JP7428859B2 - Judgment system, judgment method, and program - Google Patents

Description

The present invention relates to a determination system, a determination method, and a program.

Conventionally, a monitoring device has been disclosed that identifies commercial transactions suspected of being fraudulent based on rule information (for example, see Patent Document 1). This monitoring device updates the rule information so that the number of fraudulent transaction candidates for which check requests are made remains constant.

Japanese Patent Application Publication No. 2020-113178

However, with the above-mentioned conventional techniques, suspicion of fraud may not be accurately determined in some cases. For example, the above-mentioned monitoring device does not update the rule information when the number of fraudulent transaction candidates is constant, so if a person committing fraud identifies a rule during this period, a transaction that satisfies the conditions of the rule information will be performed. This is because there are cases.

The present invention takes such circumstances into consideration, and one of the objects of the present invention is to provide a determination system, a determination method, and a program that can accurately determine suspicion of fraud.

In one aspect of the present invention, a determination system includes an acquisition unit that acquires determination target information indicating the content of an electronic commerce transaction, and a rule used by the determination unit for determination or a determination condition included in the rule that is provided to a user conducting the electronic commerce. a changing unit that changes one or both of a rule included in a plurality of rules and a judgment condition included in the rule for each judgment process performed by the judgment unit so as not to be identified; and the judgment target information. and a determination unit that determines whether there is a suspicion of fraud in electronic commerce based on the plurality of rules including one or both of the rule changed by the change unit and the determination condition.

According to one aspect of the present invention, suspicion of fraud can be accurately determined.

1 is a diagram showing an example of a functional configuration of a determination system 1. FIG. FIG. 3 is a sequence diagram showing an example of the flow of processing. 3 is a diagram showing an example of the contents of user information 32 that the service providing device 10 has. FIG. It is a figure which shows an example of the content of determination target information. FIG. 2 is a diagram for explaining Concern 1. FIG. FIG. 3 is a diagram for explaining concern 2. FIG. This is an example of a probability distribution (normal distribution) used by the changing unit 54. 6 is a diagram showing an example of rule information 62. FIG. 6 is a diagram showing an example of rule change information included in change information 64. FIG. 6 is a diagram illustrating an example of change information 64A indicating whether the determination condition of rule 1 is adopted or rejected. FIG. 7 is a diagram illustrating an example of change information 64B of modification example 2 of countermeasure 2. FIG.

DESCRIPTION OF EMBODIMENTS Hereinafter, embodiments of a determination system, a determination method, and a program of the present invention will be described with reference to the drawings.

[overview]
A determination system according to an aspect of the present embodiment includes an acquisition unit that acquires determination target information indicating the content of an electronic commerce transaction, and a rule used by the determination unit for determination or a determination condition included in the rule that is provided to a user who conducts the electronic commerce. a changing unit that changes one or both of a rule included in a plurality of rules and a judgment condition included in the rule for each judgment process performed by the judgment unit so as not to be identified; and the judgment target information. and a determination unit that determines whether there is a suspicion of fraud in electronic commerce based on the plurality of rules including one or both of the rule changed by the change unit and the determination condition.

"Change for each judgment process" means, in principle, to change every time, every predetermined number of times, or once every time, but do not change it temporarily during the process. It may also include.

"Electronic commerce" includes, but is not limited to, at least shopping, auctions, flea markets, reservations for the use of facilities, or reservations for the provision of services. Any type of electronic commerce may be used as long as the suspicion of fraud is determined based on rules or determination conditions. Hereinafter, as an example, electronic commerce will be explained as shopping.

[Judgment system]
FIG. 1 is a diagram showing an example of the functional configuration of the determination system 1. As shown in FIG. The determination system 1 includes, for example, one or more user terminal devices U, a service providing device 10, and a determination device 50. The service providing device 10 and the determining device 50 communicate with each other via a network. The service providing device 10 and the terminal device U communicate with each other via the network NW. The network NW includes, for example, the Internet, a LAN (Local Area Network), a WAN (Wide Area Network), a cellular network, and the like.

