Information push method, device, electronic device and computer readable medium

Abstract

The present disclosure relates to an information push method, device, electronic device and computer-readable medium. The method includes: obtaining promotion parameters and restriction conditions of promotion information, wherein the restriction conditions include display quantity restriction, single display restriction and display user restriction; determining characteristic parameters of the promotion information through the promotion parameters and user parameters of registered users; determining the promotion coefficient of the promotion information according to the restriction conditions, the characteristic parameters and user parameters of the current user; and pushing the promotion information to the current user through the promotion coefficient. The information push method, device, electronic device and computer-readable medium disclosed in the present disclosure can make the contract advertising delivery strategy in a real environment more reasonable and reduce the advertising effect replenishment rate in advertising promotion.

Description

Technical Field

[0001] The present disclosure relates to the field of computer information processing, and in particular, to an information push method, device, electronic device, and computer-readable medium. Background Art

[0002] With the rapid development and popularization of the Internet, online advertising has also emerged as a new advertising industry. Among them, contract advertising is the mainstream form of displaying online advertising, that is, the media must complete the corresponding advertising contract within the specified time according to the advertiser's targeting conditions. Contract advertising is also called targeted advertising. The media matches the user's user profile with the advertiser's targeting conditions for advertising, and then pushes the advertisement to the user end that meets the targeting conditions to improve the advertiser's advertising effect.

[0003] Advertisers have proposed various targeting conditions, and the existing methods for promoting targeted advertising cannot fully meet the needs of advertisers. In the actual targeted advertising promotion process, the advertising effect compensation rate often occurs due to the failure to reach the advertising contract, which causes human and financial losses to both the media and advertisers.

[0004] Therefore, a new information push method, device, electronic device and computer-readable medium are needed.

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

[0006] In view of this, the present disclosure provides an information push method, device, electronic device and computer-readable medium, which can make the contract advertising delivery strategy in a real environment more reasonable and reduce the advertising effect replenishment rate in advertising promotion.

[0007] Other features and advantages of the present disclosure will become apparent from the following detailed description, or may be learned in part by the practice of the present disclosure.

[0008] According to one aspect of the present disclosure, an information push method is proposed, which includes: obtaining promotion parameters and restriction conditions of promotion information, the restriction conditions including display quantity restriction, single display restriction, and display user restriction; determining characteristic parameters of the promotion information through the promotion parameters and user parameters of registered users; determining a promotion coefficient of the promotion information based on the restriction conditions, the characteristic parameters, and the user parameters of the current user; and pushing the promotion information to the current user through the promotion coefficient.

[0009] In an exemplary embodiment of the present disclosure, determining characteristic parameters of promotion information through promotion parameters and user parameters of registered users includes: determining mathematical expressions of characteristic parameters of promotion information through promotion parameters and user parameters of registered users; and iteratively calculating the mathematical expressions of characteristic parameters through an optimization algorithm to obtain characteristic parameters.

[0010] In an exemplary embodiment of the present disclosure, the optimization algorithm includes: a gradient descent method and its variants, and a coordinate descent method.

[0011] In an exemplary embodiment of the present disclosure, during the iterative calculation process, if the difference between the last calculation result and the current calculation result is less than a predetermined value, the iterative process is terminated.

[0012] In an exemplary embodiment of the present disclosure, determining the distribution parameters of the promotion information according to the restriction conditions, the characteristic parameters and the user parameters of the current user includes: when the current user is not a registered user, specifying the user parameters for the current user.

[0013] In an exemplary embodiment of the present disclosure, specifying user parameters for the current user includes: according to the similarity between the current user and the registered users, taking the user parameters of the registered user with the highest similarity to the current user as the user parameters of the current user.

[0014] In an exemplary embodiment of the present disclosure, determining the distribution parameters of the promotion information according to the restriction conditions, the characteristic parameters and the user parameters of the current user includes: when the current user is a registered user, using the registered user parameters as the current user parameters.

[0015] In an exemplary embodiment of the present disclosure, determining the distribution parameters of promotion information based on restriction conditions, characteristic parameters and user parameters of the current user includes: inputting the restriction conditions, characteristic parameters and user parameters of the current user into a distribution model to determine the promotion coefficient of the promotion information, and the promotion coefficient model is established by obtaining a first optimization target with constraints.

[0016] In an exemplary embodiment of the present disclosure, it also includes: establishing a promotion information constraint model through restriction conditions and user parameters of registered users; and obtaining the optimal solution expression of the promotion information constraint model as a promotion coefficient model of the promotion information.

[0017] In an exemplary embodiment of the present disclosure, establishing a promotion information constraint model through restriction conditions and registered user characteristic parameters includes: the optimization target of the promotion information constraint model is the first optimization target; the optimization restriction conditions of the promotion information constraint model are display quantity restriction, single display restriction, and display user restriction.

[0018] In an exemplary embodiment of the present disclosure, the first optimization goal includes: the promotion information is evenly displayed among registered users.

[0019] In an exemplary embodiment of the present disclosure, the promotion information constraint model is established based on an improved bipartite graph network, and the improved bipartite graph network includes a user set, a user access set, and a display information set.

[0020] In an exemplary embodiment of the present disclosure, obtaining the optimal solution expression of the generalized information constraint model as the generalized coefficient model of the generalized information includes: obtaining the optimal solution expression of the generalized information constraint model as the generalized coefficient model of the generalized information through the Lagrangian algorithm and inequality constraints.

[0021] In an exemplary embodiment of the present disclosure, pushing promotion information to the current user through a promotion coefficient includes: when the sum of multiple promotion coefficients is greater than a predetermined value, sorting multiple promotion information according to their corresponding allocation weights; and pushing part of the multiple promotion information to the current user according to the sorting result.

[0022] In an exemplary embodiment of the present disclosure, it also includes: determining the distribution weight of the promotion information through the promotion parameters of the promotion information and the estimated display quantity.

[0023] According to one aspect of the present disclosure, an information push device is proposed, which includes: a data receiving module, used to obtain promotion parameters and restriction conditions of promotion information, the restriction conditions including display quantity restriction, single display restriction, and display user restriction; a characteristic parameter module, used to determine the characteristic parameters of the promotion information through the promotion parameters and user parameters of registered users; a promotion coefficient module, used to determine the promotion coefficient of the promotion information according to the restriction conditions, characteristic parameters and user parameters of the current user; an information promotion module, used to push the promotion information to the current user through the promotion coefficient.

[0024] According to one aspect of the present disclosure, an electronic device is proposed, which includes: one or more processors; a storage device for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the method as described above.

[0025] According to one aspect of the present disclosure, a computer-readable medium is provided, on which a computer program is stored. When the program is executed by a processor, the method described above is implemented.

[0026] According to the information push method, device, electronic device and computer-readable medium disclosed in the present invention, the contract advertising delivery strategy in a real environment can be made more reasonable and the advertising effect replenishment rate in advertising promotion can be reduced.

[0027] It is to be understood that the foregoing general description and the following detailed description are exemplary only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] The above and other objects, features and advantages of the present disclosure will become more apparent by describing in detail its exemplary embodiments with reference to the accompanying drawings. The accompanying drawings described below are only some embodiments of the present disclosure, and it is clear to a person skilled in the art that other accompanying drawings can be obtained from these accompanying drawings without creative work.

