Social network relationship evaluation method and device

Abstract

The present invention provides a method and apparatus for evaluating social network relationships. The method utilizes transaction data acquired between a first entity and a second entity in a social network relationship to be evaluated as input to a pre-established transaction closeness model, processes the transaction data using the transaction closeness model, and outputs the transaction closeness between the first and second entities. The method then determines an evaluation result of the social network relationship between the first and second entities based on the transaction closeness. In this approach, the transaction data between the first and second entities in a social network relationship to be evaluated is processed within the pre-established transaction closeness model to obtain the transaction closeness between the first and second entities. The evaluation result of the social network relationship between the first and second entities is then determined based on the obtained transaction closeness, thereby achieving a complete and comprehensive evaluation of the social network relationships between entities.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a method and device for evaluating social network relationships. Background Art

[0002] With the in-depth application of technologies such as big data in banking, bank practitioners will use big data to analyze customers as independent entities to create customer profiles. In order to better provide customers with services, bank practitioners need to further understand the social network relationships between customers, that is, the social network relationships between entities.

[0003] Currently, in the existing technology, the evaluation of social network relationships between entities mostly uses the intimacy of traditional social relationships such as couples, parents and children, and friends to evaluate the social network relationships between entities.

[0004] However, relying solely on the traditional closeness of social relationships to evaluate the social network relationships between entities is too one-sided, resulting in incomplete and incomplete evaluation of the social network relationships between entities. Summary of the Invention

[0005] In view of this, embodiments of the present invention provide a method and apparatus for evaluating social network relationships to solve the problem of incomplete and incomplete evaluation of social network relationships between entities in the prior art.

[0006] To solve the above problems, the embodiments of the present invention provide the following technical solutions:

[0007] A first aspect of an embodiment of the present invention discloses a method for evaluating social network relationships, the method comprising:

[0008] Acquire transaction data between a first entity and a second entity in a social network relationship to be evaluated;

[0009] Using the transaction data as input to a pre-established transaction closeness model, processing the transaction data using the transaction closeness model, and outputting the transaction closeness between the first entity and the second entity;

[0010] An evaluation result of the social network relationship between the first entity and the second entity is determined based on the transaction closeness.

[0011] Optionally, the process of pre-establishing the transaction closeness model includes:

[0012] Get sample transaction data details;

[0013] Determining, based on the sample transaction data details, the relationship types between the sample transaction data and the sample entities between each two sample entities;

[0014] Constructing a three-dimensional coordinate system of transaction closeness based on the relationship type of the sample entities, the transaction time range, and the sample transaction data between each pair of sample entities, wherein the sample transaction data at least includes the sample transaction amount, the number of sample transactions, and the transaction time;

[0015] Extracting sample data features based on the time range and distribution of sample transaction data in the three-dimensional transaction density coordinate system, wherein the sample data features include at least a sample transaction number score, a sample transaction amount score, a sample transaction number percentage, a sample transaction amount percentage, a time decay factor, and a transaction balance factor;

[0016] A transaction closeness model is constructed based on the sample transaction number score, sample transaction amount score, sample transaction number ratio, sample transaction amount ratio, time decay factor and transaction balance factor.

[0017] Optionally, constructing a three-dimensional coordinate system of transaction closeness based on the relationship type of the sample entities, the transaction time range, and the sample transaction data between each two sample entities includes:

[0018] A three-dimensional coordinate system is established with the transaction time range as the X-axis, the transaction characteristics as the Y-axis, and the relationship type as the Z-axis;

[0019] Dividing the three-dimensional coordinate system into five binning intervals based on data quantiles, wherein the five binning intervals are normally distributed, and each binning interval is set with a corresponding score;

[0020] Determine the types of sample entities and the relationship types between any two sample entities, wherein the types include individual types and enterprise types, and the relationship types include individual and enterprise types, individual and individual types, and enterprise and enterprise types;

[0021] Building individual and enterprise sub-models, individual and individual sub-models, and enterprise and enterprise sub-models based on the relationship types of the sample entities;

[0022] A three-dimensional coordinate system for transaction closeness is constructed using the three-dimensional coordinate system, the individual and enterprise sub-models, the individual and individual sub-models, the enterprise and enterprise sub-models, and the sample transaction data between any two sample entities.

[0023] Optionally, extracting sample data features based on the time range and distribution of sample transaction data in the three-dimensional coordinate system of transaction closeness includes:

[0024] Extracting the sample transaction number score, sample transaction amount score, sample transaction number ratio, and sample transaction amount ratio corresponding to the sample transaction data based on the binning interval to which the sample transaction amount in the sample transaction data belongs;

[0025] Determining a transaction balance factor based on the number of sample transactions or the amount of sample transactions;

[0026] Within the time range of the three-dimensional coordinate system of transaction density, a time decay factor is calculated based on the number of sample transactions and the transaction time corresponding to the number of sample transactions.

[0027] Among them, t now is the current calculation time, t is the transaction time, and λ is the decay rate.

[0028] Optionally, a transaction density model is constructed based on the number of sample transactions, the sample transaction amount score, the sample transaction number ratio, the sample transaction amount ratio, the time decay factor, and the transaction balance factor, including:

[0029] Determine the weight coefficient of the sample transaction number ratio according to the quantile range of the bin interval to which the sample transaction number ratio belongs;

[0030] Determine the weight coefficient of the sample transaction amount ratio according to the quantile range of the bin interval to which the sample transaction amount ratio belongs;

[0031] Obtaining a transaction number balance factor and a transaction amount balance factor from the transaction balance factor, determining a sample transaction number balance factor weight based on the score of the transaction number balance factor, and determining a sample transaction amount balance factor weight based on the score of the transaction amount balance factor;

[0032] Constructing a transaction closeness model t_score using the sample transaction number score, sample transaction amount score, sample transaction number proportion weight, sample transaction amount proportion weight, transaction number balance factor, transaction number balance factor weight, transaction amount balance factor, and transaction amount balance factor weight;

[0033] Where t_score=∑(f_score*f p *d f *w f +m_score*m p *d m *w m );

[0034] f_score is the score of the number of sample transactions, f p is the weight of the number of sample transactions, d f is the transaction number balance factor, w f is the transaction number balance factor weight, m_score is the sample transaction amount score, m p is the weight of the sample transaction amount, dm is the transaction amount balance factor, w m is the transaction amount balance factor weight.

[0035] A second aspect of an embodiment of the present invention discloses a social network relationship evaluation device, comprising:

[0036] An acquisition module, configured to acquire transaction data between a first entity and a second entity in a social network relationship to be evaluated;

[0037] a processing module, configured to use the transaction data as input to a pre-established transaction closeness model, process the transaction data using the transaction closeness model, and output the transaction closeness between the first entity and the second entity;

[0038] An evaluation module is configured to determine an evaluation result of the social network relationship between the first entity and the second entity based on the transaction closeness.