[Terminal device]
The terminal device U is a computer device having a communication function, such as a smartphone, a tablet terminal, or a personal computer.

[Service providing device]
The service providing device 10 is, for example, a shopping server that provides shopping-related services to users. The service providing device 10 includes, for example, an information processing section 20 and a storage section 30. The information processing unit 20 is realized, for example, by a hardware processor such as a CPU (Central Processing Unit) executing a program (software). Some or all of these components are hardware (circuit parts) such as LSI (Large Scale Integration), ASIC (Application Specific Integrated Circuit), FPGA (Field-Programmable Gate Array), and GPU (Graphics Processing Unit). (including circuitry), or may be realized by collaboration between software and hardware. The program may be stored in advance in a storage device (a storage device with a non-transitory storage medium) such as an HDD (Hard Disk Drive) or flash memory, or may be stored in a removable storage device such as a DVD or CD-ROM. It may be stored in a medium (non-transitory storage medium), and installed in the storage device by loading the storage medium into a drive device.

The storage unit 30 is realized by, for example, a ROM (Read Only Memory), a flash memory, an SD card, a RAM (Random Access Memory), an HDD (Hard Disc Drive), a register, or the like. Further, part or all of the storage unit 30 may be a NAS (Network Attached Storage), an external storage server device, or the like. The storage unit 30 stores, for example, user information 32 (described later).

The information processing unit 20 provides content related to shopping in response to a request from the terminal device U. The information processing unit 20 supports the purchase of products or services and performs payment for the purchase according to the user's operations on the content. When making a payment, the information processing unit 20 requests the determination device 50 to determine whether a fraudulent transaction is suspected. If the information processing unit 20 receives a determination result from the determination device 50 that the transaction is not fraudulent, it notifies the terminal device U that the payment has been completed and arranges for the product or service.

[Judgment device]
The determination device 50 includes, for example, an information acquisition unit (acquisition unit) 52, a change unit 54, a determination unit 56, a provision unit 58, and a storage unit 60. Some or all of the information acquisition section 52, the change section 54, the determination section 56, and the provision section 58 are realized by, for example, a hardware processor such as a CPU executing a program (software). Some or all of these components may be realized by hardware (including circuitry) such as LSI, ASIC, FPGA, or GPU, or may be realized by collaboration between software and hardware. Good too. The program may be stored in advance in a storage device such as an HDD or flash memory (storage device equipped with a non-transitory storage medium), or may be stored in a removable storage medium (non-transitory storage medium) such as a DVD or CD-ROM. The software may be installed in the storage device by attaching the storage medium to the drive device. Note that part or all of the functions of the determination device 50 may be included in another device such as the service providing device 10.

The storage unit 60 is realized by, for example, a ROM, flash memory, SD card, RAM, HDD, register, or the like. Furthermore, part or all of the storage unit 60 may be a NAS, an external storage server device, or the like. The storage unit 60 stores, for example, rule information 62 and change information 64. This information will be described later.

The information acquisition unit 52 acquires information provided to the service providing device 10. For example, the information acquisition unit 52 acquires determination target information indicating the content of electronic commerce.

The changing unit 54 changes any of the rules included in the plurality of rules for each determination process performed by the determining unit 56 so that the rules used in the determination by the determining unit 56 or the determination conditions included in the rules are not specified by users conducting electronic commerce. Change one or both of the rule and the judgment conditions included in the rule.

The determining unit 56 determines whether the electronic commerce is suspected of being fraudulent based on the determination target information and a plurality of rules including one or both of the rules and the determination conditions changed by the changing unit 54.

The providing unit 58 provides the service providing device 10 with the determination result of the determining unit 56 or the determination result of the staff. The staff member is a person who determines in detail the suspicion of fraud in an electronic commerce transaction for which the judgment result of the determination unit 56 is determined to be suspicious, or a person who determines the suspicion of fraud using rules.

