Method, device, processor and electronic device for determining a service recommendation scheme
By constructing a triplet structure and clustering algorithm, the problem of time-consuming, labor-intensive, and low-success-rate manual screening of institutions in business recommendations by financial institutions is solved, and efficient business recommendation schemes are generated.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-08-10
- Publication Date
- 2026-08-04
AI Technical Summary
In existing technologies, when financial institutions recommend businesses, manual screening of institutions consumes a lot of time and manpower, and the success rate is low.
By constructing a triplet structure, clustering algorithms are used to perform clustering analysis on entity information, relationship information, and attribute information to generate recommendation schemes for target businesses and extract target organizations and customers.
It improved the success rate of business recommendations, saved time and manpower costs, and achieved accurate business recommendations.
Smart Images
Figure CN117033789B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of financial technology, and more specifically, to a method, apparatus, processor, and electronic device for determining a business recommendation scheme. Background Technology
[0002] When financial institutions recommend products, a common method is for their sales staff to negotiate with the target billing institutions and then recommend the institution's bill collection services. These services can include assisting the billing institutions with internal or external bill payments, such as utility bills (water, electricity, gas), service fees, and property management fees. They can also assist the institutions with bill management, user payment services, payment details inquiries, refunds, and reconciliation.
[0003] Before making service recommendations, business personnel need to screen suitable institutions from the existing list of institutional clients. However, this method has many problems: due to the large number of institutions on the list, if the screening is done manually, the screening process will consume a lot of time and manpower. In addition, if the screening is done manually by business personnel, the recommendation success rate will be low due to insufficient knowledge of different institutions.
[0004] There is currently no effective solution to the problem of low success rate of business recommendations using recommendation schemes in related technologies. Summary of the Invention
[0005] The main objective of this application is to provide a method, apparatus, processor, and electronic device for determining a business recommendation scheme, in order to solve the problem of low success rate in business recommendation using recommendation schemes in related technologies.
[0006] To achieve the above objectives, according to one aspect of this application, a method for determining a business recommendation scheme is provided. The method includes: extracting business information of a target business and constructing a triplet structure from the business information, wherein the triplet structure represents the architecture between entity information, relational information, and attribute information in the business information; the entity information refers to the business type information of the target business; the relational information refers to the payment relationship information involved in the target business; the payment relationship refers to the payment relationship between the institution providing the target business and the customer; and the attribute information includes the institution's institutional attributes and the customer's customer attributes; performing cluster analysis on the entity information, relational information, and attribute information using a clustering algorithm to obtain clustering results, and determining target attribute information based on the clustering results; extracting the target institution and target customer from the target attribute information, and generating a recommendation scheme for the target business, wherein the recommendation scheme is used to recommend to the target customer to pay at the target institution to complete the target business.
[0007] Optionally, clustering algorithms are used to perform clustering analysis on entity information, relationship information, and attribute information to obtain clustering results including: assigning numerical values to entity information, relationship information, and attribute information according to an information assignment table to obtain A entity data, B relationship data, and C attribute data, where A, B, and C are positive integers; aggregating the A entity data, B relationship data, and C attribute data using a clustering model to obtain M clusters, and extracting attribute features from each cluster to obtain M candidate features, where M is a positive integer, M is less than A, M is less than B, and M is less than C; generating N feature matrices based on the M candidate features, where each feature matrix contains organizational attribute features and customer attribute features, N is a positive integer, and N is less than M.
[0008] Optionally, generating N feature matrices from M candidate features includes: performing feature aggregation on features belonging to the same business type among the M candidate features to obtain N feature matrices.
[0009] Optionally, determining the target attribute information based on the clustering results includes: calculating the similarity data between the institutional attribute features and customer attribute features in each feature matrix to obtain N similarity data; determining whether there are similarity data less than a preset threshold among the N similarity data; and if there are target similarity data less than the preset threshold among the N similarity data, using the feature matrix associated with the target similarity data as the target attribute information.
[0010] Optionally, after determining whether there is similarity data less than a preset threshold among the N similarity data, the method further includes: if there is no target similarity data less than the preset threshold among the N similarity data, adjusting the model parameters in the clustering model to obtain an updated clustering model; re-aggregating A entity data, B relation data, and C attribute data using the updated clustering model to obtain an updated cluster, and determining the feature matrix using the updated cluster, until there is at least one feature matrix associated with similarity data less than the preset threshold.
[0011] Optionally, extracting business information for the target business includes: obtaining information notes for candidate business information stored in the database, wherein the information notes include at least one of the following: business type information associated with the candidate business information, collection relationship information associated with the candidate business information, institutional attribute information associated with the candidate business information, and customer information associated with the candidate business information; determining target information notes based on the target business, and filtering business information from candidate business information based on the target information notes.
