CN115617868A - Mining method and device for customer behaviors, terminal equipment and computer storage medium - Google Patents
Mining method and device for customer behaviors, terminal equipment and computer storage medium Download PDFInfo
- Publication number
- CN115617868A CN115617868A CN202211210608.8A CN202211210608A CN115617868A CN 115617868 A CN115617868 A CN 115617868A CN 202211210608 A CN202211210608 A CN 202211210608A CN 115617868 A CN115617868 A CN 115617868A
- Authority
- CN
- China
- Prior art keywords
- mining
- loan
- customer
- client
- target
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2465—Query processing support for facilitating data mining operations in structured databases
Abstract
The invention relates to the technical field of data processing, and discloses a method and a device for mining customer behaviors, terminal equipment and a storage medium. The method comprises the steps of firstly, obtaining respective historical information of loan products of clients to be excavated; then, aggregating the historical information of the loan products and preset crowd characteristic data to divide each customer to be mined into a plurality of customer groups; and finally, carrying out borrowing and lending action mining on the plurality of passenger groups. By adopting the technical scheme of the invention, the client lending behaviors can be mined from the time dimension and the product dimension, so that the comprehensive understanding of the lending behaviors of different customer groups, the insight of the client requirements and the identification of the client lending behavior modes are facilitated, and the development of subsequent activities such as accurate marketing, risk identification and the like is facilitated.
Description
Technical Field
The present invention relates to the field of data processing technologies, and in particular, to a method and an apparatus for mining a client behavior, a terminal device, and a computer storage medium.
Background
At present, the analysis and mining of the loan behaviors of the clients by the financial institutions mainly adopt a questionnaire survey and expert analysis method and the analysis and prediction of the loan behaviors of the clients are carried out on the basis of a neural network model. However, the traditional questionnaire + expert analysis method depends heavily on questionnaire scheme design and expert experience, and the time consumption of behavior analysis is long, so that the traditional questionnaire + expert analysis method is difficult to adapt to the changing market change environment quickly. In addition, the borrowing behavior mining based on the neural network model can quickly adapt to the complex and variable customer requirements, but the model calculation process is not beneficial to understanding the internal rules of the customer borrowing behavior, so that the model prediction result has inexplicability.
In summary, the existing method for analyzing and mining the loan behaviors of the customers is difficult to meet the application requirements of the financial institutions for finding the loan behaviors of the customers for accurate marketing and risk identification.
Disclosure of Invention
The invention mainly aims to provide a mining method, a mining device, terminal equipment and a computer storage medium for customer behaviors, and aims to accurately analyze and mine loan behaviors of customers.
In order to achieve the above object, the present invention provides a method for mining customer behavior, including:
acquiring respective loan product historical information of each client to be excavated;
aggregating the historical information of the loan products with preset crowd characteristic data to divide each customer to be mined into a plurality of customer groups;
and mining the loan behaviors of a plurality of the passenger groups.
In some embodiments, said mining of lending activity for a plurality of said customer groups comprises:
in a target customer group of the plurality of customer groups, determining a current loan product transaction of a target mining customer to generate a current transaction data set;
learning the current transaction data set according to a preset association rule to generate a frequent loan product set;
mining mutual dependencies among the loan products of the target mining client based on the frequent loan product set.
In some embodiments, said mining mutual dependencies between loan products of said target mining customer based on said set of frequent loan products comprises:
generating association rules based on the frequent lending product set;
and determining the mutual dependency among the loan products of the target mining client according to the target rule which accords with the preset confidence condition in all the generated association rules.
In some embodiments, said mining of lending activity for a plurality of said customer groups comprises:
determining loan sequence data of a target mining customer in a target customer group of a plurality of customer groups;
determining a latest frequent sequence mode according to the loan sequence data;
and mining the lending behavior precedence relationship of the target mining client based on the latest frequent sequence mode.
In some embodiments, the step of determining loan sequence data for the target mining customer may include:
determining historical base data of the target mining client;
and generating loan sequence data of the target mining client according to the historical basic data.
In some embodiments, the step of aggregating the loan product history information with preset demographic data to divide each of the customers to be mined into a plurality of customer groups includes:
generating a loan history matrix according to the history information of the loan products;
and performing Gaussian clustering on the historical loan behavior vectors of the clients obtained by vectorizing the loan history matrix and preset population characteristic data to divide the clients to be mined into a plurality of client groups.
In addition, in order to achieve the above object, the present invention further provides an excavating device for customer behavior, comprising:
the information acquisition module is used for acquiring the historical information of the loan products of the customers to be excavated;
the client grouping module is used for aggregating the historical information of the loan products and preset crowd characteristic data so as to divide each client to be mined into a plurality of client groups;
and the behavior mining module is used for mining the loan behaviors of the passenger groups.
In some embodiments, the behavior mining module comprises:
the product dimension mining unit is used for determining the current loan product transaction of a target mining client in a target client group in a plurality of client groups to generate a current transaction data set; learning the current transaction data set according to a preset association rule to generate a frequent loan product set; mining the interdependency among the loan products of the target mining client based on the frequent loan product set;
the product dimension mining unit comprises:
a rule generating subunit, configured to generate an association rule based on the frequent lending product set;
the mining subunit is used for determining the mutual dependency among the loan products of the target mining client according to a target rule which accords with a preset confidence condition in all the generated association rules;
the behavior mining module further comprises:
the time dimension mining unit is used for determining loan and loan sequence data of a target mining client in a target client group in a plurality of client groups; determining a latest frequent sequence mode according to the loan sequence data; mining the sequence relation of the loan behaviors of the target mining client based on the latest frequent sequence mode;
the time dimension mining unit is further used for determining historical basic data of the target mining client; generating loan sequence data of the target mining client according to the historical basic data;
the customer grouping module comprises:
the matrix generating unit is used for generating a loan history matrix according to the loan product history information;
and the clustering processing unit is used for carrying out Gaussian clustering processing on the historical loan behavior vectors of the clients obtained by vectorizing the loan history matrix and preset population characteristic data so as to divide each client to be mined into a plurality of client groups.
The functional modules of the mining device for the customer behaviors realize the steps of the mining method for the customer behaviors when in operation.
In addition, to achieve the above object, the present invention also provides a terminal device, including: a memory, a processor and a customer behavior mining program stored on the memory and executable on the processor, the customer behavior mining program when executed by the processor implementing the steps of the customer behavior mining method as described above.
In addition, to achieve the above object, the present invention further provides a computer storage medium having a mining program of customer behavior stored thereon, wherein the mining program of customer behavior realizes the steps of the mining method of customer behavior as described above when being executed by a processor.
Furthermore, to achieve the above object, the present invention also provides a computer program product comprising a computer program which, when being executed by a processor, realizes the steps of the mining method of customer behavior as described above.
The invention provides a mining method, a mining device, terminal equipment, a computer storage medium and a computer program product for customer behaviors, which are characterized in that firstly, historical information of loan products of each customer to be mined is acquired; then, aggregating the historical information of the loan products with preset crowd characteristic data to divide each customer to be mined into a plurality of customer groups; and finally, carrying out borrowing and lending action mining on the plurality of passenger groups.
In the method, in the process of mining the borrowing and lending behaviors of the clients, historical information of the borrowing and lending products of the clients is obtained, then the historical information of the borrowing and lending products and the crowd characteristic data are aggregated to perform grouping processing on the clients to obtain a plurality of client groups, and finally the borrowing and lending behaviors are mined on the client groups. Therefore, compared with the traditional method of mining the client behaviors by adopting questionnaire and expert analysis or neural network model prediction, the method and the system have the advantages that different client groups are obtained by aggregating the historical borrowing product behaviors of the clients and the crowd characteristic data, and then the borrowing behaviors of different client groups are mined, so that the borrowing behaviors and the client requirements of different client groups can be comprehensively known, the purpose of accurately analyzing and mining the client borrowing behaviors is achieved, and the development of subsequent activities such as accurate marketing, risk identification and the like by combining analysis and mining of financial institutions is facilitated.
Drawings
Fig. 1 is a schematic diagram of an apparatus structure of a hardware operating environment of a terminal apparatus according to an embodiment of the present invention;
FIG. 2 is a flowchart illustrating a first embodiment of a mining method for customer behavior according to the present invention;
FIG. 3 is a schematic view of an application flow according to an embodiment of the mining method for customer behavior of the present invention;
FIG. 4 is a schematic view illustrating another application flow of an embodiment of the mining method for customer behavior according to the present invention;
fig. 5 is a functional block diagram of an embodiment of the mining apparatus for customer behavior according to the present invention.
The implementation, functional features and advantages of the objects of the present invention will be further explained with reference to the accompanying drawings.
Detailed Description
It should be understood that the specific embodiments described herein are merely illustrative of the invention and do not limit the invention.
Referring to fig. 1, fig. 1 is a schematic device structure diagram of a terminal device hardware operating environment according to an embodiment of the present invention.
The terminal device in the embodiment of the present invention may be a smart phone, a PC (Personal Computer), a tablet Computer, a portable Computer, a server, and the like.
As shown in fig. 1, the terminal device may include: a processor 1001, e.g. a CPU, a communication bus 1002, a user interface 1003, a network interface 1004, a memory 1005. The communication bus 1002 is used to implement connection communication among these components. The user interface 1003 may include a Display (Display), an input unit such as a Keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface, a wireless interface. The network interface 1004 may optionally include a standard wired interface, a wireless interface (e.g., a Wi-Fi interface). The memory 1005 may be a high-speed RAM memory or a non-volatile memory (e.g., a magnetic disk memory). The memory 1005 may alternatively be a storage device separate from the processor 1001.
Those skilled in the art will appreciate that the terminal device configuration shown in fig. 1 is not intended to be limiting of the terminal device and may include more or fewer components than those shown, or some components may be combined, or a different arrangement of components.
As shown in fig. 1, a memory 1005, which is a kind of computer storage medium, may include therein an operating system, a network communication module, a user interface module, and a mining program of client behavior.
In the terminal shown in fig. 1, the network interface 1004 is mainly used for connecting to a backend server and performing data communication with the backend server; the user interface 1003 is mainly used for connecting a client and performing data communication with the client; and the processor 1001 may be configured to invoke a mining routine of the customer behavior stored in the memory 1005 and perform the following operational steps:
acquiring respective loan product historical information of each client to be excavated;
aggregating the historical information of the loan products with preset crowd characteristic data to divide each customer to be mined into a plurality of customer groups;
and mining the loan behaviors of a plurality of the passenger groups.
Further, the processor 1001 may be further configured to invoke a mining program of the customer behavior stored in the memory 1005, and further perform the following operation steps:
in a target customer group of the plurality of customer groups, determining a current loan product transaction of a target mining customer to generate a current transaction data set;
learning the current transaction data set according to a preset association rule to generate a frequent loan product set;
mining mutual dependencies among the loan products of the target mining client based on the frequent loan product set.
Further, the processor 1001 may be further configured to invoke a mining procedure of the customer behavior stored in the memory 1005 and perform the following operation steps:
generating association rules based on the frequent lending product set;
and determining the mutual dependency among the loan products of the target mining client according to the target rule which accords with the preset confidence condition in all the generated association rules.
Further, the processor 1001 may be further configured to invoke a mining procedure of the customer behavior stored in the memory 1005 and perform the following operation steps:
determining loan sequence data of a target mining customer in a target customer group of a plurality of customer groups;
determining a latest frequent sequence mode according to the loan sequence data;
and mining the lending behavior precedence relationship of the target mining client based on the latest frequent sequence mode.
Further, the processor 1001 may be further configured to invoke a mining program of the client behavior stored in the memory 1005, and further perform the following operation steps:
determining historical base data of the target mining client;
and generating loan sequence data of the target mining client according to the historical basic data.
Further, the processor 1001 may be further configured to invoke a mining program of the client behavior stored in the memory 1005, and further perform the following operation steps:
generating a loan history matrix according to the history information of the loan products;
and performing Gaussian clustering on the historical loan behavior vectors of the clients obtained by vectorizing the loan history matrix and preset population characteristic data to divide the clients to be mined into a plurality of client groups.
Based on the hardware structure, the invention provides various embodiments of the mining method of the client behavior.
It should be noted that, finding the borrowing behavior pattern of the customer has important significance for the bank to identify the borrowing will of the customer, perform accurate marketing and predict repayment default intention of the customer. Different organizations have different mining of the loan behavior of customers based on different business scenarios and application purposes. At present, the method mainly comprises two categories, wherein one category is to understand and predict the borrowing behavior mode of a customer by a traditional questionnaire-expert analysis method, namely analysis is carried out by small data and expert experience; the other type is that based on the internet loan behaviors, the client loan risk is predicted through a neural network model.
However, the traditional questionnaire-expert analysis method depends heavily on questionnaire scheme design and expert experience, is time-consuming and difficult to adapt to changing market change environments quickly, and the past experience-oriented schemes are often limited and cannot identify emerging lending habits timely. Borrowing risk mining based on a neural network model can quickly adapt to complex and variable customer requirements, but the model process is a black box, is not beneficial to understanding the internal rules of the borrowing behaviors of customers and has inexplicability.
In summary, the existing method for analyzing and mining the loan behaviors of the customers is difficult to meet the application requirements of the financial institutions for finding the loan behaviors of the customers for accurate marketing and risk identification.
In order to solve the above phenomenon, the invention provides a method for mining customer behaviors, which aggregates customer historical loan product behaviors and population characteristics to obtain different customer groups by considering customer historical loan product behaviors and current loan conditions, so as to mine the precedence relationship of loan products based on a customer loan sequence and mine the contemporaneous correlation of loan products based on the customer loan behaviors at the current time point, and finally obtain loan behavior patterns of different customer groups. The method and the system have the advantages that the client lending behaviors are mined from the time dimension and the product dimension, the comprehensive understanding of the lending behaviors of different customer groups and the learning of the client requirements are facilitated, the client lending behavior mode is identified, and the development of subsequent activities such as accurate marketing, risk identification and the like is facilitated.
In other words, in the process of mining the loan behavior for the clients, the invention acquires the loan product history information of the clients, aggregates the loan product history information with the crowd characteristic data to perform grouping processing for the clients to obtain a plurality of client groups, and finally mines the loan behavior for the client groups. Therefore, compared with the traditional mode of mining the client behaviors by adopting questionnaire and expert analysis or neural network model prediction, the method and the system have the advantages that different client groups are obtained by aggregating the historical borrowing and lending product behaviors of the clients and the crowd characteristic data, and then the borrowing and lending behaviors of different client groups are mined, so that the borrowing and lending behaviors and the client requirements of different client groups can be comprehensively known, the aim of accurately analyzing and mining the client borrowing and lending behaviors is fulfilled, and the development of subsequent activities such as accurate marketing, risk identification and the like by combining analysis and mining of a financial institution is facilitated.
Referring to fig. 2, fig. 2 is a flowchart illustrating a first embodiment of a mining method for customer behavior according to the present invention. It should be noted that although a logical order is shown in the flow chart, in some cases, the mining method of client behavior of the present invention may perform the steps shown or described in a different order than here.
In this embodiment, the mining method of the client behavior of the present invention is applied to the terminal device. The mining method of the client behavior comprises the following steps:
step S10, obtaining respective loan product historical information of each client to be excavated;
in this embodiment, in the process of performing loan behavior mining on one or more clients, the terminal device firstly uses the one or more clients as clients to be mined, which currently need to mine loan product behaviors, and then obtains and obtains respective history information of loan products of each client to be mined from a financial data platform or other another data platform which is docked in advance.
Step S20, aggregating the historical information of the loan products and preset crowd characteristic data to divide each customer to be mined into a plurality of customer groups;
in this embodiment, after obtaining the respective history information of the loan product of each to-be-mined client, the terminal device immediately aggregates the history information of the loan product with one or more preset crowd characteristic data, so as to divide the one or more to-be-mined clients into a plurality of client groups.
It should be noted that, in this embodiment, the same customer to be mined may belong to different customer groups.
Further, in a possible embodiment, the step S20 may specifically include:
step S201, generating a loan history matrix according to the history information of the loan products;
in this embodiment, the loan product history information includes: during which time various types of lended products are borrowed.
In this embodiment, after obtaining the respective history information of the loan products of each client to be mined, the terminal device further constructs a two-dimensional matrix by using the history information of the loan products of each client to be mined, and names the two-dimensional matrix as the history matrix of the loan products.
Illustratively, after obtaining the loan product history information of each client to be mined, the terminal device obtains the loan product history information based on the information including: during the period, whether various types of borrowing products are borrowed or not, the terminal equipment forms a two-dimensional matrix by enabling the x-axis direction to represent the types of the borrowing products, enabling the y-axis direction to represent the period number, and enabling the specific value-yes or no to represent whether the borrowing products are borrowed or not, and then the terminal further names the two-dimensional matrix as a borrowing history matrix.
For example, assume that the loan history matrix formed by the terminal device based on the history information of the loan products of the client a to be mined is:
the matrix then represents: customer A to be excavated, borrowing product i in the 1 st stage 2 Borrowing product i in stage 2 1 Borrowing product i at the same time in stage 3 1 And i 3 。
Step S202, performing gaussian clustering on the historical loan behavior vector of the client obtained by vectorizing the loan history matrix and preset population characteristic data to divide each client to be mined into a plurality of client groups.
In this embodiment, after constructing a corresponding loan history matrix based on each client to be mined, the terminal device further vectorizes the loan history matrix to obtain a client history loan behavior vector, and performs gaussian clustering on the client history loan behavior vector and one or more preset crowd characteristic data, so as to divide the client to be mined into a plurality of client groups.
It should be noted that, in this embodiment, the preset crowd characteristic data includes: profession, marital, work units, etc.
Illustratively, as shown in the client grouping process shown in fig. 3, in this embodiment, the terminal device performs gaussian clustering on a client historical loan behavior vector obtained after vectorization of a loan history matrix constructed based on the loan product history information of the client to be mined and data of the client to be mined, such as occupation, marital, work units, and the like, as features, so as to divide the client to be mined into a plurality of client groups.
It should be noted that, in this embodiment, because gaussian clustering uses a mean and a standard deviation, and a final output result of each sample is a probability, one customer to be mined may belong to a plurality of different customer groups, and this clustering manner is closer to real life and conforms to the condition that one customer may belong to a plurality of product interest groups.
In addition, in this embodiment, gaussian clustering is performed at each K value according to a set range of the number K of clusters, and profile coefficients at different numbers K of clusters are calculated, and finally, profile coefficients of a clustering result are calculated. And drawing a line graph about K by the contour coefficient, and selecting the K value with the maximum contour coefficient as the final clustering number to finally obtain K passenger groups with repeated customers.
And step S30, carrying out borrowing behavior mining on the plurality of passenger groups.
In this embodiment, after obtaining different customer groups by aggregating the historical information of the loan products of the customers to be mined and the crowd characteristic data, the terminal device mines the precedence relationship of the loan products based on the customer loan sequence and mines the contemporaneous correlation of the loan products based on the customer loan behavior at the current time, and finally obtains the loan behavior patterns of the different customer groups.