[Sequence Diagram]
First, an example of the processing flow of the determination system will be described. FIG. 2 is a sequence diagram showing an example of the flow of processing.

When the user performs a payment operation (S10), the terminal device U transmits operation information indicating that the operation has been performed to the service providing device 10 (S12). Next, upon acquiring the operation information, the service providing device 10 generates determination target information including the user information 32 and information indicating the details of the transaction (S14), and transmits the generated determination target information to the determination device 50. (S16). The determination target information will be described later.

Next, the information acquisition unit 52 of the determination device 50 acquires the determination target information, and the modification unit 54 modifies the rule or determination condition of the rule information 62 based on the modification information 64 (S18). Next, the determination unit 56 determines whether there is a suspicion of fraud in the electronic commerce based on the changed rules or determination conditions (S20).

Next, the providing unit 58 transmits the determination result of the determining unit 56 to the service providing device 10 (S22). Next, the service providing device 10 provides the determination result provided by the determining device 50 to the terminal device U (S24).

Through this process flow, it is determined whether electronic commerce is fraudulent. Below, we will explain the user information 32 used in this process, the judgment target information including the user information 32, concerns when the changed rules are not used, and the judgment process based on transaction information and the changed rules. .

[User information]
FIG. 3 is a diagram showing an example of the contents of the user information 32 that the service providing device 10 has. The user information 32 is, for example, information in which a user ID, attribute information, usage history, and evaluation information are associated with each other. Note that some of this information may be omitted. The user ID is identification information of a user, and is, for example, identification information input when using a service or logging in. The attribute information is the user's attributes registered by the user, such as address, telephone, email address, gender, and age. In addition to the above, the attribute information may include various information such as annual income, hobbies, and family structure.

The usage history is the usage history of the user's service. For example, the information includes content browsing history, products (or services) placed in a shopping cart, purchased products, and payment methods used in the past. Evaluation information is information indicating a user's evaluation generated based on the user's behavior in using the service. For example, the evaluations of users who have written reviews or acted in accordance with the intentions of the service provider may tend to be higher than the evaluations of other users. Furthermore, the evaluation of the evaluation information may take into account evaluation results by a user different from the user.

The service providing device 10 generates determination target information including the user information 32 described above and electronic commerce transaction information (transaction information).

[Judgment target information]
FIG. 4 is a diagram showing an example of the contents of the determination target information. The determination target information includes transaction information and user information 32. The transaction information includes, for example, information on the product (or service) that the user intends to purchase, information on payment (payment method, credit card number), delivery destination, delivery date, and various other information.

The service providing device 10 transmits the determination target information to the determination device 50 and requests the determination device 50 to determine whether fraud is suspected based on the determination target information. The determination device 50 applies the changed rules or determination conditions to the acquired determination target information to determine suspicion of fraud.

[Concerns when not using the changed rules or judgment conditions]
Here, concerns will be explained when a determination of fraud is made using a predetermined rule or determination condition without changing the rule or determination condition. For example, the determination device 50 applies a plurality of rules (rules consisting of one or more determination conditions) to the determination target information, and when the determination target information satisfies each of the plurality of rules, or when the determination target information satisfies each of the plurality of rules, , it is determined that there is no suspicion of fraud. Further, the determination device 50 may apply determination target information to each of a plurality of rules to derive a score, and compare the score with a threshold value to determine suspicion of fraud. In the above process, if the rules are not changed, there is a risk that a user who commits fraud will identify the rules or judgment conditions and use the identification results to avoid suspicion of fraud.