[0039] Optionally, it further includes: a pre-built module, the pre-built module including:

[0040] The acquisition submodule is used to obtain sample transaction data details;

[0041] a determination submodule, configured to determine, based on the sample transaction data details, the relationship types between the sample transaction data and the sample entities between any two sample entities;

[0042] A first construction submodule is configured to construct a three-dimensional coordinate system for transaction closeness based on the relationship type of the sample entities, the transaction time range, and the sample transaction data between each pair of the sample entities, wherein the sample transaction data includes at least the sample transaction amount, the number of sample transactions, and the transaction time;

[0043] an extraction submodule for extracting sample data features based on the time range and distribution of sample transaction data in the three-dimensional transaction density coordinate system, wherein the sample data features include at least a sample transaction number score, a sample transaction amount score, a sample transaction number percentage, a sample transaction amount percentage, a time decay factor, and a transaction balance factor;

[0044] The second construction submodule is used to construct a transaction closeness model based on the sample transaction number score, sample transaction amount score, sample transaction number ratio, sample transaction amount ratio, time decay factor and transaction balance factor.

[0045] Optionally, the first building block includes:

[0046] A coordinate system construction unit is used to establish a three-dimensional coordinate system with the transaction time range as the X-axis, the transaction characteristics as the Y-axis, and the relationship type as the Z-axis;

[0047] a dividing unit, configured to divide the three-dimensional coordinate system into five binning intervals based on data quantiles, wherein the five binning intervals are normally distributed, and each binning interval is set with a corresponding score;

[0048] A first determining unit is configured to determine the type of sample entities and the relationship type between any two sample entities, wherein the types include individual types and enterprise types, and the relationship types include individual and enterprise types, individual and individual types, and enterprise and enterprise types;

[0049] A first construction unit is configured to construct individual and enterprise sub-models, individual and individual sub-models, and enterprise and enterprise sub-models based on the relationship types of the sample entities;

[0050] The second construction unit is used to construct a three-dimensional coordinate system of transaction closeness using the three-dimensional coordinate system, the individual and enterprise sub-models, the individual and individual sub-models, the enterprise and enterprise sub-models, and the sample transaction data between any two sample entities.

[0051] Optionally, the extraction submodule includes:

[0052] an extraction unit, configured to extract, based on the binning interval to which the sample transaction amount in the sample transaction data belongs, a sample transaction number score, a sample transaction amount score, a sample transaction number ratio, and a sample transaction amount ratio corresponding to the sample transaction data;

[0053] a second determining unit, configured to determine a transaction balance factor according to the number of sample transactions or the amount of sample transactions;

[0054] A calculation unit is configured to calculate a time decay factor based on the number of sample transactions and the transaction time corresponding to the number of sample transactions within a time range in the three-dimensional coordinate system of transaction density.

[0055] Among them, t now is the current calculation time, t is the transaction time, and λ is the decay rate.

[0056] Optionally, the second building block includes:

[0057] A third determining unit is configured to determine a weight coefficient for the sample transaction number proportion based on a quantile range of the binning interval to which the sample transaction number proportion belongs, and to determine a weight coefficient for the sample transaction amount proportion based on a quantile range of the binning interval to which the sample transaction amount proportion belongs;

[0058] a fourth determining unit, configured to obtain a transaction number balance factor and a transaction amount balance factor from the transaction balance factor, determine a weight of the sample transaction number balance factor based on the score of the transaction number balance factor, and determine a weight of the sample transaction amount balance factor based on the score of the transaction amount balance factor;

[0059] A third construction unit is configured to construct a transaction closeness model t_score using the sample transaction number score, the sample transaction amount score, the sample transaction number proportion weight, the sample transaction amount proportion weight, the transaction number balance factor, the transaction number balance factor weight, the transaction amount balance factor, and the transaction amount balance factor weight;

[0060] Where t_score=∑(f_score*f p *d f *w f +m_score*m p *d m *w m );

[0061] f_score is the score of the number of sample transactions, f p is the weight of the number of sample transactions, d f is the transaction number balance factor, W f is the transaction number balance factor weight, m_score is the sample transaction amount score, m p is the weight of the sample transaction amount, d m is the transaction amount balance factor, w m is the transaction amount balance factor weight.

[0062] Based on the above-mentioned embodiments of the present invention, a method and apparatus for evaluating social network relationships are provided. The method includes: obtaining transaction data between a first entity and a second entity whose social network relationship is to be evaluated; using the transaction data as input to a pre-established transaction closeness model, processing the transaction data using the transaction closeness model, and outputting the transaction closeness between the first and second entities; and determining an evaluation result of the social network relationship between the first and second entities based on the transaction closeness. In this solution, a transaction closeness model is pre-established, and the transaction data between the first and second entities whose social network relationship is to be evaluated is processed within the transaction closeness model to obtain the transaction closeness between the first and second entities. The evaluation result of the social network relationship between the first and second entities is determined based on the obtained transaction closeness, thereby achieving a complete and comprehensive evaluation of the social network relationship between the entities. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0064] Figure 1 A schematic diagram of a process for evaluating social network relationships provided by an embodiment of the present invention;

[0065] Figure 2 A schematic diagram of a process for pre-establishing a transaction closeness model provided by an embodiment of the present invention;

[0066] Figure 3 A schematic diagram of a process for constructing a three-dimensional coordinate system for transaction closeness provided by an embodiment of the present invention;

[0067] Figure 4 An application scenario diagram of a three-dimensional coordinate system provided by an embodiment of the present invention;

[0068] Figure 5 A schematic diagram of a process for extracting sample data features provided by an embodiment of the present invention;

[0069] Figure 6 A schematic diagram of a process for constructing a transaction closeness model provided by an embodiment of the present invention;

[0070] Figure 7 A schematic diagram of the structure of a social network relationship evaluation device provided by an embodiment of the present invention;

[0071] Figure 8 This is a schematic structural diagram of another social network relationship evaluation device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0072] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0073] In this application, the terms "comprises," "comprising," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not preclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0074] The terms "first," "second," "third," "fourth," and so on (if any) in the specification and claims of this application and the accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a particular order or sequence. It should be understood that the terms used in this manner are interchangeable where appropriate, so that the embodiments described herein can be implemented in an order other than that shown or described herein.

[0075] In order to facilitate understanding of the technical solution of the present invention, the technical terms appearing in the present invention are explained:

[0076] Entity: The two parties to a money transfer transaction, which may be a business or an individual.

[0077] Transaction data: refers to the details of funds deposited and withdrawn from bank transfer transactions conducted by enterprises / individuals.

[0078] Transaction relationship: refers to the relationship between entities established when an enterprise or individual conducts a transfer transaction in a bank.

[0079] Transaction closeness: refers to the degree of closeness of transactions between two entities, which is quantified based on transfer transaction data and is estimated from aspects such as transaction amount, frequency, and transaction time.

[0080] Transaction volume ratio: refers to the ratio of the number of transactions between two entities to the total number of transactions between one entity.

[0081] Transaction Amount Ratio: refers to the ratio of the transaction amount between two entities to the total number of transactions between one of the entities.

[0082] Transaction balance factor: refers to the adjustment factor that describes the balance of two-way transactions between the two parties (debit and credit).

[0083] Time decay factor: refers to the adjustment coefficient that adjusts the difference in the contribution of trading behaviors at different times to tightness.

[0084] As can be seen from the background technology, relying solely on the traditional social relationship intimacy to evaluate the social network relationships between entities is too one-sided, resulting in incomplete and incomplete evaluation of the social network relationships between entities.

[0085] Therefore, an embodiment of the present invention provides a method and apparatus for evaluating social network relationships. In this solution, a transaction closeness model is pre-established. Transaction data between a first entity and a second entity whose social network relationship is to be evaluated is processed in the transaction closeness model to obtain the transaction closeness between the first entity and the second entity. An evaluation result of the social network relationship between the first entity and the second entity is determined based on the obtained transaction closeness, thereby achieving a complete and comprehensive evaluation of the social network relationship between the entities.