[0012] Optionally, when there are multiple recommendation schemes, after extracting the target organization and target customer from the target attribute information and generating a recommendation scheme for the target business, the method further includes: pushing the target recommendation scheme from the multiple recommendation schemes to the customer and receiving the customer's rating data for the target recommendation scheme; if the rating data is less than the rating threshold, pushing other recommendation schemes to the customer, wherein the other recommendation schemes are recommendation schemes other than the target recommendation scheme from the multiple recommendation schemes.
[0013] To achieve the above objectives, according to another aspect of this application, a device for determining a business recommendation scheme is provided. The device includes: a first extraction unit, configured to extract business information of a target business and construct a triplet structure from the business information, wherein the triplet structure represents the architecture between entity information, relationship information, and attribute information in the business information; the entity information refers to the business type information of the target business; the relationship information refers to the payment relationship information involved in the target business; the payment relationship refers to the payment relationship between the institution providing the target business and the customer; and the attribute information includes the institution's institutional attributes and the customer's customer attributes; an analysis unit, configured to perform cluster analysis on the entity information, relationship information, and attribute information using a clustering algorithm to obtain clustering results and determine target attribute information based on the clustering results; and a second extraction unit, configured to extract the target institution and target customer from the target attribute information and generate a recommendation scheme for the target business, wherein the recommendation scheme is used to recommend to the target customer to pay at the target institution to complete the target business.
[0014] According to another aspect of the present invention, a processor is also provided, which is used to run a program, wherein the program controls a device containing a non-volatile storage medium to perform a method for determining a service recommendation scheme during runtime.
[0015] According to another aspect of the present invention, an electronic device is also provided, comprising one or more processors and a memory; the memory stores computer-readable instructions, and the processor is configured to execute the computer-readable instructions, wherein the computer-readable instructions, when executed, perform a method for determining a business recommendation scheme.
[0016] This application employs the following steps: First, business information of the target business is extracted, and a triplet structure is constructed from this information. The triplet structure represents the architecture between entity information, relationship information, and attribute information within the business information. Entity information refers to the business type information of the target business; relationship information refers to the collection relationship information involved in the target business; the collection relationship refers to the collection relationship between the institution providing the target business and the customer; attribute information includes the institution's institutional attributes and the customer's customer attributes. Then, a clustering algorithm is used to perform cluster analysis on the entity information, relationship information, and attribute information to obtain clustering results, and target attribute information is determined based on the clustering results. Next, target institutions and target customers are extracted from the target attribute information to generate a recommendation scheme for the target business. This recommendation scheme is used to recommend that target customers pay at the target institution to complete the target business. This solves the problem of low success rate in related technologies using recommendation schemes for business recommendations. By extracting business information of the target business and constructing a triplet structure from it, and then using a clustering algorithm to perform cluster analysis on the triplet structure to obtain clustering results, a recommendation scheme for the target business is generated from the clustering results, thereby improving the success rate of push notifications using recommendation schemes. Attached Figure Description
[0017] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:
[0018] Figure 1 This is a flowchart of a method for determining a business recommendation scheme according to an embodiment of this application;
[0019] Figure 2 This is a schematic diagram of a method for determining target attribute information according to an embodiment of this application;
[0020] Figure 3 This is a schematic diagram of a device for determining a business recommendation scheme according to an embodiment of this application;
[0021] Figure 4 This is a schematic diagram of an electronic device provided according to an embodiment of this application. Detailed Implementation
[0022] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.
[0023] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.
[0024] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate for the embodiments of this application described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.
[0025] It should be noted that all information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for display, data used for analysis, etc.) involved in this disclosure are information and data authorized by the user or fully authorized by all parties.
[0026] The present invention will now be described in conjunction with preferred implementation steps. Figure 1 This is a flowchart of a method for determining a business recommendation scheme according to an embodiment of this application, such as... Figure 1 As shown, the method includes the following steps:
[0027] Step S101: Extract the business information of the target business and construct a triple structure from the business information. The triple structure represents the architecture between entity information, relationship information and attribute information in the business information. Entity information refers to the business type information of the target business. Relationship information refers to the collection relationship information involved in the target business. Collection relationship refers to the collection relationship between the institution providing the target business and the customer. Attribute information includes the institution's institutional attributes and the customer's customer attributes.
[0028] Specifically, target business refers to the business services that financial institutions need to recommend to different institutions or customers. These can include bill payment services, bill management services, and commercial loan services. Business information refers to relevant information generated during the target business. For example, when the target business is a bill payment service, the business information could include the payment details and the payment amount, and could include the customer group making the payment, the institution making the payment, the payment method, and the purpose of the payment.