(Concern 1)
There is a concern that the threshold value of a rule or judgment condition may be identified by a fraudulent user making multiple attempts. FIG. 5 is a diagram for explaining concern 1. For example, it is assumed that under certain rules or judgment conditions, a payment of 30,000 yen or more is likely to be suspected of being fraudulent. In this case, if the settlement amount threshold is set low, false positives will be determined, and if it is set high, the number of false negatives will increase. Although an appropriate threshold has been set, if a fixed threshold is set, there is a possibility that the threshold will be identified if a fraudulent user makes multiple attempts with different payment amounts. If the threshold value is specified, there is a concern that the above-mentioned rules may be bypassed and fraudulent payments may be made with the amount closest to the threshold value, increasing the amount of damage caused by the fraudulent payments.

For example, when binary search is used as the search algorithm, the average number of trials and the maximum number of trials are expressed by the formula shown in FIG. Consider trial pattern 1-3 using this. In trial pattern 1 where the preference range is a list of 10 items, the average number of trials is 3.3 and the maximum error is 10,000 yen. In trial pattern 2 where the preference range is a list of 90 items, the average number of trials is 6.5 and the maximum error is 1000 yen. In trial pattern 3 where the preference range is a list of 90,000 items, the average number of trials is 16.5 and the maximum error is 0 yen.

Thus, a fraudulent user may be able to identify a threshold or approximate value, which is a matter of concern.

(Concern 2)
There is a concern that the rules or judgment conditions may be identified by a user committing fraud multiple times. FIG. 6 is a diagram for explaining concern 2. It is assumed that the judgment condition of the rule is "condition A∧condition B∧condition B". "∧" is a logical product. For example, if you remove any of Condition A, Condition B, or Condition C, the number of false positive results will increase, and if you add conditions other than Condition A, Condition B, or Condition C, the number of false positive results will increase. The number of false negative results increases. For this reason, it is desired to set an appropriate combination of conditions, ie, "condition A∧condition B∧condition B." However, if "Condition A∧Condition B∧Condition B" is set, there is a possibility that the above judgment condition will be identified by an unauthorized user trying a different combination of conditions. There is a concern that if the judgment conditions are specified, the above rules will be bypassed and the amount of damage caused by fraudulent payments will increase.

For example, when the number of conditions to be tried is "n", the maximum number of trials required to specify a condition can be determined using a formula as shown in FIG. Consider trial patterns 1 and 2 using this. In trial pattern 1 in which the number of trial conditions is three, the maximum number of trials (number of combinations) is eight. In trial pattern 2 in which the number of trial conditions is five, the maximum number of trials (number of combinations) is 32.

In this way, a user who commits fraud can identify the determination conditions by repeating trials, which is a matter of concern.

[Measures to address concerns]
As countermeasures for concerns 1 and 2 above, the changing unit 54 of the determination device 50 performs countermeasures 1 and 2.

(Measure 1)
The changing unit 54 changes the threshold value of the rule or the determination condition using the probability distribution. FIG. 7 is an example of a probability distribution (normal distribution) used by the changing unit 54. The vertical axis in FIG. 7 shows the probability, and the horizontal axis shows the threshold value of the payment amount. In FIG. 7, the average value μ is 30,000 yen. The standard deviation α is 10,000 yen.

The changing unit 54 changes the threshold based on the algorithm included in the change information 64. When the change unit 54 inputs a request, the algorithm outputs a threshold value according to the probability distribution set for the determination condition. This algorithm may be software published in an open source library. Further, the changing unit 54 may request another device to output a threshold value according to the probability distribution, and may obtain the changed threshold value from the request destination.

For example, the changing unit 54 uses a probability distribution as described above, so that the distribution of thresholds selected for each process is changed to follow the probability distribution. For example, 30,000 yen or a value close to 30,000 yen is frequently selected as a threshold value, and a threshold value far from 30,000 yen is selected less frequently. Note that, for example, even if a value far from 30,000 yen is selected as the threshold value, a rule including a determination condition different from this determination condition is also applied, so the possibility of a false positive or false negative determination is suppressed.

