and the credible subgraph and the matching degree of the second interaction and the credible subgraph;

and determining the comprehensive credibility of the undetermined service according to the reference credibility and the information of the service object triggering the undetermined service.

2. The method of claim 1, wherein the first interaction comprises an interaction of the terminal with a payment account and the second interaction comprises an interaction of the terminal with a revenue account.

3. The method of claim 1, wherein the business object comprises: a user using the terminal;

determining the comprehensive credibility of the undetermined service according to the reference credibility and the information of the service object triggering the undetermined service, wherein the comprehensive credibility of the undetermined service comprises the following steps:

Acquiring behavior information of the user in a preset time period before the pending service is triggered;

and determining the comprehensive credibility according to the difference.

4. The method of claim 1, further comprising:

sending the data of the pending service to a server, so that the server updates the credible subgraph according to the data of the pending service;

the obtaining of the credible subgraph comprises the following steps:

5. A trusted traffic determination method, the method being performed by a server, the method comprising:

generating a credible subgraph corresponding to the terminal according to the historical business data, wherein the credible subgraph comprises business objects corresponding to credible relations in the historical business data as nodes, and edges connecting the nodes are determined according to the credible relations;

Sending the generated credible subgraph to the terminal, so that when the terminal detects the undetermined service, the reference credibility degree of the undetermined service is determined according to at least one of the matching degree of a first interaction and the credible subgraph and the matching degree of a second interaction and the credible subgraph included in the undetermined service; and determining the comprehensive credibility of the undetermined service according to the benchmark credibility and the information of the service object triggering the undetermined service.

6. The method of claim 5, wherein obtaining a plurality of historical traffic data collected by a terminal comprises: acquiring a plurality of historical service data acquired by a terminal in a first preset time period;

7. The method of claim 5, wherein the plurality of historical traffic data comprises: the first service data comprise interaction data of the terminal and a first payment account; generating a credible subgraph corresponding to the terminal according to the plurality of historical service data comprises the following steps: and determining whether the first relationship between the terminal and the first payment account is a trusted relationship or not according to at least one of the quantity of the first service data, the resource corresponding to the first service data and the earliest time corresponding to the first service data, and adding the first relationship into the trusted subgraph if the first relationship is determined to be the trusted relationship.

8. The method of claim 6, wherein the plurality of historical traffic data comprises: second service data, wherein the second service data comprises interaction data of the terminal and a second revenue account; generating a credible subgraph of the terminal in a second preset time period in the future according to the plurality of historical service data, wherein the credible subgraph comprises the following steps:

determining whether a second relationship between the terminal and the second revenue account is a trusted relationship in the second preset time period according to at least one of the frequency of the second service data and the attribute of the second revenue account; or, a preset trigger probability prediction model is adopted to determine the probability that the service corresponding to the second service data is triggered in the second preset time period; determining whether the second relation is a trusted relation within the second preset time period or not according to the probability;

9. The method of claim 8, wherein determining whether the second relationship is a trusted relationship within the second predetermined period of time based on the probability comprises:

Determining a preset credible attribute of the second income account;

10. A trusted traffic determination apparatus, the apparatus being for a terminal, the apparatus comprising:

the credible subgraph acquisition unit is configured to acquire a credible subgraph after a pending service is detected, wherein the credible subgraph is acquired based on historical service data and comprises service objects corresponding to credible relations in the historical service data as nodes, and edges connecting the nodes are determined according to the credible relations; the pending service comprises a first interaction and a second interaction;

a benchmark credibility determining unit configured to determine a benchmark credibility of the service to be determined according to at least one of a matching degree of the first interaction and the credible subgraph and a matching degree of the second interaction and the credible subgraph;

11. The apparatus of claim 10, wherein the first interaction comprises an interaction of the terminal with a payment account and the second interaction comprises a second interaction of the terminal with a revenue account.

12. The apparatus of claim 10, wherein the business object comprises: a user using the terminal;

the comprehensive credibility determining unit is configured to: acquiring behavior information of the user in a preset time period before the pending service is triggered; determining the difference between the user behavior and the behavior of the user when the credible service is triggered historically according to the behavior information by adopting a pre-trained comprehensive credibility determination model; and determining the comprehensive credibility according to the difference.

13. The apparatus of claim 10, wherein the apparatus further comprises a data transmission unit configured to: sending the data of the pending service to a server, so that the server updates the credible subgraph according to the data of the pending service;

14. A trusted traffic determination apparatus, the apparatus for a server, the apparatus comprising:

A credible subgraph generating unit configured to generate a credible subgraph corresponding to the terminal according to the plurality of historical business data, wherein the credible subgraph comprises business objects corresponding to each credible relationship in the plurality of historical business data as nodes, and edges connecting the nodes are determined according to each credible relationship;

the credible subgraph sending unit is configured to send the generated credible subgraph to the terminal, so that when the terminal detects undetermined service, the reference credibility of the undetermined service is determined according to at least one of the matching degree of a first interaction and the credible subgraph and the matching degree of a second interaction and the credible subgraph; and determining the comprehensive credibility of the service to be determined according to the reference credibility and the information of the service object triggering the service to be determined.

15. The apparatus of claim 14, wherein the historical traffic data acquisition unit is configured to: acquiring a plurality of historical service data acquired by a terminal in a first preset time period;

the credible sub-graph generating unit is configured to: and generating a credible sub-graph of the terminal in a second preset time period in the future according to the plurality of historical service data, wherein the second preset time period is determined based on the first preset time period.

16. The apparatus of claim 14, wherein the plurality of historical traffic data comprises: the first service data comprise interaction data of the terminal and a first payment account; wherein the credible subgraph generation unit is configured to: and determining whether the first relationship between the terminal and the first payment account is a trusted relationship or not according to at least one of the quantity of the first service data, the resource corresponding to the first service data and the earliest time corresponding to the first service data, and adding the first relationship into the trusted subgraph if the first relationship is determined to be the trusted relationship.

17. The apparatus of claim 15, wherein the plurality of historical traffic data comprises: second service data, wherein the second service data comprises interaction data of the terminal and a second revenue account; the credible subgraph generation unit is configured to:

determining whether a second relationship between the terminal and the second revenue account is a trusted relationship in the second preset time period according to at least one of the frequency of the second service data and the attribute of the second revenue account; or, a preset trigger probability prediction model is adopted to determine the probability that the service corresponding to the second service data is triggered in the second preset time period; determining whether the second relation is a credible relation within the second preset time period according to the probability;

And in the case that the second relation is determined to be a trusted relation, adding the second relation into the trusted subgraph.

18. The apparatus according to claim 17, wherein the trusted sub-graph generating unit is configured to: determining a preset credible attribute of the second income account; and determining whether the second relation is a credible relation within the second preset time period according to the probability and the credible attribute.

19. A computer-readable storage medium, on which a computer program is stored which, when executed in a computer, causes the computer to carry out the method of any one of claims 1-9.

20. A computing device comprising a memory having executable code stored therein and a processor that, when executing the executable code, implements the method of any of claims 1-9.