Customer data analysis system and method based on big data

Abstract

The invention discloses a customer data analysis system and method based on big data. The customer data analysis system comprises an acquisition module, a storage module, an analysis module and a comparison module, the acquisition module acquires server information and user information, the server information comprises server push information and online information, the user information comprises user online information and benefit conversion information, and the information is transmitted to the storage module after acquisition; the storage module stores the information and transmits the information to the analysis module; and the analysis module receives the data, and transmits an accurate evaluation coefficient generated by server push information and online information and a user evaluation coefficient generated by user retention information and benefit conversion information to the comparison module for comparison. According to the video pushing method and device, the accuracy degree of video pushing is analyzed, when the accuracy degree of video pushing becomes poor, video pushing is updated in advance, then the importance degree of the user is analyzed, video pushing is conducted on the user with the high importance degree in a preferential batch mode, and the experience feeling of important users is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of customer data analysis, and in particular to a customer data analysis system and method based on big data. Background Art

[0002] Big data refers to a large, diverse and rapidly growing collection of data. It is characterized by high speed, high volume, high diversity and high value density. The value of big data is mainly reflected in its ability to analyze and mine information. Through the collection, storage, processing and analysis of big data, hidden patterns, trends and associations can be revealed, thereby providing deeper insights and more accurate predictions for decision-making.

[0003] A customer data analysis system is a system used to implement and manage customer data. The system usually combines customer data information and data analysis methods to provide functions for monitoring, analyzing and managing customer data.

[0004] The prior art has the following deficiencies: the prior art is a customer data analysis system based on big data, in which the video platform operator pushes videos based on customer data. When the accuracy of the pushed videos is low, it cannot meet the customer's viewing needs, resulting in different playback expectations of the pushed videos. The videos need to be pushed and updated. When the video push is updated, it is not clear which customers should be given priority for the video push update. It is not possible to push videos in batches with different priorities according to the importance of the customers, and it is not possible to provide good services to customers with high importance in a timely manner, which results in a poor experience for important customers and affects operational efficiency.

[0005] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0006] The purpose of the present invention is to provide a customer data analysis system and method based on big data. The present invention conducts a comprehensive analysis on the accuracy of video push. When the accuracy of the monitored video push deteriorates, the video push is updated in advance. Then the importance of the user is analyzed, and priority batches of videos are pushed to users with high importance, so as to solve the problems in the above-mentioned background technology.

[0007] In order to achieve the above-mentioned object, the present invention provides the following technical solutions: a customer data analysis system based on big data, comprising a collection module, a storage module, an analysis module and a comparison module;

[0008] A collection module collects server information and user information. The server information includes server push information and online information. The user information includes user online information and benefit conversion information. After collection, the server information and user information are transmitted to the storage module.

[0009] The storage module stores the data collected by the collection module, and stores threshold information and user history information, and sends the stored data to the analysis module;

[0010] The analysis module receives the data sent by the storage module and the comparison module, generates an accurate evaluation coefficient based on the server push information and online information, transmits the accurate evaluation coefficient to the comparison module, analyzes the user retention information and benefit conversion information according to the signal sent by the comparison module, and sends the analysis results to the comparison module;

[0011] The comparison module receives the data sent by the analysis module, compares the accuracy evaluation coefficient with the accuracy evaluation threshold, transmits the generated signal to the analysis module, and compares the received analysis results.

[0012] Preferably, the server push information includes the request push rate, which is marked as TS i , online information includes the average online time value, which is marked as SC i , user retention information includes user retention rate, and is marked as LC i , the conversion information includes the number of ad interactions and is marked as HD i .

[0013] Preferably, the logic for requesting push rate acquisition is as follows:

[0014] Get the user's click rate, D i =(a / A)*100%, where A represents the total number of people who received push notifications and a represents the number of people who clicked on the push notifications;

[0015] Get the user's feedback rate, F k =(b / a)*100%, where a represents the number of people who click on the push information, and b represents the number of people who give feedback on the clicked push information;

[0016] Get the user's click rate D i , feedback rate F k , the request push rate is calculated by the formula, based on the formula: TS i =D i *F k .

[0017] Preferably, the number of all users Gs within time t is obtained, the total online time Ts of all users within time t is obtained, and the average online time value is calculated by a formula based on the number of users Gs and the total online time Ts of the users, and the formula is: SCi =Ts / Gs.