As described above, by changing the threshold value of the determination condition based on the probability distribution, the changing unit 54 can change the threshold value while making use of the originally desired threshold value. As a result, it becomes difficult for an unauthorized user to specify the threshold value, and suspicion of fraud can be determined with high accuracy.

Note that the information acquisition unit 52 acquires determination target information indicating the content of the first electronic commerce related to the first service or determination target information indicating the content of the second electronic commerce related to the second service, and changes the information to the changing unit 54. changes the threshold value so that the distribution of the threshold value of the determination condition included in the rule used for the first electronic commerce follows the first probability distribution, and the distribution of the threshold value of the determination condition included in the rule used for the second electronic commerce is changed. The threshold value may be changed so that the distribution of the threshold value of the condition follows the second probability distribution.

For example, the changing unit 54 may change the probability distribution to be adopted for each type of electronic commerce. For example, different probability distributions are set for each type of electronic commerce. The type of electronic commerce is electronic commerce corresponding to the service. For example, shopping and flea market are different types. Furthermore, a probability distribution may be set depending on the type of transaction within one service. For example, a probability distribution may be set for each type of thing to be traded.

Note that in the above example, the threshold value has been described as a threshold value for an amount of money, but instead of (or in addition to) this, the threshold value for other indicators may be changed. For example, thresholds for indicators related to user behavior in electronic commerce, achievements, evaluations, and other indicators may be changed as described above.

(Measure 2)
The change unit 54 refers to the change information 64 and determines whether the rule information 62 of the rule information is valid or invalid. The changing unit 54 determines whether to adopt a predetermined rule included in the plurality of rules for each determination process performed by the determining unit 56.

FIG. 8 is a diagram showing an example of the rule information 62. The rule information 62 is information indicating the content of the rule used when the determination unit 56 determines the determination target information. Rule information 62 includes a plurality of rules. A rule includes one or more determination conditions. If the judgment conditions included in the rule are met, it indicates a tendency towards suspected fraud. For example, Rule 1 is to satisfy judgment conditions AC. Rule 2 is to satisfy judgment conditions DF. Rule 3 is that determination conditions G and H are satisfied.

The changing unit 54 determines whether or not to adopt a predetermined rule based on the degree of adoption or rejection (rule change information) set for the predetermined rule for each determination process performed by the determination unit 56. FIG. 9 is a diagram showing an example of rule change information included in the change information 64. The rule change information is information in which a valid rate or invalid rate is associated with a rule. The effective rate is the rate at which the rule is used in the determination process. The invalid rate is the rate at which the rule is not used in the determination process. For example, looking at Rule 2, Rule 2 is adopted once out of every two times.

The changing unit 54 refers to the rule change information and determines whether or not to adopt the rule. This results in different rules being adopted for each determination process.

As described above, the changing unit 54 can avoid identification of the rule by an unauthorized user by determining whether to adopt the rule and changing the rule. It is expected that repeated fraud by unauthorized users can be avoided and damage caused by unauthorized use will be suppressed.

Note that the degree to which the rule is enabled or disabled may be set depending on the characteristics of the rule. For example, rules that are likely to be easily identified may be set to have a relatively high invalidation rate. A rule may be established that does not invalidate or has a relatively low rate of invalidation. This rule is, for example, a rule that is difficult for an unauthorized person to attempt, or a rule that it is desirable to make a determination in accordance with the rule even if an unauthorized person is identified.

As described above, the determination device 50 can eliminate the concern by applying one or both of Measure 1 and Measure 2 to Concern 1 and applying Measure 2 to Concern 2. can. Thereby, the determination device 50 can accurately determine suspicion of fraud and suppress unauthorized use.

(Variation 1 of Measure 2)
In the above-mentioned measure 2, it is assumed that a rule is adopted or rejected, but instead of (or in addition to) this, it is also possible to decide whether or not a determination condition included in a rule is adopted. For each determination process performed by the determination unit 56, the changing unit 54 changes the degree of adoption or rejection of the condition based on the degree of adoption or the degree of rejection set for at least one condition among the plurality of conditions constituting the rule included in the plurality of rules. Decide whether to accept or reject the application.

