Method, device, equipment and readable storage medium for pushing recommended content

Abstract

The present application discloses a method, device, equipment and readable storage medium for pushing recommended content, which relates to the field of data processing. The method includes: obtaining historical conversion data; obtaining an expected conversion expression, wherein the expected conversion expression includes historical expected data and future expected expressions, wherein the historical expected data corresponds to the expected return of historical interactive operations before a specified time, and the future expected expression includes the correspondence between future expected data and an adjustment coefficient; determining the adjustment coefficient based on the historical conversion data and the expected conversion expression; and adjusting the specified resource quantity data with the adjustment coefficient. The adjustment coefficient is obtained by using the historical conversion data and the expected conversion expression. Since the expected conversion expression is determined based on the integration of the historical expected data and the future expected data, that is, the resource quantity control accuracy is relatively high, when the present application is applied to a cloud server, the above-mentioned conversion data is stored and read in the cloud server.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of data processing, and in particular to a method, apparatus, device, and readable storage medium for pushing recommended content. Background Art

[0002] An ad bid refers to the amount of resources an advertiser pays for a user interaction event when placing an ad, such as the amount of resources paid for one thousand impressions, one click event, or one conversion event. Typically, advertisers predetermine the total budget for a certain period of time based on the ad bid. Conversion events refer to interactive events derived from exposures, such as download events, registration events, etc.

[0003] In related technologies, when determining the conversion volume, all conversion events reported within a specified time period are determined as conversion events within the specified time period, and the cost of the conversion events within the specified time period is determined based on the number of conversion events and the total cost.

[0004] However, since there are often delays in reporting conversion events, such as a delay between a user downloading software and registering an account, the accuracy of the resource input calculated for conversion events within a specified time period is low, making it impossible to use a reference for the delivery effect of advertising. This affects the amount of resource input for advertising and deviates from the amount expected by the customer. Summary of the Invention

[0005] The embodiments of the present application provide a method, apparatus, device, and readable storage medium for pushing recommended content, which can improve the accuracy of evaluating the effectiveness of advertising delivery. The technical solution is as follows:

[0006] In one aspect, a method for pushing recommended content is provided, the method comprising:

[0007] Reading historical conversion data of recommended content from a database, where the historical conversion data is used to represent data obtained by historical interactive operations;

[0008] Obtaining an expected conversion expression, wherein the expected conversion expression includes historical expected data and a future expected expression, wherein the historical expected data corresponds to the expected return flow of the historical interactive operation before a specified time, and the future expected expression includes a correspondence between the future expected data and an adjustment coefficient, wherein the future expected data corresponds to the expected return flow of the future interactive operation;

[0009] Determining the adjustment coefficient based on the historical conversion data and the expected conversion expression;

[0010] Adjusting the specified resource quantity data using the adjustment coefficient to obtain adjusted resource quantity data of the recommended content;

[0011] The recommended contents are sorted according to the adjusted resource quantity data, and the sorting order is used to determine the recommended contents to be pushed.

[0012] In another aspect, a device for pushing recommended content is provided, the device comprising:

[0013] An acquisition module is configured to read historical conversion data of recommended content from a database, the historical conversion data being used to represent data obtained from historical interactive operation reflows; obtain an expected conversion expression, the expected conversion expression including historical expected data and a future expected expression, the historical expected data corresponding to the expected reflow of the historical interactive operation before a specified time; the future expected expression including a correspondence between future expected data and an adjustment coefficient, the future expected data corresponding to the expected reflow of future interactive operations;

[0014] a determination module, configured to determine the adjustment coefficient based on the historical conversion data and the expected conversion expression;

[0015] an adjustment module, configured to adjust the specified resource quantity data using the adjustment coefficient to obtain the adjusted resource quantity data of the recommended content;

[0016] A sorting module is used to sort the recommended content according to the adjusted resource quantity data, and the sorting order is used to determine the recommended content to be pushed.

[0017] On the other hand, a computer device is provided, comprising a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set, or an instruction set, and the at least one instruction, the at least one program, the code set, or the instruction set is loaded and executed by the processor to implement a method for pushing recommended content as described in any of the above-mentioned embodiments of the present application.

[0018] On the other hand, a computer-readable storage medium is provided, wherein the storage medium stores at least one instruction, at least one program, a code set, or an instruction set, and the at least one instruction, the at least one program, the code set, or the instruction set is loaded and executed by a processor to implement a method for pushing recommended content as described in any of the above-mentioned embodiments of the present application.

[0019] In another aspect, a computer program product or computer program is provided, comprising computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the method for pushing recommended content described in any of the above embodiments.

[0020] The beneficial effects of the technical solutions provided in the embodiments of the present application include at least:

[0021] The adjustment coefficient is obtained through the historical conversion data and the expected conversion expression, so as to adjust the specified resource quantity. Since the expected conversion expression is determined based on the integration of historical expected data and future expected data, that is, the adjustment coefficient is determined by the historical conversion data, historical expected data, and future expected data, the resource quantity data is balanced, the display accuracy of the delivery effect is high, and the control accuracy of the resource quantity input after the resource quantity data is adjusted is also high. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0023] Figure 1 This is a simplified schematic diagram of the principle of bid adjustment provided by an exemplary embodiment of the present application;

[0024] Figure 2 This is a schematic diagram of an implementation environment provided by an exemplary embodiment of the present application;

[0025] Figure 3 This is a flowchart of a method for pushing recommended content provided by an exemplary embodiment of the present application;

[0026] Figure 4 is based on Figure 3 Schematic diagram of historical conversion data, historical expected data, and future expected data provided by the illustrated embodiment;

[0027] Figure 5 is a flowchart of an advertisement data publishing process provided by an exemplary embodiment of the present application;

[0028] Figure 6 is a flowchart of a method for pushing recommended content provided by another exemplary embodiment of the present application;

[0029] Figure 7 is based on Figure 6 A flow chart of the adjustment process of the adjustment coefficient provided by the illustrated embodiment;

[0030] Figure 8 is a flowchart of a method for pushing recommended content provided by another exemplary embodiment of the present application;

[0031] Figure 9This is a schematic diagram of an advertisement delivery interface provided by an exemplary embodiment of the present application;

[0032] Figure 10 is a schematic diagram of a test result provided by an exemplary embodiment of the present application;

[0033] Figure 11 This is a structural block diagram of a device for pushing recommended content provided by an exemplary embodiment of the present application;

[0034] Figure 12 is a structural block diagram of a device for pushing recommended content provided by another exemplary embodiment of the present application;

[0035] Figure 13 This is a structural block diagram of a server provided by an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0036] In order to make the objectives, technical solutions and advantages of this application clearer, the implementation methods of this application will be further described in detail below with reference to the accompanying drawings.

