Delivery method and system based on new media customer preference data analysis

Abstract

The present invention discloses a delivery method and system based on customer preference data analysis of new media, which relates to the field of delivery analysis technology, including: based on the delivery purpose, obtaining at least one customer preference tag that affects the delivery purpose; collecting real-time customer preference data of each current customer preference tag of the new media platform as the benchmark preference data of the new media; obtaining the operation and maintenance strategy of the new media platform, and extracting several operation and maintenance features of the new media platform based on the operation and maintenance strategy of the new media platform; constructing an association model between each operation and maintenance feature and each customer preference tag; determining the data change model of each customer preference tag of the new media platform during the delivery period; and generating a delivery optimization strategy. The advantages of the present invention are: pre-adjusting the delivery strategy that is compatible with the future customer preference data, ensuring that the delivery content is highly matched with the preferences of the target user group during the actual delivery, thereby improving the delivery effect.

Description

Technical Field

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

[0002] As digital marketing becomes increasingly important, understanding and accurately grasping customer preferences has become the key to improving the competitiveness of enterprises. As the main channel for information dissemination and user interaction, new media platforms contain rich customer behavior data and preference information. By collecting and analyzing customer preference data through new media platforms, exploring and optimizing advertising delivery methods, the advertising effect and conversion rate can be improved, bringing greater commercial value to enterprises.

[0003] In the existing technology, the analysis of customer preference data for new media is usually limited to the current user group of the platform, while the delivery process is a future action, which targets the user group of the platform in the future. Changes in the operation and maintenance strategy of the new media platform usually lead to great changes in the user group of the new media platform, which in turn leads to changes in the customer preference data of the new media. However, the existing solutions lack predictive analysis of changes in customer preference data of such new media, resulting in the adjustment of delivery strategy lagging behind the changes in customer preference data of new media, affecting the actual delivery effect. Summary of the invention

[0004] In order to solve the above-mentioned technical problems, a delivery method and system based on new media customer preference data analysis is provided. This technical solution solves the problem that the above-mentioned existing solutions lack predictive analysis of changes in customer preference data of such new media, resulting in the adjustment of delivery strategies lagging behind changes in customer preference data of new media, affecting the actual delivery effect.

[0005] In order to achieve the above purpose, the technical solution adopted by the present invention is:

[0006] A delivery method based on new media customer preference data analysis, comprising:

[0007] Based on the delivery purpose, obtaining at least one customer preference tag that affects the delivery purpose;

[0008] Collect the real-time customer preference data of each customer preference tag on the new media platform as the benchmark preference data of the new media;

[0009] Obtain the operation and maintenance strategy of the new media platform, and extract the operation and maintenance characteristics of several new media platforms based on the operation and maintenance strategy of the new media platform;

[0010] Build an association model between each operation and maintenance feature and each customer preference label;

[0011] Based on the association model between each operation and maintenance feature and each customer preference tag and the benchmark preference data of the new media, determine the data change model of each customer preference tag during the launch of the new media platform;

[0012] Generate delivery optimization strategies based on the data change model of various customer preference tags during the delivery period of the new media platform.

[0013] Preferably, the construction of the association model between each operation and maintenance feature and each customer preference tag specifically includes:

[0014] From the historical operation and maintenance data of the new media platform, select the data change rates of several customer preference labels when the current operation and maintenance features to be verified are executed to different degrees, and record them as initial training data;

[0015] Combined with the historical operation and maintenance data of the new media platform, calculate the improvement rate corresponding to each initial training data;

[0016] Based on the improvement rate corresponding to each initial training data and the execution degree of the operation and maintenance features to be verified corresponding to each initial training data;

[0017] Through linear regression fitting, the correlation model between the execution degree of the operation and maintenance features to be verified and the customer preference labels is obtained;

[0018] The expression of the association model is:

[0019] ;

[0020] In the formula, is the data change rate increase rate of customer preference labels when the execution degree of the operation and maintenance feature to be verified is V, is the execution degree of the operation and maintenance feature to be verified corresponding to the jth initial training data, is the improvement rate corresponding to the jth initial training data, is the total number of initial training data.

[0021] Preferably, the step of calculating the improvement rate corresponding to each initial training data by combining the historical operation and maintenance data of the new media platform specifically includes:

[0022] Obtain all other operation and maintenance features executed at the time corresponding to the initial training data of the operation and maintenance feature to be verified at each execution level, and record them as comparative operation and maintenance features;

[0023] Filter out a number of historical operation and maintenance data that have and only have the comparative operation and maintenance features, and record them as comparative operation and maintenance data;

[0024] Extract the data change rate of customer preference labels under all comparison operation and maintenance data, and calculate the average value to obtain the comparison training data;

[0025] The difference between the initial training data and the comparison training data is used as the improvement rate corresponding to the initial training data.

[0026] Preferably, the data change model of each customer preference tag of the new media platform during the delivery period is determined based on the association model between each operation and maintenance feature and each customer preference tag and the benchmark preference data of the new media, specifically including:

[0027] Based on the association model between each operation and maintenance feature and each customer preference tag, at least one target association model corresponding to each customer preference tag is screened out;

[0028] Determine the operation and maintenance features corresponding to the target association model, record them as target operation and maintenance features, and record the remaining operation and maintenance features except the target operation and maintenance features as reference operation and maintenance features;

[0029] Based on the baseline operation and maintenance characteristics, determine the baseline improvement rate of each customer preference tag of the new media platform during the launch period;

[0030] Determine a data change model of customer preference tags of the new media platform during the delivery period based on at least one target association model and a benchmark improvement rate of each customer preference tag of the new media platform during the delivery period;

[0031] The data change model of the customer preference label of the new media platform during the delivery period is specifically as follows:

[0032] ;

[0033] In the formula, The data change model of customer preference labels during the launch of new media platforms, where t is the dependent variable representing time, Provide customer preference label benchmark preference data for new media platforms. It is the benchmark increase rate of customer preference tags during the launch period of the new media platform. is the i-th target association model, is the execution degree of the operation and maintenance feature corresponding to the i-th target association model, is the total number of target-related models.

[0034] Preferably, the step of screening out at least one target association model corresponding to each customer preference tag based on the association model between each operation and maintenance feature and each customer preference tag specifically includes:

[0035] Extract the correlation index from the association model between operation and maintenance features and customer preference labels, denoted as ;

[0036] Determine whether there is an association model with a correlation index greater than a preset value. If so, take all association models with correlation indexes greater than the preset value as the target association model corresponding to the customer preference label. If not, take the association model corresponding to the maximum correlation index as the target association model corresponding to the customer preference label.

[0037] in, .

[0038] Preferably, the determining of the benchmark improvement rate of each customer preference tag of the new media platform during the delivery period based on the benchmark operation and maintenance characteristics specifically includes:

[0039] The operation and maintenance strategy that only consists of all the baseline operation and maintenance features is recorded as the comparison operation and maintenance strategy;

[0040] From all the historical operation and maintenance strategies of the new media platform, at least one historical operation and maintenance strategy whose Jaccard similarity coefficient with the comparison operation and maintenance strategy is less than the similarity threshold is selected and recorded as the similar operation and maintenance strategy of the comparison operation and maintenance strategy;

[0041] The average data change rate of customer preference labels under all similar operation and maintenance strategies is calculated to obtain the benchmark improvement rate of customer preference labels of the new media platform during the delivery period.

[0042] Preferably, the generating of the delivery optimization strategy based on the data change model of each customer preference label during the delivery of the new media platform specifically includes:

[0043] Based on the data change model of each customer preference label during the launch period of the new media platform, predict the predicted data value of the customer preference label at each time point during the launch period of the new media platform;

[0044] Adjust the delivery strategy in advance based on the predicted data values ​​of customer preference labels at different time points, and implement the delivery strategy at the corresponding time points.

[0045] Furthermore, a delivery system based on new media customer preference data analysis is proposed, which is used to implement the delivery method based on new media customer preference data analysis as described above, including:

[0046] A delivery purpose analysis module, the delivery purpose analysis module is used to obtain at least one customer preference tag that affects the delivery purpose based on the delivery purpose;

[0047] A platform analysis module, the platform analysis module is electrically connected to the delivery purpose analysis module, the platform analysis module is used to collect the real-time customer preference data of each customer preference tag of the new media platform as the benchmark preference data of the new media, obtain the operation and maintenance strategy of the new media platform, extract a number of operation and maintenance features of the new media platform based on the operation and maintenance strategy of the new media platform, construct an association model between each operation and maintenance feature and each customer preference tag, and determine the data change model of each customer preference tag of the new media platform during the delivery period based on the association model between each operation and maintenance feature and each customer preference tag and the benchmark preference data of the new media;

[0048] The delivery adjustment module is electrically connected to the platform analysis module, and is used to generate a delivery optimization strategy based on a data change model of each customer preference label of the new media platform during the delivery period.

[0049] Optionally, the platform analysis module includes:

[0050] A benchmark collection unit, which is used to collect real-time customer preference data of each customer preference tag of the new media platform as the benchmark preference data of the new media, and obtain the operation and maintenance strategy of the new media platform, and extract several operation and maintenance features of the new media platform based on the operation and maintenance strategy of the new media platform;

[0051] An operation and maintenance analysis unit, the operation and maintenance analysis unit is used to build a correlation model between each operation and maintenance feature and each customer preference tag;

[0052] The platform comprehensive fitting unit is used to determine the data change model of each customer preference label of the new media platform during the delivery period based on the association model between each operation and maintenance feature and each customer preference label and the benchmark preference data of the new media.

[0053] Compared with the prior art, the present invention has the following beneficial effects:

[0054] The present invention proposes a delivery solution based on the analysis of customer preference data of new media. Based on the operation and maintenance strategy of new media, it dynamically predicts the future changes in user groups of new media platforms, and then analyzes the future customer preference data of new media platforms. Based on the future customer preference data, it pre-adjusts the delivery strategy that is compatible with it, ensuring that the delivered content is highly matched with the preferences of the target user group during actual delivery, thereby improving the delivery effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 This is a flow chart of the delivery method based on new media customer preference data analysis proposed by the present invention;

[0056] Figure 2 A flow chart of a method for constructing an association model between each operation and maintenance feature and each customer preference tag in the present invention;

[0057] Figure 3 A flow chart of a method for calculating the improvement rate corresponding to each initial training data in the present invention;

[0058] Figure 4 A flow chart of a method for determining a data change model of each customer preference tag of a new media platform during the delivery period in the present invention;

[0059] Figure 5 A flow chart of a method for screening out at least one target association model corresponding to each customer preference tag in the present invention;

[0060] Figure 6 A flow chart of a method for determining a benchmark improvement rate of each customer preference tag of a new media platform during a delivery period in the present invention;

[0061] Figure 7 A flow chart of the method for generating a delivery optimization strategy in the present invention;

[0062] Figure 8 This is a structural diagram of the delivery system based on new media customer preference data analysis proposed by the present invention;

[0063] Fig. 9 It is a schematic diagram of the computer-readable storage medium structure of the present invention. DETAILED DESCRIPTION

[0064] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art may think of other obvious variations.

[0065] Reference Figure 1 As shown, a delivery method based on new media customer preference data analysis includes:

[0066] Based on the purpose of delivery, at least one customer preference tag that affects the delivery purpose is obtained. It can be understood that different delivery purposes have different customer preference tag requirements. For product delivery targeting female target groups, new media customer preference tags are required, including gender, age, etc.

[0067] Collect the real-time customer preference data of each customer preference tag on the new media platform as the benchmark preference data of the new media;

[0068] Obtain the operation and maintenance strategy of the new media platform, and extract the operation and maintenance characteristics of several new media platforms based on the operation and maintenance strategy of the new media platform;

[0069] Build an association model between each operation and maintenance feature and each customer preference label;

[0070] Based on the association model between each operation and maintenance feature and each customer preference tag and the benchmark preference data of the new media, determine the data change model of each customer preference tag during the launch of the new media platform;

[0071] Generate delivery optimization strategies based on the data change model of various customer preference tags during the delivery period of the new media platform.

[0072] This solution uses the historical data of new media platforms and big data analysis technology to build a dynamic customer preference prediction and delivery optimization solution. It can monitor the operation and maintenance strategy of new media platforms in real time, predict the changing trend of future user groups, and then analyze the evolution direction of customer preferences. Based on these prediction results, the delivery strategy is adjusted to ensure that the delivery content is always consistent with the preferences of the target user group, thereby improving the click-through rate, conversion rate and overall delivery effect of the advertisement.