[0086] like Figure 1 FIG. 1 is a flow chart of a method for evaluating social network relationships according to an embodiment of the present invention. The method mainly includes the following steps:

[0087] Step S101: Acquire transaction data between a first entity and a second entity in a social network relationship to be evaluated.

[0088] In step S101 , as can be seen from the above technical terms, the entity refers to the two parties of the transfer transaction, which may be a company or an individual.

[0089] In the specific implementation of step S101 , when entities conduct a transfer transaction, a social network relationship exists between them, and a large amount of transaction data is generated during the transaction. Transaction data between a first entity and a second entity whose social network relationship is to be evaluated is obtained.

[0090] Step S102: taking the transaction data as input to a pre-established transaction closeness model, processing the transaction data using the transaction closeness model, and outputting the transaction closeness between the first entity and the second entity.

[0091] In step S102, transaction closeness is a score of 0-1000. The higher the score, the higher the intimacy, that is, the higher the closeness. Therefore, the closeness of a person's relationship with his friends in the circle of friends can be quantified from the perspective of transactions.

[0092] In the specific implementation of step S102, a transaction closeness model is pre-established, the acquired transaction data is input into the pre-established transaction closeness model, the transaction closeness model is used to adjust the input transaction data for processing, the transaction closeness between the first entity and the second entity is obtained, and the transaction closeness between the first entity and the second entity is output.

[0093] Step S103: determining an evaluation result of the social network relationship between the first entity and the second entity based on the transaction closeness.

[0094] In the specific implementation of step S103 , the evaluation result of the social network relationship between the first entity and the second entity is determined according to the output transaction closeness between the first entity and the second entity.

[0095] For example, the transaction closeness between the first entity and the second entity is 980, and the evaluation result of the social network relationship between the first entity and the second entity is determined to be high intimacy based on the transaction closeness 980 between the first entity and the second entity.

[0096] For another example, the transaction closeness between the first entity and the second entity is 420, and the evaluation result of the social network relationship between the first entity and the second entity is determined to be low intimacy based on the transaction closeness 420 between the first entity and the second entity.

[0097] Optionally, in a specific embodiment, another social network relationship evaluation method provided by an embodiment of the present invention mainly includes the following steps:

[0098] Step S11: Acquire transaction data between the first entity and the second entity in the social network relationship to be evaluated.

[0099] Step S12: taking the transaction data as input to a pre-established transaction closeness model, processing the transaction data using the transaction closeness model, and outputting the transaction closeness between the first entity and the second entity.

[0100] It should be noted that the execution principle and process of the above steps S11 to S12 are the same as those of Figure 1 The execution principle and process of step S101 to step S102 disclosed in are the same, which can be referred to and will not be repeated here.

[0101] Step S13: Compare the transaction closeness with the preset closeness.

[0102] In the specific implementation of step S13 , the obtained transaction closeness between the first entity and the second entity is compared with a preset closeness.

[0103] Step S14: Determine whether the transaction closeness is greater than a preset closeness. If so, execute step S15; if not, execute step S16.

[0104] In the process of implementing step S14, it is determined whether the transaction closeness is greater than the preset closeness. If so, it means that the transaction closeness between the first entity and the second entity is greater than the preset closeness, and step S15 is executed. If not, it means that the transaction closeness between the first entity and the second entity is less than the preset closeness, and step S16 is executed.

[0105] Step S15: Determine that the evaluation result of the social network relationship between the first entity and the second entity is high intimacy.

[0106] In the specific implementation of step S15 , it is determined that the transaction closeness between the first entity and the second entity is greater than a preset closeness, and further determined that the evaluation result of the social network relationship between the first entity and the second entity is high intimacy.

[0107] Step S16: Determine that the evaluation result of the social network relationship between the first entity and the second entity is low intimacy.

[0108] In the specific implementation of step S16 , it is confirmed that the transaction closeness between the first entity and the second entity is less than a preset closeness, and then the evaluation result of the social network relationship between the first entity and the second entity is determined to be low intimacy.

[0109] A social network relationship evaluation method provided by an embodiment of the present invention obtains transaction data between a first entity and a second entity whose social network relationship is to be evaluated; uses the transaction data as input to a pre-established transaction closeness model, processes the transaction data using the transaction closeness model, and outputs the transaction closeness between the first and second entities; and determines an evaluation result of the social network relationship between the first and second entities based on the transaction closeness. In this solution, a transaction closeness model is pre-established, and the transaction data between the first and second entities whose social network relationship is to be evaluated is processed within the transaction closeness model to obtain the transaction closeness between the first and second entities. The evaluation result of the social network relationship between the first and second entities is determined based on the obtained transaction closeness, thereby achieving a complete and comprehensive evaluation of the social network relationship between the entities.

[0110] Based on the social network relationship evaluation method provided by the embodiment of the present invention, step S101 is executed to use the transaction data as the input of the pre-established transaction closeness model, process the transaction data using the transaction closeness model, and output the transaction closeness between the first entity and the second entity. Figure 2 FIG. 1 is a flow chart of pre-establishing a transaction closeness model according to an embodiment of the present invention, which mainly includes the following steps:

[0111] Step S201: Obtain sample transaction data details.

[0112] In step S201, the sample transaction data details include at least sample transaction data.

[0113] In the specific implementation of step S201, the sample transaction data details corresponding to each sample are obtained.

[0114] Step S202: Based on the sample transaction data details, determine the relationship types between the sample transaction data and the sample entities between each two sample entities.

[0115] In step S202 , the relationship types of the sample entities include individual and enterprise types, individual and individual types, and enterprise and enterprise types.

[0116] In the specific implementation of step S202 , based on the acquired sample transaction data details, the relationship types between the sample transaction data and the sample entities between any two sample entities in the sample transaction data details are determined.

[0117] Step S203: constructing a three-dimensional coordinate system of transaction closeness based on the relationship type of the sample entities, the transaction time range, and the sample transaction data between each two sample entities.

[0118] In step S203, the sample transaction data includes at least the sample transaction amount and the number of sample transactions.

[0119] The transaction time range is any time range from 0 to 365 days, which can be 0-3 months, 3-6 months, or 6-12 months, but is not limited thereto.

[0120] In the specific implementation of step S203 , a three-dimensional coordinate system of transaction closeness is constructed using the determined relationship types of sample entities, transaction time ranges, and sample transaction data between any two sample entities.

[0121] Step S204: extracting sample data features based on the time range and distribution of sample transaction data in the three-dimensional coordinate system of transaction closeness.

[0122] In step S204, the sample data features include at least the sample transaction number score, the sample transaction amount score, the sample transaction number ratio, the sample transaction amount ratio, the time decay factor and the transaction balance factor.

[0123] In the process of implementing step S204, sample data features are extracted based on the time range in the three-dimensional coordinate system of transaction density and the distribution of sample transaction data, specifically extracting the sample transaction number score, sample transaction amount score, sample transaction number ratio, sample transaction amount ratio, time decay factor and transaction balance factor.

[0124] Step S205: Construct a transaction closeness model based on the sample transaction number score, sample transaction amount score, sample transaction number ratio, sample transaction amount ratio, time decay factor and transaction balance factor.

[0125] In the specific implementation of step S205, a transaction closeness model is constructed using the sample transaction number score, sample transaction amount score, sample transaction number ratio, sample transaction amount ratio, time decay factor and transaction balance factor extracted above.

[0126] A social network relationship evaluation method provided by an embodiment of the present invention constructs a transaction closeness model by utilizing sample data features such as the sample transaction number score, sample transaction amount score, sample transaction number ratio, and sample transaction amount ratio corresponding to the extracted sample transaction data, thereby achieving the goal of complete and comprehensive evaluation of social network relationships between entities.

[0127] Optionally, step S203 is performed to construct a three-dimensional coordinate system of transaction closeness based on the relationship type of the sample entities, the transaction time range, and the sample transaction data between each two sample entities. Figure 3 FIG. 1 is a flow chart of constructing a three-dimensional coordinate system for transaction closeness according to an embodiment of the present invention, which mainly includes the following steps:

[0128] Step S301: Establish a three-dimensional coordinate system with the transaction time range as the X-axis, the transaction characteristics as the Y-axis, and the relationship type as the Z-axis.

[0129] In step S301 , the transaction characteristics refer to the transaction amount of the sample entity and the number of transactions of the sample entity.

[0130] In the specific implementation of step S301, the transaction time range of the sample entity is used as the X-axis of the three-dimensional coordinate system, the transaction characteristics of the sample entity is used as the Y-axis of the three-dimensional coordinate system, and the relationship type of the sample entity is used as the Z-axis of the three-dimensional coordinate system to construct a three-dimensional coordinate system.

[0131] For example, Figure 4 , which is an application scenario diagram of a three-dimensional coordinate system provided by an embodiment of the present invention.

[0132] exist Figure 4 In the figure, the time range represented by 0-3 months, 3-6 months, and 6-12 months is the transaction time range of the sample entity, which is the X-axis of the three-dimensional coordinate system. The transaction characteristics represented by the transaction amount and the number of transactions are the transaction characteristics of the sample entity, which is the Y-axis of the three-dimensional coordinate system. The relationship types represented by individuals and enterprises, enterprises and enterprises, and individuals and individuals are the relationship types of the sample entities, which are the Z-axis of the three-dimensional coordinate system. Therefore, based on the above content, a three-dimensional coordinate system can be constructed.

[0133] It should be noted that the use of a three-dimensional coordinate system can divide complex transaction scenarios into 18 intervals (the product of three time ranges, two transaction characteristics, and three relationship types), reducing the complexity of analyzing transaction scenarios.

[0134] It should be noted that based on the three-dimensional coordinate system constructed above, a logical framework can be determined that uses the number of transactions (including the absolute value and proportion of the number of transactions) and the transaction amount (including the absolute value and proportion of the transaction amount) as the main characteristic variables, and uses transaction time to complete the adjustment of the number of transactions and the transaction amount.

[0135] Step S302: Divide the three-dimensional coordinate system into five bin intervals based on the data quantiles.

[0136] In step S302 , the five binning intervals are normally distributed, and each binning interval is set with a corresponding score.

[0137] In the specific implementation of step S302 , the three-dimensional coordinate system is divided into five bin intervals according to the distribution of data quantiles.

[0138] The following example illustrates the specific binning process.

[0139] For example, based on the data distribution of the adjusted transaction amount, the adjusted transaction amount is binned according to the data quantile distribution, and a corresponding score is set. The score is based on a thousand-point system and used as the basic score for the transaction amount feature. The interval division strategy is shown in Table 1:

[0140] Table 1:

[0141] interval The range of data quantile division First interval [Minimum - 10th percentile) Second interval [10th percentile - 30th percentile) The third interval [30th percentile - 70th percentile) The fourth section [70th percentile - 90th percentile) Fifth section [90th percentile - maximum value)

[0142] Based on the interval division strategy shown in Table 1, the adjusted interval division and transaction amount score are shown in Table 2 (taking the individual and individual sub-models as an example):

[0143] Table 2:

[0144] Interval (left closed, right open) Transaction amount score Minimum-10th percentile 600 10th percentile-30th percentile 700 30th percentile-70th percentile 800 70th percentile-90th percentile 900 90th percentile - maximum value 1000

[0145] Based on the interval division strategy shown in Tables 1 and 2, and based on the data of the adjusted transaction amount ratio, the adjusted transaction amount ratio is divided into bins and the corresponding weight is set. The weight is used as the weight coefficient of the adjusted transaction amount ratio, as shown in Table 3 (taking the transaction amount ratio between individuals as an example):

[0146] Table 3:

[0147] Interval (left closed, right open) Transaction amount weight coefficient Minimum-10th percentile 0.6 10th percentile-30th percentile 0.7 30th percentile-70th percentile 0.8 70th percentile-90th percentile 0.9 90th percentile - maximum value 1

[0148] For another example: Based on the data distribution of the adjusted number of transactions, the adjusted number of transactions is binned according to the data quantile distribution, and a corresponding score is set. The score is based on a thousand-point system and used as the basic score for the transaction number feature. The interval division strategy is shown in Table 4:

[0149] Table 4:

[0150] interval The range of data quantile division First interval [Minimum - 10th percentile) Second interval [10th percentile - 30th percentile) The third interval [30th percentile - 70th percentile) The fourth section [70th percentile - 90th percentile) Fifth section [90th percentile - maximum value)

[0151] Based on the interval division strategy shown in Table 4, the adjusted interval division and transaction number scores are shown in Table 5 (taking the individual and individual sub-models as an example):

[0152] Table 5:

[0153] interval Transaction Count Score Minimum-10th percentile 600 10th percentile-30th percentile 700 30th percentile-70th percentile 800 70th percentile-90th percentile 900 90th percentile - maximum value 1000

[0154] Based on the interval division strategy shown in Tables 4 and 5, and based on the data of the adjusted transaction count ratio, the adjusted transaction count ratio is divided into bins and the corresponding weight is set. The weight is used as the weight coefficient of the adjusted transaction count ratio, as shown in Table 6 (taking the transaction count ratio between individuals as an example):

[0155] Table 6:

[0156]

[0157]

[0158] Step S303: Determine the type of the sample entity and the relationship type between any two sample entities.

[0159] In step S303 , the types include individual types and enterprise types, and the relationship types include individual and enterprise types, individual and individual types, and enterprise and enterprise types.

[0160] In the specific implementation of step S303 , it is determined whether the sample entity is of the individual type or the enterprise type, and whether the relationship between any two sample entities is of the individual and enterprise type, individual and individual type, or enterprise and enterprise type.

[0161] Step S304: constructing individual and enterprise sub-models, individual and individual sub-models, and enterprise and enterprise sub-models based on the relationship types of the sample entities.

[0162] In the specific implementation of step S304, based on Figure 4 The three-dimensional coordinate system shown can divide the relationship types of sample entities into different sub-models according to the different properties of the sample entities, that is, the individual and enterprise types, individual and individual types, and enterprise and enterprise types are divided into individual and enterprise sub-models, individual and individual sub-models, and enterprise and enterprise sub-models.

[0163] Step S305: construct a three-dimensional coordinate system for transaction closeness using the three-dimensional coordinate system, individual and enterprise sub-models, individual and individual sub-models, enterprise and enterprise sub-models, and sample transaction data between any two sample entities.