[0029] In their daily operations, financial institutions need to negotiate and collaborate with organizations across various industries. During this process, financial institutions offer recommended business solutions to these organizations, which then process the business according to these solutions, achieving a win-win outcome. However, before negotiating and establishing a mutually beneficial partnership, it's necessary to select suitable institutions with a high success rate from the existing client list and then develop customized business recommendation plans for them. If this process is performed manually by business personnel, it consumes significant time and manpower. Therefore, a solution is to construct a triplet structure and build a knowledge graph based on it. This knowledge graph can then determine suitable institutions, clients, and appropriate recommendation plans.
[0030] The triple structure is obtained by processing business information using the RDF (Resource Description Framework) architecture. Therefore, in order to achieve accurate recommendations, it is first necessary to obtain the business information related to the target business of financial institutions, and then select appropriate business information from the business information to construct the triple structure.
[0031] It should be noted that the triple structure refers to the basic unit in a knowledge graph, including entities, relations, and attributes. That is, it includes entity information, relation information, and attribute information. An entity represents an individual; in this embodiment, entity information can be the business type information of the target service that a financial institution needs to push, such as information about the fuel industry. A relation represents the connection or association between entities; in this embodiment, relation information is the information about the flow of funds between a charging institution and paying customers, such as the water bill collection information between a water plant and residents in a community. An attribute represents the characteristics or attributes of an entity; in this embodiment, attribute information can refer to the nature of the institution and the attributes of customers. For example, the attribute information of an institution can be a water supply institution, and the attribute information of some residents can be a group living in a community and aged 40 or above. By constructing and analyzing triples, the relationships and attributes between entities can be revealed, thereby helping to determine recommendation schemes.
[0032] Step S102: Use a clustering algorithm to perform cluster analysis on entity information, relationship information and attribute information to obtain clustering results, and determine the target attribute information based on the clustering results.
[0033] It should be noted that clustering algorithm refers to an unsupervised learning algorithm used to divide samples in a dataset into several categories or clusters, such that samples within the same category or cluster have high similarity, while samples between different categories or clusters have low similarity. In the embodiments of this application, hierarchical clustering algorithm, density clustering algorithm, and K-means algorithm can be used.
[0034] Specifically, after determining the triple structure based on business information, a clustering algorithm is used to perform cluster analysis on the information contained in the triple structure. That is, the similarity data between different institutional attribute features and different customer attribute features in the triple structure is calculated by the clustering algorithm, and the clustering results containing target attribute information are obtained based on the similarity data. Then, the institutional attribute features and customer attribute features can be obtained from the target attribute information. Here, the target attribute information refers to the attribute information of institutions and customers who can cooperate on business recommendation solutions.
[0035] Step S103: Extract the target institution and target customer from the target attribute information, and generate a recommendation scheme for the target business. The recommendation scheme is used to recommend to the target customer to pay at the target institution to complete the target business.
[0036] Specifically, after obtaining the target attribute information through cluster analysis, the organizational attribute information representing the organizational attributes is extracted from it, that is, the target organization is extracted from it. Then, the customer attribute information representing the customer attributes is extracted, that is, the target customer is obtained.
[0037] Furthermore, the target organization and target customer are identified as the parties to be negotiated, and a recommended plan for the operation of the target organization's business is designed based on the industry information of the target organization. The designed recommended plan is then pushed to the target customer or the target organization.
[0038] The method for determining a business recommendation scheme provided in this application extracts business information of the target business and constructs a triplet structure from the business information. The triplet structure represents the architecture between entity information, relationship information, and attribute information in the business information. Entity information refers to the business type information of the target business; relationship information refers to the collection relationship information involved in the target business; the collection relationship refers to the collection relationship between the institution providing the target business and the customer; and attribute information includes the institution's institutional attributes and the customer's customer attributes. A clustering algorithm is used to perform clustering analysis on the entity information, relationship information, and attribute information to obtain clustering results, and target attribute information is determined based on the clustering results. Target institution and target customer are extracted from the target attribute information to generate a recommendation scheme for the target business. The recommendation scheme is used to recommend that the target customer pay at the target institution to complete the target business. This solves the problem of low success rate in related technologies using recommendation schemes for business recommendation. By extracting business information of the target business and constructing a triplet structure from the business information, and using a clustering algorithm to perform clustering analysis on the triplet structure to obtain clustering results, a recommendation scheme for the target business is generated from the clustering results, thereby improving the success rate of push notifications using recommendation schemes.
[0039] To obtain clustering results, the triplet structure needs to be processed. Optionally, in the method for determining the business recommendation scheme provided in this application embodiment, a clustering algorithm is used to perform clustering analysis on entity information, relationship information, and attribute information to obtain clustering results, including: assigning numerical values to entity information, relationship information, and attribute information according to an information assignment table to obtain A entity data, B relationship data, and C attribute data, where A, B, and C are positive integers; performing data aggregation processing on the A entity data, B relationship data, and C attribute data through a clustering model to obtain M clusters, and extracting attribute features from each cluster to obtain M candidate features, where M is a positive integer, M is less than A, M is less than B, and M is less than C; generating N feature matrices based on the M candidate features, where each feature matrix contains institutional attribute features and customer attribute features, N is a positive integer, and N is less than M.
[0040] Before performing cluster analysis on the information contained in the triplet structure using clustering algorithms, the information contained in the triplet structure needs to be processed. Specifically, relevant information needs to be obtained first, and feature selection and feature extraction processing need to be performed on this information. Since the entity information, relation information, and attribute information contained in the triplet structure are all character-type information, while cluster analysis requires numerical information, it is necessary to convert the entity information, relation information, and attribute information into numerical values. That is, the numerical values representing entity information, relation information, and attribute information are obtained through an information assignment table. Then, the numerical values are used to convert multiple pieces of information in the triplet structure obtained from the business information, thereby obtaining entity data, relation data, and attribute data.
[0041] Furthermore, since entity information contains a large amount of business type information, and attribute information contains a large amount of organizational and customer attribute information, preliminary data aggregation is required for multiple entity data, multiple relational data, and multiple attribute data. For example, if the attribute information includes information from different companies such as water bill payment, electricity bill payment, and gas bill payment, then the companies for water bill payment, electricity bill payment, and gas bill payment are merged and collectively referred to as companies related to utility bill payments. By aggregating the data, more business information can be obtained, laying the foundation for determining the target attribute information in the subsequent process. Then, attribute features related to the attribute information are extracted from the clusters obtained after the preliminary data aggregation process. Candidate features are obtained from the attribute features. Each cluster can include entity features obtained after processing entity data, relational features obtained after processing relational data, and attribute features obtained after processing attribute data.
[0042] Furthermore, to more accurately identify target organizations and enterprises, the candidate features need to be processed again, and the processed candidate features are represented in a vectorized form to obtain a feature matrix. This embodiment, by sequentially assigning numerical values, aggregating data, and aggregating features to the information contained in the triplet structure, can obtain a feature matrix with organizational and customer attribute information as elements, thus laying the foundation for cluster analysis to obtain target attribute information.
[0043] The feature matrix is obtained by processing multiple candidate features. Optionally, in the method for determining the business recommendation scheme provided in the embodiments of this application, generating N feature matrices based on M candidate features includes: performing feature aggregation processing on features belonging to the same business type among the M candidate features to obtain N feature matrices.
[0044] Specifically, after obtaining multiple candidate features from the cluster, the business type of these candidate features is determined, and features belonging to the same business type are selected as features for aggregation. For example, when determining the business type of the candidate features, a set of customer attribute features and institutional attribute features related to utility bill payment are obtained. Therefore, feature aggregation processing can be performed on this set of customer attribute features and institutional attribute features, that is, a feature matrix is obtained by using customer attribute features and institutional attribute features as elements of a matrix.
[0045] It should be noted that after determining that multiple candidate features can be aggregated using business types, this principle can also be used to construct a matrix generation model based on a machine learning model. The feature matrix can be directly obtained from the matrix generation model. Specifically, the candidate features obtained after data aggregation are input into the machine learning model. The matrix generation model then arranges and combines different candidate features according to business types, resulting in multiple feature matrices. Each feature matrix contains one customer attribute feature and one organization attribute feature. For example, the organization attribute features related to organization attributes include features A, B, and C, and the customer attribute feature related to customer attributes is X. Therefore, the matrix generation model can obtain three feature matrices, where the elements can be AX, BX, and CX. This embodiment uses candidate features to determine the feature matrix, thus laying the foundation for subsequent cluster analysis using the feature matrix.
[0046] Figure 2 This is a schematic diagram of a method for determining target attribute information according to an embodiment of this application, such as... Figure 2 As shown, optionally, in the method for determining a business recommendation scheme provided in this application embodiment, determining the target attribute information based on the clustering results includes:
[0047] Step S201: Calculate the similarity data between the institutional attribute features and customer attribute features in each feature matrix to obtain N similarity data.
[0048] Step S202: Determine whether there are any similarity data points among the N similarity data that are less than a preset threshold.
[0049] Step S203: If there are target similarity data with a value less than a preset threshold among the N similarity data, the feature matrix associated with the target similarity data is used as the target attribute information.
[0050] Specifically, after obtaining multiple feature matrices, the similarity function can be used to calculate the similarity data of the elements in the feature matrices, that is, to calculate the similarity data between institutional attribute features and customer attribute features.
[0051] Furthermore, after obtaining the similarity data corresponding to each feature matrix, the relationship between each similarity data and the preset threshold is determined. When there are target similarity data that are less than the preset threshold among these similarity data, it indicates that the correlation between the institutional attribute features and customer attribute features in the feature matrix associated with these similarity data is high. That is, there is a clear supply and demand relationship between the institutions and customers obtained from the institutional attribute features and customer attribute features. At this time, the target attribute information can be determined from the target similarity data that are less than the preset threshold.
[0052] Specifically, based on the similarity data of each target, a feature matrix associated with it is determined. This allows us to obtain the institutional attribute features and customer attribute features within these associated feature matrices. Then, the institutional attribute information and customer attribute information are determined from these features, and finally, these are identified as the target attribute information. This embodiment calculates the similarity data of the feature matrix and uses this similarity data to determine the target attribute information. This method provides a relatively simple way to obtain target attribute information, saving time and manpower costs in developing recommendation schemes.
[0053] Optionally, in the method for determining the business recommendation scheme provided in the embodiments of this application, after determining whether there is similarity data less than a preset threshold among the N similarity data, the method further includes: if there is no target similarity data less than the preset threshold among the N similarity data, adjusting the model parameters in the clustering model to obtain an updated clustering model; re-aggregating A entity data, B relation data, and C attribute data through the updated clustering model to obtain an updated cluster, and determining the feature matrix by the updated cluster until there is at least one feature matrix associated with similarity data less than the preset threshold.
[0054] Specifically, if the similarity data is calculated from each feature matrix and the relationship between the similarity data and the preset threshold is judged, and all similarity data are greater than the preset threshold, it indicates that the correlation between the institutional attribute features and customer attribute features in each feature matrix is low, the supply and demand relationship between the institution and the customer is not obvious, and new institutional attribute features and customer attribute features need to be found.
[0055] Specifically, when performing clustering analysis using clustering algorithms, a clustering model can be constructed using the clustering algorithm, and similarity data can be obtained using the clustering model. However, if the similarity data obtained using the clustering model is greater than a preset threshold, it indicates that the model parameters in the model do not match the target business. Therefore, by adjusting the model parameters and then using the adjusted clustering model, the data after numerical transformation obtained from the triple structure of the target business can be re-aggregated and processed for data aggregation and feature aggregation until an updated feature matrix with similarity data less than the preset threshold is obtained. The target attribute information is then obtained from the similarity data of this feature matrix. In this embodiment, by adjusting the model parameters of the clustering model, the model is used to re-aggregate the triple structure associated with the feature matrix with similarity data greater than the preset threshold, thereby obtaining a feature matrix that meets the requirements. The target attribute information is then determined from this feature matrix, thus saving time for the revision of the recommendation scheme.
[0056] Before using clustering algorithms to perform clustering analysis on relevant business information, it is necessary to filter and extract the business information. Optionally, in the method for determining the business recommendation scheme provided in this application embodiment, extracting the business information of the target business includes: obtaining information notes of candidate business information stored in the database, wherein the information notes include at least one of the following: business type information associated with the candidate business information, collection relationship information associated with the candidate business information, institutional attribute information associated with the candidate business information, and customer information associated with the candidate business information; determining the target information notes according to the target business, and filtering the business information from the candidate business information according to the target information notes.
[0057] Specifically, since the databases associated with financial institutions contain massive amounts of business information, judging each piece of data in each business information and determining the business information of the target business based on the judgment results would waste time. Therefore, in order to reduce analysis time and cost, judgment can be made based on the information notes annotated when each business information is generated.
[0058] Specifically, since the information notes contain information on the business type, collection relationship, institutional attributes, and customer information for each business, after identifying the target business, the business type, collection relationship, attributes of related institutions, and customer attributes of the target business are determined based on the target business. Then, based on the business type, collection relationship, institutional attributes, and customer attributes, relevant business information is filtered from a massive amount of candidate business information. This embodiment utilizes the information notes of the business information to determine the business information of the target business, thereby laying the foundation for formulating a recommendation scheme for the target business.
[0059] To provide better service to customers, optionally, in the method for determining a business recommendation scheme provided in this application embodiment, when there are multiple recommendation schemes, after extracting the target organization and target customer from the target attribute information and generating a recommendation scheme for the target business, the method further includes: pushing the target recommendation scheme from the multiple recommendation schemes to the customer and receiving the customer's rating data for the target recommendation scheme; if the rating data is less than the rating threshold, pushing other recommendation schemes to the customer, wherein the other recommendation schemes are recommendation schemes other than the target recommendation scheme from the multiple recommendation schemes.
[0060] Specifically, to improve the recommendation strategy and provide a better customer experience, after recommending the target recommendation scheme determined by the target attribute information obtained from the clustering algorithm to the customer, the target recommendation scheme can be adjusted by collecting customer feedback rating data. When the customer feedback rating data is higher than a preset threshold, it indicates that the customer is highly satisfied with the target recommendation scheme, and subsequent services can be carried out directly according to the target recommendation scheme. If the customer feedback rating data is lower than the preset threshold, it indicates that the target recommendation scheme is not suitable for the customer's institutional business, and other recommendation schemes need to be recommended to the customer. Among them, other recommendation schemes refer to other feature matrices obtained after undergoing the same feature aggregation processing as the target recommendation scheme.
[0061] It should be noted that even after recommending other options to customers, customer rating data regarding those recommendations can still be collected. This rating data is used to determine customer satisfaction until it exceeds a preset requirement, i.e., the rating data exceeds a preset threshold. This embodiment collects customer feedback rating data after recommending options to customers, determines the merits of each option based on the rating data, and pushes other recommendations when the rating data falls below a preset threshold. This improves customer satisfaction and provides insights for the development of subsequent recommendation schemes.
[0062] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.
[0063] This application also provides a device for determining a service recommendation scheme. It should be noted that this device can be used to execute the service recommendation scheme determination method provided in this application. The following describes the device for determining a service recommendation scheme provided in this application.
[0064] Figure 3This is a schematic diagram of a device for determining a business recommendation scheme according to an embodiment of this application, such as... Figure 3 As shown, the device includes: a first extraction unit 30, an analysis unit 31, and a second extraction unit 32.
[0065] The first extraction unit 30 is used to extract business information of the target business and construct a triple structure from the business information. The triple structure represents the architecture between entity information, relation information and attribute information in the business information. Entity information refers to the business type information of the target business, relation information refers to the collection relationship information involved in the target business, collection relationship refers to the collection relationship between the institution providing the target business and the customer, and attribute information includes the institution's institutional attributes and the customer's customer attributes.
[0066] Analysis unit 31 is used to perform cluster analysis on entity information, relationship information and attribute information using clustering algorithms, obtain clustering results, and determine target attribute information based on the clustering results;
[0067] The second extraction unit 32 is used to extract target institutions and target customers from target attribute information and generate a recommendation scheme for the target business. The recommendation scheme is used to recommend to the target customer to pay at the target institution to complete the target business.
[0068] Optionally, in the business recommendation scheme determination device provided in this application embodiment, the analysis unit 31 includes: an assignment module, used to assign numerical values to entity information, relationship information and attribute information according to an information assignment table to obtain A entity data, B relationship data and C attribute data, wherein A, B and C are positive integers; a first processing module, used to perform data aggregation processing on the A entity data, B relationship data and C attribute data through a clustering model to obtain M clusters, and extract attribute features from each cluster to obtain M candidate features, wherein M is a positive integer, M is less than A, M is less than B, and M is less than C; and a generation module, used to generate N feature matrices according to the M candidate features, wherein each feature matrix contains institutional attribute features and customer attribute features, N is a positive integer and N is less than M.
[0069] Optionally, in the business recommendation scheme determination device provided in the embodiments of this application, the analysis unit 31 includes: a second processing module, used to perform feature aggregation processing on the features belonging to the same business type among M candidate features to obtain N feature matrices.
[0070] Optionally, in the business recommendation scheme determination device provided in this application embodiment, the analysis unit 31 includes: a calculation module, used to calculate the similarity data between the institutional attribute features and customer attribute features in each feature matrix to obtain N similarity data; a judgment module, used to determine whether there is similarity data less than a preset threshold among the N similarity data; and a first determination module, used to take the feature matrix associated with the target similarity data as target attribute information when there is target similarity data less than the preset threshold among the N similarity data.
[0071] Optionally, in the business recommendation scheme determination device provided in the embodiments of this application, the device further includes: an adjustment unit, used to adjust the model parameters in the clustering model after determining whether there is similarity data less than a preset threshold among N similarity data, and in the case that there is no target similarity data less than the preset threshold among the N similarity data, to obtain an updated clustering model; and a processing unit, used to re-aggregate A entity data, B relation data, and C attribute data through the updated clustering model to obtain an updated cluster, and determine the feature matrix from the updated cluster until there is at least one feature matrix associated with similarity data less than the preset threshold.
[0072] Optionally, in the business recommendation scheme determination device provided in this application embodiment, the first extraction unit 30 includes: an acquisition module, used to acquire information notes of candidate business information stored in the database, wherein the information notes include at least one of the following: business type information associated with the candidate business information, collection relationship information associated with the candidate business information, institutional attribute information associated with the candidate business information, and customer information associated with the candidate business information; and a second determination module, used to determine target information notes according to the target business, and to filter business information from the candidate business information according to the target information notes.
[0073] Optionally, in the business recommendation scheme determination device provided in the embodiments of this application, the device further includes: a first push unit, configured to, when there are multiple recommendation schemes, after extracting the target organization and target customer from the target attribute information and generating a recommendation scheme for the target business, push the target recommendation scheme among the multiple recommendation schemes to the customer, and receive the customer's rating data for the target recommendation scheme; and a second push unit, configured to, when the rating data is less than a rating threshold, push other recommendation schemes to the customer, wherein the other recommendation schemes are recommendation schemes other than the target recommendation scheme among the multiple recommendation schemes.
[0074] The device for determining a business recommendation scheme provided in this application embodiment, through a first extraction unit 30, is used to extract business information of the target business and construct a triplet structure from the business information. The triplet structure represents the architecture between entity information, relationship information, and attribute information in the business information. Entity information refers to the business type information of the target business; relationship information refers to the collection relationship information involved in the target business; collection relationship refers to the collection relationship between the institution providing the target business and the customer; attribute information includes the institution's institutional attributes and the customer's customer attributes. The analysis unit 31 is used to perform cluster analysis on the entity information, relationship information, and attribute information using a clustering algorithm. The first unit obtains clustering results and determines target attribute information based on these results. The second extraction unit 32 extracts target institutions and target customers from the target attribute information and generates a recommendation scheme for the target business. The recommendation scheme is used to recommend that target customers pay at the target institution to complete the target business. This solves the problem of low success rate of business recommendation using recommendation schemes in related technologies. By extracting business information of the target business and constructing a triplet structure from the business information, clustering analysis is performed on the triplet structure using a clustering algorithm to obtain clustering results. The recommendation scheme for the target business is generated from the clustering results, thereby improving the success rate of push notifications using recommendation schemes.
[0075] The device for determining the business recommendation scheme includes a processor and a memory. The first extraction unit 30, the analysis unit 31, the second extraction unit 32, etc. are all stored in the memory as program units. The processor executes the program units stored in the memory to realize the corresponding functions.
[0076] The processor contains a kernel, which retrieves the corresponding program units from memory. One or more kernels can be configured, and adjusting kernel parameters can address the low success rate of business recommendations using suggested solutions in related technologies.
[0077] The memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.
[0078] This invention provides a computer-readable storage medium storing a program that, when executed by a processor, implements a method for determining a business recommendation scheme.
[0079] This invention provides a processor for running a program, wherein the program executes a method for determining a business recommendation scheme during runtime.
[0080] Figure 4 This is a schematic diagram of an electronic device provided according to an embodiment of this application, such as... Figure 4 As shown, this embodiment of the invention provides an electronic device 40, which includes a processor, a memory, and a program stored in the memory and executable on the processor. The processor is used to execute computer-readable instructions, wherein the computer-readable instructions execute a method for determining a business recommendation scheme. The device in this document can be a server, PC, PAD, mobile phone, etc.
[0081] This application also provides a computer program product, which, when executed on a data processing device, is suitable for performing a method for determining a business recommendation scheme.
[0082] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0083] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0084] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.
[0085] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.
[0086] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.
[0087] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.
[0088] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.
[0089] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.
[0090] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0091] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.
Claims
1. A method for determining a business recommendation scheme, characterized in that, include: Extract business information of the target business and construct a triple structure from the business information. The triple structure represents the architecture between entity information, relationship information and attribute information in the business information. The entity information refers to the business type information of the target business. The relationship information refers to the collection relationship information involved in the target business. The collection relationship refers to the collection relationship between the institution providing the target business and the customer. The attribute information includes the institution's institutional attributes and the customer's customer attributes. The business information is relevant information generated in the process of carrying out the target business. Clustering algorithms are used to perform cluster analysis on the entity information, the relationship information, and the attribute information to obtain clustering results, and the target attribute information is determined based on the clustering results; The target institution and target customer are extracted from the target attribute information to generate a recommendation scheme for the target service. The recommendation scheme is used to recommend to the target customer to pay at the target institution to complete the target service. Clustering algorithms are used to perform clustering analysis on the entity information, the relationship information, and the attribute information to obtain clustering results including: The entity information, relation information, and attribute information are numerically assigned according to the information assignment table to obtain A entity data, B relation data, and C attribute data, where A, B, and C are positive integers. The A entity data, B relation data, and C attribute data are aggregated using a clustering model to obtain M clusters. Attribute features are then extracted from each cluster to obtain M candidate features, where M is a positive integer, M is less than A, M is less than B, and M is less than C. N feature matrices are generated based on the M candidate features, wherein each feature matrix contains organizational attribute features and customer attribute features, and N is a positive integer, N is less than M; The target attribute information determined based on the clustering results includes: Calculate the similarity data between the institutional attribute features and customer attribute features in each feature matrix to obtain N similarity data; Determine whether there are any similarity data points among the N similarity data that are less than a preset threshold; If there is a target similarity data with a similarity value less than the preset threshold among the N similarity data, the feature matrix associated with the target similarity data is used as the target attribute information; After determining whether there are any similarity data points among the N similarity data that are less than a preset threshold, the method further includes: If there is no target similarity data less than the preset threshold among the N similarity data, the model parameters in the clustering model are adjusted to obtain an updated clustering model; The updated clustering model is used to re-aggregate the A entity data, B relation data, and C attribute data to obtain updated clusters. The updated clusters are then used to determine feature matrices until at least one feature matrix has a similarity data value less than a preset threshold.
2. The method according to claim 1, characterized in that, Generating an N feature matrix based on the M candidate features includes: The features belonging to the same business type among the M candidate features are aggregated to obtain N feature matrices.
3. The method according to claim 1, characterized in that, The business information extracted from the target business includes: Retrieve information notes of candidate business information stored in the database, wherein the information notes include at least one of the following: business type information associated with the candidate business information, collection relationship information associated with the candidate business information, organization attribute information associated with the candidate business information, and customer information associated with the candidate business information; Target information notes are determined based on the target business, and the business information is obtained by filtering from the candidate business information based on the target information notes.
4. The method according to claim 1, characterized in that, When there are multiple recommended solutions, after extracting the target organization and target customer from the target attribute information and generating a recommended solution for the target business, the method further includes: Push the target recommendation scheme from multiple recommendation schemes to the target customer, and receive the target customer's rating data for the target recommendation scheme; If the rating data is less than the rating threshold, other recommendation schemes are pushed to the target customer, wherein the other recommendation schemes are recommendation schemes other than the target recommendation scheme among the multiple recommendation schemes.
5. A device for determining a business recommendation scheme, characterized in that, include: The first extraction unit is used to extract business information of the target business and construct a triple structure from the business information. The triple structure represents the architecture between entity information, relationship information and attribute information in the business information. The entity information refers to the business type information of the target business. The relationship information refers to the collection relationship information involved in the target business. The collection relationship refers to the collection relationship between the institution providing the target business and the customer. The attribute information includes the institution's institutional attributes and the customer's customer attributes. The business information is relevant information generated in the process of carrying out the target business. The analysis unit is used to perform cluster analysis on the entity information, the relationship information and the attribute information using a clustering algorithm, obtain clustering results, and determine target attribute information based on the clustering results; The second extraction unit is used to extract the target institution and target customer from the target attribute information and generate a recommendation scheme for the target business, wherein the recommendation scheme is used to recommend to the target customer to pay at the target institution to complete the target business; The analysis unit includes: an assignment module, used to assign numerical values to entity information, relation information, and attribute information according to an information assignment table, to obtain A entity data, B relation data, and C attribute data, where A, B, and C are positive integers; a first processing module, used to perform data aggregation processing on the A entity data, B relation data, and C attribute data through a clustering model to obtain M clusters, and extract attribute features from each cluster to obtain M candidate features, where M is a positive integer, M is less than A, M is less than B, and M is less than C; and a generation module, used to generate N feature matrices based on the M candidate features, where each feature matrix contains organizational attribute features and customer attribute features, N is a positive integer, and N is less than M. The analysis unit further includes: a calculation module, used to calculate the similarity data between the institutional attribute features and the customer attribute features in each feature matrix to obtain N similarity data; a judgment module, used to determine whether there is similarity data less than a preset threshold among the N similarity data; and a first determination module, used to take the feature matrix associated with the target similarity data as target attribute information when there is target similarity data less than the preset threshold among the N similarity data. The adjustment unit is used to adjust the model parameters in the clustering model after determining whether there is similarity data less than a preset threshold among N similarity data. If there is no target similarity data less than the preset threshold among the N similarity data, the model parameters are adjusted to obtain an updated clustering model. The processing unit is used to re-aggregate A entity data, B relation data, and C attribute data through the updated clustering model to obtain an updated cluster. The updated cluster is then used to determine the feature matrix until at least one feature matrix is associated with similarity data less than the preset threshold.
6. A processor, characterized in that, The processor is used to run a program, wherein the program executes the method for determining the business recommendation scheme according to any one of claims 1 to 4 when it runs.
7. An electronic device, characterized in that, It includes one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the method for determining a business recommendation scheme as described in any one of claims 1 to 4.