CN114971726A - Advertisement point location screening method and device, computer equipment and computer readable storage medium - Google Patents
Advertisement point location screening method and device, computer equipment and computer readable storage medium Download PDFInfo
- Publication number
- CN114971726A CN114971726A CN202210615229.0A CN202210615229A CN114971726A CN 114971726 A CN114971726 A CN 114971726A CN 202210615229 A CN202210615229 A CN 202210615229A CN 114971726 A CN114971726 A CN 114971726A
- Authority
- CN
- China
- Prior art keywords
- advertisement
- point
- point location
- selection
- location
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0242—Determining effectiveness of advertisements
- G06Q30/0245—Surveys
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Strategic Management (AREA)
- Finance (AREA)
- Development Economics (AREA)
- Accounting & Taxation (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Game Theory and Decision Science (AREA)
- Data Mining & Analysis (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of offline advertisement putting, and discloses an advertisement point location screening method, a device, computer equipment and a computer readable storage medium, wherein the method comprises the steps of acquiring historical point location data matched with advertisement putting directional information according to the advertisement putting directional information of a client, calculating to obtain corresponding selected probability values according to corresponding selected times aiming at each advertisement point location recorded in the historical point location data, acquiring corresponding point location characteristic data, training a machine learning model by using the point location characteristic data and the selected probability values to obtain a selected probability prediction model, predicting to obtain the matching values of each candidate advertisement and the client by using the selected probability prediction model, realizing the purpose of automatically selecting suitable advertisement points for the client, further obviously improving the point selection efficiency and leading the point selection result to have the characteristic of high reliability, the customer satisfaction and the advertisement putting efficiency are improved.
Description
Technical Field
The invention belongs to the technical field of offline advertisement putting, and particularly relates to an advertisement point location screening method and device, computer equipment and a computer readable storage medium.
Background
The advertisement player is a new generation of intelligent equipment, forms a complete advertisement broadcasting control system through terminal software control, network information transmission and multimedia terminal display, and carries out advertisement propaganda through multimedia materials such as pictures, characters, videos and/or small plug-ins (such as weather or exchange rate). With the continuous development and application of the advertisement machine technology, the advertisement machine gradually moves into places such as office buildings, building districts and the like, the advertisement information acquired by the advertisement machine becomes an essential part in the life of people, and in order to meet the advertisement putting requirement, an advertiser (namely a client relative to an advertisement service provider) needs to select an advertisement site in advance to put the advertisement. The advertisement point location is a place where the advertisement machine is arranged, one advertisement point location corresponds to one advertisement machine, and the advertisement point location has a unique number, so that the background server can screen the specific advertisement point location through the number and can send information to the specific advertisement machine, and the advertisement machine plays the set advertisement.
The elevator advertising spots refer to advertising spots located in the elevator car. In the process of selecting elevator advertisement sites by advertisers, a traditional label screening mode is generally adopted, however, the label screening mode has the following limitations when targeted delivery is carried out: (1) existing tags cannot directly match the advertiser's ideal tags, such as: the ideal tag desired by the advertiser is a, and the point of sale provider can only provide the relevant tags a1 and a2, etc.; (2) the advertisement site location provider cooperates with a plurality of data sources, the quality of different data sources is different, and if an advertiser selects a low-quality label when selecting the label, the reliability of the overall data of the advertisement platform is influenced, and the customer experience and the repeated purchase rate are influenced; (3) when the multiple labels are screened, the label weight is difficult to determine, if the floor corresponding to the intersection of the multiple labels is simply taken, the target floor is too few to meet the delivery budget of an advertiser; (4) the label screening needs manual operation, point positions need to be adjusted repeatedly in order to meet the advertising budget of an advertiser, the point selection period is long, and the cost is high.
Disclosure of Invention
In order to solve the limitation problem of the existing label-based advertisement site selection mode during targeted delivery, the invention aims to provide a novel advertisement site selection method, a novel advertisement site selection device, a novel computer device and a novel computer-readable storage medium, which can obviously improve the site selection efficiency, enable the site selection result to have the characteristics of high reliability and the like, and are beneficial to improving the customer satisfaction and the advertisement delivery efficiency.
In a first aspect, the present invention provides a method for screening advertisement spots, including:
acquiring historical point selection data matched with the advertisement putting targeting information according to the advertisement putting targeting information of a client, wherein the historical point selection data records advertisement point location selection information generated in each manual point selection before the current moment, and the advertisement point location selection information comprises at least one advertisement point location selected in the corresponding manual point selection;
aiming at each advertisement point position recorded in the historical point selection data, calculating to obtain a corresponding selected probability value according to corresponding selected times;
acquiring corresponding point location characteristic data aiming at each advertisement point location, wherein the point location characteristic data comprises a plurality of point location characteristic values;
importing the point location feature data of each advertisement point location as input data required by model training and the selected probability value of each advertisement point location as output data required by model training into a machine learning model for training to obtain a selected probability prediction model;
obtaining a plurality of candidate advertisement spots of the client;
aiming at each candidate advertisement site in the candidate advertisement sites, importing corresponding site characteristic data into the selected probability prediction model to obtain a corresponding selected probability prediction value;
sequencing the candidate advertisement site positions in sequence according to the sequence of the selected probability predicted values from high to low to obtain a candidate advertisement site position sequence;
and selecting the first N candidate advertisement spots from the candidate advertisement spot position sequence as advertisement spot position screening results to be pushed to the client, wherein N represents a natural number and is pre-specified by the client or determined according to the advertisement delivery budget cost pre-input by the client.