[0164] In the specific implementation of step S305, a three-dimensional coordinate system of transaction closeness is constructed based on the three-dimensional coordinate system constructed above, the individual and enterprise sub-models, the individual and individual sub-models, the enterprise and enterprise sub-models, and the sample transaction data between the pairwise sample entities determined above.

[0165] A social network relationship evaluation method provided by an embodiment of the present invention constructs a three-dimensional coordinate system, individual and enterprise sub-models, individual and individual sub-models, and enterprise and enterprise sub-models, and determines sample transaction data between two sample entities to form a three-dimensional coordinate system for transaction closeness. This provides a prerequisite for subsequent evaluation of the social network relationship between the first entity and the second entity, thereby achieving the goal of complete and comprehensive evaluation of social network relationships between entities.

[0166] Based on the social network relationship evaluation method provided by the embodiment of the present invention, step S204 is executed to extract the characteristics of the sample data based on the time range and the distribution of the sample transaction data in the three-dimensional coordinate system of transaction closeness. Figure 5 FIG. 1 is a flow chart of extracting sample data features according to an embodiment of the present invention, which mainly includes the following steps:

[0167] Step S501: extract the sample transaction number score, sample transaction amount score, sample transaction number ratio, and sample transaction amount ratio corresponding to the sample transaction data according to the binning interval to which the sample transaction amount in the sample transaction data belongs.

[0168] In the process of specifically implementing step S501, based on the above-mentioned binning process and binning principle, the binning interval to which the sample transaction amount in the sample transaction data belongs is determined, and according to the binning interval to which the sample transaction amount in the sample transaction data belongs, the sample transaction number score, sample transaction amount score, sample transaction number ratio and sample transaction amount ratio corresponding to the sample transaction data are extracted.

[0169] Step S502: Determine a transaction balance factor based on the number of sample transactions or the sample transaction amount.

[0170] It should be noted that, as can be seen from the above technical terms, the transaction balance factor is used to describe the balance of two-way transactions between the two parties (debit and credit).

[0171] In the specific implementation of step S502 , the transaction balance factor is determined based on the number of sample transactions extracted, or the transaction balance factor is determined based on the amount of sample transactions extracted.

[0172] Step S503: Calculate the time decay factor based on the number of sample transactions and the transaction time corresponding to the number of sample transactions within the time range of the three-dimensional coordinate system of transaction density.

[0173] In step S503, t now is the current calculation time, t is the transaction time, and λ is the decay rate.

[0174] It should be noted that as the time difference increases, the decay rate decreases and the decay pattern shows exponential decay.

[0175] Initially, the attenuation factor μ = 1; after 90 days, the attenuation factor At 180 days, the attenuation factor μ = 0.5; at 360 days, the attenuation factor μ = 0.25.

[0176] In the specific implementation of step S503, within the time range of the three-dimensional coordinate system of transaction density, the number of sample transactions and the transaction time corresponding to the number of sample transactions are obtained, and the time decay factor is calculated using the number of sample transactions and the transaction time corresponding to the number of sample transactions.

[0177] It should be noted that before counting the number of transactions or transaction amounts, the time decay factor needs to be used to decay the transaction amount and number of transactions. In addition, the adjusted transaction amount = transaction amount * time decay factor, and the adjusted number of transactions = 1 * time decay factor.

[0178] For example, the transaction that just took place yesterday is still 1 transaction after time decay calculation, but the transaction that took place 180 days ago is equivalent to 0.5 transactions after decay, that is: 1*decay factor = 1*0.5 = 0.5.

[0179] A social network relationship evaluation method provided by an embodiment of the present invention utilizes a constructed three-dimensional transaction closeness coordinate system to extract sample data features such as the number of sample transactions, sample transaction amount scores, sample transaction number ratios, and sample transaction amount ratios corresponding to the sample transaction data. This provides prerequisites for constructing a transaction closeness model, thereby achieving the goal of complete and comprehensive evaluation of social network relationships between entities.

[0180] Based on the social network relationship evaluation method provided by the embodiment of the present invention, step S205 is executed to construct a transaction closeness model based on the sample transaction number score, sample transaction amount score, sample transaction number ratio, sample transaction amount ratio, time decay factor and transaction balance factor. Figure 6 FIG. 1 is a flow chart of a process for constructing a transaction closeness model according to an embodiment of the present invention, which mainly includes the following steps:

[0181] Step S601: Determine the weight coefficient of the sample transaction number ratio according to the quantile range of the bin interval to which the sample transaction number ratio belongs.

[0182] In the specific implementation of step S601, based on the above-mentioned binning process and binning principle, the binning interval to which the sample transaction number ratio belongs is determined, the quantile range of the binning interval to which the sample transaction number ratio belongs is obtained according to the binning interval to which the sample transaction number ratio belongs, and the weight coefficient of the sample transaction number ratio is determined according to the quantile range of the binning interval to which the sample transaction number ratio belongs.

[0183] Step S602: Determine the weight coefficient of the sample transaction amount ratio according to the quantile range of the binning interval to which the sample transaction amount ratio belongs.

[0184] In the process of specifically implementing step S602, based on the above-mentioned binning process and binning principle, the binning interval to which the sample transaction amount ratio belongs is determined, the quantile range of the binning interval to which the sample transaction amount ratio belongs is obtained according to the binning interval to which the sample transaction amount ratio belongs, and the weight coefficient of the sample transaction amount ratio is determined according to the quantile range of the binning interval to which the sample transaction amount ratio belongs.

[0185] Step S603: Obtain the transaction number balance factor and transaction amount balance factor in the transaction balance factor, determine the sample transaction number balance factor weight according to the score of the transaction number balance factor, and determine the sample transaction amount balance factor weight according to the score of the transaction amount balance factor.

[0186] In the specific implementation of step S603, based on the determined transaction balance factor, the transaction number balance factor and the transaction amount balance factor in the transaction balance factor are obtained, and the sample transaction number balance factor weight is determined according to the score of the transaction number balance factor, and the sample transaction amount balance factor weight is determined according to the score of the transaction amount balance factor.

[0187] It should be noted that the transaction amount balance factor is used to adjust the calculation of the transaction amount, and the transaction number balance factor is used to adjust the calculation of the transaction number. Their basic formulas are the same, as shown in formula (1):

[0188] d m / f =(1-dP )*0.2+0.9 (1)

[0189] in,

[0190] d m / f , C s 、D s , SUM are the transaction balance factor, debit transaction amount (number of transactions), credit transaction amount (number of transactions), and total debit and credit transaction amount (number of transactions), respectively.

[0191] From the above formula, we can see that the range of the transaction amount balance factor is [0.5, 1]. When the transaction amount balance factor is 1, there is only a one-way transaction between the debit and the credit. For example, if there is only a debit amount, the debit amount is equal to the total amount. At this time, the score of the transaction amount balance factor is

[0192] Similarly, if there is only a credit amount, the credit amount is equal to the total amount. In this case, the score of the transaction amount balance factor is also

[0193] When the transaction balance factor is 0.5, there is a two-way transaction between the debit and credit sides and the transaction amounts are equal. That is, the debit amount is equal to the credit amount, and the debit amount and the credit amount are both half of the middle amount. At this time, the transaction balance factor score is

[0194] When there is a two-way transaction between the debit and the credit and the transaction amounts are unequal, the score of the transaction amount balance factor is in the range of (0.5, 1). The closer the score of the transaction amount balance factor is to 1, the more unbalanced the transaction amounts between the debit and the credit. The closer the score of the transaction amount balance factor is to 0.5, the more balanced the transaction amounts between the debit and the credit.