[0029] Figure 1 It is a bipartite graph network diagram of contract advertisement in the prior art.

[0030] Figure 2 The present invention is a system block diagram showing a method and device for information push according to an exemplary embodiment.

[0031] Figure 3 The present invention is a schematic diagram of an application scenario of an information push method and device according to an exemplary embodiment.

[0032] Figure 4 The figure is a flow chart of an information push method according to an exemplary embodiment.

[0033] Figure 5 The figure is a flowchart of an information pushing method according to another exemplary embodiment.

[0034] Figure 6 It is a schematic diagram of an improved bipartite graph network in an information push method according to another exemplary embodiment.

[0035] Figure 7 The figure is a flowchart of an information pushing method according to another exemplary embodiment.

[0036] Figure 8 The figure is a flowchart of an information pushing method according to another exemplary embodiment.

[0037] Fig. 9 The figure is a block diagram of an information pushing device according to an exemplary embodiment.

[0038] Fig.10 The figure is a block diagram of an information pushing device according to another exemplary embodiment.

[0039] Fig.11 It is a block diagram of an electronic device according to an exemplary embodiment. DETAILED DESCRIPTION

[0040] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the embodiments set forth herein; rather, these embodiments are provided so that this disclosure will be comprehensive and complete and will fully convey the concepts of the example embodiments to those skilled in the art. The same reference numerals in the figures represent the same or similar parts, and thus their repeated description will be omitted.

[0041] In addition, the described features, structures or characteristics may be combined in one or more embodiments in any suitable manner. In the following description, many specific details are provided to provide a full understanding of the embodiments of the present disclosure. However, those skilled in the art will appreciate that the technical solutions of the present disclosure may be practiced without one or more of the specific details, or other methods, components, devices, steps, etc. may be adopted. In other cases, known methods, devices, implementations or operations are not shown or described in detail to avoid blurring the various aspects of the present disclosure.

[0042] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities may be implemented in software form, or in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices.

[0043] The flowcharts shown in the accompanying drawings are only exemplary and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps can be decomposed, and some operations / steps can be combined or partially combined, so the actual execution order may change according to actual conditions.

[0044] It should be understood that although the terms first, second, third, etc. may be used herein to describe various components, these components should not be limited by these terms. These terms are used to distinguish one component from another component. Therefore, the first component discussed below can be referred to as the second component without departing from the teachings of the disclosed concepts. As used herein, the term "and / or" includes any one of the associated listed items and all combinations of one or more.

[0045] Those skilled in the art will appreciate that the drawings are merely schematic diagrams of example embodiments, and the modules or processes in the drawings are not necessarily necessary for implementing the present disclosure, and therefore cannot be used to limit the protection scope of the present disclosure.

[0046] The terms used in this application are explained as follows:

[0047] Frequency control: Short for frequency control, it refers to controlling the number of times a user views the same advertisement so that it does not exceed the frequency limit required by the advertiser, in order to enhance the advertising value and user experience.

[0048] Multiple ads without duplication: This is a constraint on displaying ads. For a video or webpage, multiple ads can be displayed at the same time. Multiple ads without duplication means that multiple ads cannot be repeated in a video ad display, in order to improve the advertising value and user experience.

[0049] Targeting unit: refers to different columns on the website, such as sports columns or entertainment columns, each column as a targeting unit.

[0050] Targeting restriction: The number of playbacks on a specific targeting unit is limited. For example, multiple posts cannot be repeated.

[0051] Cross-targeting restriction: Cross-targeting refers to the restriction on the number of times an ad can be played when user traffic meets two or more targeting conditions at the same time, such as a user can watch both movies and sports programs, resulting in user traffic belonging to different targeting units.

[0052] The specific contents of the present invention will be described in detail below in conjunction with specific application scenarios.

[0053] The inventor of the present application has found that there are mainly two methods for promoting targeted advertising in the prior art:

[0054] 1. Allocation algorithm

[0055] Contract advertising is the mainstream form of display advertising, that is, the media (supply side, Supply) needs to complete the corresponding delivery contract within a specified time according to the targeted conditions of the advertiser (demand side, Demand). Generally speaking, this process can be regarded as an allocation problem on a bipartite graph G = (I∪J,E), such as Figure 1 I is the supplier set, which represents exposure under various directional conditions. Generally, it can reach hundreds of millions or even billions of dimensions. Each supplier node i corresponds to a different weight s. i , represents the total number of exposures under this directional condition. J is the demand side set, representing the advertiser's orders, which can generally reach thousands or tens of thousands of dimensions. Each demand side node j corresponds to a different weight d j , indicating the number of ad exposures required by the order. The line (i, j)∈E between the supply node and the demand node indicates that the supply node meets the directional conditions of the demand node.

[0056] The goal of the assignment problem is to find a feasible assignment solution where each edge has a value x ij , indicating that the traffic of node i is x ij The proportion is allocated to contract advertisement j. How to solve x ij This is the key problem of the contract advertisement allocation algorithm. In order to find the optimal x ij, which can be modeled as an optimization problem.

[0057] In the selection of objective functions, for example, maximizing advertising value and minimizing shortage can be considered at the same time: the former is to improve advertising value and advertiser satisfaction, and the latter is to ensure the revenue of the media. Generally speaking, an ideal distribution target can be set first, such as the ratio of demand to available supply, which can represent uniform exposure to the targeted audience, and then minimize x ij The distance to the assigned target.

[0058] However, a ij The resulting solution requires O(|E|) storage space. Considering that the number of nodes on the supply side is as high as billions, such a solution is difficult to implement. Moreover, the allocation solution based on the optimization model does not consider frequency control and non-repetition of multiple posts, and cannot be directly applied to actual situations.

[0059] 2. Frequency Algorithm

[0060] There is little research on the delivery algorithm under frequency control in the prior art. The calculation method of the frequency algorithm existing in the prior art is: calculate an allocation mode for each user, which is not a compact solution, consumes more memory, and the calculation of the allocation scheme is very time-consuming.

[0061] At present, the general approach to the frequency problem in the industry is the frequency ratio model. This approach is based on the fact that when an advertisement is assigned a frequency limit, only part of the available traffic that meets its targeting is available. For example, if the frequency control requirement of contract j is n days and k times, then for users with a large number of visits within n days, only the number of visits within k times is j's available inventory. The so-called frequency ratio refers to the proportion of available inventory corresponding to the frequency demand. Generally, the frequency ratio under different inventory consumption conditions can be obtained through historical statistics and dynamic simulation. In the process of inventory sales or distribution, the total inventory multiplied by the frequency ratio is the available inventory.

[0062] The frequency ratio model is currently only used in heuristic algorithms, and its allocation results are not optimal. A heuristic algorithm is an algorithm based on intuition or experience. It can give an approximate optimal solution to an instance of an optimization problem within an acceptable computational cost (computational time, space occupied, etc.). The degree of deviation of the approximate solution from the true optimal solution cannot necessarily be predicted in advance.

[0063] Moreover, the inventor of the present application has found that the general practice of the industry for allocating advertising contracts is to use the supply-side unit node to represent the traffic with the same directional conditions, rather than the traffic of a single user, which is very inconvenient for frequency control. Because under this modeling method, all users with the same directional conditions, regardless of the number of visits, use the same frequency ratio, which is unreasonable. For example, if the frequency limit of an order is 3, then for a user with 10 visits, the frequency ratio should be 0.3 (that is, 30% of the exposure can be used to display the order), and for a user with 3 or less visits, the frequency ratio should be 1. However, if these two types of users are in the same traffic unit, they will use a common frequency ratio, such as 0.5, which is obviously inappropriate for both users.

[0064] Existing frequency control algorithms also have a potential assumption, that is, a user always belongs to a unique supply-side unit. Under this assumption, the access frequency of users under each supply-side unit will be counted separately. However, in reality, this assumption is not true for some compatible targeting conditions. The so-called compatible targeting conditions refer to the attributes that a user can have at the same time, such as platform (PC or mobile), content (movie channel or sports channel), etc.; while gender, age, etc. are exclusive targeting conditions. For example, a user can access using both a PC and a mobile terminal. The traditional model will divide this user into different supply-side units for processing, that is, treat it as two users. When the targeting condition of a contract is {platform = PC + Mobile}, these "two" users can participate in advertising display, and the frequency will be calculated separately, resulting in errors in the actual frequency calculation.

[0065] The information push method of the present application improves the traditional bipartite graph model of advertising distribution, so that it can realize frequency control of cross-directional user traffic, and adds the constraint condition of non-repetition of multiple posts, and then derives the theoretical optimal solution expression under the new constraint, and gives an iterative algorithm for obtaining the optimal solution.

[0066] The specific contents of the present invention will be described in detail below with reference to the accompanying drawings.

[0067] Figure 2 The present invention is a system block diagram showing a method and device for information push according to an exemplary embodiment.

[0068] like Figure 2 As shown, the system architecture 100 may include terminal devices 101, 102, 103, a network 104 and a server 105. The network 104 is used to provide a medium for communication links between the terminal devices 101, 102, 103 and the server 105. The network 104 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0069] Users can use terminal devices 101, 102, 103 to interact with server 105 through network 104 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 101, 102, 103, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social platform software, etc.

[0070] The terminal devices 101 , 102 , and 103 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, and desktop computers.

[0071] The server 105 may be a server that provides various services, such as a background management server that supports video websites browsed by users using the terminal devices 101, 102, and 103. The server 105 may analyze and process the received data such as the user's video browsing request, and feed back the processing results (such as promotion information A, promotion information B, and promotion information D pushed to the user) to the terminal device.

[0072] The server 105 may, for example, obtain the promotion parameters and restriction conditions of the promotion information, wherein the restriction conditions include display quantity restriction, single display restriction, and display user restriction; the server 105 may, for example, determine the characteristic parameters of the promotion information through the promotion parameters and the user parameters of the registered user; the server 105 may, for example, determine the promotion coefficient of the promotion information based on the restriction conditions, the characteristic parameters, and the user parameters of the current user; the server 105 may, for example, push the promotion information to the current user through the promotion coefficient.

[0073] The server 105 can also be a server that provides various services, such as a server that supports video websites browsed by users using terminal devices 101, 102, and 103. For example, a user uses terminal device 103 (or terminal device 101 or 102) to send a video playback request to server 105. Server 105 can, for example, receive the playback request from the user, and obtain user parameters through the relevant information registered by the user on the Internet. The user parameters can include, for example, the user's age, gender, user geographic location, user feature tags, and the estimated number of user browsing times. This application is not limited to this. Any parameter value used to define user features or attributes can be used as user parameters in this application. Server 105 can also, for example, receive advertising promotion requests from advertisers, and obtain promotion parameters through the promotion request. The promotion parameters can be, for example, promotion scene restrictions for promotion information, can be, for example, geographical areas for limiting the playback of promotion information, the number of times the promotion information is played to the same user, the user characteristics of the user who plays the promotion information, and the specific unit of the promotion information played in the web page. The specific unit in the web page can be classified by the content tags displayed in the web page, and can be, for example, entertainment unit, sports unit, children's unit, etc. The promotion scenario restrictions of the promotion information may also correspond to the user parameters mentioned above, for example. The promotion information may limit the user's age, gender, user geographic location, and user feature tags, etc.

[0074] The server 105 may be a physical server, or may be composed of multiple servers. The server 105 may store offline data, or may provide real-time calculation requests. For example, some of the servers 105 provide offline data storage services, and some provide real-time calculation services. In the present application, the server 105 may store the feature parameters of the promotion information calculated in advance offline through the offline data storage service, and when the user sends a play request, the promotion coefficient of the promotion information is calculated in real time, so as to push the promotion information to the current user through the promotion coefficient.

[0075] It should be noted that the information push method provided in the embodiment of the present disclosure can be executed by the server 105, and accordingly, the information push device can be arranged in the server 105. The web page terminal provided for the user to browse the web page is generally located in the terminal devices 101, 102, and 103.

[0076] Figure 3 The present invention is a schematic diagram of an application scenario of an information push method and device according to an exemplary embodiment.

[0077] like Figure 3In the scenario shown, multiple promotion information providers submit requests for promotion information. Without loss of generality, for example, advertisers submit advertising promotion requests, which include advertising promotion parameters and restrictions, such as display to specific users, display to the same user not more than 10 times, etc.

[0078] The advertiser submits the advertising promotion request to the media party, and the media party matches the advertiser's request with the registered users. According to the advertiser's request, the feature parameters required for the actual information promotion are pre-calculated.

[0079] When a user sends a playback request or a web page opening request to the media, the media will conduct real-time matching again based on the user's characteristics and the advertiser's request for promotional information. After a successful match, the advertisements will be pushed to the users who meet the advertiser's requirements. This method can both meet the advertiser's restrictions on the advertising audience and reduce the media's advertising effect compensation rate due to the real-time advertising matching not meeting the advertiser's requirements.

[0080] Figure 4 FIG. 1 is a flow chart of an information push method according to an exemplary embodiment. Figure 4 As shown, the information push method in the present application at least includes steps S402 to S408 when implemented.

[0081] In S402, the promotion parameters and restrictions of the promotion information are obtained, and the restrictions include display quantity restriction, single display restriction, and display user restriction. The promotion information may be, for example, advertising information, and the promotion parameters may refer to the description above, for example: the total number of times the promotion information is promoted, the broadcast frequency of the promotion information, the broadcast duration of the promotion information, and the restrictions of the promotion information, which may be, for example, display quantity restriction, single display restriction, and display user restriction. Specifically, the geographical area where the promotion information is broadcast may be limited, the number of times the promotion information is broadcast to the same user may be limited, the user characteristics of the broadcast object of the promotion information may be limited, and the specific unit in the web page where the promotion information is broadcast may be limited. The specific unit in the web page may be classified by the content tags displayed in the web page, which may be, for example, entertainment unit, sports unit, children's unit, etc. The promotion scene restriction of the promotion information may also correspond to the user parameters described above, and the promotion information may limit the user's age, gender, user geographic location, and user characteristic tags, etc.

[0082] In one embodiment, the display quantity limit may be, for example, the frequency control mentioned above, which refers to controlling the number of times a user views the same advertisement so as not to exceed the upper frequency limit required by the advertiser, so as to enhance the advertising value and user experience.

