Account promotion content generation method and device, computer equipment and storage medium

Abstract

The invention relates to the technical field of data analysis, and particularly discloses an account promotion content generation method and device, computer equipment and a storage medium. According to the method, the account promotion target is identified through the data analysis model, the candidate content generation strategy is automatically matched according to the account promotion target, and the candidate strategy is objectively evaluated and screened according to the second account data, so that the high accuracy of the target content generation strategy is ensured; the publishable account promotion content is automatically generated through the content generation model, manual generation by an account manager is not needed, manual intervention is not needed, and the accuracy of a content generation strategy and the generation efficiency of the promotion content are improved. The method is applied to the online promotion marketing business of financial products (such as insurance products and financial products), the target content generation strategy conforming to the promotion target in the account stage can be quickly generated, the promotion content is automatically generated according to the target content generation strategy, and the marketing efficiency is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data analysis, in particular to a method and device for generating account promotion content, a computer device and a storage medium. BACKGROUND

[0002] With the rapid development of mobile Internet, using short videos, text and pictures, and other social media for customer acquisition and brand building has become an important way for agents in the financial and insurance industry. Through the publication of professional popular science, case sharing, product analysis, and other content, financial agents can reach a wider range of potential users and improve brand awareness and business conversion. However, agents face great challenges in content operation practice: lack of professional and efficient digital tools, resulting in low efficiency of account content production, unclear marketing goals, high compliance risks, and ultimately difficulty in effectively improving performance through online channels. In view of the above problems, at present, there are some general short video data analysis platforms or MCN (Multi-Channel Network) agency services that can provide basic data dashboard functions for account managers, such as displaying account fan numbers, like numbers, and interaction rates. However, existing platforms cannot provide content generation strategies based on data analysis results, nor can they automatically generate promotion content that meets the account promotion goals. Account managers still need to actively analyze data promotion goals to determine content generation strategies and generate account promotion content with the help of other tools. The content generation efficiency is still low, and the accuracy of the strategy is low due to human experience errors. Therefore, how to improve the accuracy of the content generation strategy and the efficiency of the promotion content generation has become a problem to be solved. SUMMARY

[0003] The present application provides a method and device for generating account promotion content, a computer device and a storage medium to improve the accuracy of the content generation strategy and the efficiency of the promotion content generation.

[0004] In a first aspect, the present application provides a method for generating account promotion content, the method comprising: collecting first account data of each social platform account of an account management personnel; analyzing the first account data based on a preset data analysis model, identifying an account promotion goal, and performing keyword matching based on the account promotion goal to obtain at least one candidate content generation strategy; obtaining second account data of each social platform account in a preset time period, and based on the second account data, evaluating the candidate content generation strategy to determine a target content generation strategy for each social platform account; processing the target content generation strategy for each social platform account based on a content generation model to generate account promotion content for each social platform account.

[0005] In a second aspect, the present application further provides a device for generating account promotion content, the device comprising: an account data acquisition module configured to collect first account data of each social platform account of an account manager; a candidate strategy matching module configured to analyze the first account data based on a preset data analysis model, identify an account promotion target, and perform keyword matching based on the account promotion target to obtain at least one candidate content generation strategy; a target strategy determination module configured to obtain second account data of each social platform account in a preset time period, and evaluate the candidate content generation strategy based on the second account data to determine a target content generation strategy for each social platform account; a promotion content generation module configured to process the target content generation strategy for each social platform account based on a content generation model to generate account promotion content for each social platform account.

[0006] In a third aspect, the present application further provides a computer device comprising a memory and a processor; the memory is configured to store a computer program; the processor is configured to execute the computer program and implement the above-mentioned method for generating account promotion content when executing the computer program.

[0007] In a fourth aspect, the present application further provides a computer readable storage medium storing a computer program, wherein the computer program is executed by a processor to make the processor implement the above-mentioned method for generating account promotion content.

[0008] The application discloses a method and device for generating account promotion content, a computer device and a storage medium. First account data of each social platform account of an account manager is collected. The first account data is analyzed based on a preset data analysis model, an account promotion target is identified, keyword matching is performed based on the account promotion target, at least one candidate content generation strategy is obtained, second account data of each social platform account in a preset time period is obtained, the candidate content generation strategy is evaluated based on the second account data, a target content generation strategy of each social platform account is determined, the target content generation strategy of each social platform account is processed based on a content generation model, and account promotion content of each social platform account is generated. The account promotion target is identified through the data analysis model, the candidate content generation strategy is automatically matched according to the account promotion target, the candidate strategy is objectively evaluated and screened according to the second account data, the high precision of the target content generation strategy is ensured, the publishable account promotion content is automatically generated through the content generation model, the account manager does not need to manually generate, and manual intervention is not needed, so that the accuracy of the content generation strategy and the generation efficiency of the promotion content are improved. BRIEF DESCRIPTION OF DRAWINGS

[0009] In order to more clearly illustrate the technical solutions of the embodiments of the application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor.

[0010] Figure 1 FIG. 1 is a first schematic flowchart of a method for generating account promotion content provided by an embodiment of the application; Figure 2 FIG. 2 is a second schematic flowchart of a method for generating account promotion content provided by an embodiment of the application; Figure 3 FIG. 3 is a third schematic flowchart of a method for generating account promotion content provided by an embodiment of the application; Figure 4 FIG. 4 is a schematic block diagram of a device for generating account promotion content provided by an embodiment of the application; Figure 5 FIG. 5 is a structural schematic block diagram of a computer device provided by an embodiment of the application. DETAILED DESCRIPTION

[0011] With reference to the drawings, the technical solutions in the embodiments of the present application will be clearly and completely described below. Obviously, the described embodiments are only some of the embodiments of the present application, but not all of them. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts are within the scope of the present application.

[0012] The flowcharts shown in the drawings are only illustrative, and do not necessarily include all contents and operations / steps, nor are they necessarily executed in the order described. For example, some operations / steps can be further decomposed, combined or partially merged, so that the actual execution order can be changed according to actual conditions.

[0013] It should be understood that the terms used in the specification of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the specification and the appended claims of the present application, unless otherwise clear from the context, the singular forms "a", "an" and "the" are intended to include the plural forms.

[0014] It should also be understood that the term "and / or" used in the specification and the appended claims of the present application means any combination of one or more of the associated listed items and all possible combinations, and includes these combinations.

[0015] Embodiments of the present application provide a method and device for generating account promotion content, computer equipment and a storage medium. The method for generating account promotion content can be applied to a server, which identifies an account promotion target through a data analysis model, automatically matches a candidate content generation strategy according to the account promotion target, objectively evaluates and filters the candidate strategy according to second account data, ensures the high precision of the target content generation strategy, automatically generates publishable account promotion content through a content generation model, does not require account managers to manually generate, does not require human intervention, and improves the accuracy of the content generation strategy and the efficiency of the promotion content generation. The server can be a standalone server or a server cluster.

[0016] Some embodiments of the present application will be described in detail below with reference to the accompanying drawings. In the case of no conflict, the embodiments described below and the features in the embodiments can be combined with each other.

[0017] Please refer to Figure 1 , Figure 1is a schematic flowchart of a method for generating account promotion content provided by an embodiment of the present application. The method for generating account promotion content can be applied in a server, for identifying an account promotion target through a data analysis model, automatically matching a candidate content generation strategy according to the account promotion target, and objectively evaluating and screening the candidate strategy according to second account data, ensuring high precision of the target content generation strategy, automatically generating publishable account promotion content through a content generation model, without manual generation by an account manager, without human intervention, and improving the accuracy of the content generation strategy and the efficiency of the promotion content generation.

[0018] As shown in Figure 1 The method for generating account promotion content specifically includes steps S101 to S104.

[0019] S101, collecting first account data of each social platform account of an account manager; In an embodiment, the account manager can be an insurance agent, and the first account data refers to original operation data (such as fans, interaction, conversion data, and exposure) of a target agent account on each social platform.

[0020] The collection is performed according to preset dimensions, covering all social platforms. The preset dimensions include but are not limited to account basic data (such as total number of fans and average daily number of new fans), content performance data (such as content and content type statistics, content form, single content exposure, complete play rate, number of likes, number of comments, number of reposts, and number of collections), user interaction data (such as comment section interaction), and conversion effect data (such as number of private message consultations, number of form submissions, and number of transaction leads).

[0021] In specific embodiments, data can be pulled through an official interface provided by a social platform, and data integration can be performed to obtain the final first account data.

[0022] S102, analyzing the first account data based on a preset data analysis model, identifying an account promotion target, and matching keywords based on the account promotion target to obtain at least one candidate content generation strategy; In an embodiment, the preset data analysis model can be a rule-based classifier or a simple machine learning model, which is used to analyze the first account data and identify the account promotion target.

[0023] Specifically, first, the first account data is analyzed by a data analysis model to identify the current stage of the account. The account stage includes a cold start period, a growth period, a trial conversion period, and a conversion amplification period. The stage division standard can be freely set by the user according to actual needs, such as a cold start period when the current number of fans is less than a first preset threshold, a growth period when the first preset threshold is less than or equal to the current number of fans and less than a second preset threshold, a trial conversion period when the second preset threshold is less than or equal to the current number of fans and less than a third preset threshold, and a conversion amplification period when the current number of fans is greater than or equal to the third preset threshold. Then, the account promotion target is matched according to the current stage of the account. For example, the cold start period target is to guide users to follow, the growth period target is to enhance the user's professional trust in the account manager (insurance agent), and the trial conversion period and conversion amplification period target is to guide private messages and conversion to promote transactions.

[0024] In one embodiment, for different account promotion targets, different content generation strategies are matched, for example, high-frequency, life-like content is recommended to be published to guide users to follow; knowledge sharing and insurance popular science are recommended to be published to enhance professional trust; case sharing and preferential activities are recommended to be published to guide private messages and conversion.

[0025] Further, the keyword matching based on the account promotion target to obtain at least one candidate content generation strategy includes: based on the account promotion target, matching at least one content generation strategy parameter from a preset action library; and based on each content generation strategy parameter, generating at least one candidate content generation strategy.

[0026] In one embodiment, the action library is a multi-dimensional action library, which contains core executable elements of account operation, such as content type, frequency, form, conversion component, etc., and each element is labeled with an applicable stage.

[0027] The content type includes but is not limited to popular science (such as insurance terminology interpretation, difference between insurance types, etc.), case (such as claim settlement case, insurance story, etc.), welfare (such as free insurance consultation), and activity (such as old customer recommendation with gifts, time-limited insurance discount, etc.). The publishing frequency includes but is not limited to high frequency (such as 3 daily updates), medium frequency (such as 2 daily updates), and low frequency (such as 3-4 weekly updates). The form includes but is not limited to short video, text and image, and live broadcast. The conversion component includes private message guidance, appointment link, form, and old customer repurchase reminder.

[0028] In an embodiment, the account promotion target output by the data analysis model is a text description, such as increasing the number of fans. Through natural language processing technology, the account promotion target text is analyzed to identify entities such as "number of fans", and intent recognition is performed to obtain core actions such as "increase" and keywords "increase" and "number of fans". According to the obtained keywords, parameter matching is performed in the action library to obtain at least one content generation strategy parameter corresponding to the content type, publishing frequency, form, and conversion component.

[0029] The matched parameters are combined according to the content type-frequency-form-conversion component to form a candidate content generation strategy.

[0030] In an embodiment, the content generation strategy parameters corresponding to the content type, publishing frequency, form, and conversion component can be randomly combined, as long as each candidate content generation strategy includes parameters corresponding to the content type, publishing frequency, form, and conversion component.

[0031] For example, when the account promotion target is to increase the number of fans, the matched content type is popular science and case, the frequency is medium, the form is short video and text, and the conversion component is private message guidance and appointment link. Popular science and case, short video and text, private message guidance and appointment link can be randomly combined to generate multiple candidate content generation strategies, which can include candidate content generation strategy 1: content type is popular science, frequency is medium, form is short video, and conversion component is private message guidance; candidate content generation strategy 2: content type is case, frequency is medium, form is short video, and conversion component is private message guidance; candidate content generation strategy 3: content type is case, frequency is medium, form is text, and conversion component is private message guidance; and candidate content generation strategy 4: content type is popular science, frequency is medium, form is text, and conversion component is private message guidance.

[0032] In the above embodiment, by using the pre-set action library, discrete operation experience is converted into standardized and reusable content generation strategy parameters, which are combined and matched according to the account stage to ensure the diversity of the generated candidate strategies, thereby ensuring that the target content generation strategy meeting the promotion target can be accurately obtained in subsequent strategy determination.

[0033] S103, acquiring second account data of each social platform account in a preset time period, and based on the second account data, evaluating the candidate content generation strategy to determine the target content generation strategy of each social platform account; In one embodiment, the second account data refers to the content effect data of the extracted preset time period (such as the last week), including interaction rate, net increase of fans, lead quantity (such as private message, leave a message), and number of insurance policies or insurance policy conversion rate, and the like content effect indicators, which are used to evaluate the promotion effect score of each candidate content generation strategy. It can be directly extracted from the first account data.

[0034] In one embodiment, the weight coefficients corresponding to the interaction rate, net increase of fans, lead quantity (such as private message, leave a message), and number of insurance policies or insurance policy conversion rate are matched according to the account promotion target, and a weighted calculation is performed to obtain the promotion effect score of each candidate content generation strategy. The target content generation strategy is determined according to the promotion effect score.

[0035] In one embodiment, as shown in Figure 2 The method for generating the account promotion content based on the second account data, evaluating the candidate content generation strategy, and determining the target content generation strategy of each social platform account, specifically includes steps S1031 to S1034.

[0036] S1031, according to each candidate content generation strategy, the second account data is divided to obtain the content effect data corresponding to each candidate content generation strategy; S1032, analyzing the content effect data to obtain the estimated value of each content effect indicator of each social platform account; S1033, matching the weight coefficients corresponding to each content effect indicator based on the account promotion target, and obtaining the promotion effect score of each candidate content generation strategy based on each weight coefficient and the estimated value of each content effect indicator; S1034, based on the preset strategy selection rule and the promotion effect score, the target content generation strategy of each social platform account is determined in each candidate content generation strategy.

[0037] In one embodiment, the content effect data of the account with the same or highly similar (such as similarity greater than a preset threshold) content generation strategy is screened out from the second account data. For example, the content effect data of the account with the same or similar content, content form, and conversion component of the candidate content generation strategy.

[0038] In one embodiment, since the same strategy has different operation effect differences on different platforms, the data is first divided according to the social platform, and then the estimated value of the content effect indicator is calculated for the data under each platform. Specifically, the average or median of each content effect indicator (such as interaction rate, net increase of fans, lead quantity, and number of insurance policies or insurance policy conversion rate) is calculated according to the content effect data corresponding to each candidate content generation strategy as the estimated value of the content effect indicator.

[0039] In an embodiment, the weight coefficient is obtained by matching in a preset stage weight mapping table according to the account promotion target. The stage weight mapping table can be set by the user according to actual conditions.

[0040] The weight coefficient is an importance weight allocated to each content effect index according to the account promotion target. For example, for attention interaction and fan increase, the weight coefficients of the interaction rate, fan net increase, lead quantity, and policy number or policy conversion rate are 0.5, 0.3, 0.2, and 0, respectively. For guiding private messages and conversion, the weight coefficients of the interaction rate, fan net increase, lead quantity, and policy number or policy conversion rate are 0.2, 0.2, 0.3, and 0.3, respectively. For attention lead and policy conversion, the weight coefficients of the interaction rate, fan net increase, lead quantity, and policy number or policy conversion rate are 0.1, 0.1, 0.3, and 0.5, respectively.

[0041] The promotion effect score of each candidate content generation strategy is obtained by weighted calculation of the weight coefficient and the estimated value corresponding to the content effect index. The promotion effect score is a comprehensive effect score of the candidate content generation strategy on a certain platform, and the promotion effect score calculation formula is:

[0042] wherein ER is the interaction rate, AF is the fan net increase, Lead is the lead quantity (private message / leave capital), Conv is the policy number or policy conversion rate, and w1, w2, w3, and w4 are the weight coefficients.

[0043] In an embodiment, based on the preset strategy selection scheme and the promotion effect score of each candidate content generation strategy, the target content generation strategy is determined. Specifically, the strategy selection rule can be to select the highest score scheme with a probability of 90% and randomly select one from the top 3 in the promotion effect score with a probability of 10%, that is, to select the current highest score action 90% of the time, and to randomly select one from the top 3 in the promotion effect score 10% of the time to be the target content generation strategy, so as to avoid falling into local optimum.

[0044] In another embodiment, the content effect data corresponding to the accounts of the same industry, the same stage, and the same strategy can also be directly pulled as the second account data through the official interface provided by the social platform.

[0045] In an embodiment, since the user habits of each platform are different, the feedback to the promotion content generated by the same strategy can be different, that is, the content effect data of the same content generation strategy on different platforms is different, and therefore the promotion effect scores of the same candidate content generation strategy on different platforms can be different. The target content generation strategy of each platform can be determined according to the promotion effect score. For example, the users of Douyin prefer short videos, and the users of Xiaohongshu prefer text and images.

[0046] In the above embodiments, by dividing the second account data according to the strategy, and rigorously analyzing the content effect data, an objective estimation value of the key content effect index is calculated for each candidate strategy, realizing the objectivity and data-driven of strategy evaluation, and avoiding subjective bias. In addition, the weight coefficient is dynamically matched based on the growth stage of the account, ensuring the accurate alignment of the strategy and the promotion goal, so as to ensure that the selected target content generation strategy can accurately match the account promotion goal, and further lay the foundation for the matching of subsequent content generation strategies.

[0047] S104, processing the target content generation strategy of each social platform account based on the content generation model, and generating the account promotion content of each social platform account.

[0048] In one embodiment, the target content generation strategy includes elements such as content type, frequency, form, and conversion components. Therefore, content materials and content generation models can be matched according to the target content generation strategy to generate scripts, images, videos, and voices, and the final account promotion content can be generated according to single or multiple combinations of the generated scripts, images, videos, and voices.

[0049] It can be understood that the target content generation strategy corresponding to different platforms may be different, and therefore the account promotion content of each social platform may also be different.

[0050] In one embodiment, as shown in Figure 3 The step S104 of the account promotion content generation method specifically includes steps S1041 to S1044.

[0051] S1041, based on the target content generation strategy, matching content materials in a preset material library and model matching in a preset model library to obtain the content generation model; S1042, processing the content materials based on the content generation model to generate at least one initial promotion content; S1043, performing compliance verification on the initial promotion content; S1044, when the initial promotion content passes the compliance verification, taking the initial promotion content as the account promotion content.

[0052] In one embodiment, the material library stores reusable and different types of content elements, such as case materials and popular science materials. The model library stores different generation models, including but not limited to text generation models, voice generation models, image generation models, and video generation models.

[0053] According to the content type in the target content generation strategy, the content material conforming to the content type is matched from the material library. Specifically, a keyword matching method can be adopted. For example, according to the keyword "case" corresponding to the "case type", the content material such as "child insurance claim case" and "vehicle insurance claim case" can be matched from the material library.

[0054] In one embodiment, different content forms need to be generated by corresponding different generation models, and therefore, the model matching can be performed in the model library according to the content form in the target content generation strategy, and the matched model is used as the content generation model. It can be understood that the matched model can not be unique, but can be two or more models. For example, for the "picture-text" content form, both image and text are needed, and therefore, the matched model includes an image generation model and a text generation model.

[0055] For example, assuming that the target content generation strategy is "case story, short video, and conversion component as private message guide", the video generation model and the voice generation model are matched in the preset model library as the content generation model according to the "short video". The content material is input into the video generation model and the voice generation model, and a short video containing video and voice is generated as the initial promotion content.

[0056] In another embodiment, the material screening can also be combined with the content form. For example, if the content type is "case type" and the content form is "picture-text", the case and the existing image are matched in the material library, so as to reduce the image generation workload of the subsequent content generation model and accelerate the generation efficiency.

[0057] In one embodiment, the target content generation strategy also includes the release frequency, and therefore, the corresponding group number of content materials can be obtained according to the release frequency. Each group of content materials can include text, image, video, etc., and each group of content materials can generate an initial promotion content. For example, when the release frequency is 2 times, two groups of content materials can be matched from the material library, and the two groups of content materials are processed by the content generation model to generate two initial promotion contents.

[0058] In one embodiment, when the initial promotion content is generated, the account manager can actively add content generation conditions (such as video voice personification), and the content generation model processes the content material in combination with the target content generation strategy and the content generation condition to generate the initial promotion content meeting the requirements of the account manager.

[0059] In an embodiment, the pre-trained compliance model is used to analyze the initial promotion content, including detecting sensitive words (such as best, absolute, stable, etc.), checking whether necessary risk prompt is included, and obtaining a risk identification result of the initial promotion content, including no risk and risk.

[0060] Further, after the compliance verification of the initial promotion content, if the compliance verification result is that there is a compliance risk, the risk content in the initial promotion content is marked, and the marked initial promotion content is returned to the content generation model to modify the risk content based on the content generation model until the compliance verification passes, and the account promotion content is obtained.

[0061] In an embodiment, the compliance model performs verification, and if a risk is identified, the risk content in the initial promotion content can be directly marked, such as highlighting the risk words.

[0062] In another embodiment, the compliance model can also generate a structured risk report according to the risk identification result, for example, risk type: illegal promise, risk position: 3rd sentence in the 2nd paragraph, original content: guarantee to get the lowest premium for you, suggestion: change to help you strive for better premium.

[0063] In an embodiment, the initial promotion content after the risk content is marked, or the initial promotion content and the structured risk report are returned to the content generation model, the content generation model modifies the risk marked content or the risk content indicated in the risk report, generates new promotion content, and the new promotion content is verified again by the compliance model. After the verification passes, it is used as the final account promotion content. Otherwise, the above steps of returning and regenerating are repeated until the compliance verification passes.

[0064] In an embodiment, if the model is modified multiple times and still fails to pass the compliance verification, a risk prompt is sent to the account manager to remind the account manager to modify the risk.

[0065] In the above embodiments, by intelligently matching the target content generation strategy with the preset material library and model library, the most suitable content material and content generation model are automatically called to quickly generate initial promotion content that meets the strategy intent, realizing the large-scale and automated production of high-quality promotion content and improving the content generation efficiency. It can realize the batch processing and unified management of millions of agent accounts, and improve the overall content generation efficiency.

[0066] Further, the step S104 is followed by further comprising: analyzing the first account data to obtain an optimal exposure time period and a sub-optimal exposure time period of each of the social platform accounts; determining a publishing time of the account promotion content on each of the social platform accounts based on the optimal exposure time period, the sub-optimal exposure time period and a publishing frequency in the target content generation strategy; publishing the account promotion content based on the publishing time and obtaining exposure data of the account promotion content after the publishing; updating the publishing frequency based on the exposure data and publishing the account promotion content based on the updated frequency.

[0067] In one embodiment, the exposure amount of the account on each platform is counted according to the first account data, and a time interval with a higher exposure amount than other time intervals and a second highest time interval are obtained as the optimal exposure time period and the sub-optimal exposure time period, respectively.

[0068] In one embodiment, the frequency is distributed to the corresponding time interval according to a preset frequency distribution principle to formulate a publishing schedule. The frequency distribution principle can be set by the user according to actual needs, for example, 70% of the frequency is distributed to the optimal time interval (to ensure core time period exposure), and 30% of the frequency is distributed to the sub-optimal time interval (to disperse risks and cover different active users).

[0069] In one embodiment, the account promotion content is published on each platform according to the publishing schedule. The exposure data is obtained by monitoring the exposure data of the promotion content, and the obtained exposure data is compared with the average exposure level in the previous period (such as the last 7 days) to calculate the exposure growth rate. If the exposure growth rate is positive, no adjustment is needed, and the publishing according to the publishing schedule continues; if the exposure growth rate is negative and less than a preset threshold, the frequency is updated, the publishing frequency is adjusted according to the preset strategy, and the publishing time is determined again according to the latest publishing frequency and the latest exposure data. The preset strategy for adjusting the publishing frequency can be freely set by the account manager according to needs. For example, if the exposure growth rate is less than 30% for the last 3 days, the frequency is automatically reduced from “high” to “medium”.

[0070] In another embodiment, the first account data can also be analyzed according to a pre-trained machine learning model to predict the exposure probability of different time periods in the future, select the time period with the highest probability as the optimal exposure time period, and select the time period with the second highest probability as the sub-optimal exposure time period.

[0071] ​Further, the analysis of the first account data to obtain the optimal exposure time period and the sub-optimal exposure time period of each social platform account comprises: based on a time series prediction model, analyzing the first account data to obtain exposure linear trend characteristics, exposure period characteristics and exposure holiday characteristics; based on the exposure linear trend characteristics, the exposure period characteristics and the exposure holiday characteristics, performing exposure state prediction to obtain the optimal exposure time period and the sub-optimal exposure time period.

[0072] In one embodiment, a pre-trained time series prediction model (such as a Facebook Prophet or an ARIMA (Autoregressive Integrated Moving Average) model) is used to analyze the first account data to obtain a time series data sequence containing time and exposure.

[0073] Specifically, the time series prediction model performs curve fitting according to the time series data sequence to extract the overall change of exposure over time and obtain exposure linear characteristics. The model analyzes the exposure according to a preset time period (such as every day, every week, etc.) to obtain the regular fluctuations of exposure within a fixed period, such as the exposure fluctuates regularly within 24 hours, and the exposure is usually much higher at 20-22 pm when the user is active than at 4-6 am. The model identifies special dates such as holidays and platform promotions according to the time series data sequence, and analyzes the exposure corresponding to the special dates to obtain exposure holiday characteristics.

[0074] In one embodiment, the exposure linear trend characteristics, the exposure period characteristics and the exposure holiday characteristics are superimposed and predicted to obtain the predicted exposure of each hour in a future period. According to the predicted exposure, the optimal exposure time period and the sub-optimal exposure time period of each day, or the optimal exposure time period and the sub-optimal exposure time period of each week are determined.

[0075] It can be understood that whether the optimal exposure time period is daily or weekly can be determined according to the publishing frequency, such as if the publishing frequency is 3 times a day, the optimal exposure time period is determined daily, and if the publishing frequency is twice a week, the optimal exposure time period and the sub-optimal exposure time period are determined weekly.

[0076] In the above embodiment, through deep analysis of content effect data, the optimal and suboptimal exposure time periods of each social platform are intelligently analyzed or predicted, and the content release time is accurately arranged accordingly, avoiding manual experience errors and effectively improving the accuracy of the strategy. The content is released in the window period of the highest user activity and platform traffic, thereby significantly improving the initial exposure amount and reach efficiency of the content. Secondly, by continuously tracking the exposure data after release, the release effect is evaluated in real time, and when the exposure effect is significantly reduced, the release frequency is automatically reduced, realizing dynamic adjustment of the strategy and further improving the accuracy of the strategy.

[0077] Please refer to Figure 4 , Figure 4 An embodiment of the present application provides a schematic block diagram of an account promotion content generation device, which is used for executing the account promotion content generation method. The account promotion content generation device can be configured in a server.

[0078] As Figure 4 shown, the account promotion content generation device 400 comprises: An account data acquisition module 401 is configured to collect first account data of each social platform account of an account manager. A candidate strategy matching module 402 is configured to analyze the first account data based on a preset data analysis model, identify an account promotion target, perform keyword matching based on the account promotion target, and obtain at least one candidate content generation strategy. A target strategy determination module 403 is configured to obtain second account data of each social platform account in a preset time period, evaluate the candidate content generation strategies based on the second account data, and determine a target content generation strategy of each social platform account. A promotion content generation module 404 is configured to process the target content generation strategy of each social platform account based on a content generation model, and generate account promotion content of each social platform account.

[0079] Further, the target strategy determination module 403 comprises: A data division unit is configured to divide the second account data according to each candidate content generation strategy, and obtain content effect data corresponding to each candidate content generation strategy. An estimated value obtaining unit is configured to analyze the content effect data, and obtain an estimated value corresponding to each content effect index of each social platform account. The score obtaining unit is configured to match a weight coefficient corresponding to each content effect index based on the account promotion target, and obtain a promotion effect score of each candidate content generation strategy based on each weight coefficient and an estimated value corresponding to each content effect index. The strategy determining unit is configured to determine a target content generation strategy of each social platform account among the candidate content generation strategies based on a preset strategy selection rule and the promotion effect score.

[0080] Further, the promotion content generation module 404 comprises: The content generation model matching unit is configured to match a content material in a preset material library and perform model matching in a preset model library based on the target content generation strategy, and obtain the content generation model. The initial promotion content generation unit is configured to process the content material based on the content generation model to generate at least one initial promotion content. The compliance verification unit is configured to perform compliance verification on the initial promotion content. The account promotion content determining unit is configured to determine the initial promotion content as the account promotion content when the initial promotion content passes the compliance verification.

[0081] Further, the promotion content generation module 404 further comprises: The risk marking unit is configured to mark a risk content in the initial promotion content if the compliance verification result is that there is a compliance risk. The risk correction unit is configured to return the marked initial promotion content to the content generation model, so as to correct the risk content based on the content generation model until the compliance verification passes, and obtain the account promotion content.

[0082] Further, the account promotion content generation apparatus 400 further comprises a promotion content publishing module, and the promotion content publishing module comprises: The time period determining unit is configured to analyze the first account data to obtain an optimal exposure time period and a suboptimal exposure time period of each social platform account. The publishing time determining unit is configured to determine a publishing time of the account promotion content on each social platform account based on the optimal exposure time period, the suboptimal exposure time period, and a publishing frequency in the target content generation strategy. The exposure data obtaining unit is configured to publish the account promotion content based on the publishing time, and obtain exposure data of the account promotion content after being published. The frequency update unit is used to update the frequency of publication based on the exposure data, and to publish the account promotion content based on the updated frequency.

[0083] Furthermore, the time period determination unit includes: The feature acquisition subunit is used to analyze the first account data based on the time series prediction model to obtain exposure linear trend features, exposure cycle features, and exposure holiday features; An exposure state prediction unit is used to predict the exposure state based on the exposure linear trend characteristics, the exposure cycle characteristics, and the exposure holiday characteristics, and to obtain the optimal exposure time period and the second-optimal exposure time period.

[0084] Furthermore, the candidate strategy matching module 402 includes: The content generation strategy parameter matching unit is used to match at least one content generation strategy parameter from a preset action library based on the account promotion goal. A candidate content generation strategy generation unit is used to generate at least one candidate content generation strategy based on each of the content generation strategy parameters.

[0085] It should be noted that those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the above-described apparatus and modules can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0086] The aforementioned apparatus can be implemented as a computer program, which can be used in, for example... Figure 5 It runs on the computer device shown.

[0087] Please see Figure 5 , Figure 5 This is a schematic block diagram illustrating the structure of a computer device according to an embodiment of this application. The computer device may be a server.

[0088] See Figure 5 The computer device includes a processor, memory, and network interface connected via a system bus, wherein the memory may include non-volatile storage media and internal memory.

[0089] Non-volatile storage media can store operating systems and computer programs. These computer programs include program instructions that, when executed, cause a processor to perform any method for generating account promotional content.

[0090] The processor provides computing and control capabilities, supporting the operation of the entire computer device.

[0091] Internal memory provides an environment for the execution of computer programs on non-volatile storage media. When these computer programs are executed by a processor, the processor can perform any method of generating account promotion content.

[0092] This network interface is used for network communication, such as sending assigned tasks. Those skilled in the art will understand that... Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0093] It should be understood that the processor can be a Central Processing Unit (CPU), but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among these, a general-purpose processor can be a microprocessor or any conventional processor.

[0094] In one embodiment, the processor is configured to run a computer program stored in memory to perform the following steps: Collect primary account data from each social media platform account managed by account administrators; The data of the first account is analyzed based on a preset data analysis model to identify the account promotion target, and keyword matching is performed based on the account promotion target to obtain at least one candidate content generation strategy. Second account data of each of the social media platform accounts within a preset time period is obtained, and the candidate content generation strategy is evaluated based on the second account data to determine the target content generation strategy for each of the social media platform accounts. The target content generation strategy of each social platform account is processed based on the content generation model to generate account promotion content for each social platform account.

[0095] In one embodiment, when the processor evaluates the candidate content generation strategies based on the second account data and determines the target content generation strategies for each of the social platform accounts, it is configured to: Based on each of the candidate content generation strategies, the second account data is divided to obtain the content effect data corresponding to each of the candidate content generation strategies; The content performance data is analyzed to obtain estimated values ​​for each content performance indicator for each of the aforementioned social media platform accounts. Based on the account promotion goals, the weight coefficients corresponding to each of the content performance indicators are matched, and based on the weight coefficients and the estimated values ​​corresponding to each of the content performance indicators, the promotion performance score of each of the candidate content generation strategies is obtained. Based on the preset strategy selection rules and the promotion effect score, the target content generation strategy for each social platform account is determined from the candidate content generation strategies.

[0096] In one embodiment, when the processor processes the target content generation strategy for each of the social media platform accounts based on the content generation model to generate account promotion content for each of the social media platform accounts, it is configured to: Based on the target content generation strategy, content materials are matched in a preset material library, and model matching is performed in a preset model library to obtain the content generation model; The content materials are processed based on the content generation model to generate at least one initial promotional content. The initial promotional content was subject to compliance verification. When the initial promotional content passes compliance verification, the initial promotional content will be used as the account's promotional content.

[0097] In one embodiment, after performing compliance verification on the initial promotional content, the processor is further configured to: If the compliance verification result indicates that there is a compliance risk, then the risky content in the initial promotional content will be marked. The initial promotional content after being marked is sent back to the content generation model, so that the risky content can be corrected based on the content generation model until the compliance verification is passed, and the account promotional content is obtained.

[0098] In one embodiment, after the processor processes the target content generation strategy for each of the social media platform accounts based on the content generation model and generates account promotion content for each of the social media platform accounts, it is further configured to: Analyze the data of the first account to obtain the optimal and second-optimal exposure time periods for each of the aforementioned social media platform accounts; Based on the optimal exposure time period, the second-best exposure time period, and the publication frequency in the target content generation strategy, the publication time of the account promotion content on each of the social media platform accounts is determined. Based on the publication time, publish the account promotion content and obtain the exposure data after the account promotion content is published; The posting frequency is updated based on the exposure data, and the account promotion content is posted based on the updated frequency.

[0099] In one embodiment, the processor, in addition to analyzing the first account data to obtain the optimal and second-optimal exposure time periods for each of the social media platform accounts, is also configured to: Based on the time series prediction model, the data of the first account is analyzed to obtain the linear trend characteristics of exposure, the periodic characteristics of exposure, and the holiday characteristics of exposure. Exposure status is predicted based on the exposure linear trend characteristics, the exposure cycle characteristics, and the exposure holiday characteristics to obtain the optimal exposure time period and the second-optimal exposure time period.

[0100] In one embodiment, when the processor performs keyword matching based on the account promotion goal to obtain at least one candidate content generation strategy, it is configured to: Based on the account promotion goals, at least one content generation strategy parameter is matched from the preset action library; At least one candidate content generation strategy is generated based on the parameters of each of the content generation strategies.

[0101] The embodiments of this application also provide a computer-readable storage medium storing a computer program, the computer program including program instructions, and the processor executing the program instructions to implement any of the account promotion content generation methods provided in the embodiments of this application.

[0102] The computer-readable storage medium may be an internal storage unit of the computer device described in the foregoing embodiments, such as the hard disk or memory of the computer device. The computer-readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk, SmartMedia Card (SMC), Secure Digital (SD) card, or Flash Card equipped on the computer device.

[0103] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for generating account promotion content, characterized by, The method comprises the following steps: Collecting first account data of each social platform account managed by an account manager; Analyzing the first account data based on a preset data analysis model, identifying an account promotion target, and matching keywords based on the account promotion target to obtain at least one candidate content generation strategy; Obtaining second account data of each social platform account within a preset time period, and evaluating the candidate content generation strategy based on the second account data to determine a target content generation strategy for each social platform account; Processing the target content generation strategy of each social platform account based on a content generation model to generate account promotion content for each social platform account. 2.The method of claim 1, wherein, The evaluation of the candidate content generation strategy based on the second account data to determine the target content generation strategy for each social platform account comprises: Dividing the second account data according to each candidate content generation strategy to obtain content effect data corresponding to each candidate content generation strategy; Analyzing the content effect data to obtain an estimated value of each content effect indicator corresponding to each social platform account; Matching a weight coefficient corresponding to each content effect indicator based on the account promotion target, and obtaining a promotion effect score of each candidate content generation strategy based on each weight coefficient and the estimated value of each content effect indicator; Determining the target content generation strategy for each social platform account from each candidate content generation strategy based on a preset strategy selection rule and the promotion effect score. 3.The method of claim 1, wherein, The processing of the target content generation strategy of each social platform account based on the content generation model to generate the account promotion content for each social platform account comprises: Matching content materials in a preset material library based on the target content generation strategy, and matching models in a preset model library to obtain the content generation model; Processing the content materials based on the content generation model to generate at least one initial promotion content; Conducting compliance verification on the initial promotion content; If the initial promotion content passes the compliance verification, the initial promotion content is used as the account promotion content.

4. The method of claim 3, wherein, After the compliance verification of the initial promotion content, the method further comprises: If the compliance verification result indicates that there is a compliance risk, marking the risk content in the initial promotion content; Returning the marked initial promotion content to the content generation model to modify the risk content based on the content generation model until the compliance verification passes to obtain the account promotion content.

5. The method of claim 1, wherein, After the processing of the target content generation strategy of each social platform account based on the content generation model to generate the account promotion content for each social platform account, the method further comprises: Analyzing the first account data to obtain an optimal exposure time period and a sub-optimal exposure time period for each social platform account; Determining the publication time of the account promotion content in each social platform account based on the optimal exposure time period, the sub-optimal exposure time period, and the publication frequency in the target content generation strategy. Based on the publishing time, the account promotion content is published, and exposure data after publishing the account promotion content is obtained; Based on the exposure data, the publishing frequency is updated, and the account promotion content is published based on the updated frequency.

6. The method of claim 5, wherein, The analysis of the first account data obtains the optimal exposure time period and the sub-optimal exposure time period of each social platform account, including: Based on the time series prediction model, the first account data is analyzed to obtain the exposure linear trend feature, the exposure period feature, and the exposure holiday feature; Based on the exposure linear trend feature, the exposure period feature, and the exposure holiday feature, the exposure state is predicted to obtain the optimal exposure time period and the sub-optimal exposure time period.

7. The method of claim 1 to 6, wherein, The keyword matching based on the account promotion target obtains at least one candidate content generation strategy, including: Based on the account promotion target, at least one content generation strategy parameter is matched from a preset action library; Based on each content generation strategy parameter, at least one candidate content generation strategy is generated.

8. An account promotion content generation apparatus characterized by comprising: Including: An account data acquisition module is configured to collect first account data of each social platform account of an account manager; A candidate strategy matching module is configured to analyze the first account data based on a preset data analysis model, identify an account promotion target, and perform keyword matching based on the account promotion target to obtain at least one candidate content generation strategy; A target strategy determination module is configured to obtain second account data of each social platform account in a preset time period, and evaluate the candidate content generation strategy based on the second account data to determine a target content generation strategy for each social platform account; A promotion content generation module is configured to process the target content generation strategy of each social platform account based on a content generation model to generate account promotion content for each social platform account.

9. A computer device, comprising: The computer device includes a memory and a processor; The memory is configured to store a computer program; The processor is configured to execute the computer program and implement the account promotion content generation method of any one of claims 1-7 when executing the computer program.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, and the computer program is executed by the processor to make the processor implement the account promotion content generation method of any one of claims 1-7. The computer readable storage medium stores a computer program, and the computer program is executed by the processor to make the processor implement the account promotion content generation method of any one of claims 1-7.

Images