Further, in a possible embodiment, the step S30 may specifically include:
step S301, in a target customer group in a plurality of customer groups, determining a current loan product transaction of a target mining customer to generate a current transaction data set;
in this embodiment, when performing loan behavior mining on a plurality of customer groups, the terminal device sequentially determines a target customer group among the plurality of customer groups, and then determines, among a plurality of customers to be mined in the target customer group, a current node product transaction of a customer who needs to perform loan product behavior mining currently to generate a corresponding current transaction data set.
Illustratively, the guest groups as shown in FIG. 4In the borrowing behavior mining process, in this embodiment, the terminal device determines a target customer group, namely a customer group X, which is currently subjected to borrowing behavior mining among a plurality of customer groups, and then, the terminal device sequentially takes each customer to be mined in the customer group as a target mining customer so as to sequentially obtain current borrowing and lending product data of the target mining customer from product dimensions, thereby forming a customer borrowing and lending product transaction set T. That is, for customer C in customer group X, the current loan product transaction I can be expressed as: i = { I = } { (I) 1 ，i 2 ，i 3 ，...，i m Each i k (k =1,2, 3.. Eta., m) represents the borrowed product of the current customer C, and if there are one, the current transaction data set D = { I = { I = 1 ，I 2 ，...，I n }。
Step S302, learning the current transaction data set according to a preset association rule to generate a frequent loan product set;
in this embodiment, in the process of mining the loan behavior of the terminal device for the target customer group, after the terminal device generates the current transaction data set, the terminal device further performs association rule analysis on the data set, that is, the terminal device learns that the data set has generated a corresponding frequent loan product set based on the preset association rule.
It should be noted that, in the present embodiment, the preset association rule is a mining rule including support and confidence for representing interdependencies of loan products.
Illustratively, as shown in fig. 4, the terminal device sets the support degree and the confidence degree of the association rule to support the frequency of the item set appearing in the transaction and to represent the confidence degree of the association rule with the confidence degree, so that i 1 ＝>i 2 Indicates that i is contained in the transaction data set D 1 Also includes i in how likely it is in the transaction of 2 . Based on this, the terminal device learns and generates the transaction data set D through the association rule, and thus a frequent item set, i.e., a frequent loan product set, in the transaction data set D can be obtained.
And step S303, mining the interdependencies among the loan products of the target mining client based on the frequent loan product set.
In this embodiment, after the terminal device learns the current transaction data set through the preset association rule to generate the corresponding frequent loan product set, the terminal device further mines the mutual dependency between the loan products of the target mining client based on the frequent loan product set.
Further, in a possible embodiment, the step S303 may specifically include:
step S3031, generating association rules based on the frequent loan product set;
step S3032, determining the mutual dependency between the loan products of the target mining client according to the target rule which accords with the preset confidence coefficient condition in all the generated association rules.
Illustratively, as shown in fig. 4, in the present embodiment, the terminal device further mines the mutual dependency of the loan products based on the frequent loan product set after analyzing by the association-based rule: namely, the terminal equipment firstly finds out all non-empty subsets of the maximum frequent loan product set; then generating all possible association rules based on all non-empty subsets; and finally, calculating the confidence degrees of all the rules in turn, thereby finding out all the rules R which are larger than the confidence degree threshold value, wherein the rules R represent the interdependence relationship among the loan products, such as { i 1 ,i 4 }＝>i 1 Represents lending i 1 ,i 4 Customers of products are highly likely to loan i 1 And (5) producing the product.
Further, in another possible embodiment, the step S30 may further include:
step S304, determining loan sequence data of a target mining client in a target client group in a plurality of client groups;
in this embodiment, when performing loan behavior mining on a plurality of customer groups, the terminal device sequentially determines a target customer group among the plurality of customer groups, and then determines loan sequence data of a target mining customer currently needing loan product behavior mining among a plurality of to-be-mined customers in the target customer group.
Further, in a possible embodiment, the step S304 may specifically include:
step S3041, determining historical basic data of the target mining client;
step S3042, generating loan sequence data of the target mining client according to the historical basic data.
For example, in the mining process of the loan behavior of the customer group as shown in fig. 4, in this embodiment, when determining the loan sequence data of the target mining customer, the terminal device first obtains the history data of the loan product of the target mining customer, and further forms the history basic data of the target mining customer, which includes the customer ID, the time of the loan product, and the code number of the loan product. Then, the terminal device sets the statistical period of the loan products, so as to combine the records with the same customer ID (if monthly statistics indicates that the loan products in the same month are in the same set), and arrange the records according to the loan time sequence, so as to obtain the loan sequence data S of the target mining customer.
For example, assume that the loan sequence data of each customer to be mined in the current target customer group determined by the terminal device is shown in the following table:
wherein the lending sequence<{i 1 ,i 6 },{i 2 }>Indicating that client C001 loaned i within one cycle 1 ,i 6 Product, in a later period loan i 2 And (5) producing the product.
Step S305, determining a latest frequent sequence mode according to the loan sequence data;
in the embodiment, after determining the loan sequence data of the target mining client, the terminal device further mines the loan sequence data by using the SPADE algorithm to determine the latest frequent sequence pattern.
Illustratively, as shown in fig. 4, the terminal device mines loan sequence data of the target mining client by using the SPADE algorithm, wherein the SPADE algorithm decomposes an original problem into sub-problems which can be solved in a main memory by using a combinatorial property, and a search mode and a connection operation based on a sequence lattice are adopted. The specific process is as follows: (1) Generating all 1-sequences, the sequence composed of the individual terms being the 1-sequence, counting<{i 1 }><{i 2 }>...<{i m }>The sequence support degree corresponding to each sequence is a frequent 1-sequence when the sequence is larger than the support degree threshold value, and k =1. (2) And generating a candidate sequence of the 2-sequence through the self-connection operation of the frequent 1-sequence, and further obtaining the frequent 2-sequence, wherein k =2, and the time sequence of occurrence of loan behaviors is considered in association. (3) Candidate sequences for the 3-sequence are generated by frequent 2-sequence self-join operations. There are three possible ways for the self-connection of 2-frequent itemsets, one is the self-connection of itemset atomic items, the other is the connection of itemset atomic items with sequence atomic items, and the other is the self-connection of sequence atomic items. In order to improve the operation efficiency, the SPADE algorithm blocks the sequence in an equivalence class mode, wherein the equivalence class refers to the sequence with the same prefix. (4) k is increased progressively, and a candidate k-sequence S is generated by self-connection based on equivalence class by taking the frequent k-1 sequence as an atomic item k If k-sequence S k And if the sequence is an empty set, stopping the operation, wherein the frequent subsequence found by the k-1 is the latest frequent sequence mode.
And step S306, mining the lending behavior precedence relationship of the target mining client based on the latest frequent sequence mode.
For example, as shown in fig. 4, after determining the latest frequent sequence pattern, the terminal device may further mine the loan precedence relationship based on the latest frequent sequence pattern. That is, the terminal device first generates all possible sequence combinations based on the latest frequent sequence pattern; then, the confidence degrees of all the sequence combinations are calculated in sequence, and all the rules RS which are larger than the threshold value of the confidence degree are found out, wherein the rules RS represent the precedence relationship between loan behaviors, for example, { i { I } 5 }→{i 7 Denotes that i is borrowed 5 Customers of productsThe user is likely to borrow and loan i in the future 7 And (5) producing the product.
Therefore, the preference information of different customer groups is obtained according to the future borrowing and lending intentions of the different customer groups and the interdependence relationship among the borrowing and lending products, so that the method is beneficial to understanding the borrowing and lending behaviors of the customers, and assists in accurate marketing and risk identification.
In this embodiment, in the method for mining customer behaviors provided by the embodiment of the present invention, during a process of mining a loan behavior for one or more customers by using a terminal device, the one or more customers are first used as customers to be mined for currently mining a loan product behavior, then respective loan product history information of each customer to be mined is obtained from a financial data platform or other alternative data platforms which are butted in advance, and then the terminal device immediately aggregates the loan product history information with one or more preset crowd characteristic data, so as to divide the one or more customers to be mined into a plurality of customer groups. And finally, the terminal equipment excavates the precedence relationship of the loan products based on the client loan sequence and excavates the synchronization correlation of the loan products based on the client loan behaviors at the current time point, and finally obtains the loan behavior modes of different customer groups.
Therefore, compared with the traditional mode of mining the client behaviors by adopting questionnaire and expert analysis or neural network model prediction, the method integrates the client loan product history and the characteristics of client groups for grouping, mines the loan intention and the loan behavior sequence relation of different client groups from the time dimension and the product dimension, and can automatically obtain the loan preference of different groups.
In addition, the invention excavates the precedence relationship of the loan products based on the customer loan sequence and excavates the synchronization correlation of the loan products based on the customer loan behaviors at the current time point, and finally obtains the loan behavior patterns of different customer groups. In other words, in the technical scheme of the invention, the time dimension change of the clients is considered in the clustering and loan behavior mining of the clients, so that the multidimensional loan behavior pattern of the clients can be fully identified, and the mining result is more accurate.
In addition, the invention also provides a digging device for the customer behavior.
Referring to fig. 5, fig. 5 is a functional module diagram of an embodiment of a mining device for customer behavior according to the present invention.
As shown in fig. 5, the mining apparatus for client behavior according to the present invention includes:
the information acquisition module 10 is used for acquiring respective history information of the loan products of the clients to be excavated;
the customer grouping module 20 is configured to aggregate the historical information of the loan products with preset crowd characteristic data to divide each customer to be mined into a plurality of customer groups;
and the behavior mining module 30 is used for mining the loan behaviors of a plurality of the passenger groups.
Further, the behavior mining module includes:
the product dimension mining unit is used for determining the current loan product transaction of a target mining client in a target client group in a plurality of client groups to generate a current transaction data set; learning the current transaction data set according to a preset association rule to generate a frequent loan product set; and mining mutual dependencies among the loan products of the target mining client based on the frequent loan product set.
Further, the product dimension mining unit includes:
a rule generating subunit, configured to generate an association rule based on the frequent loan product set;
and the mining subunit is used for determining the mutual dependency among the loan products of the target mining client according to the target rule which accords with the preset confidence condition in all the generated association rules.
Further, the behavior mining module further includes:
the time dimension mining unit is used for determining loan and loan sequence data of a target mining client in a target client group in a plurality of client groups; determining a latest frequent sequence mode according to the loan sequence data; and mining the lending behavior precedence relationship of the target mining client based on the latest frequent sequence mode.
Further, the time dimension mining unit is further configured to determine historical basic data of the target mining client; and generating loan sequence data of the target mining client according to the historical basic data.
Further, the customer clustering module includes:
the matrix generating unit is used for generating a loan history matrix according to the history information of the loan products;
and the clustering processing unit is used for carrying out Gaussian clustering processing on the historical loan behavior vectors of the clients obtained by vectorizing the loan history matrix and preset population characteristic data so as to divide each client to be mined into a plurality of client groups.
The function implementation of each module in the mining device for the customer behavior corresponds to each step in the mining method embodiment for the customer behavior, and the function and implementation process thereof are not described in detail herein.
The present invention also provides a computer storage medium, on which a mining program of customer behavior is stored, and when being executed by a processor, the mining program of customer behavior implements the steps of the mining method of customer behavior according to any one of the above embodiments.
The specific embodiment of the computer storage medium of the present invention is basically the same as the embodiments of the mining method for the client behavior described above, and is not described herein again.
The invention also provides a computer program product comprising a computer program which, when executed by a processor, implements the steps of the mining method of customer behaviour as described in any one of the above embodiments.
The specific embodiment of the computer storage medium of the present invention is substantially the same as the embodiments of the mining method for the client behavior described above, and details are not described herein.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or system. Without further limitation, an element defined by the phrases "comprising a," "8230," "8230," or "comprising" does not exclude the presence of other like elements in a process, method, article, or system comprising the element.
The above-mentioned serial numbers of the embodiments of the present invention are merely for description and do not represent the merits of the embodiments.
Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium (e.g., ROM/RAM, magnetic disk, optical disk) as described above and includes instructions for enabling a terminal device (e.g., a mobile phone, a computer, a server, or a network device) to execute the method according to the embodiments of the present invention.
The above description is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention, and all equivalent structures or equivalent processes performed by the present invention or directly or indirectly applied to other related technical fields are also included in the scope of the present invention.
Claims (10)
1. A mining method for customer behaviors is characterized by comprising the following steps:
acquiring respective loan product historical information of each client to be excavated;
aggregating the lending product history information and preset crowd characteristic data to divide each customer to be mined into a plurality of customer groups;
and mining the lending behaviors of a plurality of the passenger groups.
2. The method for mining customer behavior according to claim 1, wherein the step of mining lending behavior for a plurality of the customer groups comprises:
determining, in a target customer group of the plurality of customer groups, a current loan product transaction of a target mining customer to generate a current transaction data set;
learning the current transaction data set according to a preset association rule to generate a frequent loan product set;
mining mutual dependencies among the loan products of the target mining customer based on the frequent loan product set.
3. The method of mining customer behavior according to claim 2, wherein said step of mining mutual dependencies between lending products of said target mining customer based on said set of frequent lending products comprises:
generating association rules based on the frequent lending product set;
and determining the mutual dependency among the loan products of the target mining client according to the target rule which accords with the preset confidence condition in all the generated association rules.
4. The method of mining customer behavior according to claim 1, wherein the step of mining loan behavior for a plurality of said customer groups comprises:
determining loan sequence data of a target mining customer in a target customer group of the plurality of customer groups;
determining a latest frequent sequence mode according to the loan sequence data;
and mining the loan behavior precedence relationship of the target mining client based on the latest frequent sequence pattern.
5. The method of mining customer behavior according to claim 4, wherein the step of determining loan sequence data for the target mining customer comprises:
determining historical base data of the target mining client;
and generating loan sequence data of the target mining client according to the historical basic data.
6. The method for mining customer behavior according to claim 1, wherein the step of aggregating the history information of lending products with preset population characteristic data to divide each customer to be mined into a plurality of customer groups comprises:
generating a loan history matrix according to the loan product history information;
and performing Gaussian clustering on the historical loan behavior vectors of the clients obtained by vectorizing the loan history matrix and preset population characteristic data to divide the clients to be mined into a plurality of client groups.
7. A mining device for customer behavior, characterized in that the mining device for customer behavior comprises:
the information acquisition module is used for acquiring respective historical information of the borrowed and lended products of the clients to be excavated;
the client grouping module is used for aggregating the historical information of the loan products and preset crowd characteristic data so as to divide each client to be mined into a plurality of client groups;
and the behavior mining module is used for mining the loan behaviors of the passenger groups.
8. The client behavior mining device of claim 7, wherein the behavior mining module comprises:
the product dimension mining unit is used for determining the current loan product transaction of a target mining client in a target client group in a plurality of client groups to generate a current transaction data set; learning the current transaction data set according to a preset association rule to generate a frequent loan product set; mining the interdependency among the loan products of the target mining client based on the frequent loan product set;
the product dimension mining unit comprises:
a rule generating subunit, configured to generate an association rule based on the frequent lending product set;
the mining subunit is used for determining the mutual dependency among the loan products of the target mining client according to a target rule which accords with a preset confidence condition in all the generated association rules;
the behavior mining module further comprises:
the time dimension mining unit is used for determining loan sequence data of a target mining client in a target client group in a plurality of client groups; determining a latest frequent sequence mode according to the loan sequence data; mining the sequence relation of the loan behaviors of the target mining client based on the latest frequent sequence mode;
the time dimension mining unit is further used for determining historical basic data of the target mining client; generating loan sequence data of the target mining client according to the historical basic data;
the customer clustering module comprises:
the matrix generating unit is used for generating a loan history matrix according to the loan product history information;
and the clustering processing unit is used for carrying out Gaussian clustering processing on the historical loan behavior vectors of the clients obtained by vectorizing the loan history matrix and preset population characteristic data so as to divide each client to be mined into a plurality of client groups.
9. A terminal device, characterized in that the terminal device comprises: a memory, a processor and a mining program of customer behaviour stored on said memory and executable on said processor, said mining program of customer behaviour when executed by said processor implementing the steps of the mining method of customer behaviour according to any one of claims 1 to 6.
10. A computer storage medium, characterized in that the computer storage medium has stored thereon a mining program of customer behavior, which when executed by a processor implements the steps of the mining method of customer behavior according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211210608.8A CN115617868A (en) | 2022-09-30 | 2022-09-30 | Mining method and device for customer behaviors, terminal equipment and computer storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211210608.8A CN115617868A (en) | 2022-09-30 | 2022-09-30 | Mining method and device for customer behaviors, terminal equipment and computer storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115617868A true CN115617868A (en) | 2023-01-17 |
Family
ID=84859825
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211210608.8A Pending CN115617868A (en) | 2022-09-30 | 2022-09-30 | Mining method and device for customer behaviors, terminal equipment and computer storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115617868A (en) |
-
2022
- 2022-09-30 CN CN202211210608.8A patent/CN115617868A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20210287048A1 (en) | System and method for efficient generation of machine-learning models | |
| US20220374746A1 (en) | Model interpretation | |
| US20190370695A1 (en) | Enhanced pipeline for the generation, validation, and deployment of machine-based predictive models | |
| Song et al. | A comparative study of dimensionality reduction techniques to enhance trace clustering performances | |
| US9147273B1 (en) | System and method for modeling and analyzing data via hierarchical random graphs | |
| US20130204831A1 (en) | Identifying associations in data | |
| US7636698B2 (en) | Analyzing mining pattern evolutions by comparing labels, algorithms, or data patterns chosen by a reasoning component | |
| EP3405872A1 (en) | Dynamically optimizing user engagement | |
| Martınez-Plumed et al. | Accounting for the neglected dimensions of ai progress | |
| US20220207414A1 (en) | System performance optimization | |
| US20230419134A1 (en) | Methods of explaining an individual predictions made by predictive processes and/or predictive models | |
| CN116882520A (en) | Prediction method and system for predetermined prediction problem | |
| Liu et al. | An efficient smart data mining framework based cloud internet of things for developing artificial intelligence of marketing information analysis | |
| Keshavarzi et al. | Clustering of large scale QoS time series data in federated clouds using improved variable Chromosome Length Genetic Algorithm (CQGA) | |
| US11144938B2 (en) | Method and system for predictive modeling of consumer profiles | |
| US20230032739A1 (en) | Propensity modeling process for customer targeting | |
| Lukita et al. | Predictive and Analytics using Data Mining and Machine Learning for Customer Churn Prediction | |
| CN115617868A (en) | Mining method and device for customer behaviors, terminal equipment and computer storage medium | |
| Taranto et al. | Uncertain Graphs meet Collaborative Filtering. | |
| JP2022531480A (en) | Visit prediction | |
| Demir et al. | Customer churn prediction with machine learning methods in telecommunication industry | |
| US11822564B1 (en) | Graphical user interface enabling interactive visualizations using a meta-database constructed from autonomously scanned disparate and heterogeneous sources | |
| US11928128B2 (en) | Construction of a meta-database from autonomously scanned disparate and heterogeneous sources | |
| US20230419195A1 (en) | System and Method for Hierarchical Factor-based Forecasting | |
| Amrou M'hand et al. | FuSTM: ProM plugin for fuzzy similar tasks mining based on entropy measure |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |