Intelligent data matching recommendation method, system and device and storage medium

Abstract

The invention discloses an intelligent data matching recommendation method, which comprises the following steps of: S1, burying points on a user browsing page, and transmitting consumption behavior data of a user to a server for data collection; s2, performing data cleaning processing on the acquired user consumption data according to an RFM model; s3, performing fuzzy clustering on the data by using an FMC algorithm after the data cleaning processing, and searching an optimal central value of the fuzzy clustering in the data after the data cleaning processing by improving a greedy algorithm; and S4, calculating the optimal operation promotion data of the current user, and displaying the optimal operation promotion data to the user to browse so as to improve the conversion rate of the operation data. According to the method and the system, the basic data acquisition of the user and the operation data of the promotion are closely combined, so that the user data and the operation promotion data can be subjected to optimal intimacy matching, and the conversion effect of the operation data is further improved.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to an intelligent data matching recommendation method, device and storage medium. Background Art

[0002] Data is the core capital of an enterprise. In order to efficiently convert data into profits, enterprises need to accurately differentiate data and push different information to target users in a targeted manner. Therefore, enterprises must be able to accurately analyze users' consumption preferences and their consumption tendencies for current marketing products.

[0003] Most of the current advertising promotions only apply some type labels to users, and do not best match the user's consumption preferences with the pushed advertising information. This will result in a very low conversion rate for clicks and views of the pushed information.

[0004] To this end, this application specifically proposes an intelligent data matching recommendation method to solve the above technical problems. Summary of the invention

[0005] The main purpose of the present invention is to provide an intelligent data matching recommendation method to solve the technical problems raised in the background technology.

[0006] The present invention adopts the following technical solutions to solve the above technical problems:

[0007] An intelligent data matching recommendation method comprises the following steps:

[0008] S1. Place a point on the user's browsing page and transmit the user's consumption behavior data to the server for data collection;

[0009] S2. The acquired user consumption data is cleaned and processed according to the RFM model;

[0010] S3. After data cleaning, the FMC algorithm is used to perform fuzzy clustering on the data, and the optimal center value of the fuzzy clustering is found in the data after data cleaning by improving the greedy algorithm;

[0011] S4. Calculate the optimal operation and promotion data for the current user and display it to the user for browsing to improve the conversion rate of operation data.

[0012] Preferably, the RFM model in step S2 includes: time statistics from the last consumption of the user to the present, the number of consumptions of the user within a certain period, and the consumption amount of the user within a certain period;

[0013] The time statistics from the user's last consumption to the present time are used as an indicator to measure the time of the customer's most recent purchase. The closer the consumption behavior is to the present time, the higher the customer's loyalty to the brand. Otherwise, there is a risk of churn.

[0014] Preferably, the specific operation process of finding the optimal center value of fuzzy clustering in step S3 includes:

[0015] S31. Preprocess the collected data to improve the quality of the data;

[0016] S32. Through feature engineering, extract key features related to consumer goals in marketing plans and combine them to generate new composite features to enhance the performance of the RFM model;

[0017] S33. Divide the entire marketing plan into several stages, each stage corresponds to a sub-problem, that is, there is a sub-problem P 1 , P 2 , P 3 ,…,P m For each sub-problem, the improved greedy algorithm is used to process the sub-problem data and obtain the solution result solve(P i ),exist:

[0018] S P =merge(solve(P 1 ), solve(P 2 ), solve(P 3 ),…,solve(P m ))

[0019] The combined result S P It is the optimal center of fuzzy clustering;

[0020] S34. For all solve(P i ), and use the following expression for iterative optimization:

[0021] solve′(P i+1 )=solve(P i+1 )+max(V[P i ]+V[P i+1 ],V[P i+1 ])

[0022] Where m-1≥i≥0, V[P i+1 ] is represented as subproblem P i+1 Similar to the previous subproblem P i Solve the relevance of the results, V[P 0 ]=V[P 1 ]=0.

[0023] Preferably, the specific operation steps of the pretreatment in step S31 include:

[0024] Use the Z-score method to identify and remove illogical or obviously erroneous data points in the data set;

[0025] Estimate missing data by means and perform data filling;

[0026] The padded data is normalized or standardized to ensure that all features are at the same magnitude for easy comparison and analysis.

[0027] Preferably, the specific operation steps of step S32 include:

[0028] The threshold K is preset, and the Pearson correlation coefficient is used to select features with high correlation. There is an expression:

[0029]

[0030] Among them, r xy is the correlation coefficient between features x and y, and are the means of x and y respectively;

[0031] Select the features with the highest correlation coefficient and combine them as new composite features. Recalculate the feature correlation coefficients of the new composite features and other original features.

[0032] The calculation is repeated iteratively until the correlation coefficients of all features are less than the threshold K.

[0033] Preferably, the specific operation flow of improving the greedy algorithm to process sub-problem data in step S33 includes:

[0034] L1. Use a distributed architecture to divide the data set into multiple subsets, each subset is assigned to a different computing node, each node runs the greedy algorithm independently, processes multiple candidate solutions in parallel, and merges the results. There is an evaluation index J i :

[0035]

[0036] Among them C i is the candidate point set of the i-th node, and f(c) is the screening function for condition c;

[0037] L2. Evaluation index J i After generation, the annealing simulation algorithm is used to search and find the local optimal solution J A , at this time J A That is the subproblem P i The optimal solution for data processing (Pi ).

[0038] Preferably, the standard function expression for selecting the local optimal solution during the search process of the annealing simulation algorithm in the L2 step is:

[0039] h(c)=ω 1 g(cost)+ω 2 g(revenue)+ω 3 g(risk)

[0040] where ω 1 ,ω 2 ,ω 3 For the adjustment variable, g(d) is expressed as an evaluation function for the conditional parameter d.

[0041] An intelligent data matching recommendation system is used to implement any of the above-mentioned intelligent data matching recommendation methods, and the system architecture includes:

[0042] The data collection layer collects user browsing content data through the K8S data collection cluster;

[0043] The data processing layer stores the RFM model and the FMC algorithm. The RFM model is formed after data cleaning based on the collected data. The FMC algorithm is used to process the data collected by the RFM model and select the optimal solution that meets the preset conditions.

[0044] The business logic layer internally stores a selected data model and a greedy algorithm, wherein the greedy algorithm is used to adjust the data processing method of the FMC algorithm, and the selected data model is composed of the optimal solution obtained by the FMC algorithm;

[0045] The data presentation layer is used to push the data of the selected data model to the specified user.

[0046] On the other hand, the present invention further discloses a computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor executes the steps of the above method.

[0047] On the other hand, the present invention further discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the above method.

[0048] It can be seen from the above technical solution that the present invention provides an intelligent data matching recommendation method. Compared with the prior art, the present invention has the following advantages:

[0049] 1. The present invention closely combines the basic data collection of users with the promoted operational data, uses the FMC algorithm to fuzzy cluster the data, and uses the improved greedy algorithm to find the optimal center value of the fuzzy clustering in the data, so as to make the best intimacy match between user data and operational promotion data, thereby improving the conversion effect of operational data.

[0050] 2. The present invention performs unified data cleaning processing on the collected data by constructing an RFM model, and can stratify the data based on aspects such as user activity, purchasing preferences, and consumption power, thereby facilitating data stratification processing, and can deduce the marketing direction for users, further optimize the formulation of marketing strategies, and improve the page click rate and browsing rate of designated users after data conversion.

[0051] 3. The present invention adopts a centralized data standard method and combines it with a distributed node architecture to perform greedy algorithm operations. It can select the centralized marketing plan that users are most interested in and has the highest transaction rate from each data of all current marketing plans, and make overall scheduling of the marketing plan selection in combination with the marketing plans in the current marketing library to obtain the optimal recommendation of two-way data.

[0052] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present invention, nor is it intended to limit the scope of the present invention. Other features of the present invention will become easy to understand through the following description. Of course, it is not necessary to achieve all of the advantages described above simultaneously for any product implementing the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] The drawings constituting a part of the present application are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0054] Figure 1 It is a schematic diagram of the overall process of the method of the present invention;

[0055] Figure 2 It is a schematic diagram of the framework structure of the system of the present invention;

[0056] Figure 3 This is an example diagram of the greedy algorithm selection process of the present invention. DETAILED DESCRIPTION

[0057] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. In the absence of conflict, the embodiments in this application and the features in the embodiments can be combined with each other. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0058] In the embodiment, see Figures 1 to 3 .

[0059] like Figure 1 As shown, the embodiment of the present invention proposes an intelligent data matching recommendation method, which closely combines the basic data collection of users with the promoted operational data, uses the FMC algorithm to perform fuzzy clustering on the data, and uses the improved greedy algorithm to find the optimal center value of the fuzzy clustering in the data, so that the user data and the operational promotion data can be matched with the best intimacy, thereby improving the conversion effect of the operational data. Specifically, the method includes the following steps:

[0060] S1. Place a point on the user's browsing page and transmit the user's consumption behavior data to the server for data collection;

[0061] S2. The acquired user consumption data is cleaned and processed according to the RFM model;

[0062] The RFM model here is composed of three data:

[0063] R (Recentness): The time statistics from the user's last consumption to the current time (date interval / day);

[0064] F (Frequency): the number of times a user consumes within a certain period (statistical period / times);

[0065] M (Monetary): the amount of money a user spends in a certain period (statistical period / yuan);

[0066] The time from the last consumption to the present is used as an indicator to measure the time of the customer's most recent purchase. The closer the consumption behavior is to the present, the higher the customer's loyalty to the brand. Otherwise, there may be a risk of loss.

[0067] In addition, the RFM model can be used to uniformly clean the collected data, so as to better stratify the data based on user activity, purchase preferences, consumption power, etc.

[0068] At this time, by building an RFM model to perform unified data cleaning processing on the collected data, the data can be stratified and divided according to the user's activity, purchasing preferences, consumption power, etc., which facilitates data stratification processing, can deduce the marketing direction for users, further optimize the formulation of marketing strategies, and improve the page click-through rate and browsing rate of designated users after data conversion.

[0069] S3. After data cleaning, the FMC algorithm is used to perform fuzzy clustering on the data, and the optimal center value of the fuzzy clustering is found in the data after data cleaning by improving the greedy algorithm;

[0070] The FMC algorithm here is essentially a local search hill climbing method, which randomly selects the center value of the initial cluster for calculation and is easy to fall into a local extreme point. Therefore, it is very sensitive to initialization, especially when the data volume is large and the data dimensions are large, it is difficult to search for the desired optimal result. To address this problem, the present application combines an improved greedy algorithm and takes the best or optimal (i.e., most favorable) choice in the current state in each step of selection, thereby hoping to lead to the best or optimal algorithm in the end. The basic principle is to make the best choice at each step, without considering the overall optimal solution, but to achieve the overall optimal solution through a series of local optimal choices;

[0071] Specifically, the specific operation process of finding the optimal center value of fuzzy clustering includes:

[0072] S31. Preprocess the collected data to improve the quality of the data. The specific steps include:

[0073] (1) Use the Z-score method to identify and remove illogical or obviously erroneous data points in the data set;

[0074] (2) Estimate missing data by the mean and perform data filling;

[0075] (3) The padded data is normalized or standardized to ensure that all features are at the same magnitude for easy comparison and analysis;

[0076] S32. Through feature engineering, extract key features related to consumer goals in the marketing plan and combine them to generate new composite features to enhance the performance of the RFM model. The specific steps include:

[0077] (a) Preset the threshold K and use the Pearson correlation coefficient to select features with high correlation. There is an expression:

[0078]

[0079] Among them, r xyis the correlation coefficient between features x and y, and are the means of x and y respectively;

[0080] (b) Select the features with the highest correlation coefficient and combine them as a new composite feature. Recalculate the feature correlation coefficient of the new composite feature and other original features;

[0081] (c) Repeat the calculation iteratively until the correlation coefficient of all features is less than the threshold K

[0082] S33. Divide the entire marketing plan into several stages, each stage corresponds to a sub-problem, that is, there is a sub-problem P 1 , P 2 , P 3 ,…,P m For each sub-problem, the improved greedy algorithm is used to process the sub-problem data and obtain the solution result solve(P i ),exist:

[0083] S P =merge(solve(P 1 ), solve(P 2 ), solve(P 3 ),…,solve(P m ))

[0084] The combined result S P It is the optimal center of fuzzy clustering;

[0085] It should be noted here that the marketing plan is broken down into multiple stages, and the improved greedy algorithm is used to make decisions in each stage, and adjustments are made in the subsequent stages based on the decision results of the previous stage; the conventional greedy selection process here refers to Figure 3 ;

[0086] At this point, it is necessary to further explain that the specific operation process of improving the greedy algorithm for sub-problem data processing includes:

[0087] L1. Use a distributed architecture to divide the data set into multiple subsets, each subset is assigned to a different computing node, each node runs the greedy algorithm independently, processes multiple candidate solutions in parallel, and merges the results. There is an evaluation index J i :

[0088]

[0089] Among them C i is the candidate point set of the i-th node, and f(c) is the screening function for condition c;

[0090] At this time, parallel computing technology is used to parallelize each decision step of the greedy algorithm to improve the efficiency of the algorithm. At the same time, in a distributed computing environment, the calculation tasks of the greedy algorithm are distributed to multiple nodes, which can facilitate the processing of large-scale data and complex problems.

[0091] L2. Evaluation index J i After generation, the annealing simulation algorithm is used to search and find the local optimal solution J A , at this time J A That is the subproblem P i The optimal solution for data processing (P i );

[0092] Furthermore, the standard function expression for selecting the local optimal solution during the search process of the annealing simulation algorithm is:

[0093] h(c)=ω 1 h(cost)+ω 2 h(return)+ω 3 h(risk)

[0094] where ω 1 ,ω 2 ,ω 3 To adjust the variables, g(d) is expressed as the evaluation function for the condition parameter d;

[0095] Here, by introducing multi-objective optimization technology into the greedy algorithm, multiple objectives including risk, benefit, and cost can be considered to find a more comprehensive solution;

[0096] Furthermore, a feedback mechanism is introduced here to dynamically adjust the greedy algorithm according to the performance in actual applications. 1 ,ω 2 ,ω 3 Parametric variables can balance the relationship between multiple objectives.

[0097] S34. For all solve(P i ), and use the following expression for iterative optimization:

[0098] solve′(P i+1 )=solve(P i+1 )+max(V[P i ]+V[P i+1 ],V[P i+1 ])

[0099] Where m-1≥i≥0, V[P i+1 ] is represented as subproblem P i+1 Similar to the previous subproblem P i Solve the relevance of the results, V[P 0]=V[P 1 ]=0;

[0100] Here, the performance of the greedy algorithm is gradually optimized through multiple iterations, which can make it perform better in practical applications;

[0101] At this point, a backtracking mechanism can be introduced to allow the system to go back to the previous stage and reselect in some cases, thus avoiding falling into a local optimum.

[0102] In summary, by adopting a centralized data standard method and combining it with a distributed node architecture for greedy algorithm calculations, we can select the centralized marketing plan that users are most interested in and has the highest transaction rate from each piece of data of all current marketing plans, and make overall scheduling of the marketing plan selection in combination with the marketing plans in the current marketing library to obtain the best recommendation for two-way data.

[0103] Therefore, this method can effectively improve the performance of the greedy algorithm and the quality of the solution, making it perform well in a wider range of application scenarios. At the same time, it combines the advantages of multiple algorithms to obtain all the better solutions, making data push richer.

[0104] S4. Calculate the optimal operation and promotion data for the current user and display it to the user for browsing to improve the conversion rate of operation data.

[0105] On the other hand, the present invention also discloses an intelligent data matching recommendation system, which is used to implement the above intelligent data matching recommendation method. Figure 2 As shown, the system architecture includes:

[0106] The data collection layer collects user browsing content data through the K8S data collection cluster;

[0107] The data processing layer stores the RFM model and FMC algorithm. The RFM model is formed after data cleaning based on the collected data. The FMC algorithm is used to process the data collected by the RFM model and select the optimal solution that meets the preset conditions.

[0108] The business logic layer stores the selected data model and the greedy algorithm. The greedy algorithm is used to adjust the data processing method of the FMC algorithm. The selected data model consists of the optimal solution obtained by the FMC algorithm.

[0109] The data presentation layer is used to push the data of the selected data model to the specified user.

[0110] On the other hand, the present invention further discloses a computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor executes the steps of the above method.

[0111] On the other hand, the present invention further discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the above method.

[0112] In another embodiment provided by the present application, a computer program product including instructions is also provided, which, when executed on a computer, enables the computer to execute any one of the intelligent data matching recommendation methods in the above embodiments.

[0113] It is understandable that the system provided by the embodiment of the present invention corresponds to the method provided by the embodiment of the present invention, and the explanation, examples and beneficial effects of the relevant contents can refer to the corresponding parts in the above method.

[0114] The embodiment of the present application also provides an electronic device, including a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other through the communication bus.

[0115] Memory, used to store computer programs;

[0116] The processor is used to implement the above-mentioned intelligent data matching recommendation method when executing the program stored in the memory.

[0117] The communication bus mentioned in the above electronic device may be a Peripheral Component Interconnect (PCI) bus or an Extended Industrial Standard Architecture (EISA) bus, etc. The communication bus may be divided into an address bus, a data bus, a control bus, etc.

[0118] The communication interface is used for communication between the above electronic device and other devices.

[0119] The memory may include a random access memory (RAM) or a non-volatile memory (NVM), such as at least one disk memory. Optionally, the memory may also be at least one storage device located away from the aforementioned processor.

[0120] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0121] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) mode. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server, data center, etc. that contains one or more available media integrated. The available medium can be a magnetic medium, (e.g., a floppy disk, a hard disk, a tape), an optical medium, or a semiconductor medium, etc.

[0122] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

[0123] In addition, it should be noted that if the embodiments of the present invention involve directional indications (such as up, down, left, right, front, back, etc.), the directional indications are only used to explain the relative position relationship, movement status, etc. between the components in a certain specific posture. If the specific posture changes, the directional indication will also change accordingly.

[0124] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of the present invention, the descriptions of "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the meaning of "and / or" appearing in the full text includes three parallel schemes. Taking "A and / or B" as an example, it includes scheme A, or scheme B, or schemes that A and B meet at the same time. In addition, in the embodiments of the present invention, "multiple" refers to more than two. In addition, the technical solutions between the various embodiments can be combined with each other, but it must be based on the ability of ordinary technicians in the field to implement. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

Claims

1. An intelligent data matching recommendation method, characterized in that: The following steps are involved: S1. Place a point on the user's browsing page and transmit the user's consumption behavior data to the server for data collection; S2. The acquired user consumption data is cleaned and processed according to the RFM model; S3. After data cleaning, the FMC algorithm is used to perform fuzzy clustering on the data, and the optimal center value of the fuzzy clustering is found in the data after data cleaning by improving the greedy algorithm; S4. Calculate the optimal operation and promotion data for the current user and display it to the user for browsing to improve the conversion rate of operation data.

2. The intelligent data matching recommendation method according to claim 1, characterized in that: The RFM model in step S2 includes: time statistics from the last consumption of the user to the present, the number of consumptions of the user in a certain period, and the consumption amount of the user in a certain period; The time statistics from the user's last consumption to the present time are used as an indicator to measure the time of the customer's most recent purchase. The closer the consumption behavior is to the present time, the higher the customer's loyalty to the brand. Otherwise, there is a risk of churn.

3. The intelligent data matching recommendation method according to claim 1, characterized in that: The specific operation process of finding the optimal center value of fuzzy clustering in step S3 includes: S31. Preprocess the collected data to improve the quality of the data; S32. Through feature engineering, extract key features related to consumer goals in marketing plans and combine them to generate new composite features to enhance the performance of the RFM model; S33. Divide the entire marketing plan into several stages, each stage corresponds to a sub-problem, that is, there are sub-problems P1, P2, P3, ..., P m For each sub-problem, the improved greedy algorithm is used to process the sub-problem data and obtain the solution result solve(P i ),exist: S P =merge(solve(P1),solve(P2),solve(P3),…,solve(P m )) The combined result S P It is the optimal center of fuzzy clustering.

4. The intelligent data matching recommendation method according to claim 3, characterized in that: The S3 step also includes: S34. For all solve(P i ), and use the following expression for iterative optimization: solve′(P i+1 )=solve(P i+1 )+max(V[P i ]+V[P i+1 ],V[P i+1 ]) Where m-1≥i≥0, V[P i+1 ] is represented as subproblem P i+1 Similar to the previous subproblem P i The correlation of the solution results is V[P0]=V[P1]=0.

5. The intelligent data matching recommendation method according to claim 3, characterized in that: The specific operation steps of step S32 include: The threshold K is preset, and the Pearson correlation coefficient is used to select features with high correlation. There is an expression: Among them, r xy is the correlation coefficient between features x and y, and are the means of x and y respectively; Select the features with the highest correlation coefficient and combine them as new composite features. Recalculate the feature correlation coefficients of the new composite features and other original features. The calculation is repeated iteratively until the correlation coefficients of all features are less than the threshold K.

6. The intelligent data matching recommendation method according to claim 3, characterized in that: The specific operation flow of improving the greedy algorithm to process sub-problem data in step S33 includes: L1. Use a distributed architecture to divide the data set into multiple subsets, each subset is assigned to a different computing node, each node runs the greedy algorithm independently, processes multiple candidate solutions in parallel, and merges the results. There is an evaluation index J i : Among them C i is the candidate point set of the i-th node, and f(c) is the screening function for condition c; L2. Evaluation index J i After generation, the annealing simulation algorithm is used to search and find the local optimal solution J A , at this time J A This is the subproblem P i The optimal solution for data processing (P i ).

7. The intelligent data matching recommendation method according to claim 6, characterized in that: The standard function expression for selecting the local optimal solution during the search process of the annealing simulation algorithm in the L2 step is: h(c)=ω1g(cost)+ω2g(benefit)+ω3g(risk) Where ω1, ω2, and ω3 are adjustment variables, and g(d) is the evaluation function for the conditional parameter d.

8. An intelligent data matching recommendation system, used to implement the intelligent data matching recommendation method according to any one of claims 1 to 7, characterized in that: The system architecture includes: The data collection layer collects user browsing content data through the K8S data collection cluster; The data processing layer stores the RFM model and the FMC algorithm. The RFM model is formed after data cleaning based on the collected data. The FMC algorithm is used to process the data collected by the RFM model and select the optimal solution that meets the preset conditions. The business logic layer internally stores a selected data model and a greedy algorithm, wherein the greedy algorithm is used to adjust the data processing method of the FMC algorithm, and the selected data model is composed of the optimal solution obtained by the FMC algorithm; The data presentation layer is used to push the data of the selected data model to the specified user.

9. A computer-readable storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a processor, the processor is caused to perform the steps of the method according to any one of claims 1 to 8.

10. A computer device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the method according to any one of claims 1 to 8.

Images