[0037] First, a brief introduction to the terms involved in the embodiments of this application is given:

[0038] Ad bid: refers to the amount of resources an advertiser pays for a user interaction event when placing an ad, such as the amount of resources paid for 1,000 impressions / 1 click event / 1 conversion event. For example, 15 yuan for 1,000 impressions (Cost Per Mille, CPM), 0.1 cent per click (Cost Per Click, CPC), 10 yuan per activation (Cost Per Action, CPA). Ad bid is used to represent the advertiser's required cost.

[0039] A price adjustment strategy involves adjusting the bid price based on the ad's bid price to achieve a cost target within a specific timeframe. For example, an advertiser bids 10 yuan per activation. At 12:00 PM, they spend 100 yuan and receive 5 activations, resulting in a real-time cost of 20 yuan. To ensure the ad's cost reaches 10 yuan per activation by 12:00 PM, the system adjusts the bid price downward based on the strategy, for example, to 8 yuan. The bid price affects the priority of the ad during impressions.

[0040] For illustration, please refer to Figure 1According to the consumption column 110, activation quantity 120, activation cost 130 and price adjustment method 140, Ad A started running on the 30th with a bid of 90 yuan / activation. On the 30th, it consumed 3000 yuan of the budget and obtained 15 activations at a cost of 200 yuan / activation. The remaining 15 activations were returned on the 1st. On the 1st, it consumed 2000 yuan of the budget and brought 13 activations on the same day at a cost of 71.11 yuan / activation. The remaining 6 activations were returned on the 2nd, and so on. Figure 1 As shown, since the cost of No. 30 is reflected as 200 yuan / activation, the bid is lowered; since the cost of No. 1 is reflected as 71.11 yuan / activation, the bid is raised; since the cost of No. 2 is reflected as 111 yuan / activation, the bid is lowered.

[0041] Since advertisers typically lack access to forward-looking data such as impressions and clicks, they typically measure the effectiveness of their advertising campaigns by reporting conversions. This means that the conversions reported over a period of time are considered the total conversions for that period, and the effectiveness of advertising is determined by the number and timing of conversions.

[0042] However, there's often a delay in reporting conversion events. For example, a user might download adware on the first day of an ad campaign and then register an account within it on the third day. This means a conversion request is generated on the first day, but the conversion is completed and the user returns on the third day. This results in lower accuracy when calculating the cost of conversion events within a specified time period, leading to lower accuracy in the delivery of ads.

[0043] Cloud technology: refers to a hosting technology that unifies hardware, software, network and other resources within a wide area network or local area network to achieve data computing, storage, processing and sharing. Cloud technology is a general term for network technology, information technology, integration technology, management platform technology, application technology, etc. based on the cloud computing business model. It can form a resource pool that can be used on demand and is flexible and convenient. The backend services of the technical network system require a large amount of computing and storage resources. For example, during the quality inspection process of dual-recorded videos, it is necessary to store the dual-recorded videos and perform AI calculations on different key video clips in the dual-recorded videos. With the rapid development and application of the Internet industry, each item may have its own identification mark in the future, and all of them need to be transmitted to the backend system for logical processing. Data of different levels will be processed separately. All types of industry data require strong system backing support, which can only be achieved through cloud computing.

[0044] In combination with the above-mentioned noun introduction, the implementation environment of the embodiment of the present application is explained.

[0045] For illustration, please refer to Figure 2 , the implementation environment includes a terminal 210, a server 220 and a communication network 230;

[0046] Terminal 210 has an application installed therein that can receive advertising messages delivered by advertisers. When a user interacts with the advertising message, an interaction event is generated based on the interaction and reported to server 220. Server 220 is a server corresponding to an advertising delivery platform, or a server corresponding to an advertiser.

[0047] Optionally, the interactive operation includes at least one of a click operation, a link jump operation, a purchase operation, a software download operation, and a registration operation.

[0048] In response to server 220 being the server corresponding to the advertising delivery platform, server 220 predicts the cost of advertising delivery based on historical conversion events and expected conversion events, thereby adjusting the bid based on the advertiser's bid. In response to server 220 being the server corresponding to the advertiser, upon receiving the conversion event, server 220 reports it to the server corresponding to the advertising delivery platform. The server corresponding to the advertising delivery platform then predicts the cost of advertising delivery based on historical conversion events and expected conversion events, thereby adjusting the bid based on the advertiser's bid.

[0049] It is worth noting that the above-mentioned server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms. The terminal can be a smart phone, tablet computer, laptop computer, desktop computer, smart speaker, smart watch, etc., but is not limited to this. The terminal and the server can be directly or indirectly connected through wired or wireless communication, and this application does not limit this.

[0050] Illustratively, taking the application of the method for pushing recommended content provided in the embodiments of the present application in a cloud server as an example, cloud computing refers to the delivery and use model of Internet technology (IT) infrastructure, which refers to obtaining required resources through the network in an on-demand, easily scalable manner; in a broad sense, cloud computing refers to the delivery and use model of services, which refers to obtaining required services through the network in an on-demand, easily scalable manner. Such services can be IT and software, Internet-related, or other services. Cloud computing is the product of the integration of the development of traditional computer and network technologies such as grid computing, distributed computing, parallel computing, utility computing, network storage technologies, virtualization, and load balancing.

[0051] Illustratively, in the embodiment of the present application, calculations of different transformation data are processed in parallel through parallel computing, thereby improving the calculation efficiency of the transformation data.

[0052] Cloud computing has rapidly grown, driven by the internet, real-time data streams, the diversification of connected devices, and the growing demand for search services, social networks, mobile commerce, and open collaboration. Unlike previous parallel and distributed computing approaches, the emergence of cloud computing will fundamentally revolutionize the entire internet and enterprise management model.

[0053] In combination with the above-mentioned noun introduction and implementation environment, the method for pushing recommended content provided in the embodiment of this application is described. Figure 3 This is a flow chart of a method for pushing recommended content provided by an exemplary embodiment of the present application, which is described by taking the method applied to a server as an example. Figure 3 As shown, the method includes:

[0054] Step 301 : Read historical conversion data of recommended content from a database. The historical conversion data is used to represent data obtained by historical interactive operation reflow.

[0055] Historical conversion data is used to represent the historical interactive operations of recommended content within the historical time period, and the data obtained by reflowing within the required time period. For example, taking the current time of the current date as the benchmark, historical conversion data is used to represent the historical interactive operations before the current time, and the data obtained by reflowing before the current time of the current date.

[0056] In some embodiments, the historical conversion data includes data obtained by reflowing historical interactive operations before the current moment on the current date, and data obtained by reflowing historical interactive operations before the current date before the current moment on the current date.

[0057] For example, if the current time is 3:36 PM on August 20th, historical conversion data includes historical interactions that occurred before 3:36 PM on August 20th, as well as data that flowed back before 3:36 PM. For example, historical interactions that occurred at 3:20 PM and flowed back at 3:32 PM. Historical conversion data also includes historical interactions that occurred on August 19th and flowed back at 3:30 PM on August 20th.

[0058] Optionally, when the historical conversion data includes data obtained by reflowing historical interactive operations before the current date and before the current moment within the current date, the historical interactive operations are before the current date and after the specified date, such as: historical interactive operations generated after August 15 and before August 20, and data obtained by reflowing before 15:36 on August 20.

[0059] It is worth noting that in the above examples, date is used as the time division granularity. In actual operation, the time division granularity can also be based on hours, two days, weeks, etc., and the embodiments of the present application are not limited to this.

[0060] A reflux refers to the process where, after a user sends a conversion request and then performs the corresponding interactive action, the user's action is reported to the server. For example, a user clicks an ad, downloads adware through the ad interface, registers a user account within the adware, and reports the registration event to the server. This registration event is considered a conversion reflux, with the reflux time window being the time between the software download and the user account registration.

[0061] Step 302: Obtain expected conversion expression, which includes historical expected data and future expected expression.

[0062] The historical expectation data corresponds to the expected return of historical interactive operations before a specified time; the future expectation expression includes the correspondence between the future expectation data and the adjustment coefficient, and the future expectation data corresponds to the expected return of future interactive operations.

[0063] That is, the historical expected data is the expected return data of historical interactive operations that have not yet returned; and the future expected data is the expected return data of future interactive operations that have not yet occurred.

[0064] Optionally, the historical expected data includes: data obtained by reflowing historical interactive operations before the current date after the current moment in the current date, and data obtained by reflowing historical interactive operations before the current moment in the current date after the current moment in the current date.

[0065] For example, the current time is 15:36 on August 20, and the historical expected data includes historical interactive operations generated before August 20 and data expected to flow back after 15:36, such as historical interactive operations generated on August 19 and data expected to flow back after 15:36.

[0066] Indicatively, the historical expected data also includes historical interactive operations generated before 15:36 on August 20, and data expected to flow back after 15:36 on August 20.

[0067] Optionally, when the historical expected data includes data obtained by reflowing historical interactive operations before the current date after the current moment within the current date, the historical interactive operations are before the current date and after the specified date, such as: historical interactive operations generated after August 15 and before August 20, and the data expected to be reflowed after 15:36 on August 20.

[0068] It is worth noting that in the above examples, date is used as the time division granularity. In actual operation, the time division granularity can also be based on hours, two days, weeks, etc., and the embodiments of the present application are not limited to this.

[0069] The future expected data includes data obtained by reflowing future interactive operations after the current moment on the current date.

[0070] For example, the current time is 15:36 on August 20, and the expected future data includes the future interactive operations expected to occur after 15:36 on August 20, and the data expected to flow back after 15:36, for example, the interactive operations generated at 15:40 on August 20, and the data expected to flow back at 16:36.

[0071] For illustration, please refer to Figure 4 , which shows a schematic diagram of the conversion data in the related art and the conversion data after optimization in the embodiment of the present application, such as Figure 4 As shown, in the corresponding situation 410 in the related art, the conversion data includes conversion data a, conversion data c, and conversion data e; while in the corresponding situation 420 of the embodiment of the present application, the conversion data includes conversion data a, conversion data b, conversion data c, conversion data d, and conversion data e.

[0072] Among them, conversion data a represents the data obtained by the historical interactive operations before the current date that flowed back before the current moment within the current date (belonging to historical conversion data); conversion data b represents the data obtained by the historical interactive operations before the current date that flowed back after the current moment within the current date (belonging to historical expected data); conversion data c represents the data obtained by the historical interactive operations before the current moment within the current date that flowed back (belonging to historical conversion data); conversion data d represents the data obtained by the historical interactive operations before the current moment within the current date that flowed back after the current moment within the current date (belonging to historical expected data); conversion data e represents the data obtained by the future interactive operations after the current moment within the current date that flowed back after the current moment within the current date (belonging to future expected data).

[0073] That is, based on the relevant technology, the embodiment of the present application adds consideration of conversion data b and conversion data d, and uses the data obtained by flowing back the historical interactive operations before the current date after the current moment on the current date, and the data obtained by flowing back the historical interactive operations before the current moment on the current date after the current moment on the current date as reference data for bid adjustment.

[0074] Optionally, historical conversion data is determined based on the above-mentioned conversion data a and conversion data c, and expected conversion expression is determined based on the above-mentioned conversion data b, conversion data d and conversion data e, wherein the historical expected data (conversion data b and conversion data d) can be inferred through historical conversion data, and the future expected expression includes the correspondence between the future expected data (conversion data e) and the adjustment coefficient.

[0075] When determining the expected conversion expression, first determine the historical expected data based on the historical conversion data and the return rate corresponding to the historical conversion data, including the above-mentioned conversion data b and conversion data d. For example, taking the return situation of two consecutive days as an example to determine the conversion data, please refer to the following formula 1:

[0076] Historical expected data = (Conversion number 1 + Conversion number 2) / Return rate - (Conversion number 1 + Conversion number 2)

[0077] Conversion number 1 represents the number of conversions on the day before the current date, conversion number 2 represents the number of conversions before the current time on the current date, and return rate represents the overall return rate of return flows on the current date and the day before the current date.

[0078] In the process of determining future expectations, the corresponding relationship between future expectations and adjustment coefficients is determined based on consumption data and historical adjustment coefficients, where consumption data is used to indicate the total resource input. For example, taking the return flow situation for two consecutive days as an example to determine the conversion data, please refer to the following formula 2:

[0079] Future expected data = expected consumption / ((historical consumption / request caliber conversion) / historical adjustment coefficient × adjustment coefficient)

[0080] Among them, expected consumption is used to indicate the expected resource investment specified by the advertiser, historical consumption is used to indicate the resource consumption in the past two days, request-caliber conversion is used to indicate conversion events with request interaction events as conversion conditions, and historical adjustment coefficient indicates the adjustment coefficient that is likely to be generated on the current date and the day before the current date.

[0081] Optionally, the reflux rate is updated and counted in an hourly period.

[0082] The return rate also needs to be adjusted based on the coefficient of variation. Specifically, the coefficient of variation within a preset historical time period is obtained. This coefficient of variation indicates the degree of dispersion of the return rate within that time period, and the return rate corresponding to the historical conversion data is adjusted based on the coefficient of variation. The return rate is downgraded by determining whether the coefficient of variation reaches a threshold.

[0083] If the coefficient of variation of the ad-based reflow rate exceeds the threshold, it is downgraded to the account-based reflow rate. If the coefficient of variation of the account-based reflow rate exceeds the threshold, it is downgraded to the product-based reflow rate. For each reflow rate dimension, a linear mapping is performed based on the coefficient of variation. A larger coefficient of variation indicates a lower confidence level and a larger minimum cutoff value for the corresponding reflow rate.

[0084] Step 303: Determine an adjustment coefficient based on historical conversion data and expected conversion expression.

[0085] Based on the relationship between historical conversion data, expected conversion expression and cost data, and the correspondence between estimated resource quantity data and adjustment coefficients, the estimated resource quantity data is adjusted using the adjustment coefficients to obtain the adjustment coefficients when the estimated resource quantity data meets the requirements of the specified resource quantity data.

[0086] Step 304 : Adjust the designated resource quantity data using the adjustment coefficient to obtain adjusted resource quantity data of the recommended content.

[0087] Optionally, specify a target cost corresponding to the resource quantity data.

[0088] Illustratively, the adjustment coefficient is 0.8, and the designated resource quantity data is 100. After adjusting the designated resource quantity data using the adjustment coefficient, the advertiser's bid is 80.

[0089] Optionally, the target cost is adjusted using the adjustment coefficient; or the cost obtained by adjusting the historical adjustment coefficient is adjusted using the adjustment coefficient.

[0090] It is worth noting that in the above embodiment, the resource quantity data is implemented as cost as an example for explanation. The resource quantity data can also be implemented as other forms of resource input, such as: virtual prop input, virtual bill input, etc.

[0091] Step 305: Sort the recommended content according to the adjusted resource quantity data.

[0092] The sorting order is used to determine the recommended content to be pushed. Optionally, the bidding data affects the advertiser's ad exposure priority.

[0093] Optionally, when sorting recommended content, the bid score of the recommended content is calculated, such as: advertising resource acquisition (effective Cost Per Mile, eCPM), and the estimated conversion rate is estimated by a lightweight conversion estimation model. The product of the advertising bid, estimated conversion rate, estimated click-through rate and adjustment coefficient is determined as the bid score of the recommended content, and then sorted according to the bid score.

[0094] Illustratively, bid score = ad bid × estimated conversion rate × estimated click-through rate × adjustment factor. Optionally, the order of recommended content is sorted according to a pre-set sorting rule, such as sorting the order of recommendations in ascending order based on bid scores; or sorting the order of recommendations in descending order based on bid scores.

[0095] In some embodiments, since the recommended content adjusted toward a lower price has a cost price higher than the bid, the recommendation quantity needs to be increased, while the recommended content adjusted toward a higher price has a cost price lower than the bid, the recommendation quantity can be reduced. Therefore, the recommendation order of the recommended content is arranged in reverse order of the bid scores.

[0096] In some embodiments, the top m recommended contents are determined as the recommended contents to be pushed according to the sorting order, such as determining the first-ranked recommended content as the recommended content to be pushed.

[0097] To sum up, the method provided in this embodiment obtains an adjustment coefficient through historical conversion data and expected conversion expression, thereby adjusting the specified resource quantity. Since the expected conversion expression is determined based on the integration of historical expected data and future expected data, that is, the adjustment coefficient is determined by historical conversion data, historical expected data, and future expected data, the resource quantity data is balanced, the display accuracy of the delivery effect is high, and the control accuracy of the resource quantity input after the resource quantity data is adjusted is also high.

[0098] For illustration, the resource amount data adjustment method provided in the embodiment of the present application is applied to the process of adjusting the advertising bid as an example. Figure 5The process includes: step 501, real-time advertising data processing. Step 502, estimating the advertising conversion return rate. Estimating and adjusting the return rate based on the coefficient of variation. Step 503, obtaining real-time advertising conversion data and consumption data. The real-time advertising conversion data includes historical conversion data, historical expected data, and future expected data. Step 504, determining the real-time price adjustment strategy based on the advertising bid information. Optionally, the adjustment coefficient is determined based on the real-time advertising conversion data, consumption data, and advertising bid information. Step 505, publishing the data to the database. Step 506, sorting the advertising releases based on the data in the database.

[0099] In an optional embodiment, the bid adjustment is achieved by constructing an estimated resource quantity expression, adjusting the adjustment coefficient, and solving the optimal solution. Figure 6 This is a flow chart of a method for pushing recommended content provided by another exemplary embodiment of the present application. Taking the method applied to a server as an example, Figure 5 As shown, the method includes:

[0100] Step 601 : Read historical conversion data of recommended content from a database. The historical conversion data is used to represent data obtained by historical interactive operation reflow.

[0101] Historical conversion data is used to represent the historical interactive operations of the recommended content within the historical time period, and the data obtained by reflowing within the required time period. For example, taking the current time of the current date as the benchmark, historical conversion data is used to represent the historical interactive operations before the current time, and the data obtained by reflowing before the current time of the current date.

[0102] In some embodiments, the historical conversion data includes data obtained by reflowing historical interactive operations before the current moment on the current date, and data obtained by reflowing historical interactive operations before the current date before the current moment on the current date.

[0103] Step 602: Obtain expected conversion expression, which includes historical expected data and future expected expression.

[0104] The historical expectation data corresponds to the expected return of historical interactive operations before a specified time; the future expectation expression includes the correspondence between the future expectation data and the adjustment coefficient, and the future expectation data corresponds to the expected return of future interactive operations.

[0105] That is, the historical expected data is the expected return data of historical interactive operations that have not yet returned; and the future expected data is the expected return data of future interactive operations that have not yet occurred.

[0106] Optionally, the historical expected data includes: data obtained by reflowing historical interactive operations before the current date after the current moment in the current date, and data obtained by reflowing historical interactive operations before the current moment in the current date after the current moment in the current date.

[0107] Step 603: Obtain an estimated resource quantity expression based on the historical conversion data and the expected conversion expression. The estimated resource quantity expression includes a corresponding relationship between the estimated resource quantity data and the adjustment coefficient.

[0108] The ratio of the sum of the historical resource data and the expected resource data to the sum of the historical conversion data and the expected conversion expression is determined as the estimated resource expression.

[0109] Optionally, the request-based conversion for calculating the expected cost for the entire day and the historical expected cost can be replaced with the report-based conversion. For the calculation of the estimated resource volume data, refer to the following formula 3:

[0110] cpa_e=(historyCost+expectCost) / (historyConversion+expectConversion)

[0111] Among them, cpa_e represents the estimated cost, that is, the estimated resource quantity data, historyCost represents the historical resource quantity data, expectCost represents the expected resource quantity data, that is, the cost data corresponding to the expected conversion in the future, historyConversion represents the historical conversion data, and expectConversion represents the expected conversion data and the future expected data.

[0112] Here, expectConversion=expectCost / (expectOriginCpa×λ), λ represents the adjustment coefficient, and expectOriginCpa represents the original expected cost, that is, the specified resource quantity data.

[0113] For illustration, please refer to Figure 7 , which shows a flow chart of the adjustment process of the adjustment coefficient provided by an exemplary embodiment of the present application, including: Step 701, advertising data collection. That is, obtaining feedback data generated after the advertisement is released, such as conversion data, etc. Step 702, judging whether a cold start is required. Among them, a cold start refers to a startup method after clearing the data record. Step 703, determining a cold start strategy. Step 704, determining a future consumption estimate. Optionally, traverse the adjustment coefficient and the consumption when the adjustment coefficient is 1, and make an estimate of future consumption based on the ratio of the consumption at a single moment to the consumption throughout the day. Schematically, Among them, t represents the time, λ represents the adjustment coefficient, costλ Indicates traversing the consumption corresponding to λ, cost 1.0 Indicates the consumption when the adjustment coefficient is 1, cost 全天 represents the total consumption for the whole day, and ε is a pre-set parameter. Step 705, determine the cost smoothing coefficient. Optionally, the smoothing coefficient w is calculated based on the pre-determined conversion coefficient. Schematically, Among them, t total Indicates the total duration, and conversion' is a pre-set conversion coefficient. Step 706, future expected cost estimation. Optionally, the future expected cost estimation is calculated based on the future consumption estimation and the smoothing coefficient. Schematically, CPA1.0 represents the predicted CPA when the adjustment coefficient is 1, and targetCPA represents the target bid. Step 707: Calculate the optimal bid. The optimal bid is determined by adjusting the adjustment coefficient. Step 708: Determine the adjustment coefficient.

[0114] Step 604 : Taking the designated resource quantity data as the adjustment target, determine the value of the adjustment coefficient in the estimated resource quantity expression.

[0115] Optionally, the values ​​of the adjustment coefficient are traversed within the price adjustment range of the adjustment coefficient, and in response to the estimated result expressed by the estimated resource quantity matching the specified resource quantity data and meeting the resource quantity requirements, the value of the adjustment coefficient corresponding to the estimated result is determined.

[0116] Indicatively, the values ​​of the adjustment coefficient are traversed within the adjustment range of the adjustment coefficient. When the difference between the estimated result and the specified resource quantity data is within the required difference range, the value of the adjustment coefficient corresponding to the estimated result is determined.

[0117] Step 605 : Adjust the designated resource quantity data using the adjustment coefficient to obtain the adjusted resource quantity data of the recommended content.

[0118] Step 606: Sort the recommended content according to the adjusted resource quantity data.

[0119] The sorting order is used to determine the recommended content to be pushed. Optionally, the bidding data affects the advertiser's ad exposure priority.

[0120] To sum up, the method provided in this embodiment obtains an adjustment coefficient through historical conversion data and expected conversion expression, thereby adjusting the specified resource quantity. Since the expected conversion expression is determined based on the integration of historical expected data and future expected data, that is, the adjustment coefficient is determined by historical conversion data, historical expected data, and future expected data, the resource quantity data is balanced, the display accuracy of the delivery effect is high, and the control accuracy of the resource quantity input after the resource quantity data is adjusted is also high.

[0121] The method provided in this embodiment obtains an estimated resource quantity expression based on historical conversion data and expected conversion expression, and determines the adjustment coefficient based on the estimated result of the estimated resource quantity expression as the adjustment target, thereby improving the accuracy of determining the adjustment coefficient and improving the accuracy of cost estimation.

[0122] In an optional embodiment, there is still a problem of starting volume in conversion reporting, that is, there is a certain difference between the return rate in the initial stage of advertising delivery and the return rate in the stable period of delivery. For different stages, the correspondence between future expected data and the adjustment coefficient is determined through different conversion calibers. Figure 8 is a flowchart of a method for pushing recommended content provided by another exemplary embodiment of the present application. Figure 8 As shown, taking the method applied to a server as an example, the method includes:

[0123] Step 801 : Read historical conversion data of recommended content from a database. The historical conversion data is used to represent data obtained by historical interactive operation reflow.

[0124] Historical conversion data is used to represent the historical interactive operations of the recommended content within the historical time period, and the data obtained by reflowing within the required time period. For example, taking the current time of the current date as the benchmark, historical conversion data is used to represent the historical interactive operations before the current time, and the data obtained by reflowing before the current time of the current date.

[0125] In some embodiments, the historical conversion data includes data obtained by reflowing historical interactive operations before the current moment on the current date, and data obtained by reflowing historical interactive operations before the current date before the current moment on the current date.

[0126] Step 802 : Determine historical expected data based on the historical conversion data and the return rate corresponding to the historical conversion data.

[0127] Step 803, in response to being in the start-up phase, the corresponding relationship between future expected data and the adjustment coefficient is determined based on the consumption data, historical adjustment coefficient and request caliber conversion.

[0128] Optionally, the start-up phase is used to represent the initial stage, schematically, the initial stage of advertising delivery, such as the first three days of advertising delivery, and the stable phase represents the period after the end of the start-up phase.

[0129] Optionally, during the initial stage, request-based conversion is used to determine the conversion data.

[0130] Step 804, in response to being in a stable stage, the corresponding relationship between future expected data and the adjustment coefficient is determined based on the consumption data, historical adjustment coefficient and reporting caliber conversion.

[0131] Optionally, starting from the stable phase, the request-caliber conversion is transitioned to the reporting-caliber conversion to determine the conversion data.

[0132] Indicatively, to address the issue of reported conversion volume, the price adjustment from the first to third day of ad delivery is switched to the request-based conversion mode. Starting from the fourth day, the price adjustment transitions from request-based conversion to reported-based conversion. The conversion data during the transition period is obtained using a weighted method, as shown in Formula 4 below:

[0133] Conversion number = d(t) × estimated conversion number + (1-d(t)) × reported conversion number

[0134] Where d(t) is the weight, and d(t)=max(min(e -bt , 1.0)0.0), t∈[0,1440], where t represents the time and b is an auxiliary parameter in the weight calculation process. Estimated conversion number = requested conversion number / return rate.

[0135] Step 805: Determine an adjustment coefficient based on historical conversion data and expected conversion expression.

[0136] Based on the relationship between historical conversion data, expected conversion expression and cost data, and the correspondence between estimated resource quantity data and adjustment coefficients, the estimated resource quantity data is adjusted using the adjustment coefficients to obtain the adjustment coefficients when the estimated resource quantity data meets the requirements of the specified resource quantity data.

[0137] Step 806 : Adjust the designated resource quantity data using the adjustment coefficient to obtain adjusted resource quantity data of the recommended content.

[0138] Step 807: Sort the recommended content according to the adjusted resource quantity data.

[0139] The sorting order is used to determine the recommended content to be pushed. Optionally, the bidding data affects the advertiser's ad exposure priority.

[0140] To sum up, the method provided in this embodiment obtains an adjustment coefficient through historical conversion data and expected conversion expression, thereby adjusting the specified resource quantity. Since the expected conversion expression is determined based on the integration of historical expected data and future expected data, that is, the adjustment coefficient is determined by historical conversion data, historical expected data, and future expected data, the resource quantity data is balanced, the display accuracy of the delivery effect is high, and the control accuracy of the resource quantity input after the resource quantity data is adjusted is also high.

[0141] The method provided in this embodiment addresses the problem of reporting conversion volume, and estimates conversion data through request-caliber conversion and reporting-caliber conversion in the initial stage and stable stage respectively, thereby avoiding the problem of low conversion data estimation accuracy in the initial stage due to the characteristics of the initial stage.

[0142] For example, when advertisers check the effect of advertising, they can check it through the advertising interface. Please refer to Figure 9 The delivery interface 900 includes a conversion caliber option 910, which includes a play option 911 and a report option 912. The play option 911 is used to indicate that the request caliber conversion is used as the calculation caliber of the conversion data; the report option 912 is used to indicate that the report caliber conversion is used as the calculation caliber of the conversion data.

[0143] In the indicator list, various indicators of the advertising effect are displayed, including conversion volume 921 and conversion cost 922, which are used to reflect the conversion volume and conversion cost under different calculation calibers.

[0144] Optionally, the indicator list also includes exposure 923, click volume 924, and click-through rate 925 to provide advertisers with a more comprehensive display of advertising effects.

[0145] For illustration, please refer to Figure 10 After Account 1 starts testing at the target time, when the initial phase ends, the consumption achievement rate 1010 begins to rise and reaches a stable state at around achievement rate 1. The consumption 1020 gradually increases after a certain period of time after the initial phase ends.

[0146] The consumption achievement rate is equal to the ratio of the sum of the consumption of advertisements with cost achievement to the total consumption; cost achievement refers to the achievement of cost deviation within plus or minus 20%; cost deviation = (CPA-TargetCPA) / TargetCPA, where TargetCPA refers to the user's target bid, and its value is the average of TargetCPA in the click data; CPA = consumption / optimization target conversion number.

[0147] Figure 11 This is a structural block diagram of a device for pushing recommended content provided by an exemplary embodiment of the present application. Figure 11 As shown, the device includes:

[0148] Acquisition module 1110 is configured to read historical conversion data of recommended content from a database, the historical conversion data representing data obtained from historical interactive operation reflows; obtain an expected conversion expression, the expected conversion expression including historical expected data and a future expected expression, the historical expected data corresponding to the expected reflow of the historical interactive operation before a specified time; the future expected expression including a correspondence between future expected data and an adjustment coefficient, the future expected data corresponding to the expected reflow of future interactive operations;

[0149] A determination module 1120 is configured to determine the adjustment coefficient based on the historical conversion data and the expected conversion expression;

[0150] An adjustment module 1130, configured to adjust the specified resource quantity data using the adjustment coefficient to obtain the adjusted resource quantity data of the recommended content;

[0151] The sorting module 1140 is configured to sort the recommended content according to the adjusted resource quantity data, and the sorting order is used to determine the recommended content to be pushed.

[0152] In an optional embodiment, the determining module 1120 is further configured to determine the historical expected data based on the historical conversion data and the reflux rate corresponding to the historical conversion data;

[0153] The determination module 1120 is further configured to determine a corresponding relationship between the future expected data and the adjustment coefficient based on consumption data and a historical adjustment coefficient, wherein the consumption data is used to indicate a total resource input amount.

[0154] In an optional embodiment, the determination module 1120 is also used to obtain an estimated resource quantity expression based on the historical conversion data and the expected conversion expression, wherein the estimated resource quantity expression includes a correspondence between the estimated resource quantity data and the adjustment coefficient; and determine the value of the adjustment coefficient in the estimated resource quantity expression with the specified resource quantity data as the adjustment target.

[0155] In an optional embodiment, the acquisition module 1110 is further configured to acquire historical resource quantity data and expected resource quantity data;

[0156] The determination module 1120 is further configured to determine the estimated resource expression as the ratio of the sum of the historical resource quantity data and the expected resource quantity data to the sum of the historical conversion data and the expected conversion expression.

[0157] In an optional embodiment, if Figure 12 As shown, the determining module 1120 includes:

[0158] A traversal unit 1121 is configured to traverse the values ​​of the adjustment coefficient within the price adjustment range of the adjustment coefficient;

[0159] The determination unit 1122 is configured to determine a value of the adjustment coefficient corresponding to the estimated result in response to the estimated result of the estimated resource quantity expression matching the specified resource quantity data and meeting the resource quantity requirement.

[0160] In an optional embodiment, the acquisition module 1110 is further configured to acquire a coefficient of variation within a historical preset time period, where the coefficient of variation is used to indicate a degree of dispersion of the reflux rate within the historical preset time period.

[0161] The adjustment module 1130 is further configured to adjust the reflux rate corresponding to the historical conversion data according to the coefficient of variation.

[0162] In an optional embodiment, the determination module 1120 is further configured to, in response to being in the scale-up phase, determine a correspondence between the future expected data and the adjustment coefficient based on the consumption data, the historical adjustment coefficient, and a request caliber conversion, where the request caliber conversion is used to represent a conversion event with a request interaction event as a conversion condition;

[0163] The determination module 1120 is also used to determine the correspondence between the future expected data and the adjustment coefficient in response to being in a stable stage based on the consumption data, the historical adjustment coefficient and the reporting caliber conversion, and the reporting caliber conversion is used to represent a conversion event with the reported interactive event as a conversion condition.

[0164] In an optional embodiment, the historical conversion data includes: data obtained by reflowing the historical interactive operations before the current moment on the current date, and data obtained by reflowing the historical interactive operations before the current date before the current moment on the current date;

[0165] The historical expected data includes: data obtained by reflowing the historical interactive operations before the current date after the current moment in the current date, and data obtained by reflowing the historical interactive operations before the current moment in the current date after the current moment in the current date;

[0166] The future expected data includes: data obtained by reflowing the future interactive operations after the current moment in the current date after the current moment in the current date.

[0167] To sum up, the device provided in this embodiment obtains an adjustment coefficient through historical conversion data and expected conversion expression, thereby adjusting the specified resource amount. Since the expected conversion expression is determined based on the integration of historical expected data and future expected data, that is, the adjustment coefficient is determined by historical conversion data, historical expected data, and future expected data, the resource amount data is balanced, the display accuracy of the input effect is high, and the control accuracy of the resource amount input after the resource amount data is adjusted is also high.

[0168] It should be noted that the recommended content push device provided in the above embodiment is merely an example of the division of the aforementioned functional modules. In actual applications, the aforementioned functions can be assigned to different functional modules as needed, i.e., the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. Furthermore, the recommended content push device provided in the above embodiment and the recommended content push method embodiment are based on the same concept. The specific implementation process is detailed in the method embodiment and will not be repeated here.

[0169] Figure 13 The following is a schematic diagram showing the structure of a server provided by an exemplary embodiment of the present application. Specifically:

[0170] The server 1300 includes a central processing unit (CPU) 1301, a system memory 1304 including a random access memory (RAM) 1302 and a read-only memory (ROM) 1303, and a system bus 1305 connecting the system memory 1304 and the CPU 1301. The server 1300 also includes a mass storage device 1306 for storing an operating system 1313, application programs 1314, and other program modules 1315.

[0171] The mass storage device 1306 is connected to the central processing unit 1301 through a mass storage controller (not shown) connected to the system bus 1305. The mass storage device 1306 and its associated computer-readable media provide non-volatile storage for the server 1300. In other words, the mass storage device 1306 may include a computer-readable medium (not shown) such as a hard disk or a compact disc read-only memory (CD-ROM) drive.

[0172] Without loss of generality, computer-readable media may include computer storage media and communication media. Computer storage media include volatile and non-volatile, removable and non-removable media implemented by any method or technology for storing information such as computer-readable instructions, data structures, program modules or other data. Computer storage media include RAM, ROM, Erasable Programmable Read Only Memory (EPROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other solid-state storage technologies, CD-ROM, Digital Versatile Disc (DVD) or other optical storage, tape cassettes, magnetic tape, disk storage or other magnetic storage devices. Of course, those skilled in the art will appreciate that computer storage media are not limited to the above-mentioned ones. The above-mentioned system memory 1304 and mass storage device 1306 can be collectively referred to as memory.

[0173] According to various embodiments of the present application, the server 1300 may also be connected to a remote computer on a network such as the Internet for operation. That is, the server 1300 may be connected to the network 1312 via the network interface unit 1311 connected to the system bus 1305, or the network interface unit 1311 may be used to connect to other types of networks or remote computer systems (not shown).

[0174] The memory also includes one or more programs, which are stored in the memory and configured to be executed by the CPU.

[0175] An embodiment of the present application also provides a computer device, which includes a processor and a memory, wherein the memory stores at least one instruction, at least one program, code set or instruction set, and the at least one instruction, at least one program, code set or instruction set is loaded and executed by the processor to implement the recommended content push method provided by the above-mentioned method embodiments.

[0176] An embodiment of the present application also provides a computer-readable storage medium, which stores at least one instruction, at least one program, code set or instruction set, and the at least one instruction, at least one program, code set or instruction set is loaded and executed by a processor to implement the recommended content push method provided by the above-mentioned method embodiments.

[0177] Embodiments of the present application also provide a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the method for pushing recommended content described in any of the above embodiments.

[0178] Optionally, the computer-readable storage medium may include: a read-only memory (ROM), a random access memory (RAM), a solid-state drive (SSD), or an optical disk. Among them, the random access memory may include a resistance random access memory (ReRAM) and a dynamic random access memory (DRAM). The serial numbers of the above embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0179] Those skilled in the art will understand that all or part of the steps to implement the above embodiments may be accomplished by hardware, or by a program to instruct the relevant hardware, and the program may be stored in a computer-readable storage medium, which may be a read-only memory, a disk, or an optical disk, etc.

[0180] The above description is merely an optional embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A method for pushing recommended content, characterized in that: Applied to a server, the method includes: Reading historical conversion data of recommended content from a database, where the historical conversion data is used to represent data obtained by historical interactive operations; Obtaining an expected conversion expression, wherein the expected conversion expression includes historical expected data and a future expected expression, wherein the historical expected data corresponds to the expected return flow of the historical interactive operation before a specified time, and the future expected expression includes a correspondence between the future expected data and an adjustment coefficient, wherein the future expected data corresponds to the expected return flow of the future interactive operation; Determining an estimated resource expression as a ratio of the sum of the historical resource data and the expected resource data to the sum of the historical conversion data and the expected conversion expression, wherein the estimated resource expression includes a correspondence between the estimated resource data and the adjustment coefficient; Traversing the values ​​of the adjustment coefficient within the price adjustment range of the adjustment coefficient; In response to an estimated result of the estimated resource expression matching the specified resource data and meeting the resource requirement, determining a value of the adjustment coefficient corresponding to the estimated result; Adjusting the specified resource quantity data using the adjustment coefficient to obtain adjusted resource quantity data of the recommended content; The recommended contents are sorted according to the adjusted resource quantity data, and the sorting order is used to determine the recommended contents to be pushed.

2. The method according to claim 1, characterized in that The obtaining of the desired conversion expression includes: Determining the historical expected data according to the historical conversion data and the reflux rate corresponding to the historical conversion data; The corresponding relationship between the future expected data and the adjustment coefficient is determined based on the consumption data and the historical adjustment coefficient, wherein the consumption data is used to indicate the total resource input amount.

3. The method according to claim 2, characterized in that Before determining the historical expected data according to the historical conversion data and the reflux rate corresponding to the historical conversion data, the method further includes: Obtaining a coefficient of variation within a historical preset time period, where the coefficient of variation is used to indicate a degree of dispersion of the reflux rate within the historical preset time period; The reflux rate corresponding to the historical conversion data is adjusted according to the coefficient of variation.

4. The method according to claim 2, characterized in that The determining, based on the consumption data and the historical adjustment coefficient, the corresponding relationship between the future expected data and the adjustment coefficient includes: In response to being in the scale-up phase, determining a correspondence between the future expected data and the adjustment coefficient based on the consumption data, the historical adjustment coefficient, and a request-caliber conversion, wherein the request-caliber conversion is used to represent a conversion event with a request interaction event as a conversion condition; In response to being in a stable stage, the correspondence between the future expected data and the adjustment coefficient is determined based on the consumption data, the historical adjustment coefficient and the reporting caliber conversion, and the reporting caliber conversion is used to represent a conversion event with the reported interactive event as a conversion condition.

5. The method according to any one of claims 1 to 4, characterized in that: The historical conversion data includes: data obtained by reflowing the historical interactive operations before the current moment on the current date, and data obtained by reflowing the historical interactive operations before the current date before the current moment on the current date; The historical expected data includes: data obtained by reflowing the historical interactive operations before the current date after the current moment in the current date, and data obtained by reflowing the historical interactive operations before the current moment in the current date after the current moment in the current date; The future expected data includes: data obtained by reflowing the future interactive operations after the current moment in the current date after the current moment in the current date.

6. A device for pushing recommended content, characterized in that: The device comprises: An acquisition module is configured to read historical conversion data of recommended content from a database, the historical conversion data being used to represent data obtained from historical interactive operation reflows; obtain an expected conversion expression, the expected conversion expression including historical expected data and a future expected expression, the historical expected data corresponding to the expected reflow of the historical interactive operation before a specified time; the future expected expression including a correspondence between future expected data and an adjustment coefficient, the future expected data corresponding to the expected reflow of future interactive operations; a determination module configured to determine an estimated resource expression by taking a ratio of a sum of the historical resource quantity data and the expected resource quantity data to a sum of the historical conversion data and the expected conversion expression, wherein the estimated resource expression includes a correspondence between the estimated resource quantity data and the adjustment coefficient; The determining module includes: A traversal unit, configured to traverse the values ​​of the adjustment coefficient within the price adjustment range of the adjustment coefficient; a determining unit, configured to determine a value of the adjustment coefficient corresponding to the estimated result in response to an estimated result of the estimated resource quantity expression matching the specified resource quantity data and meeting the resource quantity requirement; The device further comprises: an adjustment module, configured to adjust the specified resource quantity data using the adjustment coefficient to obtain the adjusted resource quantity data of the recommended content; A sorting module is used to sort the recommended content according to the adjusted resource quantity data, and the sorting order is used to determine the recommended content to be pushed.

7. The device according to claim 6, characterized in that The determining module is further configured to determine the historical expected data based on the historical conversion data and the reflux rate corresponding to the historical conversion data; The determination module is further configured to determine a corresponding relationship between the future expected data and the adjustment coefficient based on consumption data and a historical adjustment coefficient, wherein the consumption data is used to indicate a total resource input amount.

8. A computer device, characterized in that: The computer device includes a processor and a memory, wherein the memory stores at least one program, and the at least one program is loaded and executed by the processor to implement the method for pushing recommended content according to any one of claims 1 to 5.

9. A computer-readable storage medium, characterized in that The storage medium stores at least one program, and the at least one program is loaded and executed by the processor to implement the method for pushing recommended content according to any one of claims 1 to 5.

Images