FIG. 10 is a diagram illustrating an example of change information 64A indicating whether the determination condition of rule 1 is adopted or rejected. For example, the degree of acceptance or rejection is associated with each determination condition. For example, determination condition A and determination condition C are valid 100% of the time, and determination condition C is valid 50% of the time. Therefore, when making a judgment using Rule 1, the judgment unit 56 employs judgment condition C once in every two times.

As described above, by the changing unit 54 deciding whether to adopt or reject the determination conditions, suspicion of fraud can be determined with high accuracy.

(Variation 2 of Measure 2)
In Modification 2 of Countermeasure 2, the determination is made taking into consideration a group including a plurality of rules. For example, the plurality of rules includes a plurality of predetermined rules forming a set, and the changing unit 54 employs at least one predetermined rule among the plurality of predetermined rules in the target determination process, At least one predetermined rule different from the adopted predetermined rule is not adopted in the determination process.

Further, for example, the plurality of predetermined rules include a first predetermined rule and a second predetermined rule that form a pair, and the changing unit 54 selects the first predetermined rule and the second predetermined rule in the target determination process. One of the two predetermined rules is adopted, and the other rule is not adopted.

Each of the plurality of predetermined rules is a rule that has a mutually complementary relationship, and a complementary relationship is defined as a rule in which the determination result using one rule and the determination result using the other rule are similar to each other to a degree greater than or equal to a threshold value. It is a relationship that results.

FIG. 11 is a diagram illustrating an example of change information 64B of Modification 2 of Countermeasure 2. For example, rule 4 and rule 5 are a pair. For example, the constraint condition when rule 4 is invalid is that rule 5 is applied. The changing unit 54 enables rule 5 when rule 4 is invalidated, and enables rule 4 when rule 5 is invalidated. Rule 4 and Rule 5 have a complementary relationship, and by applying either rule, it is possible to accurately determine fraud, and it is also possible to avoid the rule from being identified by a fraudulent user. .

Further, for example, rule 6 to rule 8 is a group. For example, the constraint when rule 6 is invalid is that rule 7 or 8 is applied. The changing unit 54 enables rule 7 or 8 when rule 6 is disabled, enables rule 6 or 8 when rule 7 is disabled, and further enables rule 6 or 7 when rule 8 is disabled. Enable. There is a complementary relationship with rules 6-8, and by applying either rule, it is possible to accurately determine fraud, and furthermore, it is possible to prevent the rule from being identified by a fraudulent user. Note that if the combination of rules 6 and 7 is complementary to rule 8, if rule 8 is invalid, rules 6 and 7 may be validated.

As described above, by enabling one of the rules in a complementary relationship and disabling the other, the changing unit 54 can accurately determine fraud, and furthermore, it is possible to accurately determine a rule by a fraudulent user. This can be avoided.

Furthermore, the determining unit 56 may cause the storage unit 60 to store the processing result of the changing unit 54 described above as log information. The log information is later associated with the actual result of the electronic commerce (information indicating whether the electronic commerce was fraudulent or not). Using this log information, the administrator may adjust the information used when changing the threshold, or adjust the rules or judgment conditions for invalidation and the degree of invalidation.

In addition, if the e-commerce is other e-commerce such as an auction or a flea market, various information such as the number of characters in comments, the number of items listed, image information, etc. will be included in the judgment target information, and thresholds and rules for judging this will be included. may be changed using the above method.

According to the embodiment described above, the determination system acquires determination target information indicating the content of an electronic commerce transaction, and prevents the user conducting the electronic commerce from specifying the rules used for determination or the determination conditions included in the rules. , for each judgment process, change one or both of the rules included in the plurality of rules and the judgment conditions included in the rule, and change the judgment target information and one or both of the changed rule and judgment conditions. By determining the suspicion of fraud in electronic commerce based on a plurality of rules including the following, it is possible to accurately determine the suspicion of fraud in electronic commerce.