Based on the invention, an advertisement site selection scheme based on a machine learning model and historical manual site selection data is provided, namely, historical site selection data matched with advertisement putting directional information is obtained according to advertisement putting directional information of a client, then corresponding selected probability values are obtained through calculation according to corresponding selected times aiming at each advertisement site recorded in the historical site selection data, corresponding site characteristic data are obtained, then the machine learning model is trained by using the site characteristic data and the selected probability values to obtain a selected probability prediction model, finally the matching values of each candidate advertisement site and the client are obtained through prediction of the selected probability prediction model, the purpose of automatically selecting proper advertisement sites for the client is realized, and further, the advertisement site labels or target population labels do not need to be manually appointed, and in the point selection process, repeated checking and point location adjustment are not needed, the point selection efficiency can be obviously improved, the point selection result has the characteristics of high reliability and the like, and the customer satisfaction and the advertisement putting efficiency are favorably improved. In addition, the method also establishes the machine learning algorithm model for each industry respectively, so that the model has strong interpretability and higher customer acceptance.
In one possible design, obtaining historical click data matched with the advertisement delivery targeting information according to the advertisement delivery targeting information of the client includes:
acquiring advertisement putting targeted information of a client, wherein the advertisement putting targeted information comprises the affiliated/appointed industry and/or the located/appointed area of the client;
and extracting historical point selection data matched with the affiliated/appointed industry and/or the affiliated/appointed region from a database, wherein the historical point selection data records advertisement point location selection information generated during each manual point selection before the current time, and the advertisement point location selection information comprises at least one advertisement point location selected during corresponding manual point selection.
In one possible design, for each advertisement spot location recorded in the historical spot location data, calculating a corresponding selected probability value according to a corresponding selected number of times, including:
aiming at each advertisement site recorded in the historical site selection data, according to the corresponding selected times, calculating to obtain the corresponding selected probability value according to the following formula:
wherein s represents a positive integer, P s A selected probability value, S, representing the S-th advertisement spot recorded in said historical spot selection data s Representing the selected number of times of said s-th spot, C s Representing the total number of manual hits or the total number of candidates for the s-th spot in the historical hit data.
In one possible design, when the advertisement spots are all elevator advertisement machines of a building, corresponding spot location feature data is obtained for each advertisement spot, including:
aiming at each advertisement point location, extracting the floor survey data corresponding to each time of participating in manual point selection from the database, wherein the floor survey data comprises floor attribute information, floor crowd attribute information and/or floor crowd preference information of the corresponding floor, the attribute information of the building comprises the city level of the building, the price of the building, the age of the building, the number of floors, the greening area, the parking number and/or the matching label around the building, the attribute information of the building crowd comprises the income level of the building crowd, the consumption level of the building crowd, the age distribution of the building crowd and/or the academic distribution of the building crowd, the building crowd preference information comprises a building crowd sports preference label, a building crowd financial preference label, a building crowd tourism preference label and/or a building crowd gourmet preference label;
aiming at each advertisement point location, converting a plurality of pieces of information recorded by the floor survey data corresponding to each time of participation in manual point selection into a plurality of point location characteristic values by adopting a one-bit effective coding processing mode and/or a normalization processing mode to obtain point location characteristic data corresponding to each time of participation in manual point selection, wherein the plurality of pieces of information correspond to the plurality of point location characteristic values one by one;
and aiming at each advertisement point location, point location characteristic data corresponding to each time of manual point selection is processed through equalization, and corresponding point location characteristic data are obtained comprehensively.
In one possible design, the machine learning model employs a machine learning model based on a GBDT algorithm or an XGBoost algorithm.
In one possible design, obtaining a plurality of candidate advertisement spots for the customer includes:
acquiring a target area and a future target period which are specified in advance by the client;
and taking all the advertisement spots which are positioned in the target area and are not selected in the future target time period as a plurality of candidate advertisement spots of the client.
In one possible design, after obtaining the candidate advertisement site location sequence and before selecting the first N candidate advertisement site locations from the candidate advertisement site location sequence as advertisement site location screening results and pushing the advertisement site location screening results to the client, the advertisement site location screening method further includes:
acquiring an advertisement delivery budget cost input by the client in advance;
according to the budget cost of advertisement delivery, determining the value of N according to the following steps S801-S803:
s801, initializing the number K of available advertising points to 1, and then executing the step S802;
s802. judging
If the cost is larger than the budget cost of advertisement putting, if so, determining the value of N as K-1, otherwise, executing a step S803, wherein K represents a positive integer, v represents a positive integer k Indicating that the k-th candidate advertisement spot is not located in the sequence of candidate advertisement spots in the order from front to backA cost of advertising from a target time period, wherein the future target time period is pre-specified by the customer;
s803, adding 1 to the available advertisement point digit K, and then executing the step S802.
In a second aspect, the invention provides an advertisement site location screening device, which comprises a site location data acquisition module, a probability value calculation module, a feature data acquisition module, a model training module, a candidate site location acquisition module, a probability value estimation module, an advertisement site location sorting module and an advertisement site location selection module;
the system comprises a point selection data acquisition module, a point selection data acquisition module and a point selection data acquisition module, wherein the point selection data acquisition module is used for acquiring historical point selection data matched with advertisement putting targeting information according to the advertisement putting targeting information of a client, the historical point selection data records advertisement point location selection information generated in each manual point selection before the current moment, and the advertisement point location selection information comprises at least one advertisement point location selected in the corresponding manual point selection;
the probability value calculating module is in communication connection with the point selection data acquiring module and is used for calculating corresponding selected probability values according to corresponding selected times aiming at advertisement points recorded in the historical point selection data;
the characteristic data acquisition module is in communication connection with the point selection data acquisition module and is used for acquiring corresponding point location characteristic data aiming at each advertisement point location, wherein the point location characteristic data comprises a plurality of point location characteristic values;
the model training module is respectively in communication connection with the probability value calculation module and the feature data module and is used for importing the point location feature data of each advertisement point location as input data required by model training and selected probability values of each advertisement point location as output data required by model training into a machine learning model for training to obtain a selected probability prediction model;
the candidate advertisement site obtaining module is used for obtaining a plurality of candidate advertisement site positions of the client;
the probability value estimation module is respectively in communication connection with the model training module and the candidate point location acquisition module, and is used for importing corresponding point location characteristic data into the selected probability prediction model aiming at each candidate advertisement point location in the candidate advertisement point locations to obtain a corresponding selected probability prediction value;
the advertisement point location sequencing module is in communication connection with the probability value pre-estimating module and is used for sequencing the candidate advertisement point locations in sequence according to the sequence of the selected probability predicted values from high to low to obtain a candidate advertisement point location sequence;
the advertisement site selection module is in communication connection with the advertisement site ordering module and is used for selecting the first N candidate advertisement sites from the candidate advertisement site sequence as an advertisement site screening result and pushing the advertisement site screening result to the client, wherein N represents a natural number and is pre-specified by the client or determined according to an advertisement delivery budget cost pre-input by the client.