[0195] In the embodiment of the present invention, the distribution of the transaction balance factor weight coefficient is controlled to be within the range of [0.9, 1]. When the transaction is in the most balanced situation (the amount (number of transactions) balance factor is 0.5), the transaction balance factor weight coefficient is 1 and does not decay; when the transaction is in the most unbalanced situation (the amount balance factor is 1), the transaction balance factor weight coefficient is 0.9 and decays by 0.9.

[0196] Step S604: Construct a transaction closeness model t_score using the sample transaction number score, sample transaction amount score, sample transaction number proportion weight, sample transaction amount proportion weight, transaction number balance factor, transaction number balance factor weight, transaction amount balance factor, and transaction amount balance factor weight.

[0197] Where t_score=∑(f_score*fp *d f *w f +m_score*m p *d m *w m );

[0198] f_score is the score of sample transactions, f p is the weight of the number of sample transactions, d f is the transaction number balance factor, w f is the transaction number balance factor weight, m_score is the sample transaction amount score, m p is the weight of the sample transaction amount, d m is the transaction amount balance factor, w m is the transaction amount balance factor weight.

[0199] In the specific implementation of step S604, the transaction closeness model t_score is constructed using the sample transaction number score, sample transaction amount score, sample transaction number proportion weight, sample transaction amount proportion weight, transaction number balance factor, transaction number balance factor weight, transaction amount balance factor and transaction amount balance factor weight obtained above.

[0200] A social network relationship evaluation method provided by an embodiment of the present invention constructs a transaction closeness model by utilizing sample data features such as the sample transaction number score, sample transaction amount score, sample transaction number ratio, and sample transaction amount ratio corresponding to the obtained sample transaction data, thereby achieving the goal of complete and comprehensive evaluation of social network relationships between entities.

[0201] In order to better understand the social network relationship evaluation method provided by the above embodiment of the present invention, an example is given below for detailed description.

[0202] Taking the communication relationship as an example, the closeness of the relationship between two entities is estimated through factors such as call duration, number of calls, call duration ratio, call number ratio, call balance factor, and time decay factor.

[0203] Among them, call duration is analogous to transaction amount, number of calls is analogous to number of transactions, call duration ratio is analogous to transaction amount ratio, number of calls ratio is analogous to transaction ratio, call balance factor is analogous to transaction balance factor. The influence of time on relationship in call scenario is similar to that of transaction, and time decay factor is used.

[0204] The details are as follows:

[0205] First, the time decay factor is determined.

[0206] Specifically, the time decay factor is calculated using the call time corresponding to the call duration and the number of calls.

[0207] Among them, t now is the current calculation time; t is the call time, t now -t is the time difference; λ is the decay rate. Every λ days, the transaction decays to the equivalent of 0.5 transactions. λ is tentatively set to 180 days.

[0208] Before processing the call duration and the number of calls, a time decay factor is used for decay processing. Moreover, the adjusted call duration = call duration * time decay factor, and the adjusted number of calls = 1 * time decay factor.

[0209] Secondly, before aggregating the call duration and number of calls between two entities, a time decay factor is used to adjust each call.

[0210] Next, calculate the call balance factor d m / f =(1-d P )*0.2+0.9.

[0211] in,

[0212] d m / f , C s 、D s , SUM are respectively the call balance factor, outgoing call duration (number of times), incoming call duration (number of times), and the total call duration (number of times) of incoming and outgoing calls.

[0213] The call duration and the number of calls are measured respectively to obtain the call duration balance factor and the call number balance factor.

[0214] Next, the binning process is completed to obtain the basic score of the adjusted call duration (number of times) and the weight value of the adjusted call duration (number of times) ratio.

[0215] Finally, based on the call density measurement model, the call density measurement is completed.

[0216] Specifically: the call duration (number of times) score multiplied by the call duration (number of times) proportion weight coefficient, the call duration (number of times) balance factor weight coefficient and the call duration (number of times) weight.

[0217] The call density calculation model is as follows:

[0218] t_score=∑(f_score*f p *d f *w f +m_score*m p *dm *w m )

[0219] Among them: f_score, f p d f 、w f They are respectively the call count score, call count proportion weight, call count balance factor, and call count weight; m_score, m p d m 、w m They are call duration score, call duration weight, call duration balance factor, and call duration weight.

[0220] According to a social network relationship evaluation method provided by an embodiment of the present invention, a transaction closeness model is pre-established. Transaction data between a first entity and a second entity whose social network relationship is to be evaluated is processed in the transaction closeness model to obtain the transaction closeness between the first entity and the second entity. An evaluation result of the social network relationship between the first entity and the second entity is determined based on the obtained transaction closeness, thereby achieving a complete and comprehensive evaluation of the social network relationship between the entities.

[0221] Corresponding to the social network relationship evaluation method shown in the above embodiment of the present invention, the embodiment of the present invention also provides a social network relationship evaluation device, such as Figure 7 As shown, the social network relationship evaluation device includes: an acquisition module 71 , a processing module 72 and an evaluation module 73 .

[0222] The acquisition module 71 is configured to acquire transaction data between a first entity and a second entity in a social network relationship to be evaluated.

[0223] The processing module 72 is configured to use the transaction data as input to a pre-established transaction closeness model, process the transaction data using the transaction closeness model, and output the transaction closeness between the first entity and the second entity.

[0224] The evaluation module 73 is configured to determine an evaluation result of the social network relationship between the first entity and the second entity based on the transaction closeness.

[0225] It should be noted that the specific principles and execution processes of each module or unit in the social network relationship evaluation device disclosed in the above embodiment of the present invention are the same as those of the social network relationship evaluation method implemented in the above embodiment of the present invention. Please refer to the corresponding parts of the social network relationship evaluation method disclosed in the above embodiment of the present invention, and no further details will be given here.

[0226] According to an embodiment of the present invention, a social network relationship evaluation device is provided. This device obtains transaction data between a first entity and a second entity whose social network relationship is to be evaluated; uses the transaction data as input to a pre-established transaction closeness model, processes the transaction data using the transaction closeness model, and outputs the transaction closeness between the first and second entities; and determines an evaluation result of the social network relationship between the first and second entities based on the transaction closeness. In this solution, a pre-established transaction closeness model is used to process the transaction data between the first and second entities whose social network relationship is to be evaluated, thereby obtaining the transaction closeness between the first and second entities. The evaluation result of the social network relationship between the first and second entities is then determined based on the obtained transaction closeness, thereby achieving a complete and comprehensive evaluation of the social network relationship between the entities.

[0227] Optional, based on the above Figure 7 The social network relationship evaluation device shown is combined with Figure 7 ,like Figure 8 As shown, the social network relationship evaluation device is further provided with a pre-construction module 74, which includes an acquisition submodule, a determination submodule, a first construction submodule, an extraction submodule and a second construction submodule.

[0228] The acquisition submodule is used to obtain sample transaction data details.

[0229] The determination submodule is used to determine the relationship type between the sample transaction data and the sample entities between each two sample entities based on the sample transaction data details.