Although the mode for implementing the present invention has been described above using embodiments, the present invention is not limited to these embodiments in any way, and various modifications and substitutions can be made without departing from the gist of the present invention. can be added.

1. Judgment system includes 10. Service providing device 20. Information processing section 50. Judgment device 52. Information acquisition section 54. Modification section 56. Judgment section 58. Providing section 60. Storage section 62. Rule information 64. Modification information.

Claims (13)

an acquisition unit that acquires judgment target information indicating the content of the electronic commerce;
For each judgment process performed by the judgment unit, one rule included in a plurality of rules and one of the rules used for judgment by the judgment unit or the judgment conditions included in the rules should not be identified by the user conducting the e-commerce transaction. a changing section that changes one or both of the judgment conditions included in the rule;
a determination unit that determines whether there is a suspicion of fraud in electronic commerce based on the determination target information and the plurality of rules including one or both of the rule changed by the change unit and the determination condition ; ,
The acquisition unit acquires determination target information indicating the content of the first electronic commerce related to the first service or determination target information indicating the content of the second electronic commerce related to the second service,
The change section is
changing the threshold so that the distribution of the threshold of the determination condition included in the rule used for the first electronic commerce follows a first probability distribution;
changing the thresholds so that the distribution of the thresholds of the determination conditions included in the rules used for the second electronic commerce follows a second probability distribution;
Judgment system. The changing unit determines whether to adopt a predetermined rule included in the plurality of rules for each determination process performed by the determining unit.
The determination system according to claim 1. The changing unit determines whether or not to adopt the predetermined rule based on a degree of adoption or a degree of rejection set for the predetermined rule for each determination process performed by the determination unit.
The determination system according to claim 2. an acquisition unit that acquires judgment target information indicating the content of the electronic commerce;
For each judgment process performed by the judgment unit, one rule included in a plurality of rules and one of the rules used for judgment by the judgment unit or the judgment conditions included in the rules should not be identified by the user conducting the e-commerce transaction. a changing section that changes one or both of the judgment conditions included in the rule;
a determination unit that determines whether there is a suspicion of fraud in electronic commerce based on the determination target information and the plurality of rules including one or both of the rule changed by the change unit and the determination condition ; ,
The changing unit determines whether or not to adopt the predetermined rule based on a degree of adoption or a degree of rejection set for a predetermined rule included in the plurality of rules for each determination process performed by the determination unit,
The plurality of rules include a plurality of predetermined rules forming a set,
In the target determination process, the changing unit adopts at least one predetermined rule among the plurality of predetermined rules in the determination process, and adopts at least one predetermined rule different from the adopted predetermined rule. The rule is not adopted in the judgment process,
Judgment system. The plurality of predetermined rules include a first predetermined rule and a second predetermined rule that form a pair,
The changing unit adopts one of the first predetermined rule and the second predetermined rule and does not adopt the other rule in the target determination process.
The determination system according to claim 4 . Each of the plurality of predetermined rules is a rule that is complementary to each other,
The complementary relationship is a relationship in which the determination result using one rule and the determination result using the other rule are similar to a degree equal to or higher than a threshold value.
The determination system according to claim 4 or 5 . The changing unit changes the adoption degree or rejection degree set to at least one condition among a plurality of conditions constituting a rule included in the plurality of rules for each determination process performed by the determination unit. Deciding whether to accept or reject the conditions,
The determination system according to any one of claims 1 to 6 . The electronic commerce includes at least shopping, auctions, flea markets, reservations for the use of facilities, or reservations for the provision of services.
The determination system according to any one of claims 1 to 7 . The acquisition unit acquires the determination target information, attribute information of the user, history of use of the service by the user, and information indicating evaluation within the service from a service providing device that provides the electronic commerce service. ,
The determination unit determines whether there is a suspicion of fraud in the electronic commerce based on the plurality of rules changed by the change unit for each determination process and the information acquired by the acquisition unit.
The determination system according to any one of claims 1 to 8 . The computer is
an acquisition process for acquiring judgment target information indicating the content of the electronic commerce;
In order to prevent the user who conducts the electronic commerce from identifying the rules used in the determination process or the determination conditions included in the rules, for each determination process, one of the rules included in the plurality of rules and the determination conditions included in the rule are specified. a change process that changes one or both of the above;
the determination process of determining the suspicion of fraud in the electronic commerce based on the determination target information and the plurality of rules including one or both of the rule changed in the change process and the determination condition;
A process of acquiring determination target information indicating the content of the first electronic commerce related to the first service or determination target information indicating the content of the second electronic commerce related to the second service;
A process of changing the threshold value so that the distribution of the threshold value of the determination condition included in the rule used for the first electronic commerce follows a first probability distribution;
A process of changing the threshold value so that the distribution of the threshold value of the determination condition included in the rule used for the second electronic commerce follows a second probability distribution;
Judgment method to perform. to the computer,
an acquisition process for acquiring judgment target information indicating the content of the electronic commerce;
In order to prevent the user who conducts the electronic commerce from identifying the rules used in the determination process or the determination conditions included in the rules, for each determination process, one of the rules included in the plurality of rules and the determination conditions included in the rule are specified. a change process that changes one or both of the above;
the determination process of determining the suspicion of fraud in the electronic commerce based on the determination target information and the plurality of rules including one or both of the rule changed in the change process and the determination condition;
A process of acquiring determination target information indicating the content of the first electronic commerce related to the first service or determination target information indicating the content of the second electronic commerce related to the second service;
A process of changing the threshold value so that the distribution of the threshold value of the determination condition included in the rule used for the first electronic commerce follows a first probability distribution;
A process of changing the threshold value so that the distribution of the threshold value of the determination condition included in the rule used for the second electronic commerce follows a second probability distribution;
A program to run. The computer is
an acquisition process for acquiring judgment target information indicating the content of the electronic commerce;
In order to prevent the user who conducts the electronic commerce from identifying the rules used in the determination process or the determination conditions included in the rules, for each determination process, one of the rules included in the plurality of rules and the determination conditions included in the rule are specified. a change process that changes one or both of the above;
the determination process of determining the suspicion of fraud in the electronic commerce based on the determination target information and the plurality of rules including one or both of the rule changed in the change process and the determination condition;
for each of the determination processes, a process of determining whether to adopt the predetermined rule based on a degree of adoption or a degree of rejection set in the predetermined rule included in the plurality of rules;
The plurality of rules include a plurality of predetermined rules forming a set,
In the determination process of the target, at least one predetermined rule among the plurality of predetermined rules is adopted in the determination process, and at least one predetermined rule different from the adopted predetermined rule is used in the determination process. Processing that is not adopted for processing,
Judgment method to perform. to the computer,
an acquisition process for acquiring judgment target information indicating the content of the electronic commerce;
In order to prevent the user who conducts the electronic commerce from identifying the rules used in the determination process or the determination conditions included in the rules, for each determination process, one of the rules included in the plurality of rules and the determination conditions included in the rule are specified. a change process that changes one or both of the above;
the determination process of determining the suspicion of fraud in the electronic commerce based on the determination target information and the plurality of rules including one or both of the rule changed in the change process and the determination condition;
for each of the determination processes, a process of determining whether to adopt the predetermined rule based on a degree of adoption or a degree of rejection set in the predetermined rule included in the plurality of rules;
The plurality of rules include a plurality of predetermined rules forming a set,
In the determination process of the target, at least one predetermined rule among the plurality of predetermined rules is adopted in the determination process, and at least one predetermined rule different from the adopted predetermined rule is used in the determination process. Processing that is not adopted for processing,
A program to run.

Images