[0018] Preferably, the user retention information and the benefit conversion information are analyzed based on the user retention rate in the user retention information and the number of advertisement interactions in the benefit conversion information to generate a user evaluation coefficient.

[0019] Preferably, the logic for obtaining the user retention rate is as follows:

[0020] Set the calculated retention cycle duration t, divide the historical time period into x time periods with t as the retention cycle duration, and set a rated usage duration Si to obtain the actual usage duration SSi of the users in each divided time period. Take the ratio of the rated usage duration to the actual usage duration in each time period as the duration proportion value ZBi=SSi / Si, and perform statistics on the duration proportion values ​​of each time period to obtain the user retention rate. The specific calculation formula is as follows:

[0021]

[0022] ; Wherein, i is a positive integer.

[0023] Preferably, the logic for obtaining the number of ad interactions is as follows:

[0024] Obtain the user's advertising interaction behavior, create a unique tracking tag or parameter for each advertisement, and embed it into the ad link or ad creative to identify the user's interactive behavior. When the user interacts with the advertisement, the tracking tag is used to collect relevant data, analyze the relevant data, and accumulate and count the satisfied data as the number of ad interactions HD i .

[0025] Preferably, the accurate evaluation coefficient is compared with the accurate evaluation threshold, and the generated signal is transmitted to the analysis module. The specific process is as follows:

[0026] If the accuracy evaluation coefficient is greater than the accuracy evaluation threshold, a push stability signal is generated through the comparison module;

[0027] If the accuracy assessment coefficient is less than or equal to the accuracy assessment threshold, a push warning signal is generated through the comparison module and the push warning signal is transmitted to the analysis module.

[0028] Preferably, the received analysis results are compared, and the specific comparison process is as follows:

[0029] If the user evaluation coefficient is greater than the user evaluation threshold, the user is marked as an important user;

[0030] If the user evaluation coefficient is less than or equal to the user evaluation threshold, the user is marked as a normal user.

[0031] A customer data analysis method based on big data comprises the following steps:

[0032] Collect server information and user information. Server information includes server push information and online information. User information includes user online information and benefit conversion information.

[0033] Generate an accurate evaluation coefficient from the server push information and online information, and compare the accurate evaluation coefficient with an accurate evaluation threshold;

[0034] According to the comparison results, the user online information and benefit conversion information are used to generate a user evaluation coefficient;

[0035] Compare the generated user evaluation coefficient with the user evaluation threshold and mark the user as an important user or an ordinary user;

[0036] According to the size of the user evaluation coefficient, priority video updates are given to important users.

[0037] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0038] The present invention establishes an accurate evaluation coefficient by collecting server push information and online information, compares the accurate evaluation coefficient with an accurate evaluation threshold, generates a push stability signal and a push warning signal according to the comparison result, updates the push video when the push warning signal is generated, and in the update of the push video, establishes a user evaluation coefficient through the collected user retention information and benefit conversion information, analyzes the user, compares the user evaluation coefficient of the user with the user evaluation threshold, marks the user as an important user and an ordinary user, implements different measures according to the comparison result, gives priority to video push to important users, and enables important users to get a better video experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0040] Figure 1 The module diagram of a customer data analysis system and method based on big data of the present invention.

[0041] Figure 2 The present invention is a method flow chart of a customer data analysis system and method based on big data. DETAILED DESCRIPTION

[0042] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of the present disclosure will be more comprehensive and complete, and the concept of the example embodiments will be fully conveyed to those skilled in the art.

[0043] Embodiment 1: The present invention provides Figure 1 A customer data analysis system based on big data is shown, comprising a collection module, a perception model building module, a comparison module and an analysis module;

[0044] server,

[0045] A collection module collects server information and user information. The server information includes server push information and online information. The user information includes user online information and benefit conversion information. After collection, the server information and user information are transmitted to the storage module.

[0046] The storage module stores the data collected by the collection module, and stores threshold information and user history information, and sends the stored data to the analysis module;

[0047] The analysis module generates an accurate evaluation coefficient from the server push information and online information in the server information, and transmits the accurate evaluation coefficient to the comparison module;

[0048] The server push information includes the request push rate, and the online information includes the average online time. After collection, the collection module marks the request push rate and server overload times as TS i , SC i ;

[0049] Video platform operators push videos to users based on their preferences, but real-time updates of video pushes cause the server to be fully loaded for a long time, increasing costs. However, there are limitations to regular updates of video pushes, which can easily cause user loss. Therefore, based on the collection of big data, automatic updates of video pushes can be achieved, thus achieving a better solution.