[0230] The first construction submodule is used to construct a three-dimensional coordinate system of transaction closeness based on the relationship type of sample entities, the transaction time range and the sample transaction data between each two sample entities. The sample transaction data at least includes the sample transaction amount, the number of sample transactions and the transaction time.

[0231] The extraction submodule is used to extract sample data features based on the time range and distribution of sample transaction data in the three-dimensional coordinate system of transaction density. The sample data features include at least the sample transaction number score, the sample transaction amount score, the sample transaction number ratio, the sample transaction amount ratio, the time decay factor and the transaction balance factor.

[0232] The second construction submodule is used to construct a transaction closeness model based on the sample transaction number score, the sample transaction amount score, the sample transaction number ratio, the sample transaction amount ratio, the time decay factor and the transaction balance factor.

[0233] A social network relationship evaluation device provided by an embodiment of the present invention constructs a transaction closeness model by utilizing sample data features such as the sample transaction number score, sample transaction amount score, sample transaction number ratio, and sample transaction amount ratio corresponding to the extracted sample transaction data, thereby achieving the purpose of complete and comprehensive evaluation of social network relationships between entities.

[0234] Optional, based on the above Figure 8 The first construction submodule shown further includes: a coordinate system construction unit, a division unit, a first determination unit, a first construction unit, and a second construction unit.

[0235] The coordinate system construction unit is used to establish a three-dimensional coordinate system with the transaction time range as the X-axis, the transaction characteristics as the Y-axis, and the relationship type as the Z-axis.

[0236] The division unit is used to divide the three-dimensional coordinate system into five binning intervals based on the data quantile. The five binning intervals are normally distributed, and each binning interval is set with a corresponding score.

[0237] The first determination unit is used to determine the type of the sample entity and the relationship type between any two sample entities. The types include individual type and enterprise type. The relationship types include individual and enterprise type, individual and individual type, and enterprise and enterprise type.

[0238] The first construction unit is used to construct individual and enterprise sub-models, individual and individual sub-models, and enterprise and enterprise sub-models based on the relationship types of the sample entities.

[0239] The second construction unit is used to construct a three-dimensional coordinate system for transaction closeness using a three-dimensional coordinate system, individual and enterprise sub-models, individual and individual sub-models, enterprise and enterprise sub-models, and sample transaction data between any two sample entities.

[0240] A social network relationship evaluation device provided by an embodiment of the present invention constructs a three-dimensional coordinate system, individual and enterprise sub-models, individual and individual sub-models, and enterprise and enterprise sub-models, and determines sample transaction data between two sample entities to form a three-dimensional coordinate system for transaction closeness. This provides a prerequisite for subsequent evaluation of the social network relationship between a first entity and a second entity, thereby achieving the goal of complete and comprehensive evaluation of social network relationships between entities.

[0241] Optional, based on the above Figure 8 The extraction submodule shown further includes: an extraction unit, a second confirmation unit and a calculation unit.

[0242] The extraction unit is used to extract the sample transaction number score, sample transaction amount score, sample transaction number ratio and sample transaction amount ratio corresponding to the sample transaction data according to the binning interval to which the sample transaction amount in the sample transaction data belongs.

[0243] The second determining unit is used to determine the transaction balance factor according to the number of sample transactions or the sample transaction amount.

[0244] A calculation unit for calculating a time decay factor based on the number of sample transactions and the transaction time corresponding to the number of sample transactions within the time range of the three-dimensional coordinate system of transaction density.

[0245] Among them, t now is the current calculation time, t is the transaction time, and λ is the decay rate.

[0246] A social network relationship evaluation device provided by an embodiment of the present invention utilizes a constructed three-dimensional transaction closeness coordinate system to extract sample data features such as the number of sample transactions, sample transaction amount scores, sample transaction number ratios, and sample transaction amount ratios corresponding to the sample transaction data, thereby providing prerequisites for constructing a transaction closeness model and achieving the goal of complete and comprehensive evaluation of social network relationships between entities.

[0247] Optional, based on the above Figure 8 The second construction submodule shown further includes: a third determining unit, a fourth determining unit and a third construction unit.

[0248] The third determination unit is used to determine the weight coefficient of the sample transaction number ratio according to the quantile range of the binning interval to which the sample transaction number ratio belongs, and to determine the weight coefficient of the sample transaction amount ratio according to the quantile range of the binning interval to which the sample transaction amount ratio belongs.

[0249] The fourth determination unit is used to obtain the transaction number balance factor and the transaction amount balance factor in the transaction balance factor, determine the sample transaction number balance factor weight according to the score of the transaction number balance factor, and determine the sample transaction amount balance factor weight according to the score of the transaction amount balance factor.

[0250] The third construction unit is used to construct a transaction closeness model t_score using the sample transaction number score, the sample transaction amount score, the sample transaction number proportion weight, the sample transaction amount proportion weight, the transaction number balance factor, the transaction number balance factor weight, the transaction amount balance factor, and the transaction amount balance factor weight.

[0251] Where t_score=∑(f_score*f p *d f*w f +m_score*m p *d m *w m );

[0252] f_score is the score of sample transactions, f p is the weight of the number of sample transactions, d f is the transaction number balance factor, w f is the transaction number balance factor weight, m_score is the sample transaction amount score, m p is the weight of the sample transaction amount, d m is the transaction amount balance factor, w m is the transaction amount balance factor weight.

[0253] A social network relationship evaluation device provided by an embodiment of the present invention constructs a transaction closeness model by utilizing sample data features such as the sample transaction number score, sample transaction amount score, sample transaction number ratio, and sample transaction amount ratio corresponding to the obtained sample transaction data, thereby achieving the purpose of complete and comprehensive evaluation of social network relationships between entities.

[0254] Each embodiment in this specification is described in a progressive manner. The same or similar parts between the embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments. In particular, for system or system embodiments, since they are basically similar to method embodiments, the description is relatively simple. For relevant parts, refer to the partial description of the method embodiment. The system and system embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. A person of ordinary skill in the art can understand and implement it without expending creative work.

[0255] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the components and steps of each example according to their functions. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0256] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A social network relationship evaluation method, characterized in that: The method comprises: Acquire transaction data between a first entity and a second entity in a social network relationship to be evaluated; Using the transaction data as input to a pre-established transaction closeness model, processing the transaction data using the transaction closeness model, and outputting the transaction closeness between the first entity and the second entity; determining an evaluation result of the social network relationship between the first entity and the second entity based on the transaction closeness; The process of pre-establishing the transaction closeness model includes: Get sample transaction data details; Determining, based on the sample transaction data details, the relationship types between the sample transaction data and the sample entities between each two sample entities; Constructing a three-dimensional coordinate system of transaction closeness based on the relationship type of the sample entities, the transaction time range, and the sample transaction data between each pair of sample entities, wherein the sample transaction data at least includes the sample transaction amount, the number of sample transactions, and the transaction time; Extracting sample data features based on the time range and distribution of sample transaction data in the three-dimensional transaction density coordinate system, wherein the sample data features include at least a sample transaction number score, a sample transaction amount score, a sample transaction number percentage, a sample transaction amount percentage, a time decay factor, and a transaction balance factor; A transaction closeness model is constructed based on the sample transaction number score, sample transaction amount score, sample transaction number ratio, sample transaction amount ratio, time decay factor and transaction balance factor.