In a third aspect, the present invention provides a computer device, including a memory and a processor, which are communicatively connected, where the memory is used to store a computer program, and the processor is used to read the computer program and execute the method for screening a spot location according to the first aspect or any possible design of the first aspect.
In a fourth aspect, the present invention provides a computer-readable storage medium, where instructions are stored, and when the instructions are run on a computer, the method for screening a spot location according to the first aspect or any possible design of the first aspect is performed.
In a fifth aspect, the present invention provides a computer program product comprising instructions which, when run on a computer, cause the computer to perform the spot location screening method as described in the first aspect or any of the possible designs of the first aspect.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a schematic flow chart of the advertisement site location screening method provided by the present invention.
Fig. 2 is a schematic structural diagram of an advertisement site location screening device provided by the present invention.
Fig. 3 is a schematic structural diagram of a computer device provided by the present invention.
Detailed Description
The invention is further described with reference to the following figures and specific embodiments. It should be noted that the description of the embodiments is provided to help understanding of the present invention, but the present invention is not limited thereto. Specific structural and functional details disclosed herein are merely representative of exemplary embodiments of the invention. This invention may, however, be embodied in many alternate forms and should not be construed as limited to the embodiments set forth herein.
It will be understood that, although the terms first, second, etc. may be used herein to describe various objects, these objects should not be limited by these terms. These terms are only used to distinguish one object from another. For example, a first object may be referred to as a second object, and similarly, a second object may be referred to as a first object, without departing from the scope of example embodiments of the present invention.
It should be understood that, for the term "and/or" as may appear herein, it is merely an associative relationship that describes an associated object, meaning that three relationships may exist, e.g., a and/or B may mean: a exists alone, B exists alone or A and B exist at the same time; for the term "/and" as may appear herein, which describes another associative object relationship, it means that two relationships may exist, e.g., a/and B, may mean: a exists singly or A and B exist simultaneously; in addition, with respect to the character "/" which may appear herein, it generally means that the former and latter associated objects are in an "or" relationship.
As shown in fig. 1, the advertisement site selection method provided in the first aspect of this embodiment may be, but not limited to, executed by a Computer device with certain computing resources, for example, executed by an electronic device such as a Personal Computer (PC, which refers to a multipurpose Computer with a size, price, and performance suitable for Personal use, a desktop Computer, a notebook Computer, a small notebook Computer, a tablet Computer, a super Computer, etc. all belong to a Personal Computer), a smart phone, a Personal digital assistant (PAD), or a wearable device, so as to perform, based on history manual site selection data, abstraction, generalization, and scale replication of a large amount of history customer or expert site selection experiences through a machine learning model, and predict a matching value between each candidate advertisement site and a customer, so as to achieve the purpose of automatically selecting a suitable advertisement site for the customer, and further without manually designating an advertisement tag or a target population tag, and in the point selection process, repeated checking and point location adjustment are not needed, the point selection efficiency can be obviously improved, the point selection result has the characteristics of high reliability and the like, and the customer satisfaction and the advertisement putting efficiency are favorably improved. As shown in fig. 1, the method for screening advertisement spots may include, but is not limited to, the following steps S1 to S8.
S1, obtaining historical point selection data matched with the advertisement putting targeting information according to the advertisement putting targeting information of a client, wherein the historical point selection data records but is not limited to advertisement point location selection information generated in each manual point selection before the current moment, and the advertisement point location selection information includes but is not limited to at least one advertisement point location selected in the corresponding manual point selection.
In step S1, the client is an advertiser who currently has a demand for advertisement placement. The advertisement placement targeting information is used for representing advertisement placement targeting content of the client, and may include, but is not limited to, an affiliated/designated industry and/or a location/designated area of the client, and the like, where the affiliated industry (e.g., an affiliated catering industry, sports industry, or game industry, etc.) and the location (e.g., a located city, etc.) may be acquired when the client registers, and the designated industry (e.g., an appointed placement catering industry, sports industry, or game industry, etc.) and the designated area (e.g., an appointed placement city, etc.) may be input by the client on a human-computer interaction interface for this point selection, so that a matching relationship between the historical point selection data and the advertisement placement targeting information may specifically be: the historical point selection data is a point selection result obtained by a historical client or an expert through an advertisement point location manual screening mode before the current moment for the purpose of putting the advertisements belonging to the affiliated/appointed industries in the affiliated/appointed regions. The advertisement spots are used for representing the minimum selectable units of the advertisement machines, and can be but are not limited to all elevator advertisement machines of a building. In addition, the historical click point data can also record but not be limited to click point location reference information generated in each manual click point before the current time, wherein the click point location reference information comprises but is not limited to a plurality of click points which are referenced in corresponding historical click points (at the moment, the click point location reference information comprises but is not limited to at least one click point which is selected from the plurality of click points in corresponding historical click points).