[0083] In one embodiment, the single display restriction may be, for example, the above-mentioned non-repetition restriction, which means that in one web page advertisement display, multiple advertisements cannot be repeated to enhance the advertisement value and user experience.

[0084] In one embodiment, the display user restriction may be, for example, the directional restriction and cross-directional restriction mentioned above, wherein the directional restriction refers to: the number of playbacks on a specific directional unit. For example, multiple posts are not repeated. Cross-directional restriction refers to when the user traffic meets two or more directional conditions at the same time, such as a user can watch both movies and sports programs, resulting in user traffic belonging to different directional units. For such users, the frequency control restriction needs to be combined with the directional unit calculation.

[0085] It is worth mentioning that the cross-directional unit restrictions in this embodiment include but are not limited to frequency control, and the directional unit restrictions include but are not limited to non-repetition of multiple posts.

[0086] In S404, characteristic parameters of the promotion information are determined by the promotion parameters and the user parameters of the registered user. For example, a mathematical expression of the characteristic parameters of the promotion information can be determined by the promotion parameters and the user parameters of the registered user, and then the mathematical expression of the characteristic parameters is iteratively calculated by an optimization algorithm to obtain the characteristic parameters. The relevant content of optimizing and solving the mathematical expression will be described in detail later.

[0087] In one embodiment, the optimization algorithm solves the mathematical expression of the characteristic parameter including: optimizing and solving by gradient descent method or coordinate descent method.

[0088] Among them, gradient descent is a kind of iterative method, which can be used to solve least squares problems (both linear and nonlinear). When solving the model parameters of machine learning algorithms, that is, unconstrained optimization problems, gradient descent is one of the most commonly used methods. When solving the minimum value of the loss function, the gradient descent method can be used to iterate step by step to obtain the minimized loss function and model parameter values. In machine learning, two gradient descent methods have been developed based on the basic gradient descent method, namely the stochastic gradient descent method and the batch gradient descent method. The gradient descent method involved in this application can be implemented by the stochastic gradient descent method, and can also be implemented by the batch gradient descent method. The gradient descent method in this application can also be obtained by other variant methods, and this application is not limited thereto.

[0089] Coordinate descent is a non-gradient optimization algorithm. In each iteration, the coordinate descent method performs a one-dimensional search along a coordinate direction at the current point to obtain a local minimum of a function. Different coordinate directions are used cyclically throughout the process. For inseparable functions, the algorithm may not be able to obtain the optimal solution in a smaller number of iterations. In order to accelerate convergence, an appropriate coordinate system can be used, for example, a new coordinate system in which the coordinates are as uncorrelated as possible can be obtained through principal component analysis. Three descent methods have been developed based on the basic coordinate descent method, namely, the cyclic coordinate descent method, the adaptive coordinate descent method, and the ring coordinate descent method. The coordinate descent method involved in this application can be implemented by the cyclic coordinate descent method, and can also be implemented by the adaptive coordinate descent method, or the ring coordinate descent method.

[0090] In one embodiment, during the process of iteratively calculating and solving the mathematical expression of the characteristic parameter, if the difference between the last calculation result and the current calculation result is less than a predetermined value, the iterative process is terminated.

[0091] In one embodiment, the user parameters of a registered user can be obtained, for example, based on the user's registration data, and can include, for example, the user's age, gender, income, points of interest, etc. The user parameters can also be obtained in real time through the user's network environment, and can include, for example, the user's geographic location, user network characteristics, etc. The user parameters can also be obtained, for example, through the user's historical browsing data, and can include, for example, user feature tags, estimated number of user browsing times, etc.

[0092] In one embodiment, for example, the user's operation preferences and attention content can be extracted through the user's historical browsing data of the content in the webpage, and then a feature label can be determined for the user. The feature label can be, for example, a feature label of the user's habits, or a feature label of the user's operation, such as "star chaser", "fitness fanatic", "homebody", etc. The average number of views of the user can also be estimated, for example, based on the number of historical views of a certain type of attention unit by the user. For example, if the average number of historical views of a user for sports programs is 2 times a day, and the average number of historical views of entertainment programs is 5 times a day, then 2 and 5 can be used as the estimated number of views of the user respectively.

[0093] In one embodiment, after determining the characteristic parameters of the promotion information through the promotion parameters and the user parameters of the registered users, the characteristic parameters corresponding to each user are stored in the server as offline data.

[0094] In S406, the promotion coefficient of the promotion information is determined according to the restriction condition, the characteristic parameter and the user parameter of the current user. For example, a promotion information constraint model can be established through the restriction condition and the user parameter of the registered user; and the optimal solution expression of the promotion information constraint model is obtained as the promotion coefficient model of the promotion information. Then, when the user sends a play request, the restriction condition, the characteristic parameter and the user parameter of the current user are input into the allocation model in real time to determine the promotion coefficient of the promotion information.

[0095] In one embodiment, the optimization goal of the promotion information constraint model is the first optimization goal; the optimization constraint conditions of the promotion information constraint model are display quantity restriction, single display restriction, and display user restriction. Among them, in this embodiment, the first optimization goal may be, for example, that the promotion information is evenly displayed among registered users. The first optimization goal may also be, for example, that the promotion information obtains the most display opportunities within the agreed time, that the promotion information obtains the most display population among registered users, and that the promotion information obtains the maximum display revenue, etc., and this application is not limited to this.

[0096] In one embodiment, the promotion information constraint model is established based on an improved bipartite graph network, which includes a user set, a user access set, and a display information set. The optimal solution expression of the promotion information constraint model is obtained by Lagrangian algorithm and inequality constraint conditions to serve as the promotion coefficient model of the promotion information. The relevant content of the promotion information constraint model will be described in detail later.

[0097] In S408, the promotion information is pushed to the current user according to the promotion coefficient. For example, the promotion information can be pushed to the current user in sequence according to the promotion coefficient. For example, the promotion information can be sorted according to the promotion coefficient and the weights of different promotion information providers, and part of the promotion information in the plurality of promotion information can be pushed to the current user according to the sorting result.

[0098] According to the information push method disclosed in the present invention, by comprehensively considering user characteristics, promotion information characteristics, and restriction conditions in the promotion information when solving the promotion coefficient of the promotion information, the solved promotion system can meet the needs of contract advertising delivery in a real environment, and can also make the contract advertising delivery strategy in a real environment more reasonable, thereby reducing the advertising effect replenishment rate in advertising promotion.

[0099] It should be clearly understood that the present disclosure describes how to form and use specific examples, but the principles of the present disclosure are not limited to any details of these examples. On the contrary, based on the teachings of the contents disclosed in the present disclosure, these principles can be applied to many other embodiments.

[0100] Figure 5 FIG. 1 is a flow chart of an information push method according to an exemplary embodiment. Figure 5 Steps S502-S506 are as follows: Figure 4 In step S406 shown, further description is given of "determining the promotion coefficient of the promotion information according to the restriction condition, the characteristic parameter and the user parameter of the current user".

[0101] like Figure 5 As shown, in S502, a promotion information constraint model is established through the restriction condition and the user parameters of the registered user. The promotion information constraint model is established based on an improved bipartite graph network, which includes a user set, a user access set and a display information set.

[0102] In common online display advertising scenarios, each user may visit the media (video or web page) multiple times; each visit may generate multiple ad exposure opportunities (the same video or web page usually has multiple patch ads). In this application, for example, each independent user's visit can be called a UV (User View); each independent media visit is called a Slot, also known as a slot; each independent ad exposure is called an impression. For frequency control constraints (cross-directional restrictions), restrictions need to be made at the user level; for multi-post non-repetition constraints (single-directional restrictions), restrictions need to be made at the slot level.

[0103] For this reason, in this application, Figure 1 The traditional advertising allocation bipartite graph G = (I∪J, E) shown in the figure is improved by adding a virtual user set K, where each user node k represents a UV, u k Represents all the inventory of the user; each supplier node i represents a slot, s i Indicates the number of exposures corresponding to the slot. Figure 6 FIG. 1 is a schematic diagram of an improved bipartite graph network in an information push method according to another exemplary embodiment. Figure 6 For convenience, Ω(k) is used to represent the child node Slot set of user k, and P(i) is the parent node of i, that is, the user to which it belongs.

[0104] A variety of corresponding relationships can be obtained by improving the bipartite graph network:

[0105] 1. The correspondence between order and slot (Supply node) E order-slot ;

[0106] 2. Correspondence between UV and Slot E user-slot ;

[0107] 3. The correspondence between orders and UVs order-user .

[0108] Moreover, a Slot belongs to and only belongs to one UV.

[0109] The benefit of Supply nodes representing Slots is that in the online phase, i That is, the number of exposures of the current slot on user i, which is a known quantity. In addition, this setting can constrain multiple posts from being repeated.

[0110] Let f kj represents the frequency requirement of order j on user k, that is, the maximum number of times user k views order j is f kj , then for the number of visits exceeding f kj For users, only part of the corresponding inventory is available. Considering the situation where users cross traffic units, the following frequency control constraints (cross-directional restrictions) can be added to the promotion information constraint model in this application:

[0111]

[0112] Where k = p(i), that is, the user to which slot node i belongs. In addition, the following unidirectional restriction constraint can be obtained:

[0113]

[0114] In summary, the promotion information constraint model (1) in this application is:

[0115]

[0116]

[0117]

[0118]

[0119]

[0120]

[0121]

[0122] Among them: Γ(i) and Γ(j) represent the neighbor node sets of i and j respectively, s i Indicates the number of exposures corresponding to the Slot. To allocate the target, S j =∑ i∈Γ(j) s i Indicates d j Total available flow, V j represents the importance of order j, x ijIndicates that the traffic of node i is x ij The proportion is allocated to contract advertising j, p j represents the penalty coefficient when order j is out of stock, u j Represents the amount of compensation for less airing of advertisements.

[0123] The first inequality constraint is a demand constraint, the second inequality constraint is a supply constraint, the third and fourth inequality constraints are non-negative constraints, the fifth inequality constraint is a cross-directional constraint, and the last inequality constraint is a single-directional constraint.

[0124] In S504, the optimal solution expression of the promotion information constraint model is obtained as the promotion coefficient model of the promotion information. The optimization target of the promotion information constraint model is the first optimization target; the optimization constraint conditions of the promotion information constraint model are the display quantity limit, the single display limit, and the display user limit. The optimal solution expression of the promotion information constraint model is obtained by Lagrangian algorithm and inequality constraint conditions as the promotion coefficient model of the promotion information.

[0125] In one embodiment, according to the optimization theory, the optimal solution of the model, that is, the optimal allocation scheme, is derived. j represents the demand-side duality, β i represents the supply-side duality, γ ij Represents x ij The non-negative dual of j Indicates u j Non-negative dual, λ kj represents the cross-directed restriction duality, η ij represents a single directional restricted dual, then according to the KKT (Karush-Kuhn-Tucker) condition, x ij The corresponding stability condition (stationarity) is:

[0126] For all

[0127] γ ij , kj and η ij The corresponding complementary slackness condition is:

[0128] For all i, j, either γ ij =0 or x ij =0

[0129] For all i, j, either η ij =0 or s i x ij =fij

[0130] For all k, j, or

[0131] γ ij , kj and η ij The corresponding dual feasibility condition is:

[0132] γ ij ≥0,η ij ≥0,λ kj ≥0

[0133] According to the stability condition,

[0134]

[0135] In one embodiment, γ ij and η ij All are eliminated, only α is retained j , β i and λ kj , thus obtaining a compact allocation scheme. First, according to γ ij The corresponding complementary relaxation conditions are either

[0136] γ ij =0 (3)

[0137] or

[0138] x ij =0

[0139] Therefore, when x ij When γ ij will increase so that x ij =0, that is

[0140]

[0141] At the same time, At that time, η ij will decrease so that

[0142] In summary, there are

[0143]

[0144] make but

[0145] x ij =h ij (α j -β i -λ kj) (4)

[0146] The mathematical expression corresponding to Formula 4 is the allocation scheme under the constraints of cross-directional unit frequency control and directional unit restriction, where α j represents the competitiveness of the order, β i represents the scarcity of the dimension, λ kj Indicates the importance of the order to the user.

[0147] According to u j The stability condition and α j , β i , kj The complementary relaxation condition can also lead to the following conclusions: Assuming α j , β i , kj They are demand duality, supply duality, and user frequency duality respectively, and their optimal values ​​satisfy:

[0148] For all j, 0≤α j ≤p j . And, either α j =p j , or j∈Γ(j) s i x ij =d j ;

[0149] For all i, β i ≥ 0. And, either β i =0, or∑ j∈Γ(i) x ij =1.

[0150] For all (k,j), and either λ kj =0, or∑ i∈Γ(k，j) S i xi j =f kj .

[0151] In one embodiment, the mathematical expression corresponding to Formula 4 is used as a generalized coefficient model for subsequent processing.

[0152] In S506, the restriction condition, the characteristic parameter and the user parameter of the current user are input into a promotion coefficient model to determine the promotion coefficient of the promotion information. For example, when the user sends a play request, the promotion coefficient of the promotion information can be calculated in real time through the promotion coefficient model and the restriction condition, the characteristic parameter and the user parameter of the current user.

[0153] In one embodiment, determining the distribution parameters of the promotion information according to the restriction condition, the characteristic parameter and the user parameter of the current user includes: when the current user is not a registered user, specifying the user parameter for the current user. Specifically, for example, according to the similarity between the current user and the registered user, the user parameter of the registered user with the highest similarity to the current user is used as the user parameter of the current user.

[0154] According to the information push method disclosed in the present invention, a promotion information constraint model is established by comprehensively considering user characteristics, promotion information characteristics, and restriction conditions in the promotion information, and the optimal solution obtained by the model is used as the promotion coefficient model, and then the subsequent promotion coefficient calculation method is performed. This method can fully and truly reflect the actual characteristics of the contract advertisement through a mathematical model, and obtain a more practical promotion coefficient through the model.

[0155] Figure 7 FIG. 1 is a flow chart of an information push method according to an exemplary embodiment. Figure 7 Steps S702-S704 shown are for Figure 4 In step S404 shown, further description is given on "determining characteristic parameters of the promotion information by using the promotion parameters and user parameters of registered users".

[0156] like Figure 7 As shown, in S702, a mathematical expression of a characteristic parameter of the promotion information is determined by using the promotion parameter and a user parameter of a registered user.

[0157] α j Represents the competitiveness of the order, and the characteristic expression can be, for example: α j =α j +ε j *(d j -∑ i∈Γ(j) s i x ij );

[0158] β i Indicates the scarcity of the dimension. The characteristic expression can be, for example: β i =β i +ε i *(∑ j∈Γ(i) s i x ij -s i );

[0159] λ kj Indicates the importance of the user to the order. The feature expression can be, for example: kj =λ kj +ε kj *(∑ i∈Γ(k，j) si x ij -f kj ).

[0160] In S704, the mathematical expression of the characteristic parameter is iteratively calculated by the optimization algorithm to obtain the characteristic parameter. In the iterative calculation process, if the difference between the last calculation result and the current calculation result is less than a predetermined value, the iterative process is terminated. The optimization algorithm includes: gradient descent method and coordinate descent method. In this embodiment, the gradient descent method can be used to iterate a certain number of times to calculate a set of α j , β i , kj , as a better solution.

[0161] When optimizing, we first perform an initialization step: for all orders j, let α j =0; for all slots i, let β i =0; for all user order pairs (k, j), let λ kj =0.

[0162] Run the following steps in parallel and iterate until the difference between the last calculation result and the current calculation result is less than the predetermined value:

[0163] For each slot i, β is calculated by the gradient update formula i =β i +ε i *(∑ j∈Γ(i) s i x ij -s i ).

[0164] If β i <0, then let β i -0;

[0165] For each order j, α is calculated by the gradient update formula j =α j +ε j *(d j -∑ i∈Γ(j) s i X ij ). If α j >p j , then let α j =p j If α j <0, then let α j =0;

[0166] For each user order pair (k, j), calculate λ using the gradient update formula kj =λ kj +εkj *(∑ i∈Γ(k，j) s i x ij -f kj );

[0167] If λ kj <0, then let λ kj =0

[0168] After the iterative calculation is completed, output α j , β i ,λ kj .like Figure 6 As shown, α can be calculated, for example, in an offline phase j , β i ,λ kj And save, and further obtain the optimal allocation solution x according to formula (4) in the online stage ij .

[0169] Figure 8 FIG. 1 is a flow chart of an information push method according to an exemplary embodiment. Figure 8 Steps S802-S808 shown are for Figure 4 In step S408 shown, further description is given on "pushing the promotion information to the current user through the promotion coefficient".

[0170] In some cases, for an exposure i, all x ij The sum (j∈Γ(i)) is not equal to 1. ij If the sum is less than 1, it means that a suitable user may not be selected for display, resulting in empty orders; when x ij If the sum is greater than 1, it is necessary to select some important users first and then promote them. To this end, in this application, when the sum of the promotion coefficients of all the information to be promoted is greater than the predetermined value, the priority is first defined for the user, and then the promotion is carried out according to the priority. This solution uses the ratio of demand to available traffic under various directional restrictions as the allocation priority: j / S j ′.

[0171] like Figure 8 As shown, in S802, the generalization coefficients of all information to be generalized are determined. For example, for each j∈Γ(i), x is calculated. ij =h ij (α j -β i -λ kj )

[0172] In S804, it is determined whether the sum of the promotion coefficients of all the information to be promoted is greater than a predetermined value.

[0173] In S806, when the sum of the promotion coefficients of all the information to be promoted is greater than a predetermined value, the allocation weight of the promotion information is determined by the promotion parameters of the promotion information and the estimated display quantity.

[0174] In S8062, the promotion information is sorted according to its corresponding allocation weight. j∈Γ(i) x ij > 1, for example, for all j∈Γ(i) according to d j / S j 'Sort from largest to smallest.

[0175] In S8064, some of the promotion information in the plurality of promotion information is pushed to the current user according to the sorting result. For example, the first l orders are selected so that ∑ l x ij =1.

[0176] In S808, when the sum of the promotion coefficients of all the information to be promoted is less than or equal to the predetermined value, the promotion information is pushed to the current user. ij The probability of selecting order j for display.

[0177] According to the information push method disclosed in the present invention, the user's cross-directional frequency control constraints and slot constraints are taken into account in the modeling, which can make the contract advertising delivery strategy in the real environment more reasonable, reduce the order replenishment rate, and enhance the overall interests of the company and advertisers.

[0178] Those skilled in the art will appreciate that all or part of the steps to implement the above embodiments are implemented as a computer program executed by a CPU and a GPU. When the computer program is executed by the CPU and the GPU, the above functions defined by the above method provided by the present disclosure are performed. 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.

[0179] In addition, it should be noted that the above figures are only schematic illustrations of the processes included in the method according to the exemplary embodiments of the present disclosure, and are not intended to be limiting. It is easy to understand that the processes shown in the above figures do not indicate or limit the time sequence of these processes. In addition, it is also easy to understand that these processes can be performed synchronously or asynchronously, for example, in multiple modules.

[0180] The following are embodiments of the device disclosed herein, which can be used to execute the method embodiments disclosed herein. For details not disclosed in the device embodiments disclosed herein, please refer to the method embodiments disclosed herein.

[0181] Fig. 9 FIG. 1 is a block diagram of an information push device according to an exemplary embodiment. Fig. 9As shown, the information pushing device 90 includes: a data receiving module 902 , a characteristic parameter module 904 , a promotion coefficient module 906 , and an information promotion module 908 .

[0182] The data receiving module 902 is used to obtain promotion parameters and restriction conditions of the promotion information, wherein the restriction conditions include display quantity restriction, single display restriction, and display user restriction.

[0183] In one embodiment, the display quantity limit of the data receiving module 902 may be, for example, the frequency control mentioned above, which refers to controlling the number of times a user views the same advertisement so as not to exceed the frequency limit required by the advertiser, so as to enhance the advertising value and user experience.

[0184] In one embodiment, the single display restriction of the data receiving module 902 may be, for example, the above-mentioned non-repetitive restriction, which means that in one web page advertisement display, multiple advertisements cannot be repeated to enhance the advertisement value and user experience.

[0185] In one embodiment, the display user restriction of the receiving data module 902 may be, for example, the directional restriction and cross-directional restriction mentioned above, wherein the directional restriction refers to: the number of playbacks on a specific directional unit. For example, multiple posts are not repeated. Cross-directional restriction refers to when the user traffic meets two or more directional conditions at the same time, such as a user can watch both movies and sports programs, resulting in user traffic belonging to different directional units. For such users, frequency control restrictions need to merge directional unit calculations.

[0186] The characteristic parameter module 904 is used to determine the characteristic parameters of the promotion information through the promotion parameters and the user parameters of the registered users. For example, the mathematical expression of the characteristic parameters of the promotion information can be determined through the promotion parameters and the user parameters of the registered users, and then the mathematical expression of the characteristic parameters is iteratively calculated through an optimization algorithm to obtain the characteristic parameters. The relevant content about optimizing and solving the mathematical expression will be described in detail later.

[0187] In one embodiment, the optimization algorithm of the feature parameter module 904 solves the mathematical expression of the feature parameter by: optimizing and solving by a gradient descent method or a coordinate descent method.

[0188] In one embodiment, during the process of iteratively calculating and solving the mathematical expression of the characteristic parameter, if the difference between the last calculation result and the current calculation result is less than a predetermined value, the iterative process is terminated.

[0189] In one embodiment, the characteristic parameter module 904 may determine the characteristic parameters of the promotion information by, for example, using the promotion parameters and the user parameters of the registered users, and then save the characteristic parameters corresponding to each user as offline data in the server.

[0190] The promotion coefficient module 906 is used to determine the promotion coefficient of the promotion information according to the restriction condition, the characteristic parameter and the user parameter of the current user.

[0191] For example, the promotion coefficient module 906 establishes a promotion information constraint model through restriction conditions and user parameters of registered users; and obtains the optimal solution expression of the promotion information constraint model as the promotion coefficient model of the promotion information. Then, when the user sends a play request, the restriction conditions, feature parameters and user parameters of the current user are input into the allocation model in real time to determine the promotion coefficient of the promotion information.

[0192] In one embodiment, the optimization target of the promotion information constraint model in the promotion coefficient module 906 is the first optimization target; the optimization restriction conditions of the promotion information constraint model are display quantity restriction, single display restriction, and display user restriction. Among them, in this embodiment, the first optimization target may be, for example, that the promotion information is evenly displayed among registered users. The first optimization target may also be, for example, that the promotion information obtains the most display opportunities within the agreed time, that the promotion information obtains the most display population among registered users, and that the promotion information obtains the maximum display revenue, etc., and the present application is not limited thereto.

[0193] In one embodiment, the promotion information constraint model in the promotion coefficient module 906 is established based on an improved bipartite graph network, and the improved bipartite graph network includes a user set, a user access set, and a display information set. The optimal solution expression of the promotion information constraint model is obtained by the Lagrangian algorithm and the inequality constraint condition to serve as the promotion coefficient model of the promotion information. The relevant content of the promotion information constraint model will be described in detail later.

[0194] The information promotion module 908 is used to push the promotion information to the current user according to the promotion coefficient. For example, the promotion information can be pushed to the current user in sequence according to the promotion coefficient. For example, the promotion information can be sorted according to the promotion coefficient and the weights of different promotion information providers, and part of the promotion information in the multiple promotion information can be pushed to the current user according to the sorting result.

[0195] According to the information push device disclosed in the present invention, by comprehensively considering user characteristics, promotion information characteristics, and restriction conditions in the promotion information when solving the promotion coefficient of the promotion information, the solved promotion system can meet the needs of contract advertising in a real environment, and can also make the contract advertising strategy in the real environment more reasonable, thereby reducing the advertising effect replenishment rate in advertising promotion.

[0196] Fig.10 1 is a block diagram of an information push device according to another exemplary embodiment. The information push device 100 includes the receiving data module 902, the characteristic parameter module 904, the promotion coefficient module 906, and the information promotion module 908 in the information push device 90, and further includes: a characteristic parameter optimization module 1002, a similarity calculation module 1004, a promotion coefficient optimization module 1006, and a judgment and sorting module 1008.

[0197] The characteristic parameter optimization module 1002 is used to determine the mathematical expression of the characteristic parameter of the promotion information through the promotion parameter and the user parameter of the registered user; and to iteratively calculate the mathematical expression of the characteristic parameter through an optimization algorithm to obtain the characteristic parameter.

[0198] The similarity calculation module 1004 is used to specify user parameters for the current user when the current user is not a registered user. According to the similarity between the current user and the registered users, the user parameters of the registered user with the highest similarity to the current user are used as the user parameters of the current user.

[0199] The promotion coefficient optimization module 1006 is used to establish a promotion information constraint model through the restriction conditions and the user parameters of the registered users; and to obtain the optimal solution expression of the promotion information constraint model as the promotion coefficient model of the promotion information.

[0200] The judgment sorting module 1008 is used to sort the multiple promotion information according to their corresponding allocation weights when the sum of the multiple promotion coefficients is greater than a predetermined value; and push part of the multiple promotion information to the current user according to the sorting result.

[0201] Fig.11 It is a block diagram of an electronic device according to an exemplary embodiment.

[0202] Refer to the following Fig.11 1100 according to this embodiment of the present disclosure is described. Fig.11 The electronic device 1100 shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0203] like Fig.11 As shown, the electronic device 1100 is in the form of a general computing device. The components of the electronic device 1100 may include, but are not limited to: at least one processing unit 1110, at least one storage unit 1120, a bus 1130 connecting different system components (including the storage unit 1120 and the processing unit 1110), a display unit 1140, etc.

[0204] The storage unit stores a program code, which can be executed by the processing unit 1110, so that the processing unit 1110 executes the steps according to various exemplary embodiments of the present disclosure described in the electronic prescription circulation processing method section of this specification. For example, the processing unit 1110 can execute the following steps: Figure 4 , Figure 5 , Figure 7 ,as well as Figure 8 Follow the steps shown in .

[0205] The storage unit 1120 may include a readable medium in the form of a volatile storage unit, such as a random access memory unit (RAM) 11201 and / or a cache memory unit 11202 , and may further include a read-only memory unit (ROM) 11203 .

[0206] The storage unit 1120 may also include a program / utility 11204 having a set (at least one) of program modules 11205, such program modules 11205 including but not limited to: an operating system, one or more application programs, other program modules and program data, each of which or some combination may include the implementation of a network environment.

[0207] Bus 1130 may represent one or more of several types of bus structures, including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.

[0208] The electronic device 1100 may also communicate with one or more external devices 1200 (e.g., keyboards, pointing devices, Bluetooth devices, etc.), may also communicate with one or more devices that enable a user to interact with the electronic device 1100, and / or communicate with any device that enables the electronic device 1100 to communicate with one or more other computing devices (e.g., routers, modems, etc.). Such communication may be performed via an input / output (I / O) interface 1150. Furthermore, the electronic device 1100 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) via a network adapter 1160. The network adapter 1060 may communicate with other modules of the electronic device 1100 via a bus 1130. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device 1100, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0209] Through the description of the above implementation, it is easy for those skilled in the art to understand that the example implementation described here can be implemented by software, or by software combined with necessary hardware. Therefore, the technical solution according to the implementation of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which can be a personal computer, a server, or a network device, etc.) to execute the above method according to the implementation of the present disclosure.

[0210] As another aspect, the present disclosure further provides a computer-readable medium, which may be included in the electronic device described in the above embodiment; or may exist independently without being assembled into the electronic device. The above computer-readable medium carries one or more programs, and when the above one or more programs are executed by an electronic device, the electronic device implements the method described in the above embodiment. For example, the electronic device may implement the steps shown above.

[0211] Those skilled in the art will appreciate that the above modules can be distributed in the device according to the description of the embodiment, or can be changed accordingly and only used in one or more devices different from the embodiment. The modules of the above embodiments can be combined into one module, or further divided into multiple sub-modules.

[0212] Through the description of the above embodiments, it is easy for those skilled in the art to understand that the example embodiments described here can be implemented by software, or by software combined with necessary hardware. Therefore, the technical solution according to the embodiment of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, and includes several instructions to enable a computing device (which can be a personal computer, a server, a mobile terminal, or a network device, etc.) to execute the method according to the embodiment of the present disclosure.

[0213] The exemplary embodiments of the present disclosure are specifically shown and described above. It should be understood that the present disclosure is not limited to the detailed structures, configurations or implementations described herein; on the contrary, the present disclosure is intended to cover various modifications and equivalent configurations included in the spirit and scope of the appended claims.

[0214] In addition, the structures, proportions, sizes, etc. shown in the drawings of the specification are only used to match the contents disclosed in the specification for the understanding and reading of those skilled in the art, and are not used to limit the limiting conditions for the implementation of the present disclosure, so they have no substantial technical significance. Any modification of the structure, change of the proportional relationship or adjustment of the size, without affecting the technical effects and purposes that can be achieved by the present disclosure, should still fall within the scope of the technical contents disclosed by the present disclosure. At the same time, the terms such as "on", "first", "second" and "one" quoted in this specification are only for the convenience of description, and are not used to limit the scope of the implementation of the present disclosure. The change or adjustment of their relative relationship should also be regarded as the scope of the implementation of the present disclosure without substantially changing the technical contents.

Claims

1. An information push method, characterized in that: include: Obtaining promotion parameters and restriction conditions of promotion information, wherein the restriction conditions include display quantity restriction, single display restriction, and display user restriction; the display user restriction includes directional restriction and cross-directional restriction, wherein the directional restriction indicates the limit on the number of times a specific directional unit is played, and the cross-directional restriction indicates the limit on the number of times a user's traffic meets two or more directional conditions at the same time; Determining characteristic parameters of the promotion information through the promotion parameters and the user parameters of the registered user; wherein the promotion information is used as an order of the demander, and the characteristic parameters include competitiveness of the order of the demander, scarcity of the dimension, and importance of the order to the user; A promotion information constraint model is established through the constraint conditions and the user parameters of the registered user; the optimization constraint conditions of the promotion information constraint model include a cross-directional constraint constraint and a unidirectional constraint constraint corresponding to the display user constraint; the cross-directional constraint constraint is obtained based on the correspondence between the demander's order and the user and the frequency requirement of the order on the user of the supplier, and the unidirectional constraint is obtained based on the correspondence between the order and the supplier and the frequency requirement of the order on the supplier; Obtaining the optimal solution expression of the generalization information constraint model as the generalization coefficient model of the generalization information; The restriction condition, the characteristic parameter and the user parameter of the current user are input into the promotion coefficient model to determine the promotion coefficient of the promotion information. The promotion coefficient model is established by obtaining the first optimization target with constraint conditions. The promotion coefficient is used to indicate that the flow of the supplier has x. ij The proportion of the order is allocated to the demander; When the sum of the plurality of promotion coefficients is greater than a predetermined value, using the ratio of the number of advertisement exposures required by the order to the available traffic under various directional restrictions as the allocation priority of the order, and sorting the plurality of orders according to the allocation priority; and According to the sorting result, the first l orders are selected and pushed to the current user, so that the sum of the multiple promotion coefficients corresponding to the first l orders is equal to a predetermined value; When the sum of the plurality of promotion coefficients is less than or equal to a predetermined value, i j The probability of selecting an order is pushed to the current user for display.

2. The method according to claim 1, characterized in that Determining the characteristic parameters of the promotion information by using the promotion parameters and the user parameters of the registered user includes: Determine a mathematical expression of a characteristic parameter of the promotion information by using the promotion parameter and a user parameter of a registered user; and The mathematical expression of the characteristic parameter is iteratively calculated through an optimization algorithm to obtain the characteristic parameter.

3. The method according to claim 1, characterized in that The restriction condition, the characteristic parameter and the user parameter of the current user are input into a promotion coefficient model to determine the promotion coefficient of the promotion information: When the current user is not a registered user, specify user parameters for the current user.

4. The method according to claim 3, characterized in that Specifying user parameters for the current user includes: According to the similarity between the current user and the registered users, the user parameters of the registered user with the highest similarity to the current user are used as the user parameters of the current user.

5. The method according to claim 1, characterized in that Establishing a promotion information constraint model through the restriction condition and the user parameters of the registered user includes: The optimization target of the promotion information constraint model is the first optimization target; The optimization restriction conditions of the promotion information constraint model also include the display quantity restriction and the single display restriction.

6. The method according to claim 1, characterized in that The promotion information constraint model is established based on an improved bipartite graph network, and the improved bipartite graph network includes a user set, a user access set and a display information set.

7. The method according to claim 1, characterized in that Obtaining the optimal solution expression of the generalization information constraint model as the generalization coefficient model of the generalization information includes: The optimal solution expression of the generalized information constraint model is obtained through the Lagrangian algorithm and the inequality constraint condition to serve as the generalized coefficient model of the generalized information.

8. An information push device, characterized in that: include: A data receiving module is used to obtain promotion parameters and restriction conditions of promotion information, wherein the restriction conditions include display quantity restriction, single display restriction, and display user restriction; the display user restriction includes directional restriction and cross-directional restriction, wherein the directional restriction indicates the limit on the number of times a specific directional unit is played, and the cross-directional restriction indicates the limit on the number of times a user's traffic meets two or more directional conditions at the same time; A characteristic parameter module, used to determine the characteristic parameters of the promotion information through the promotion parameters and the user parameters of the registered users; wherein the promotion information is used as the order of the demander, and the characteristic parameters include the competitiveness of the order of the demander, the scarcity of the dimension and the importance of the user to the order; A promotion coefficient module is used to establish a promotion information constraint model through the restriction conditions and the user parameters of the registered users; the optimization restriction conditions of the promotion information constraint model include the cross-directional restriction constraints and the unidirectional restriction constraints corresponding to the display user restriction; the cross-directional restriction constraints are obtained according to the corresponding relationship between the demander's order and the user and the frequency requirement of the order on the user of the supplier, and the unidirectional restriction constraints are obtained according to the corresponding relationship between the order and the supplier and the frequency requirement of the order on the supplier; the optimal solution expression of the promotion information constraint model is obtained as the promotion coefficient model of the promotion information; the restriction conditions, the characteristic parameters and the user parameters of the current user are input into the promotion coefficient model to determine the promotion coefficient of the promotion information, and the promotion coefficient model is established by obtaining the first optimization target with constraint conditions, and the promotion coefficient is used to indicate that the supplier's traffic has x ij The proportion of the order is allocated to the demander; The information promotion module is used to use the ratio of the number of advertisement exposures required by the order to the available traffic under various directional restrictions as the allocation priority of the order when the sum of the multiple promotion coefficients is greater than a predetermined value, sort the multiple orders according to the allocation priority; and select the first l orders according to the sorting result and push them to the current user, so that the sum of the multiple promotion coefficients corresponding to the first l orders is equal to the predetermined value; when the sum of the multiple promotion coefficients is less than or equal to the predetermined value, select the first l orders according to the sorting result and push them to the current user; i j The probability of selecting an order is pushed to the current user for display.

9. An electronic device, characterized in that: include: one or more processors; A storage device for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 7.

10. A computer readable medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

11. A software product, characterized in that: The method comprises several instructions to enable a computing device to execute the method according to any one of claims 1 to 7.

Images