[0073] Reference Figure 2 As shown in the figure, building the association model between each operation and maintenance feature and each customer preference label specifically includes:

[0074] From the historical operation and maintenance data of the new media platform, select the data change rates of several customer preference labels when the current operation and maintenance features to be verified are executed to different degrees, and record them as initial training data;

[0075] Combined with the historical operation and maintenance data of the new media platform, calculate the improvement rate corresponding to each initial training data;

[0076] Based on the improvement rate corresponding to each initial training data and the execution degree of the operation and maintenance features to be verified corresponding to each initial training data;

[0077] Through linear regression fitting, the correlation model between the execution degree of the operation and maintenance features to be verified and the customer preference labels is obtained;

[0078] The expression of the association model is:

[0079] ;

[0080] In the formula, is the data change rate increase rate of customer preference labels when the execution degree of the operation and maintenance feature to be verified is V, is the execution degree of the operation and maintenance feature to be verified corresponding to the jth initial training data, is the improvement rate corresponding to the jth initial training data, is the total number of initial training data.

[0081] Reference Figure 3As shown in the figure, combined with the historical operation and maintenance data of the new media platform, the calculation of the improvement rate corresponding to each initial training data includes:

[0082] Obtain all other operation and maintenance features executed at the time corresponding to the initial training data of the operation and maintenance feature to be verified at each execution level, and record them as comparative operation and maintenance features;

[0083] Filter out a number of historical operation and maintenance data that have and only have the comparative operation and maintenance features, and record them as comparative operation and maintenance data;

[0084] Extract the data change rate of customer preference labels under all comparison operation and maintenance data, and calculate the average value to obtain the comparison training data;

[0085] The difference between the initial training data and the comparison training data is used as the improvement rate corresponding to the initial training data.

[0086] It is understandable that different operation and maintenance characteristic behaviors can welcome different user groups on the platform. For example, the video platform focuses on the introduction of animation, which will lead to the joining of young user groups. Changes in the user groups of these platforms will cause changes in the overall customer preference data of the platform. In this solution, by analyzing the operation and maintenance characteristic behaviors under different investment levels and performing linear regression analysis with the data change rate of customer preference labels, the correlation attributes between different operation and maintenance characteristic behaviors and customer preference labels are intuitively displayed, providing a calculation basis for the subsequent prediction and analysis of the customer preference data of the platform.

[0087] Reference Figure 4 As shown, based on the association model between each operation and maintenance feature and each customer preference tag and the benchmark preference data of the new media, the data change model of each customer preference tag of the new media platform during the delivery period is determined to specifically include:

[0088] Based on the association model between each operation and maintenance feature and each customer preference tag, at least one target association model corresponding to each customer preference tag is screened out;

[0089] Determine the operation and maintenance features corresponding to the target association model, record them as target operation and maintenance features, and record the remaining operation and maintenance features except the target operation and maintenance features as reference operation and maintenance features;

[0090] Based on the baseline operation and maintenance characteristics, determine the baseline improvement rate of each customer preference tag of the new media platform during the launch period;

[0091] Determine a data change model of customer preference tags of the new media platform during the delivery period based on at least one target association model and a benchmark improvement rate of each customer preference tag of the new media platform during the delivery period;

[0092] The data change model of customer preference labels during the launch of new media platforms is as follows:

[0093] ;

[0094] In the formula, The data change model of customer preference labels during the launch of new media platforms, where t is the dependent variable representing time, Provide customer preference label benchmark preference data for new media platforms. It is the benchmark increase rate of customer preference tags during the launch period of the new media platform. is the i-th target association model, is the execution degree of the operation and maintenance feature corresponding to the i-th target association model, is the total number of target-related models.

[0095] Reference Figure 5 As shown, based on the association model between each operation and maintenance feature and each customer preference tag, screening out at least one target association model corresponding to each customer preference tag specifically includes:

[0096] Extract the correlation index from the association model between operation and maintenance features and customer preference labels, denoted as ;

[0097] Determine whether there is an association model with a correlation index greater than a preset value. If so, take all association models with correlation indexes greater than the preset value as the target association model corresponding to the customer preference label. If not, take the association model corresponding to the maximum correlation index as the target association model corresponding to the customer preference label.

[0098] in, .

[0099] Reference Figure 6 As shown, based on the benchmark operation and maintenance characteristics, the benchmark improvement rate of each customer preference tag of the new media platform during the delivery period is determined specifically including:

[0100] The operation and maintenance strategy that only consists of all the baseline operation and maintenance features is recorded as the comparison operation and maintenance strategy;

[0101] From all the historical operation and maintenance strategies of the new media platform, at least one historical operation and maintenance strategy whose Jaccard similarity coefficient with the comparison operation and maintenance strategy is less than the similarity threshold is selected and recorded as the similar operation and maintenance strategy of the comparison operation and maintenance strategy;

[0102] The average data change rate of customer preference labels under all similar operation and maintenance strategies is calculated to obtain the benchmark improvement rate of customer preference labels of the new media platform during the delivery period.

[0103] It is understandable that in the actual analysis process, if the correlation index is too small, it means that the change in the investment level of the operation and maintenance strategy will not cause the data change of the new media under the customer preference label. In order to reduce the complexity of the prediction model, these operation and maintenance strategies are used as benchmark data support during the analysis, and the overall change rate brought about by these operation and maintenance strategies is analyzed based on historical operation and maintenance data. On this basis, the dynamic changes of the target association model that analyzes whether the change in the investment level of the operation and maintenance strategy will cause a large data change of the new media under the customer preference label can effectively improve the computational complexity of the data change model of the customer preference label of the new media platform during the delivery period, and improve the analysis efficiency of the data change model while ensuring the accuracy of the model.

[0104] Reference Figure 7 As shown in the figure, based on the data change model of each customer preference label during the delivery period of the new media platform, the delivery optimization strategy generated specifically includes:

[0105] Based on the data change model of each customer preference label during the launch period of the new media platform, predict the predicted data value of the customer preference label at each time point during the launch period of the new media platform;

[0106] Adjust the delivery strategy in advance based on the predicted data values ​​of customer preference labels at different time points, and implement the delivery strategy at the corresponding time points.

[0107] Based on the operation and maintenance strategy of new media, dynamically predict the future changes in the user groups of new media platforms, and then analyze the future customer preference data of new media platforms. Based on the future customer preference data, pre-adjust the corresponding delivery strategy to ensure that the actual delivery content is highly matched with the preferences of the target user group, thereby improving the delivery effect.

[0108] For further information, see Figure 8 As shown, based on the same inventive concept as the above-mentioned delivery method based on new media customer preference data analysis, this solution also proposes a delivery system based on new media customer preference data analysis, including:

[0109] A delivery purpose analysis module, the delivery purpose analysis module is used to obtain at least one customer preference tag that affects the delivery purpose based on the delivery purpose;

[0110] A platform analysis module, the platform analysis module is electrically connected to the delivery purpose analysis module, and is used to collect the real-time customer preference data of each customer preference tag of the new media platform as the benchmark preference data of the new media, obtain the operation and maintenance strategy of the new media platform, extract a number of operation and maintenance features of the new media platform based on the operation and maintenance strategy of the new media platform, construct an association model between each operation and maintenance feature and each customer preference tag, and determine the data change model of each customer preference tag of the new media platform during the delivery period based on the association model between each operation and maintenance feature and each customer preference tag and the benchmark preference data of the new media;

[0111] The delivery adjustment module is electrically connected to the platform analysis module and is used to generate a delivery optimization strategy based on a data change model of each customer preference label of the new media platform during the delivery period.

[0112] The platform analysis modules include:

[0113] A benchmark collection unit is used to collect the real-time customer preference data of each customer preference tag of the new media platform as the benchmark preference data of the new media, and obtain the operation and maintenance strategy of the new media platform, and extract the operation and maintenance features of several new media platforms based on the operation and maintenance strategy of the new media platform;

[0114] Operation and maintenance analysis unit, which is used to build a correlation model between each operation and maintenance feature and each customer preference label;

[0115] The platform comprehensive fitting unit is used to determine the data change model of each customer preference label of the new media platform during the delivery period based on the association model between each operation and maintenance feature and each customer preference label and the benchmark preference data of the new media.

[0116] The process of using the above system is as follows:

[0117] Step 1: The delivery purpose analysis module obtains at least one customer preference tag that affects the delivery purpose based on the delivery purpose;

[0118] Step 2: The benchmark collection unit collects the real-time customer preference data of each customer preference tag of the new media platform as the benchmark preference data of the new media, obtains the operation and maintenance strategy of the new media platform, and extracts several operation and maintenance features of the new media platform based on the operation and maintenance strategy of the new media platform;

[0119] Step 3: The operation and maintenance analysis unit constructs a correlation model between each operation and maintenance feature and each customer preference label;

[0120] Step 4: The platform comprehensive fitting unit determines the data change model of each customer preference label during the launch of the new media platform based on the association model between each operation and maintenance feature and each customer preference label and the benchmark preference data of the new media;

[0121] Step 5: The delivery adjustment module generates a delivery optimization strategy based on the data change model of each customer preference label of the new media platform during the delivery period.

[0122] Fig. 9 Schematic diagram of the computer-readable storage medium structure provided by an embodiment of the present application. Fig. 9 As shown, a computer-readable storage medium 600 according to an embodiment of the present application is shown. Computer-readable instructions are stored on the computer-readable storage medium 600. When the computer-readable instructions are executed by the processor, a delivery method based on new media customer preference data analysis according to an embodiment of the present application described with reference to the above figures can be executed. The storage medium 600 includes, but is not limited to, for example, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory (cache). Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0123] In summary, the advantages of the present invention are: pre-adjusting the delivery strategy adapted to the future customer preference data to ensure that the delivery content is highly matched with the preferences of the target user group during actual delivery, thereby improving the delivery effect.

[0124] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions only describe the principles of the present invention. The present invention may be subject to various changes and improvements without departing from the spirit and scope of the present invention. These changes and improvements fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the attached claims and their equivalents.

Claims

1. A method for delivering content based on new media customer preference data analysis, characterized in that: include: Based on the delivery purpose, obtaining at least one customer preference tag that affects the delivery purpose; Collect real-time customer preference data for each customer preference tag on the new media platform as the benchmark preference data for the new media; Obtain the operation and maintenance strategy of the new media platform, and extract several operation and maintenance features of the new media platform based on the operation and maintenance strategy of the new media platform; Build an association model between each operation and maintenance feature and each customer preference label; Based on the association model between each operation and maintenance feature and each customer preference tag and the baseline preference data of the new media, determine the data change model of each customer preference tag during the launch period of the new media platform; Generate delivery optimization strategies based on the data change model of each customer preference tag during the delivery period of the new media platform; The method of determining the data change model of each customer preference tag during the launch of the new media platform based on the association model between each operation and maintenance feature and each customer preference tag and the baseline preference data of the new media specifically includes: Based on the association model between each operation and maintenance feature and each customer preference tag, at least one target association model corresponding to each customer preference tag is screened out; Determine the operation and maintenance features corresponding to the target association model, record them as target operation and maintenance features, and record the remaining operation and maintenance features except the target operation and maintenance features as baseline operation and maintenance features; Based on the baseline operation and maintenance characteristics, determine the baseline lift rate of each customer preference tag during the launch period of the new media platform; Determining a data change model of customer preference tags of the new media platform during the delivery period based on at least one target association model and a benchmark lift rate of each customer preference tag of the new media platform during the delivery period; The data change model of the customer preference tags of the new media platform during the delivery period is specifically as follows: ； Where, The data change model of customer preference labels during the launch of new media platforms, where t is the dependent variable representing time, Provide customer preference label benchmark preference data for new media platforms, The benchmark improvement rate of customer preference tags for new media platforms during the launch period. is the i-th target association model, is the execution degree of the operation and maintenance feature corresponding to the i-th target association model, is the total number of target association models; The step of screening out at least one target association model corresponding to each customer preference tag based on the association model between each operation and maintenance feature and each customer preference tag specifically includes: Extract the correlation index from the association model between operation and maintenance features and customer preference labels, denoted as ; Determine whether there is an association model with a correlation index greater than a preset value. If so, use all association models with correlation indexes greater than the preset value as target association models corresponding to the customer preference label. If not, use the association model corresponding to the maximum correlation index as the target association model corresponding to the customer preference label. in, ; is the execution degree of the operation and maintenance feature to be verified corresponding to the jth initial training data, is the improvement rate corresponding to the j-th initial training data, is the total number of initial training data; Determining the benchmark improvement rate of each customer preference tag of the new media platform during the delivery period based on the benchmark operation and maintenance characteristics specifically includes: The operation and maintenance strategy that only has all the baseline operation and maintenance features is recorded as the comparison operation and maintenance strategy; From all historical operation and maintenance strategies of the new media platform, at least one historical operation and maintenance strategy whose Jaccard similarity coefficient with the comparison operation and maintenance strategy is less than the similarity threshold is selected and recorded as a similar operation and maintenance strategy to the comparison operation and maintenance strategy; Calculate the average data change rate of customer preference tags under all similar operation and maintenance strategies to obtain the benchmark improvement rate of customer preference tags of the new media platform during the launch period.

2. The method for delivering content based on new media customer preference data analysis according to claim 1, characterized in that: The construction of the association model between each operation and maintenance feature and each customer preference tag specifically includes: From the historical operation and maintenance data of the new media platform, select the data change rates of several customer preference labels when the current operation and maintenance features to be verified are executed to different degrees, and record them as initial training data; Combined with the historical operation and maintenance data of the new media platform, calculate the improvement rate corresponding to each initial training data; Based on the improvement rate corresponding to each initial training data and the execution degree of the operation and maintenance features to be verified corresponding to each initial training data; Through linear regression fitting, the correlation model between the execution degree of the operation and maintenance feature to be verified and the customer preference label is obtained; The expression of the association model is: ； Where, It is the data change rate improvement rate of customer preference labels when the execution level of the operation and maintenance feature to be verified is V.

3. The method for delivering content based on new media customer preference data analysis according to claim 2, characterized in that: Calculating the improvement rate corresponding to each initial training data by combining the historical operation and maintenance data of the new media platform specifically includes: Obtain all other operation and maintenance features executed at the time corresponding to the initial training data of the operation and maintenance feature to be verified at each execution level, and record them as comparative operation and maintenance features; Filter out a number of historical operation and maintenance data that have and only have the comparative operation and maintenance features, and record them as comparative operation and maintenance data; Extract the data change rate of customer preference labels under all comparison operation and maintenance data, calculate the average value, and obtain the comparison training data; Subtract the comparison training data from the initial training data, and the difference is used as the improvement rate corresponding to the initial training data.

4. The method for delivering content based on new media customer preference data analysis according to claim 3, characterized in that: The data change model of each customer preference tag during the new media platform delivery period is used to generate a delivery optimization strategy, which specifically includes: Based on the data change model of each customer preference label during the launch period of the new media platform, the predicted data value of the customer preference label at each time point during the launch period of the new media platform is predicted; Adjust the delivery strategy in advance based on the predicted data values ​​of customer preference labels at different time points, and implement the delivery strategy at the corresponding time points.

5. A delivery system based on new media customer preference data analysis, characterized in that: A method for implementing the new media-based customer preference data analysis method according to any one of claims 1 to 4, comprising: a delivery purpose analysis module, the delivery purpose analysis module being configured to obtain, based on the delivery purpose, at least one customer preference tag that has an impact on the delivery purpose; a platform analysis module, the platform analysis module being electrically connected to the delivery purpose analysis module, the platform analysis module being used to collect real-time customer preference data for each current customer preference tag of the new media platform as the benchmark preference data of the new media, obtain the operation and maintenance strategy of the new media platform, extract a number of operation and maintenance features of the new media platform based on the operation and maintenance strategy of the new media platform, construct an association model between each operation and maintenance feature and each customer preference tag, and determine a data change model for each customer preference tag of the new media platform during the delivery period based on the association model between each operation and maintenance feature and each customer preference tag and the benchmark preference data of the new media; The delivery adjustment module is electrically connected to the platform analysis module and is used to generate a delivery optimization strategy based on a data change model of each customer preference label during the delivery period of the new media platform.

6. The delivery system based on new media customer preference data analysis according to claim 5, characterized in that: The platform analysis module includes: A benchmark collection unit, which is used to collect real-time customer preference data for each customer preference tag of the new media platform as the benchmark preference data of the new media, obtain the operation and maintenance strategy of the new media platform, and extract several operation and maintenance features of the new media platform based on the operation and maintenance strategy of the new media platform; An operation and maintenance analysis unit, configured to construct an association model between each operation and maintenance feature and each customer preference tag; The platform comprehensive fitting unit is used to determine the data change model of each customer preference label of the new media platform during the delivery period based on the association model between each operation and maintenance feature and each customer preference label and the benchmark preference data of the new media.

7. A computer-readable storage medium storing a computer-readable program, characterized in that: When the computer-readable program is executed by a processor, the delivery method based on new media customer preference data analysis according to any one of claims 1 to 4 is implemented.

Images