[0050] The request push rate refers to the number of requests sent by users to the short video server during the short video browsing process. The short video performs different operations on the user based on the request. The requests include video playback requests, video upload requests, like requests, comment requests, search requests and other requests that require the participation of the server. These requests represent different behaviors and needs of users in video applications. By analyzing and processing these requests, the platform can provide personalized recommended content, enhance user interactive experience, maintain platform order and security, etc. The process in which multiple users send requests at the same time is called a concurrent request process;

[0051] Concurrent requests refer to multiple requests received by the system at the same time in the same time period. In the video push scenario, when a large number of users initiate video push requests at the same time, concurrent requests will be generated. If the number of concurrent requests is too large, the server may not be able to process these requests in time, resulting in increased server pressure.

[0052] The server analyzes different requests and pushes them to users. For push of relevant videos and advertisements, when the server is in normal state, it can send the generated correct push information to users at the specified time.

[0053] The request push rate has an important impact on the server status evaluation. The request push rate is an indicator used to measure whether the server can correctly push after the user's request. A low request push rate means that the server push information does not meet the standard, that is, there is an error between the request and the push, which will lead to the following situations:

[0054] Irrelevant push notifications: Users receive push notifications that are irrelevant to their interests or preferences, which leads to disinterest or distraction, wasting server computing power and user time;

[0055] Inaccurate content: The content of the push notification does not match the user's actual needs, interests or behaviors, resulting in disinterest or dissatisfaction, causing the user to feel bored with the video;

[0056] Repeated push notifications: Users receive the same or similar push notifications, resulting in repeated and redundant information. Users receive too many push notifications, causing information overload and fatigue.

[0057] Therefore, obtaining the server's request push rate can be used to analyze the accuracy of pushed videos;

[0058] The logic for requesting push rate is as follows:

[0059] Get the user's click rate, which means the proportion of users who actually click to open the push after receiving it. i =(a / A)*100%, where A represents the total number of people who received push notifications and a represents the number of people who clicked on the push notifications;

[0060] Get the user's feedback rate, which indicates the proportion of users who give feedback on the pushed information, F k =(b / a)*100%, where a represents the number of people who click on the push information, and b represents the number of people who give feedback on the clicked push information;

[0061] Get the user's click rate D i , feedback rate F k , the request push rate is calculated by the formula, based on the formula: TS i =D i *Fk ;

[0062] It should be noted that feedback includes users’ actions such as liking, collecting, and commenting on the pushed information. When users perform the above actions, it is considered as feedback on the pushed information;

[0063] Online information includes average online time;

[0064] The average online duration refers to the average online duration of users in video applications. When users browse videos, the more accurate the video push is, the more interested they are in the video, and the longer they stay online, so the average online duration of users is greater.

[0065] The average online time of users indicates whether the recommended videos are attractive enough to users. When the average online time of users is too low, it may be caused by the following reasons:

[0066] Lack of interest: Users have low interest in the video content or platform, which fails to keep them on the app for a long time. This may be because the video content does not match the user's preferences, or the platform's content quality does not meet the user's expectations;

[0067] Unattractive content: The video’s content quality or relevance is not high enough to attract users to watch for a long time. The video may be poorly produced, have a boring storyline, or not match the user’s interests.

[0068] Therefore, the average online duration value SC is obtained i , the accuracy of pushed videos can be analyzed;

[0069] The logic for obtaining the average online duration value is as follows:

[0070] Get the number of all users Gs within time t, get the total online duration Ts of all users at different times within time t, and calculate the average online duration value based on the number of users Gs and the total online duration Ts of the users. The formula is: SC i =Ts / Gs;

[0071] It should be noted that the user's click rate, feedback rate, number of users and total online time of the user are data stored in the storage module after the data is aggregated. When obtaining, it is obtained through the set timestamp, user history information and other information. When obtaining the number of users, some users whose online time is illogical or whose online time is extremely short are first eliminated to reduce calculation errors and improve accuracy. The specific elimination criteria are set according to the actual scenario.

[0072] The analysis module generates an accurate evaluation coefficient from the server push information and the online information, and transmits the accurate evaluation coefficient to the comparison module;

[0073] The analysis module obtains the request push rate TS i , average online time value SC i After that, the accurate evaluation coefficient is generated and calibrated as Zp according to the following formula:

[0074] Where α and β are the request push rates TS i , Server overload times CS i The preset proportional coefficient, and α, β, are all greater than 0;

[0075] From the formula, we can see that the larger the request push rate and the larger the average online time value, that is, the larger the performance value of the accuracy evaluation coefficient Zp, the higher the accuracy of the server's video push and the more interested the user is in the pushed video content. The smaller the request push rate and the smaller the average online time value, that is, the smaller the performance value of the accuracy evaluation coefficient Zp, the lower the accuracy of the server's video push and the less interested the user is in the pushed video content.

[0076] The comparison module compares the generated accuracy evaluation coefficient with the accuracy evaluation threshold, generates a push stability signal and a push warning signal, and transmits the generated signals to the analysis module;

[0077] After the comparison module obtains the generated accuracy evaluation coefficient, it compares the generated accuracy evaluation coefficient with the accuracy evaluation threshold. If the accuracy evaluation coefficient is greater than the accuracy evaluation threshold, it indicates that the video pushed by the server is accurate, and a push stability signal is generated through the comparison module. If the accuracy evaluation coefficient is less than or equal to the accuracy evaluation threshold, it indicates that the server push accuracy is poor, and a push warning signal is generated through the comparison module, and the push warning signal is transmitted to the analysis module;

[0078] The analysis module, after receiving the push warning signal sent by the comparison module, analyzes the online users and classifies the users according to their importance based on the real-time analysis results;

[0079] User retention information includes user retention rate, and benefit conversion information includes the number of advertising interactions. The collection module marks the user retention rate and the number of advertising interactions as LC i , HD i ;

[0080] The analysis module generates a user evaluation coefficient from the user retention information and benefit conversion information of the user information, and transmits the user evaluation coefficient to the comparison module;

[0081] User retention rate is an indicator to measure user importance. It reflects the proportion of users who continue to use the platform over a period of time. A higher retention rate means that users are more sticky to the platform content and services and have higher importance. The user retention rate affects the platform in the following aspects:

[0082] User loyalty: A high retention rate indicates that users are highly loyal to the platform. They are willing to continue using the platform and are likely to become long-term active users, which is crucial for the long-term development of the application and the steady growth of the user base.

[0083] User value: Users with high retention rates usually have higher user value and are more likely to participate in paid activities, make purchases, participate in social interactions, etc., thus bringing more revenue and value to the platform;

[0084] Data analysis and personalized recommendations: A high retention rate means that the platform has more user data to analyze, thereby improving personalized recommendation algorithms and user experience, which helps provide more accurate and targeted content and recommendations, further improving user satisfaction and engagement;

[0085] User retention rate LC i The acquisition logic is as follows:

[0086] Set the calculated retention cycle duration t, divide the historical time period into x time periods with t as the retention cycle duration, and set a rated usage duration Si to obtain the actual usage duration SSi of the users in each divided time period. Take the ratio of the rated usage duration to the actual usage duration in each time period as the duration proportion value ZBi=SSi / Si, and perform statistics on the duration proportion values ​​of each time period to obtain the user retention rate. The specific calculation formula is as follows:

[0087]

[0088] ; Where i is a positive integer;

[0089] It should be noted that the retention period can be days, weeks or months, depending on business needs. No further restrictions will be made here. For the division of historical time periods, data can be collected based on the latest time or based on a specific time period.

[0090] The number of ad interactions has a significant impact on the user's judgment of importance. The number of ad interactions reflects the user's interest and engagement. The user's interactive behavior on the ad, such as clicks, shares, comments, etc., can reflect the user's attention and engagement with the ad. Actively interacting users may have a greater impact on the advertiser's brand promotion and effectiveness, and thus the user's importance is higher. A higher number of ad interactions has an impact on the following aspects:

[0091] Advertising effectiveness evaluation: Ad interaction counts can be used to evaluate the effectiveness of an ad. If the number of user interactions is high, it usually means that the ad is able to attract the user's attention and prompt them to interact with the ad. This is a positive signal for advertisers, indicating that the ad is able to effectively convey information and elicit a positive response from users;

[0092] User value: The number of ad interactions can indirectly reflect the user's purchase intention and brand loyalty. If a user interacts with an ad multiple times, it may mean that they are interested in the product or brand in the ad and may become a potential consumer or loyal user. This is of great significance to advertisers because these users may bring more commercial value and returns;

[0093] Personalized recommendations and targeted advertising: Ad interaction times can provide valuable data for personalized recommendations and targeted advertising. By analyzing users' ad interaction behaviors, the platform can understand users' interests, preferences, and consumption habits, thereby providing more accurate and personalized ad recommendations. This helps improve user experience and increase ad click-through rates and conversion rates;

[0094] The more times users interact with ads, the higher the advertising profits for video platform operators will be, making the video platform more profitable and with higher advertising conversion rates, thus receiving more advertising business promotions.

[0095] Ad interactions HD i The acquisition logic is as follows:

[0096] Obtain the user's advertising interaction behavior, create a unique tracking tag or parameter for each advertisement, and embed it into the ad link or ad creative to identify the user's interactive behavior. When the user interacts with the advertisement, the tracking tag is used to collect relevant data, analyze the relevant data, and accumulate and count the satisfied data as the number of ad interactions HD i ;

[0097] It should be noted that advertising interaction behaviors can be clicking on ads, watching advertising videos, filling out questionnaires, etc. According to the needs and the functions of the advertising platform, select the appropriate interactive behavior indicators. The tracking tag will collect relevant data to understand the user's interactive behavior based on the information in the tag, such as the number of ad clicks, video views, etc.

[0098] The analysis module obtains the user retention rate LC i , Ad Interactions HD i , and generate the user evaluation coefficient, which is calibrated as Yg, based on the formula:

[0099] Yg=μ*LC i+θ*HD i In the formula, μ and θ are the user retention rates LC respectively. i , Ad Interactions HD i The preset proportional coefficient, and μ and θ are both greater than 0;

[0100] From the formula, we can see that the greater the user retention rate and the greater the number of ad interactions, that is, the greater the performance value of the user evaluation coefficient Yg, the more important the user is; the smaller the user retention rate and the smaller the number of ad interactions, that is, the smaller the performance value of the user evaluation coefficient Yg, the less important the user is;

[0101] The comparison module compares the obtained user evaluation coefficient with the user evaluation threshold to analyze the importance of the user;

[0102] After the comparison module obtains the user evaluation coefficient, the generated user evaluation coefficient is compared with the user evaluation threshold;

[0103] If the user evaluation coefficient is greater than the user evaluation threshold, the user is marked as an important user, indicating that the user is more important. When the video push update is performed, the video push update is given priority;

[0104] If the user evaluation coefficient is less than or equal to the user evaluation threshold, the user is marked as a common user, indicating that the user is less important and priority video push updates may not be performed when video push updates are performed;

[0105] A ranking table may be generated in reverse order according to the size of the user evaluation coefficient, so that priority video push is performed according to the ranking table.

[0106] The present invention establishes an accurate evaluation coefficient by collecting server push information and online information, compares the accurate evaluation coefficient with an accurate evaluation threshold, generates a push stability signal and a push warning signal according to the comparison result, updates the push video when the push warning signal is generated, and in the update of the push video, establishes a user evaluation coefficient through the collected user retention information and benefit conversion information, analyzes the user, compares the user evaluation coefficient of the user with the user evaluation threshold, marks the user as an important user and an ordinary user, implements different measures according to the comparison result, gives priority to video push to important users, and enables important users to get a better video experience.

[0107] The present invention provides Figure 2 A customer data analysis method based on big data is shown, comprising the following steps:

[0108] Collect server information and user information. Server information includes server push information and online information. User information includes user online information and benefit conversion information.

[0109] Generate an accurate evaluation coefficient from the server push information and online information, and compare the accurate evaluation coefficient with an accurate evaluation threshold;

[0110] According to the comparison results, the user online information and benefit conversion information are used to generate a user evaluation coefficient;

[0111] Compare the generated user evaluation coefficient with the user evaluation threshold and mark the user as an important user or an ordinary user;

[0112] According to the size of the user evaluation coefficient, priority video updates are given to important users.

[0113] A customer data analysis method based on big data provided in an embodiment of the present invention is implemented by the above-mentioned customer data analysis system based on big data. The specific method and process of the customer data analysis method based on big data are detailed in the embodiment of the above-mentioned customer data analysis system based on big data, which will not be repeated here.

[0114] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0115] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

[0116] It should be noted that, in this article, if there are relational terms such as first and second, etc., they are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0117] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0118] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0119] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0120] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0121] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0122] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A customer data analysis system based on big data, characterized in that: It includes an acquisition module, a storage module, an analysis module and a comparison module; A collection module collects server information and user information. The server information includes server push information and online information. The user information includes user online information and benefit conversion information. After collection, the server information and user information are transmitted to the storage module. The storage module stores the data collected by the collection module, and stores threshold information and user history information, and sends the stored data to the analysis module; The analysis module receives the data sent by the storage module and the comparison module, generates an accurate evaluation coefficient based on the server push information and online information, transmits the accurate evaluation coefficient to the comparison module, analyzes the user retention information and benefit conversion information according to the signal sent by the comparison module, and sends the analysis results to the comparison module; The comparison module receives the data sent by the analysis module, compares the accuracy evaluation coefficient with the accuracy evaluation threshold, transmits the generated signal to the analysis module, and compares the received analysis results.

2. A customer data analysis system based on big data according to claim 1, characterized in that: Server push information includes the request push rate, which is marked as TS i , online information includes the average online time value, which is marked as SC i , user retention information includes user retention rate, and is marked as LC i , the conversion information includes the number of ad interactions and is marked as HD i .

3. A customer data analysis system based on big data according to claim 2, characterized in that: The logic for requesting push rate is as follows: Get the user's click rate, D i =(a / A)*100%, where A represents the total number of people who received push notifications and a represents the number of people who clicked on the push notifications; Get the user's feedback rate, F k =(b / a)*100%, where a represents the number of people who click on the push information, and b represents the number of people who give feedback on the clicked push information; Get the user's click rate D i , feedback rate F k , the request push rate is calculated by the formula, based on the formula: TS i =D i *F k .

4. The customer data analysis system based on big data according to claim 2, characterized in that: The logic for obtaining the average online duration value is as follows: Get the number of all users Gs within time t, get the total online time Ts of all users within time t, and calculate the average online time value based on the number of users Gs and the total online time Ts of the users. The formula is: SC i =Ts / Gs.

5. The customer data analysis system based on big data according to claim 2, characterized in that: The user retention information and benefit conversion information are analyzed by analyzing the user retention rate in the user retention information and the number of advertising interactions in the benefit conversion information to generate a user evaluation coefficient.

6. The customer data analysis system based on big data according to claim 2, characterized in that: The logic for obtaining user retention rate is as follows: Set the calculated retention cycle duration t, divide the historical time period into x time periods with t as the retention cycle duration, and set a rated usage duration Si to obtain the actual usage duration SSi of the users in each divided time period. Take the ratio of the rated usage duration to the actual usage duration in each time period as the duration proportion value ZBi=SSi / Si, and perform statistics on the duration proportion values ​​of each time period to obtain the user retention rate. The specific calculation formula is as follows: ; Wherein, i is a positive integer.

7. The customer data analysis system based on big data according to claim 2, characterized in that: The logic for obtaining the number of ad interactions is as follows: Obtain the user's advertising interaction behavior, create a unique tracking tag or parameter for each advertisement, and embed it into the ad link or ad creative to identify the user's interactive behavior. When the user interacts with the advertisement, the tracking tag is used to collect relevant data, analyze the relevant data, and accumulate and count the satisfied data as the number of ad interactions HD i .

8. The customer data analysis system based on big data according to claim 7, characterized in that: The accurate evaluation coefficient is compared with the accurate evaluation threshold, and the generated signal is passed to the analysis module. The specific process is as follows: If the accuracy evaluation coefficient is greater than the accuracy evaluation threshold, a push stability signal is generated through the comparison module; If the accuracy assessment coefficient is less than or equal to the accuracy assessment threshold, a push warning signal is generated through the comparison module and the push warning signal is transmitted to the analysis module.

9. The customer data analysis system based on big data according to claim 8, characterized in that: And compare the received analysis results. The specific comparison process is as follows: If the user evaluation coefficient is greater than the user evaluation threshold, the user is marked as an important user; If the user evaluation coefficient is less than or equal to the user evaluation threshold, the user is marked as a normal user.

10. A customer data analysis method based on big data, implemented by a customer data analysis system based on big data as claimed in any one of claims 1 to 9, characterized in that: The following steps are involved: Collect server information and user information. Server information includes server push information and online information. User information includes user online information and benefit conversion information. Generate an accurate evaluation coefficient from the server push information and online information, and compare the accurate evaluation coefficient with an accurate evaluation threshold; According to the comparison results, the user online information and benefit conversion information are used to generate a user evaluation coefficient; Compare the generated user evaluation coefficient with the user evaluation threshold and mark the user as an important user or an ordinary user; According to the size of the user evaluation coefficient, priority video updates are given to important users.

Images