2. The method according to claim 1, characterized in that A three-dimensional coordinate system of transaction closeness is constructed based on the relationship type of the sample entities, the transaction time range, and the sample transaction data between each two sample entities, including: A three-dimensional coordinate system is established with the transaction time range as the X-axis, the transaction characteristics as the Y-axis, and the relationship type as the Z-axis; Dividing the three-dimensional coordinate system into five binning intervals based on data quantiles, wherein the five binning intervals are normally distributed, and each binning interval is set with a corresponding score; Determine the types of sample entities and the relationship types between any two sample entities, wherein the types include individual types and enterprise types, and the relationship types include individual and enterprise types, individual and individual types, and enterprise and enterprise types; Building individual and enterprise sub-models, individual and individual sub-models, and enterprise and enterprise sub-models based on the relationship types of the sample entities; A three-dimensional coordinate system for transaction closeness is constructed using the three-dimensional coordinate system, the individual and enterprise sub-models, the individual and individual sub-models, the enterprise and enterprise sub-models, and the sample transaction data between any two sample entities.

3. The method according to claim 2, characterized in that The extracting of sample data features based on the time range and distribution of sample transaction data in the three-dimensional coordinate system of transaction density includes: Extracting the sample transaction number score, sample transaction amount score, sample transaction number ratio, and sample transaction amount ratio corresponding to the sample transaction data based on the binning interval to which the sample transaction amount in the sample transaction data belongs; Determining a transaction balance factor based on the number of sample transactions or the amount of sample transactions; Within the time range of the three-dimensional coordinate system of transaction density, a time decay factor is calculated based on the number of sample transactions and the transaction time corresponding to the number of sample transactions. Among them, t now is the current calculation time, t is the transaction time, and λ is the decay rate.

4. The method according to claim 3, characterized in that A transaction density model is constructed based on the number of sample transactions, the sample transaction amount score, the sample transaction number ratio, the sample transaction amount ratio, the time decay factor, and the transaction balance factor, including: Determine the weight coefficient of the sample transaction number ratio according to the quantile range of the bin interval to which the sample transaction number ratio belongs; Determine the weight coefficient of the sample transaction amount ratio according to the quantile range of the bin interval to which the sample transaction amount ratio belongs; Obtaining a transaction number balance factor and a transaction amount balance factor from the transaction balance factor, determining a sample transaction number balance factor weight based on the score of the transaction number balance factor, and determining a sample transaction amount balance factor weight based on the score of the transaction amount balance factor; Constructing a transaction closeness model t_score using the sample transaction number score, sample transaction amount score, sample transaction number proportion weight, sample transaction amount proportion weight, transaction number balance factor, transaction number balance factor weight, transaction amount balance factor, and transaction amount balance factor weight; where t_score = ∑(f_score * f p * d f * w f + m_score * m p * d m * w m ); f_score is the score of the number of sample transactions, f p is the weight of the number of sample transactions, d f is the transaction number balance factor, w f is the transaction number balance factor weight, m_score is the sample transaction amount score, m p is the weight of the sample transaction amount, d m is the transaction amount balance factor, w m is the transaction amount balance factor weight.

5. A social network relationship evaluation device, characterized in that: The device comprises: An acquisition module, configured to acquire transaction data between a first entity and a second entity in a social network relationship to be evaluated; a processing module, configured to use the transaction data as input to a pre-established transaction closeness model, process the transaction data using the transaction closeness model, and output the transaction closeness between the first entity and the second entity; an evaluation module, configured to determine an evaluation result of the social network relationship between the first entity and the second entity based on the transaction closeness; Pre-built modules include: The acquisition submodule is used to obtain sample transaction data details; a determination submodule, configured to determine, based on the sample transaction data details, the relationship types between the sample transaction data and the sample entities between any two sample entities; A first construction submodule is configured to construct a three-dimensional coordinate system for transaction closeness based on the relationship type of the sample entities, the transaction time range, and the sample transaction data between each pair of the sample entities, wherein the sample transaction data includes at least the sample transaction amount, the number of sample transactions, and the transaction time; an extraction submodule for extracting sample data features based on the time range and distribution of sample transaction data in the three-dimensional transaction density coordinate system, wherein the sample data features include at least a sample transaction number score, a sample transaction amount score, a sample transaction number percentage, a sample transaction amount percentage, a time decay factor, and a transaction balance factor; The second construction submodule is used to construct a transaction closeness model based on the sample transaction number score, sample transaction amount score, sample transaction number ratio, sample transaction amount ratio, time decay factor and transaction balance factor.

6. The device according to claim 5, characterized in that The first building block includes: A coordinate system construction unit is used to establish a three-dimensional coordinate system with the transaction time range as the X-axis, the transaction characteristics as the Y-axis, and the relationship type as the Z-axis; a dividing unit, configured to divide the three-dimensional coordinate system into five binning intervals based on data quantiles, wherein the five binning intervals are normally distributed, and each binning interval is set with a corresponding score; A first determining unit is configured to determine the type of sample entities and the relationship type between any two sample entities, wherein the types include individual types and enterprise types, and the relationship types include individual and enterprise types, individual and individual types, and enterprise and enterprise types; A first construction unit is configured to construct individual and enterprise sub-models, individual and individual sub-models, and enterprise and enterprise sub-models based on the relationship types of the sample entities; The second construction unit is used to construct a three-dimensional coordinate system of transaction closeness using the three-dimensional coordinate system, the individual and enterprise sub-models, the individual and individual sub-models, the enterprise and enterprise sub-models, and the sample transaction data between any two sample entities.

7. The device according to claim 6, characterized in that The extraction submodule includes: an extraction unit, configured to extract, based on the binning interval to which the sample transaction amount in the sample transaction data belongs, a sample transaction number score, a sample transaction amount score, a sample transaction number ratio, and a sample transaction amount ratio corresponding to the sample transaction data; a second determining unit, configured to determine a transaction balance factor according to the number of sample transactions or the amount of sample transactions; A calculation unit is configured to calculate a time decay factor based on the number of sample transactions and the transaction time corresponding to the number of sample transactions within a time range in the three-dimensional coordinate system of transaction density. Among them, t now is the current calculation time, t is the transaction time, and λ is the decay rate.

8. The device according to claim 7, characterized in that The second building block includes: A third determining unit is configured to determine a weight coefficient for the sample transaction number proportion based on a quantile range of the binning interval to which the sample transaction number proportion belongs, and to determine a weight coefficient for the sample transaction amount proportion based on a quantile range of the binning interval to which the sample transaction amount proportion belongs; a fourth determining unit, configured to obtain a transaction number balance factor and a transaction amount balance factor from the transaction balance factor, determine a weight of the sample transaction number balance factor based on the score of the transaction number balance factor, and determine a weight of the sample transaction amount balance factor based on the score of the transaction amount balance factor; A third construction unit is configured to construct a transaction closeness model t_score using the sample transaction number score, the sample transaction amount score, the sample transaction number proportion weight, the sample transaction amount proportion weight, the transaction number balance factor, the transaction number balance factor weight, the transaction amount balance factor, and the transaction amount balance factor weight; where, t_score = Σ(f_score * f p * d f * w f + m_score * m p * d m * w m ); f_score is the score of the number of sample transactions, f p is the weight of the number of sample transactions, d f is the transaction number balance factor, w f is the transaction number balance factor weight, m_score is the sample transaction amount score, m p is the weight of the sample transaction amount, d m is the transaction amount balance factor, w m is the transaction amount balance factor weight.

Images