In step S1, specifically, the historical setpoint data matched with the advertisement placement targeting information is obtained according to the advertisement placement targeting information of the client, which includes but is not limited to the following steps S11 to S12: s11, obtaining advertisement putting directional information of a client, wherein the advertisement putting directional information comprises but is not limited to the affiliated/appointed industry and/or the located/appointed area of the client and the like; s12, extracting historical point selection data matched with the affiliated/appointed industry and/or the affiliated/appointed region from a database, wherein the historical point selection data records advertisement point selection information generated in each manual point selection before the current time, and the advertisement point selection information comprises at least one advertisement point selected in the corresponding manual point selection. The database may be, but is not limited to, for collecting the click-through results obtained by a historical client or expert or the like by manually screening the spot location patterns before the current time.
And S2, aiming at each advertisement site recorded in the historical site selection data, calculating to obtain a corresponding selected probability value according to the corresponding selected times.
In step S2, specifically, for each advertisement spot recorded in the historical spot data, according to the corresponding selected times, a corresponding selected probability value may be calculated according to the following formula:
wherein s represents a positive integer, P s A selected probability value, S, representing the S-th advertisement spot recorded in said historical spot selection data s Representing the selected number of times of said s-th spot, C s Representing the total number of manual hits or the total number of candidates for the s-th spot in the historical hit data.
And S3, acquiring corresponding point location characteristic data aiming at each advertisement point location, wherein the point location characteristic data comprises but is not limited to a plurality of point location characteristic values.
In step S3, the point location feature data is used to reflect the relevant factors considered by the historical customer or expert based on the point selection experience during the manual point selection. Specifically, when the advertisement spots are all elevator advertisement machines of a building, the spot feature data can be presented through relevant data of the building, that is, corresponding floor survey data is extracted from the database for each advertisement spot, wherein the floor survey data includes, but is not limited to, floor attribute information, floor crowd attribute information, and/or floor crowd preference information, etc., of the corresponding building, the floor attribute information includes, but is not limited to, a city level (e.g., first-line city, new-line city, second-line city, third-line city, etc.), a floor price, a building age, a floor number, a greening area, parking space, and/or a building periphery matching tag (e.g., near commercial street tag, near subway station tag, etc.), etc., and the floor crowd attribute information includes, but is not limited to, income of the building crowd horizontal level, etc., of the building crowd, The method comprises the steps of obtaining the consumption level of the building crowd, the age distribution of the building crowd and/or the academic distribution of the building crowd, wherein the preference information of the building crowd comprises but is not limited to a building crowd sports preference label, a building crowd financial preference label, a building crowd tourism preference label and/or a building crowd gourmet preference label and the like; and then, converting a plurality of pieces of information (namely the floor attribute information, the floor crowd attribute information and/or the floor crowd preference information and the like) recorded by the corresponding floor survey data into a plurality of point location characteristic values by adopting a one-bit effective coding processing mode and/or a normalization processing mode to obtain corresponding point location characteristic data, wherein the plurality of pieces of information correspond to the plurality of point location characteristic values one by one.
In step S3, it is further considered that when the advertisement spots correspond to each time of manual selection (that is, when the advertisement spots are selected in each manual experience, there is a case that they are not selected by other customers in advance, and therefore "participating" here is for the advertisement spots, not for "spot feature data", and historical customers/experts may use the "spot feature data" when performing manual selection according to the selection experience, or may not use the "spot feature data" when performing manual selection according to the selection experience, and may have different spot feature data (for example, spot feature values obtained based on floor rate or floor age, etc.) due to large time difference (for example, every few years), the corresponding spot feature data may be obtained by averaging the spot feature data corresponding to each time of manual selection, so as to synthesize the corresponding spot feature data, that is, preferably, the point feature data corresponding to each advertisement point is obtained, including but not limited to the following steps S31 to S33.
S31, extracting the corresponding floor survey data when corresponding to each manual point selection from the database aiming at each advertisement point position, wherein the corresponding floor survey data comprise but are not limited to floor attribute information, floor crowd attribute information and/or floor crowd preference information and the like corresponding to the corresponding floor, the floor attribute information comprises but is not limited to city level, floor room price, floor age, floor number, greening area, parking number and/or a floor periphery matching label and the like, the floor crowd attribute information comprises but is not limited to floor crowd income level, floor crowd consumption level, floor crowd age distribution and/or floor crowd academic distribution and the like, and the floor crowd preference information comprises but is not limited to floor crowd sports preference label, floor crowd financial channel preference label, floor crowd preference label and the like, The preference labels for tourism of the building crowd and/or the preference labels for cate of the building crowd and the like.
In step S31, the database may also be used, but is not limited to, collecting the floor survey data at each manual pick before the current time.
And S32, aiming at each advertisement point location, converting a plurality of pieces of information recorded by the floor survey data corresponding to each time of participation in manual point selection into a plurality of point location characteristic values by adopting a one-bit effective coding processing mode and/or a normalization processing mode to obtain point location characteristic data corresponding to each time of participation in manual point selection, wherein the plurality of pieces of information correspond to the plurality of point location characteristic values one by one.
In the step S32, the One-bit effective coding processing manner (i.e., One-hot coding manner) and the normalization processing manner are both conventional numerical processing manners, for example, for a matching label around a building, a corresponding point location characteristic value can be obtained through the One-bit effective coding processing manner: 0 or 1, and for the price of the building, etc., the corresponding point characteristic value belonging to the interval [0,1] can be obtained by the normalization processing mode.
And S33, aiming at each advertisement point location, carrying out averaging processing on the point location characteristic data corresponding to each time of manual point selection, and comprehensively obtaining corresponding point location characteristic data.
In step S33, a specific equalizing processing manner is to perform equalizing processing on different point location feature values when the point location feature values respectively participate in manual point selection for each time, for each point location feature value in the plurality of point location feature values. For example, for a certain advertisement site, if the site feature value corresponding to each time of participating in manual site selection on the building rate has x 1 ,x 2 ,…,x m ,…,x M Then obtained and corresponding participantThe point location eigenvalues at the time of point selection will average to
Wherein m represents a positive integer, x m And M represents the total times of participating in the manual point selection.
And S4, importing the point location feature data of each advertisement point location as input data required by model training, and importing the selected probability value of each advertisement point location as output data required by model training into a machine learning model for training to obtain a selected probability prediction model.
In step S4, it is considered that the advertisement site supplier accumulates manual selection data of different industries and different historical customers in the business operation process, and for the same segment industry, different historical customers can select an ideal floor according to the judgment of experience on the lower floor, and then put in corresponding advertisements. For example: in the "drinking water" segment industry, the relevant factors implicitly considered by different historical consumers in advertising spots may be: the historical customer A is high in income level, new in building age, large in greening area and the like, and the historical customer B is: the "core area", "high school calendar population occupation ratio", and "office population occupation ratio", etc., history client C: the "young people high-occupancy ratio", "like fitness and sports", and "like food and drink", etc., the historical customer Y: the parking space is large, the housing price is high, the periphery of the building is complete, and the like. In the above point selection experiences of different experts in the same segment industry, the point selection experiences can be fitted by inputting the point location feature data and the selected probability value based on a machine learning model, and an optimal point selection model (namely, the selected probability prediction model) of the segment industry is trained, so that the purpose of automatically selecting an appropriate advertisement point location for the client subsequently can be realized by abstracting, generalizing and copying the historical client or expert point selection experiences in a large scale. Preferably, the machine learning model may be, but is not limited to, a machine learning model based on a GBDT (Gradient Boosting Decision Tree, which uses Boosting idea) algorithm or an XGBoost algorithm (a Tree-based Boosting algorithm). Because the model and the model are all integrated learning models of Boosting series, the model fitting effect is better compared with other traditional algorithm models (such as logistic regression, SVM and the like), the model effect can be compared and optimized by calculating the Pearson correlation coefficient, the mean square error and the like, and then the optimal selected point model is trained. In addition, the specific training mode is the conventional mode, and is not described herein again.
And S5, acquiring a plurality of candidate advertisement spots of the client.
In step S5, in order to ensure that the candidate spots can satisfy the basic requirements of the customer, specifically, the obtaining of the candidate spots of the customer includes, but is not limited to, the following steps S51 to S52: s51, acquiring a target area (such as a certain city jurisdiction of a certain city) and a future target time period (such as tomorrow, next week or next month) which are specified in advance by the client; s52, all the advertisement spots which are positioned in the target area and are not selected in the future target time interval are used as a plurality of candidate advertisement spots of the client.
S6, aiming at each candidate advertisement site in the candidate advertisement sites, importing the corresponding site characteristic data into the selected probability prediction model to obtain a corresponding selected probability prediction value.
In the step S6, the point location feature data of each candidate advertisement point is preferably the point location feature data at the current time, and a specific obtaining manner thereof may be obtained by conventional derivation with reference to the step S3, which is not described herein again. In addition, the selected probability prediction model is a result of abstracting, generalizing and massively copying historical customer or expert point selection experience, so that the selected probability prediction value of each candidate advertisement point can be used as a matching value with the customer.
And S7, sequencing the plurality of candidate advertisement site positions in sequence according to the sequence of the selected probability predicted values from high to low to obtain a candidate advertisement site position sequence.
S8, selecting the first N candidate advertisement spots from the candidate advertisement spot sequence as advertisement spot screening results and pushing the advertisement spot screening results to the client, wherein N represents a natural number and is pre-specified by the client.
In the step S8, for example, if the customer specifies in advance that N is 10, the first 10 candidate advertisement spots in the candidate advertisement spot sequence may be pushed to the customer as an advertisement spot screening result, so that the customer may perform advertisement delivery. In addition, if the total number of the candidate advertisement spots is less than or equal to N, the aforementioned steps S1-S4 and S6-S8 may be skipped, and the candidate advertisement spots are directly pushed to the client as the advertisement spot screening result.
Therefore, based on the advertisement site selection method described in the foregoing steps S1-S8, an advertisement site selection scheme based on a machine learning model and historical artificial site selection data is provided, that is, historical site selection data matched with the advertisement placement targeting information is obtained according to the advertisement placement targeting information of the client, then, corresponding selected probability values are calculated according to corresponding selected times for each advertisement site recorded in the historical site selection data, corresponding site characteristic data are obtained, then, the machine learning model is trained by using the site characteristic data and the selected probability values to obtain a selected probability prediction model, finally, the matching values of each candidate advertisement site and the client are obtained by prediction through the selected probability prediction model, the purpose of automatically selecting proper advertisement sites for the client is realized, and further, manual designation of advertisement site labels or target population labels is not needed, and in the point selection process, repeated checking and point location adjustment are not needed, the point selection efficiency can be obviously improved, the point selection result has the characteristics of high reliability and the like, and the customer satisfaction and the advertisement putting efficiency are favorably improved. In addition, the method also establishes the machine learning algorithm model for each industry respectively, so that the model has strong interpretability and higher customer acceptance.
In this embodiment, on the basis of the technical solution of the first aspect, a first possible design is provided for how to determine the N value according to the advertisement placement budget cost, that is, after obtaining the candidate advertisement site location sequence and before selecting the first N candidate advertisement site locations from the candidate advertisement site location sequence as advertisement site location screening results and pushing the advertisement site location screening results to the client, where the advertisement site location screening method further includes, but is not limited to: firstly, acquiring an advertisement delivery budget cost input by the client in advance; and then, according to the budget cost of advertisement delivery, determining the value of N according to the following steps S801-S803.
S801, initializing the number K of available advertising point bits to 1, and then executing the step S802.
S802. judging
If the cost is larger than the budget cost of advertisement putting, if so, determining the value of N as K-1, otherwise, executing a step S803, wherein K represents a positive integer, v represents a positive integer k Representing the advertisement placement cost of the kth candidate spot in the sequence of candidate spots in a future target slot in a front-to-back order, wherein the future target slot is pre-specified by the customer.
S803, adding 1 to the available advertisement dot digit K, and then executing the step S802.
Therefore, based on the possible design one described in the previous step, the advertisement site selection result which ensures that the budget is not exceeded can be selected for the client according to the budget cost for advertisement delivery input by the client in advance, and the method is particularly suitable for the condition that different advertisement sites have different pricing. In addition, when the value of N is determined to be zero, the current budget cost for advertisement delivery is seriously insufficient, and a corresponding prompt message can be output so as to remind a client to increase the budget.
As shown in fig. 2, a second aspect of this embodiment provides a virtual device for implementing the advertisement site location screening method in any one of the first aspect or the first aspect, including a site location data obtaining module, a probability value calculating module, a feature data obtaining module, a model training module, a candidate site location obtaining module, a probability value estimating module, an advertisement site location sorting module, and an advertisement site location selecting module;
the system comprises a point selection data acquisition module, a point selection data acquisition module and a point selection data acquisition module, wherein the point selection data acquisition module is used for acquiring historical point selection data matched with advertisement putting targeting information according to the advertisement putting targeting information of a client, the historical point selection data records advertisement point location selection information generated in each manual point selection before the current moment, and the advertisement point location selection information comprises at least one advertisement point location selected in the corresponding manual point selection;
the probability value calculating module is in communication connection with the point selection data acquiring module and is used for calculating corresponding selected probability values according to corresponding selected times aiming at advertisement points recorded in the historical point selection data;
the characteristic data acquisition module is in communication connection with the point selection data acquisition module and is used for acquiring corresponding point location characteristic data aiming at each advertisement point location, wherein the point location characteristic data comprises a plurality of point location characteristic values;
the model training module is respectively in communication connection with the probability value calculation module and the feature data module and is used for importing the point location feature data of each advertisement point location as input data required by model training and selected probability values of each advertisement point location as output data required by model training into a machine learning model for training to obtain a selected probability prediction model;
the candidate advertisement site obtaining module is used for obtaining a plurality of candidate advertisement site positions of the client;
the probability value estimation module is respectively in communication connection with the model training module and the candidate point location acquisition module, and is used for importing corresponding point location characteristic data into the selected probability prediction model aiming at each candidate advertisement point location in the candidate advertisement point locations to obtain a corresponding selected probability prediction value;
the advertisement point location sequencing module is in communication connection with the probability value pre-estimating module and is used for sequencing the candidate advertisement point locations in sequence according to the sequence of the selected probability predicted values from high to low to obtain a candidate advertisement point location sequence;
the advertisement site selection module is in communication connection with the advertisement site ordering module and is used for selecting the first N candidate advertisement sites from the candidate advertisement site sequence as an advertisement site screening result and pushing the advertisement site screening result to the client, wherein N represents a natural number and is pre-specified by the client or determined according to an advertisement delivery budget cost pre-input by the client.
For the working process, working details and technical effects of the foregoing apparatus provided in the second aspect of this embodiment, reference may be made to the advertisement site selection method possibly designed in any one of the first aspect or the first aspect, which is not described herein again.
As shown in fig. 3, a third aspect of this embodiment provides a computer device for executing the method for screening advertisement spots possibly designed in any one of the first aspect or the first aspect, where the computer device includes a memory and a processor, the memory is used for storing a computer program, and the processor is used for reading the computer program and executing the method for screening advertisement spots possibly designed in any one of the first aspect or the first aspect. For example, the Memory may include, but is not limited to, a Random-Access Memory (RAM), a Read-Only Memory (ROM), a Flash Memory (Flash Memory), a First-in First-out (FIFO), and/or a First-in Last-out (FILO), and the like; the processor may be, but is not limited to, a microprocessor of the model number STM32F105 family. In addition, the computer device may also include, but is not limited to, a power module, a display screen, and other necessary components.
For the working process, the working details, and the technical effects of the foregoing computer device provided in the third aspect of this embodiment, reference may be made to the advertisement site selection method described in the first aspect or any one of the first aspects that may be designed, which is not described herein again.
A fourth aspect of this embodiment provides a computer-readable storage medium storing instructions including instructions for implementing the advertisement site location screening method according to any one of the first aspect and the possible designs of the advertisement site location screening method according to the first aspect, that is, the instructions are stored on the computer-readable storage medium, and when the instructions are run on a computer, the advertisement site location screening method according to any one of the first aspect and the possible designs of the first aspect is implemented. The computer-readable storage medium refers to a carrier for storing data, and may include, but is not limited to, a computer-readable storage medium such as a floppy disk, an optical disk, a hard disk, a flash Memory, a flash disk and/or a Memory Stick (Memory Stick), and the computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.
For a working process, working details, and technical effects of the foregoing computer-readable storage medium provided in the fourth aspect of this embodiment, reference may be made to the first aspect or any one of the first aspects that may be designed to provide the advertisement site location screening method, which is not described herein again.
A fifth aspect of the present invention provides a computer program product comprising instructions, which when run on a computer, cause the computer to perform the spot location screening method according to the first aspect or any one of the first aspect as may be devised. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable devices.
Finally, it should be noted that the present invention is not limited to the above alternative embodiments, and that various other forms of products can be obtained by anyone in light of the present invention. The above detailed description should not be taken as limiting the scope of the invention, which is defined in the claims, and which the description is intended to be interpreted accordingly.
Claims (10)
1. An advertisement site location screening method is characterized by comprising the following steps:
acquiring historical point selection data matched with the advertisement putting targeting information according to the advertisement putting targeting information of a client, wherein the historical point selection data records advertisement point location selection information generated in each manual point selection before the current moment, and the advertisement point location selection information comprises at least one advertisement point location selected in the corresponding manual point selection;
aiming at each advertisement point position recorded in the historical point selection data, calculating to obtain a corresponding selected probability value according to corresponding selected times;
acquiring corresponding point location characteristic data aiming at each advertisement point location, wherein the point location characteristic data comprises a plurality of point location characteristic values;
importing the point location feature data of each advertisement point location as input data required by model training and the selected probability value of each advertisement point location as output data required by model training into a machine learning model for training to obtain a selected probability prediction model;
obtaining a plurality of candidate advertisement spots of the client;
aiming at each candidate advertisement site in the candidate advertisement sites, importing corresponding site characteristic data into the selected probability prediction model to obtain a corresponding selected probability prediction value;
sequencing the candidate advertisement site positions in sequence according to the sequence of the selected probability predicted values from high to low to obtain a candidate advertisement site position sequence;
and selecting the first N candidate advertisement spots from the candidate advertisement spot position sequence as advertisement spot position screening results to be pushed to the client, wherein N represents a natural number and is pre-specified by the client or determined according to the advertisement delivery budget cost pre-input by the client.
2. The method for screening advertisement spots according to claim 1, wherein the step of obtaining historical spot selection data matched with the advertisement placement targeting information according to the advertisement placement targeting information of the client comprises:
acquiring advertisement putting targeted information of a client, wherein the advertisement putting targeted information comprises the affiliated/appointed industry and/or the located/appointed area of the client;
and extracting historical point selection data matched with the affiliated/appointed industry and/or the affiliated/appointed region from a database, wherein the historical point selection data records advertisement point location selection information generated during each manual point selection before the current time, and the advertisement point location selection information comprises at least one advertisement point location selected during corresponding manual point selection.
3. The method for screening advertisement spots according to claim 1, wherein the step of calculating corresponding selected probability values according to corresponding selected times for each advertisement spot recorded in the historical spot data comprises:
aiming at each advertisement site recorded in the historical site selection data, according to the corresponding selected times, calculating to obtain the corresponding selected probability value according to the following formula:
wherein s represents a positive integer, P s A selected probability value, S, representing the S-th advertisement spot recorded in said historical spot selection data s Representing the selected number of times of said s-th spot, C s Representing the total number of manual hits or the total number of candidates for the s-th spot in the historical hit data.
4. The method for screening advertisement spots according to claim 1, wherein when the advertisement spots are all elevator advertisement machines of a building, acquiring corresponding spot feature data for each advertisement spot comprises:
aiming at each advertisement point, extracting the floor survey data corresponding to each time of participating in manual point selection from the database, wherein the floor survey data comprises floor attribute information, floor crowd attribute information and/or floor crowd preference information of the corresponding floor, the attribute information of the building comprises the city level of the building, the price of the building, the age of the building, the number of floors, the greening area, the parking number and/or the matching label around the building, the attribute information of the building crowd comprises the income level of the building crowd, the consumption level of the building crowd, the age distribution of the building crowd and/or the academic distribution of the building crowd, the building crowd preference information comprises a building crowd sports preference label, a building crowd financial preference label, a building crowd tourism preference label and/or a building crowd gourmet preference label;
aiming at each advertisement point, converting a plurality of pieces of information recorded by the floor survey data corresponding to each time of participation in manual point selection into a plurality of point characteristic values by adopting a one-bit effective coding processing mode and/or a normalization processing mode to obtain point characteristic data corresponding to each time of participation in manual point selection, wherein the plurality of pieces of information correspond to the plurality of point characteristic values one by one;
and aiming at each advertisement point location, point location characteristic data corresponding to each time of manual point selection is processed through equalization, and corresponding point location characteristic data are obtained comprehensively.
5. The method for screening the advertisement site location according to claim 1, wherein the machine learning model adopts a machine learning model based on a GBDT algorithm or an XGBoost algorithm.
6. The spot location screening method of claim 1, wherein obtaining a plurality of candidate spot locations for the customer comprises:
acquiring a target area and a future target period which are specified in advance by the client;
and taking all the advertisement spots which are positioned in the target area and are not selected in the future target time period as a plurality of candidate advertisement spots of the client.
7. The method of claim 1, wherein after obtaining the candidate advertisement spot location sequence and before selecting the first N candidate advertisement spots from the candidate advertisement spot location sequence as advertisement spot location screening results and pushing the advertisement spot location screening results to the client, the method further comprises:
acquiring an advertisement delivery budget cost input by the client in advance;
according to the budget cost of advertisement delivery, determining the value of N according to the following steps S801-S803:
s801, initializing the number K of available advertising points to 1, and then executing the step S802;
s802. judging
If the cost is larger than the budget cost of advertisement putting, if so, determining the value of N as K-1, otherwise, executing a step S803, wherein K represents a positive integer, v represents a positive integer k Representing advertisement placement costs for a kth candidate spot in a sequence of candidate spots in a future target slot, wherein the future target slot is pre-specified by the customer;
s803, adding 1 to the available advertisement dot digit K, and then executing the step S802.
8. An advertisement site selection device is characterized by comprising a site selection data acquisition module, a probability value calculation module, a characteristic data acquisition module, a model training module, a candidate site acquisition module, a probability value estimation module, an advertisement site ordering module and an advertisement site selection module;
the system comprises a point selection data acquisition module, a point selection data acquisition module and a point selection data acquisition module, wherein the point selection data acquisition module is used for acquiring historical point selection data matched with advertisement putting targeting information according to the advertisement putting targeting information of a client, the historical point selection data records advertisement point location selection information generated in each manual point selection before the current moment, and the advertisement point location selection information comprises at least one advertisement point location selected in the corresponding manual point selection;
the probability value calculating module is in communication connection with the point selection data acquiring module and is used for calculating corresponding selected probability values according to corresponding selected times aiming at advertisement points recorded in the historical point selection data;
the feature data acquisition module is in communication connection with the point selection data acquisition module and is used for acquiring corresponding point location feature data aiming at each advertisement point location, wherein the point location feature data comprises a plurality of point location feature values;
the model training module is respectively in communication connection with the probability value calculation module and the feature data module and is used for importing the point location feature data of each advertisement point as input data required by model training and the selected probability value of each advertisement point as output data required by the model training into a machine learning model for training to obtain a selected probability prediction model;
the candidate point location obtaining module is used for obtaining a plurality of candidate advertisement point locations of the client;
the probability value estimation module is respectively in communication connection with the model training module and the candidate point location acquisition module, and is used for importing corresponding point location characteristic data into the selected probability prediction model aiming at each candidate advertisement point location in the candidate advertisement point locations to obtain a corresponding selected probability prediction value;
the advertisement point location sequencing module is in communication connection with the probability value pre-estimating module and is used for sequencing the candidate advertisement point locations in sequence according to the sequence of the selected probability predicted values from high to low to obtain a candidate advertisement point location sequence;
the advertisement site selection module is in communication connection with the advertisement site ordering module and is used for selecting the first N candidate advertisement sites from the candidate advertisement site sequence as an advertisement site screening result and pushing the advertisement site screening result to the client, wherein N represents a natural number and is pre-specified by the client or determined according to an advertisement delivery budget cost pre-input by the client.
9. A computer device, comprising a memory and a processor, wherein the memory is connected in communication, the memory is used for storing a computer program, and the processor is used for reading the computer program and executing the advertisement site location screening method according to any one of claims 1 to 7.
10. A computer-readable storage medium having stored thereon instructions for performing the spot location screening method according to any one of claims 1 to 7 when the instructions are run on a computer.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210615229.0A CN114971726A (en) | 2022-05-31 | 2022-05-31 | Advertisement point location screening method and device, computer equipment and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210615229.0A CN114971726A (en) | 2022-05-31 | 2022-05-31 | Advertisement point location screening method and device, computer equipment and computer readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114971726A true CN114971726A (en) | 2022-08-30 |
Family
ID=82960442
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210615229.0A Pending CN114971726A (en) | 2022-05-31 | 2022-05-31 | Advertisement point location screening method and device, computer equipment and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114971726A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117132330A (en) * | 2023-10-23 | 2023-11-28 | 蓝色火焰科技成都有限公司 | Intelligent advertisement method, device, equipment and storage medium based on big data |
-
2022
- 2022-05-31 CN CN202210615229.0A patent/CN114971726A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117132330A (en) * | 2023-10-23 | 2023-11-28 | 蓝色火焰科技成都有限公司 | Intelligent advertisement method, device, equipment and storage medium based on big data |
| CN117132330B (en) * | 2023-10-23 | 2024-01-30 | 蓝色火焰科技成都有限公司 | Intelligent advertisement method, device, equipment and storage medium based on big data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110570232B (en) | Internet advertisement putting method and device, server and storage medium | |
| CN106570718B (en) | Information delivery method and delivery system | |
| CN111143738B (en) | Resource display method and device, electronic equipment and storage medium | |
| KR20190075141A (en) | Multi-dimensional advertisement bidding | |
| CN107526810B (en) | Method and device for establishing click rate estimation model and display method and device | |
| CN107767175A (en) | A kind of information based on Information rate launches processing method and processing device | |
| US11227309B2 (en) | Method and system for optimizing user grouping for advertisement | |
| CN116137004B (en) | Attribution method, attribution system and attribution computer for advertisement putting effect | |
| CN105654198B (en) | Brand advertisement effect optimization method capable of realizing optimal threshold value selection | |
| CN110515904B (en) | Quality prediction model training method, quality prediction method and device for media file | |
| CN108665083A (en) | A kind of method and system for advertisement recommendation for dynamic trajectory model of being drawn a portrait based on user | |
| CN108076387A (en) | Business object method for pushing and device, electronic equipment | |
| Wang | A survey of online advertising click-through rate prediction models | |
| CN106296287A (en) | For the method and apparatus adjusting the bid of promotion message | |
| CN114331543A (en) | Advertisement propagation method for large-scale crowd orientation and dynamic scene matching | |
| CN114971726A (en) | Advertisement point location screening method and device, computer equipment and computer readable storage medium | |
| CN113434755A (en) | Page generation method and device, electronic equipment and storage medium | |
| CN110992083A (en) | Data processing method and device for determining fare, storage medium and terminal | |
| CN113159809B (en) | Object processing method and device, electronic equipment and computer readable storage medium | |
| CN107507023B (en) | Information delivery method and device | |
| CN112954066A (en) | Information pushing method and device, electronic equipment and readable storage medium | |
| CN116167803A (en) | Advertisement putting method and device based on signaling data | |
| CN112989174A (en) | Information recommendation method and device, medium and equipment | |
| CN115809889A (en) | Intelligent passenger group screening method, system, medium and equipment based on marketing effect | |
| CN111091218A (en) | Method and device for generating bidding prediction model and automatically bidding advertisement delivery |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |