Intelligent instruction issuing system of advertisement all-in-one machine

Abstract

The invention discloses an intelligent instruction issuing system of an advertisement all-in-one machine, relates to the technical field of advertisement equipment, and solves the technical problem that an existing instruction issuing system is difficult to dynamically optimize different application scenes and user requirements, so that the display effect and the user experience are poor. The data acquisition module is used for acquiring environment data in real time; the advertisement obtaining module is used for obtaining advertisement characteristic data; the data processing module is used for generating time period characteristic data of each time period according to the historical record data; generating a time period adaptability score corresponding to each time period according to the advertisement characteristic data and the characteristic data of each time period; selecting an advertisement putting time period according to the time period adaptability score; acquiring environment information, and generating a parameter adjustment signal according to the environment information; the advertisement putting module is used for acquiring an advertisement putting time period and putting an advertisement according to the advertisement putting time period; acquiring a parameter adjustment signal, and performing parameter adjustment on the advertising machine according to the parameter adjustment signal; and the user experience is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of advertising equipment, and relates to advertising equipment technology, in particular to an intelligent issuing instruction system of an integrated advertising machine. Background Art

[0002] With the rapid development of information technology and multimedia technology, the way of advertising has also undergone major changes. Although traditional advertising methods such as printing, television, and radio are still widely used, their limitations are gradually emerging. For example, the cost remains high, and with the fragmentation of modern people's lives, the publicity effect of advertising has also been significantly reduced. Therefore, current advertisements are played in the elevators of shopping malls and residential buildings through advertising all-in-one machines to improve the efficiency of advertising dissemination.

[0003] However, existing integrated advertising machines usually configure display content and parameters manually, which makes it difficult to dynamically optimize according to different application scenarios and user needs, resulting in poor display effects and user experience. With the diversification and complexity of advertising market demand, there is an urgent need for a system that can intelligently and automatically adjust display content and parameters to improve advertising display effects and user satisfaction. Summary of the invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes an intelligent instruction publishing system for an all-in-one advertising machine, which is used to solve the technical problem that the existing instruction publishing system is difficult to dynamically optimize for different application scenarios and user needs, resulting in poor display effects and user experience.

[0005] To achieve the above-mentioned purpose, the first aspect of the present invention provides an intelligent publishing instruction system for an integrated advertising machine, comprising: a data collection module, an advertisement acquisition module, a data processing module, an advertisement delivery module and a database;

[0006] The data acquisition module is used to acquire environmental data in real time through a data acquisition device connected thereto;

[0007] The advertisement acquisition module is used to acquire advertisement characteristic data through the advertisement cloud platform connected thereto;

[0008] The data processing module: obtains historical record data of each time period in the delivery area through a database connected thereto; generates time period characteristic data of each time period according to the historical record data; obtains advertisement characteristic data, generates time period adaptability scores corresponding to each time period according to the advertisement characteristic data and each time period characteristic data; selects an advertisement delivery time period according to the time period adaptability scores; and obtains environmental information, and generates parameter adjustment signals according to the environmental information;

[0009] The advertisement delivery module: obtains an advertisement delivery period, and delivers advertisements according to the advertisement delivery period; obtains a parameter adjustment signal, and adjusts parameters of the advertisement machine according to the parameter adjustment signal.

[0010] The present application generates time period characteristic data of each time period by acquiring several historical record data of each time period in the advertising machine delivery area; generates a time period adaptability score of the corresponding advertisement in each delivery time period according to the time period characteristic data and the advertisement characteristic data, and selects the corresponding delivery time period according to the time period adaptability score; at the same time, when delivering the advertisement, the environmental data of the advertising machine delivery area is acquired, a parameter adjustment signal is generated according to the environmental data, and the corresponding parameters of the advertising machine are adjusted according to the parameter adjustment signal; the advertisement is delivered according to the characteristics of different time periods in the advertising machine delivery area, so that the advertisement delivery is more suitable for the characteristics of each time period; at the same time, the corresponding parameters of the advertising machine are adjusted according to the environmental data around the advertising machine when the advertisement is delivered; so that the advertisement delivery adapts to the real-time environment and increases the user experience.

[0011] Preferably, generating the period characteristic data of each period according to the historical record data includes:

[0012] Extracting the recording date, recording period and passenger image set from the historical record data; inputting each person image in the passenger image set into the person classification model in sequence to obtain the type label corresponding to each person image; integrating the type labels corresponding to several person images in the passenger image set into a type label set corresponding to the recording period; the person classification model is obtained by training the artificial intelligence model;

[0013] The type tag set of each recording period is obtained in sequence; the tag quantity of each same type tag in several type tag sets corresponding to the same recording period of different recording dates, and the type tag and its corresponding tag quantity are integrated into the period characteristic data of the recording period.

[0014] Preferably, the passenger image set is obtained in the following manner: images of persons entering the delivery area are acquired in real time by an image acquisition device, and images of different persons acquired during the same recording period are integrated into the passenger image set.

[0015] Preferably, the personnel classification model is obtained by training an artificial intelligence model, including:

[0016] Acquire a number of person images and their corresponding type labels from a database; integrate the person images and their corresponding type labels into a number of sets of training data and test data;

[0017] The artificial intelligence model is trained using training data, and the trained artificial intelligence model is tested using test data to obtain a personnel classification model whose input is a personnel image and whose output is a corresponding type label; wherein the artificial intelligence model includes a BP neural network model and an RBF neural network model.

[0018] Preferably, generating the time period adaptability score corresponding to each time period according to the advertisement characteristic data and the characteristic data of each time period includes:

[0019] Extract the product type from the advertisement characteristic data and mark it as ML; obtain each type label from the time period characteristic data and mark it as BLk, where k is the number of the type label;

[0020] The time period adaptability score SPt of each time period is calculated by the formula SPt=∑(Wtk×max(Pk(ML, BLk))), where k=1, 2, ..., N; N is the total number of type tags; Wtk is the weight coefficient of the type tag numbered k in the recording period numbered t; Pk(ML, BLk) is the degree of matching between the type tag numbered k and the product type.

[0021] Preferably, the weight coefficient is obtained by: obtaining the number of tags of each type in the period characteristic data of each recording period, and marking it as BStk;

[0022] The proportional coefficient Btk of the type tag numbered k in the recording period numbered t is calculated by the formula Btk=[ZBt / ∑(ZBt)]×(BStk / ZBt), where t=1, 2, ..., M; M is the total number of recording periods in a recording date; ZBt is the total number of type tags in the recording period numbered t, that is, ZBt=∑(BStk), k=1, 2, ..., N;

[0023] The weight coefficient Wtk is calculated by the formula Wtk=Btk / ∑(Btk), where k=1, 2,…,N.

[0024] Preferably, the matching degree between the type label and the product type is obtained by:

[0025] Get product keywords for product types and personnel keywords for each type of tag;

[0026] Product keywords and personnel keywords are input into a matching generation model to obtain the matching degree between the product type and the type label of the corresponding personnel; the matching generation model is obtained through artificial intelligence model training.

[0027] Preferably, the matching degree generation model is obtained by training an artificial intelligence model, including:

[0028] Obtaining from the database a number of product types and their corresponding product keywords, a number of type tags and their corresponding personnel keywords, and a matching score between the product keywords and the personnel keywords; the matching score is an average value of scores given by multiple experts on the degree of relevance between the product keywords and the personnel keywords;

[0029] Integrate several of the product keywords, personnel keywords and matching scores into several groups of training data and test data; use the training data to train the artificial intelligence model, and use the test data to test the artificial intelligence model, the input obtained is the product keywords and personnel keywords, the output is the artificial intelligence model with the corresponding matching scores, the matching scores are output as the matching degree, and finally a matching degree generation model is obtained; the artificial intelligence model includes a BP neural network model and a RBF neural network model.

[0030] Preferably, the step of selecting an advertisement delivery time period according to the time period adaptability score includes:

[0031] The time period adaptability score SPt corresponding to each recording time period is obtained, and the time period adaptability scores are sorted from high to low to obtain a score sorting table; the recording time period ranked first in the sorting table is used as the advertising delivery time period.

[0032] Preferably, generating a parameter adjustment signal according to environmental information includes:

[0033] Extracting light intensity and sound intensity from environmental information; generating a brightness adjustment signal according to the light intensity; generating a volume adjustment signal according to the sound intensity, wherein the parameter adjustment signal includes a brightness adjustment signal and a volume adjustment signal.

[0034] Preferably, generating a brightness adjustment signal according to light intensity includes:

[0035] The light intensity is marked as GQ; the adjusted brightness PL is calculated by the formula PL = BPL × α1 × GQ / BGQ; where BGQ is the set standard light intensity, BPL is the standard brightness under the set standard light intensity, α1 is the adjustment coefficient, α1>0; the specific value is set according to experience;

[0036] Determine whether the adjusted brightness is greater than the upper limit of the display brightness; if so, set the adjusted brightness to the upper limit of the display brightness; if not, when the adjusted brightness is less than the lower limit of the display brightness, set the adjusted brightness to the lower limit of the display brightness;

[0037] A brightness adjustment signal is generated according to the adjusted brightness.

[0038] Preferably, generating a volume adjustment signal according to the sound intensity includes:

[0039] The sound intensity is marked as SQ; the adjusted volume SY is calculated by the formula SY=BSY×α2×SQ / BSQ; wherein BSQ is the set standard sound intensity, BSY is the standard volume under the set standard sound intensity, α2 is the adjustment coefficient, α2>0; the specific value is set according to experience;

[0040] Determine whether the adjusted volume is greater than the upper limit of the device volume; if so, set the adjusted volume to the upper limit of the device volume; if not, when the adjusted volume is less than the lower limit of the device volume, set the adjusted volume to the lower limit of the device volume;

[0041] A volume adjustment signal is generated according to the adjusted volume.

[0042] Preferably, the step of adjusting the parameters of the advertising machine according to the parameter adjustment signal includes:

[0043] A brightness adjustment signal is obtained, and the screen brightness is adjusted according to the adjustment brightness in the brightness adjustment signal; a volume adjustment signal is obtained, and the device volume is adjusted according to the adjustment volume in the volume adjustment signal.

[0044] This application adjusts the corresponding parameters of the advertising machine through actual environmental data, so that the display brightness and device volume of the advertising machine can adapt to the real-time delivery environment, making the delivery effect of the advertising machine more reasonable.

[0045] Preferably, the advertisement acquisition module is further used to obtain delivery feedback through an advertisement cloud platform, and adjust the matching degree generation model according to the delivery feedback.

[0046] Preferably, the step of adjusting the matching degree generation model according to delivery feedback includes:

[0047] Extract the delivery revenue TS from the delivery feedback; obtain the estimated revenue TT;

[0048] The score correction coefficient PX is calculated by the formula PX = β × arctan ((TS-TT) / TT), where β is the adjustment coefficient and 0<β<1;

[0049] Obtain the matching score PF between the corresponding product keywords and each personnel keyword in the database; calculate the updated matching score GPF by the formula GPF=(PX+1)×PF, update the corresponding matching score in the database by updating the matching score, and use the product keywords, personnel keywords and the updated matching score to retrain the matching generation model.

[0050] Compared with the prior art, the present invention has the following beneficial effects:

[0051] 1. This application generates time period characteristic data for each time period by acquiring several historical record data of each time period in the advertising machine delivery area; generates a time period adaptability score of the corresponding advertisement in each delivery time period according to the time period characteristic data and the advertisement characteristic data, and selects the corresponding delivery time period according to the time period adaptability score; at the same time, when delivering advertisements, obtains environmental data of the advertising machine delivery area, generates a parameter adjustment signal according to the environmental data, and adjusts the corresponding parameters of the advertising machine according to the parameter adjustment signal; delivers advertisements according to the characteristics of different time periods in the advertising machine delivery area, so that the advertisement delivery is more suitable for the characteristics of each time period; and enhances the user experience.

[0052] 2. This application adjusts the corresponding parameters of the advertising machine through actual environmental data, so that the display brightness and device volume of the advertising machine can adapt to the real-time delivery environment, making the delivery effect of the advertising machine more reasonable. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0054] Figure 1 It is a schematic diagram of the principle of the intelligent instruction publishing system in the present invention;

[0055] Figure 2 The figure is a schematic diagram of the working process of the intelligent instruction publishing system in the present invention. DETAILED DESCRIPTION

[0056] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0057] See also Figure 1-Figure 2 The first aspect of the present invention provides an intelligent publishing instruction system for an integrated advertising machine, including: a data acquisition module, an advertisement acquisition module, a data processing module, an advertisement delivery module and a database.

[0058] Data acquisition module: acquires environmental data in real time through the data acquisition device connected to it; the data acquisition device includes image acquisition device, light acquisition device and sound acquisition device, etc. The environmental data includes light intensity and sound intensity, etc. The light intensity is the light intensity in the delivery area corresponding to the advertising machine, and the sound intensity is the sound intensity in the delivery area corresponding to the advertising machine;

[0059] Advertisement acquisition module: acquires advertisement characteristic data through the advertisement cloud platform connected thereto; advertisement characteristic data refers to the characteristics of the advertisement to be placed, including product types such as food, beverages and animal products, etc., provided by advertisers; the advertisement cloud platform is a platform for acquiring advertisement information;

[0060] Data processing module: obtain historical record data of each time period in the delivery area through the database connected to it; historical record data is the data recorded in each time period in the delivery area of ​​the advertising machine, such as the personnel data recorded by the advertising machine in the elevator between 10 o'clock and 11 o'clock; generate time period characteristic data of each time period according to the historical record data; the time period characteristic data is the personnel type corresponding to the recorded time period; obtain advertising characteristic data, and generate the time period adaptability score corresponding to each time period according to the advertising characteristic data and the characteristic data of each time period; the time period adaptability score can be understood as the matching degree between the advertisement and each time period; select the advertising delivery time period according to the time period adaptability score; and obtain environmental information, and generate parameter adjustment signals according to the environmental information; the parameter adjustment signals are used to adjust the corresponding parameters of the advertising machine, such as display brightness, device sound, etc.; the parameter adjustment signals include brightness adjustment signals and volume adjustment signals;

[0061] Advertisement delivery module: obtains the advertisement delivery time period, and delivers advertisements according to the advertisement delivery time period; obtains the parameter adjustment signal, and adjusts the parameters of the advertisement machine according to the parameter adjustment signal.

[0062] In this embodiment, by acquiring several historical record data of each time period in the advertising machine delivery area, the time period characteristic data of each time period is generated; the time period adaptability score of the corresponding advertisement in each delivery time period is generated according to the time period characteristic data and the advertisement characteristic data, and the corresponding delivery time period is selected according to the time period adaptability score; at the same time, when delivering the advertisement, the environmental data of the advertising machine delivery area is acquired, and a parameter adjustment signal is generated according to the environmental data, and the corresponding parameters of the advertising machine are adjusted according to the parameter adjustment signal; the advertisement is delivered according to the characteristics of different time periods in the advertising machine delivery area, so that the advertisement delivery is more suitable for the characteristics of each time period; at the same time, the corresponding parameters of the advertising machine are adjusted according to the environmental data around the advertising machine when the advertisement is delivered; so that the advertisement delivery adapts to the real-time environment and increases the user experience.

[0063] The period characteristic data of each period is generated according to the historical record data, including:

[0064] Extract the recording date, recording period and passenger image set from the historical record data; the passenger image set is obtained in the following manner: the images of people entering the delivery area are obtained in real time by the image acquisition device, and the images of different people collected in the same recording period are integrated into the passenger image set; that is, the passenger image set represents the situation of people in the delivery area corresponding to the advertising machine in a certain recording period on a certain recording date; each person image in the passenger image set is sequentially input into the person classification model to obtain the type label corresponding to each person image; the type labels corresponding to several person images in the passenger image set are integrated into the type label set corresponding to the recording period; the person classification model is obtained by training the artificial intelligence model;

[0065] The type tag set of each recording period is obtained in sequence; the tag quantity of each same type tag in several type tag sets corresponding to the same recording period of different recording dates, and the type tag and its corresponding tag quantity are integrated into the period characteristic data of the recording period.

[0066] This embodiment counts the personnel types in each recording period in the historical record data through the above steps to obtain the period characteristic data of each recording period, which is convenient for subsequent data analysis.

[0067] The personnel classification model is obtained through artificial intelligence model training, including: obtaining a number of personnel images and corresponding type labels from a database; the type label is the type of personnel in the personnel image set by relevant experts for the personnel image, such as children, teenagers, the elderly, workers, people who keep pets, etc.; integrating the personnel images and their corresponding type labels into several groups of training data and test data;

[0068] The artificial intelligence model is trained using training data, and the trained artificial intelligence model is tested using test data. Specifically, the personnel images in the test data are input into the trained artificial intelligence model, and a type label is output. When the output type label is consistent with the type label recorded in the test data, the next set of test data is used for testing; otherwise, the relevant parameters of the artificial intelligence are adjusted, and the test is performed again using the current set of test data; until a set proportion of the test data passes the test, a personnel classification model is finally obtained with the personnel image as input and the corresponding type label as output; wherein the artificial intelligence model includes a BP neural network model and an RBF neural network model.

[0069] The time period adaptability score corresponding to each time period is generated according to the advertisement characteristic data and the characteristic data of each time period, including:

[0070] Extract the product type from the advertisement characteristic data and mark it as ML; obtain each type label from the time period characteristic data and mark it as BLk, where k is the number of the type label;

[0071] The time period adaptability score SPt of each time period is calculated by the formula SPt=∑(Wtk×max(Pk(ML, BLk))), where k=1, 2, ..., N; N is the total number of type tags; Wtk is the weight coefficient of the type tag numbered k in the recording period numbered t; Pk(ML, BLk) is the degree of matching between the type tag numbered k and the product type.

[0072] In this embodiment, the time period adaptability score is calculated by the above formula. When the product type matches the type label of the person appearing in this time period more closely, it means that the person in the corresponding time period is more likely to have a demand for the content in the advertisement, and the corresponding time period adaptability score is higher; conversely, it means that the person in the corresponding time period is less likely to have a demand for the content in the advertisement, and the corresponding time period adaptability score is lower.

[0073] The weight coefficient is obtained by: obtaining the number of tags of each type in the period characteristic data of each recording period, and marking it as BStk;

[0074] The proportional coefficient Btk of the type tag numbered k in the recording period numbered t is calculated by the formula Btk=[ZBt / ∑(ZBt)]×(BStk / ZBt), where t=1, 2, ..., M; M is the total number of recording periods in a recording date; ZBt is the total number of type tags in the recording period numbered t, that is, ZBt=∑(BStk), k=1, 2, ..., N;

[0075] The weight coefficient Wtk is calculated by the formula Wtk=Btk / ∑(Btk), where k=1, 2, ..., N. The above formula is only applicable to the period when personnel are monitored in the delivery area; therefore, there is no need to limit ZBt>0.

[0076] In this embodiment, the weight coefficient of the matching degree between each type label and the product type of the advertisement product is set according to the distribution of type labels corresponding to personnel in each recording period. When there is no person of the corresponding type in the corresponding recording period, that is, the number of corresponding type labels is zero, the weight coefficient in this period is zero; when the number of corresponding type labels is greater or the proportion is greater, the corresponding weight coefficient is greater; by setting the weight coefficients of different type labels according to the actual personnel situation in each period, the time period matching coefficient obtained by calculation is more in line with the actual situation and more accurate.

[0077] The degree of matching between the type label and the product type is obtained in the following manner: obtaining product keywords of the product type, which are keywords extracted from the product introduction that can express the characteristics of the product; and personnel keywords of each type label, which are keywords extracted from the personnel introduction corresponding to the type label that can express the characteristics of the corresponding personnel type; inputting the product keywords and personnel keywords into the matching generation model to obtain the degree of matching between the product type and the type label of the corresponding personnel; the matching degree is the degree of adaptation between the type labels of the product type and the corresponding personnel, such as if the keywords of the advertisement include pet supplies, then the degree of matching between the product type and the type label of people who keep pets is higher; the matching generation model is obtained through training of an artificial intelligence model.

[0078] The matching degree generation model is obtained through artificial intelligence model training, including: obtaining a number of product types and their corresponding product keywords, a number of type tags and their corresponding personnel keywords, and the matching score between the product keywords and the personnel keywords through the database; the matching score is the average value of the scores given by multiple experts on the relevance between the product keywords and the personnel keywords;

[0079] Integrate several of the product keywords, personnel keywords and matching scores into several groups of training data and test data; use the training data to train the artificial intelligence model, and use the test data to test the artificial intelligence model. Specifically, input the product keywords and personnel keywords in the test data into the trained artificial intelligence model to obtain a matching score output. When the difference between the matching score in the corresponding test data and the output matching score is less than the set threshold, the next group of test data is tested. Otherwise, adjust the relevant parameters of the artificial intelligence and re-test the current group of test data until a set proportion of the test data passes the test. Only then can an artificial intelligence model with product keywords and personnel keywords as input and corresponding matching scores as output be obtained, and the matching scores are output as the matching degree, and finally a matching degree generation model is obtained; the artificial intelligence model includes a BP neural network model and an RBF neural network model.

[0080] Selecting an advertisement delivery period according to the period adaptability score includes: obtaining the period adaptability score corresponding to each recorded period, sorting the period adaptability score from high to low to obtain a score ranking table; and selecting the recorded period ranked first in the ranking table as the advertisement delivery period.

[0081] This embodiment selects the time period with the highest matching degree with the product corresponding to the advertisement as the corresponding advertisement delivery time period, so as to increase the delivery efficiency of the advertisement.

[0082] Generating a parameter adjustment signal according to the environmental information includes: extracting light intensity and sound intensity from the environmental information;

[0083] Generate a brightness adjustment signal according to the light intensity, including: marking the light intensity as GQ; calculating the adjustment brightness PL by the formula PL=BPL×α1×GQ / BGQ; wherein BGQ is the set standard light intensity, BPL is the standard brightness under the set standard light intensity, α1 is the adjustment coefficient, α1>0; the specific value is set according to experience;

[0084] Determine whether the adjusted brightness is greater than the upper limit of the display brightness; if yes, set the adjusted brightness to the upper limit of the display brightness; if no, when the adjusted brightness is less than the lower limit of the display brightness, set the adjusted brightness to the lower limit of the display brightness; generate a brightness adjustment signal according to the adjusted brightness; the upper limit of the display brightness is the upper limit of the brightness that the advertising machine display screen can adjust; the lower limit of the display brightness is the lower limit of the brightness that the advertising machine display screen can adjust; this embodiment calculates the adjusted brightness through the above formula, when the collected light intensity is stronger than the standard light intensity, the display brightness is adjusted higher accordingly; when the collected light intensity is weaker than the standard light intensity, the display brightness is adjusted lower accordingly; so that the display brightness of the advertising machine can adapt to the real-time delivery environment, so that the delivery effect of the advertising machine is more reasonable.

[0085] Generating a volume adjustment signal according to the sound intensity includes: marking the sound intensity as SQ; calculating the adjustment volume SY by the formula SY=BSY×α2×SQ / BSQ; wherein BSQ is the set standard sound intensity, BSY is the standard volume under the set standard sound intensity, α2 is the adjustment coefficient, α2>0; the specific value is set according to experience;

[0086] Determine whether the adjusted volume is greater than the upper limit of the device volume; if yes, set the adjusted volume to the upper limit of the device volume; if no, when the adjusted volume is less than the lower limit of the device volume, set the adjusted volume to the lower limit of the device volume; generate a volume adjustment signal according to the adjusted volume; the upper limit of the device volume is the upper limit of the adjustable volume of the playback device corresponding to the advertising machine; the lower limit of the device volume is the lower limit of the adjustable volume of the playback device corresponding to the advertising machine; this embodiment calculates the adjusted volume through the above formula, when the collected sound intensity is stronger than the standard sound intensity, the device volume is correspondingly increased; when the collected sound intensity is weaker than the standard sound intensity, the device volume is correspondingly decreased; so that the device volume of the advertising machine can adapt to the real-time delivery environment, so that the delivery effect of the advertising machine is more reasonable.

[0087] The parameter adjustment signal includes a brightness adjustment signal and a volume adjustment signal.

[0088] The parameters of the advertising machine are adjusted according to the parameter adjustment signal, including: obtaining a brightness adjustment signal, and adjusting the screen brightness according to the adjustment brightness in the brightness adjustment signal; obtaining a volume adjustment signal, and adjusting the device volume according to the adjustment volume in the volume adjustment signal.

[0089] The advertisement acquisition module is also used to obtain delivery feedback through the advertisement cloud platform and adjust the matching generation model according to the delivery feedback, including:

[0090] Extract the delivery revenue TS from the delivery feedback; obtain the estimated revenue TT;

[0091] The score correction coefficient PX is calculated by the formula PX=β×arctan((TS-TT) / TT), where β is the adjustment coefficient and 0<β<1; in this embodiment, β is set to a number much smaller than 1; in this embodiment, the score correction coefficient is calculated by the above formula, and the score correction coefficient is generated according to the gap between the actual investment return and the estimated return, thereby increasing the accuracy of the matching score between the product keywords and each personnel keyword.

[0092] Obtain the matching score PF between the corresponding product keywords and each personnel keyword in the database; calculate the updated matching score GPF by the formula GPF=(PX+1)×PF, update the corresponding matching score in the database by updating the matching score, and use the product keywords, personnel keywords and the updated matching score to retrain the matching generation model.

[0093] In this embodiment, when the investment revenue is higher than the estimated revenue, it means that the product type and the corresponding personnel in the corresponding recording period have a high degree of match, and the current matching degree is lower than the actual one; an upward correction should be made based on the matching score between the corresponding product keyword and each personnel keyword. When the investment revenue is lower than the estimated revenue, it means that the product type and the corresponding personnel in the corresponding recording period have a low degree of match, and the current matching degree is higher than the actual one; a downward correction should be made based on the matching score between the corresponding product keyword and each personnel keyword to increase the accuracy of the matching score between the product keyword and each personnel keyword.

[0094] Part of the data in the above formula is calculated by removing the dimension and taking its numerical value. The formula is a formula closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0095] Working principle of the present invention:

[0096] The present application generates time period characteristic data of each time period by acquiring several historical record data of each time period in the advertising machine delivery area; generates a time period adaptability score of the corresponding advertisement in each delivery time period according to the time period characteristic data and the advertisement characteristic data, and selects the corresponding delivery time period according to the time period adaptability score; at the same time, when delivering the advertisement, the environmental data of the advertising machine delivery area is acquired, a parameter adjustment signal is generated according to the environmental data, and the corresponding parameters of the advertising machine are adjusted according to the parameter adjustment signal; the advertisement is delivered according to the characteristics of different time periods in the advertising machine delivery area, so that the advertisement delivery is more suitable for the characteristics of each time period; at the same time, the corresponding parameters of the advertising machine are adjusted according to the environmental data around the advertising machine when the advertisement is delivered; so that the advertisement delivery adapts to the real-time environment and increases the user experience.

[0097] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. The intelligent publishing instruction system of the advertising machine includes: The data collection module, the advertisement acquisition module, the data processing module, the advertisement delivery module and the database are characterized in that: The data acquisition module is used to acquire environmental data in real time through a data acquisition device connected thereto; The advertisement acquisition module is used to acquire advertisement characteristic data through the advertisement cloud platform connected thereto; The data processing module: obtains historical record data of each time period in the delivery area through a database connected thereto; generates time period characteristic data of each time period according to the historical record data; obtains advertisement characteristic data, generates time period adaptability scores corresponding to each time period according to the advertisement characteristic data and each time period characteristic data; selects an advertisement delivery time period according to the time period adaptability scores; and obtains environmental information, and generates parameter adjustment signals according to the environmental information; The advertisement delivery module: obtains an advertisement delivery period, and delivers advertisements according to the advertisement delivery period; obtains a parameter adjustment signal, and adjusts parameters of the advertisement machine according to the parameter adjustment signal.

2. The intelligent instruction publishing system of the integrated advertising machine according to claim 1 is characterized in that: The period characteristic data of each period is generated according to the historical record data, including: Extracting the recording date, recording period and passenger image set from the historical record data; inputting each person image in the passenger image set into the person classification model in sequence to obtain the type label corresponding to each person image; integrating the type labels corresponding to several person images in the passenger image set into a type label set corresponding to the recording period; the person classification model is obtained by training the artificial intelligence model; The type tag set of each recording period is obtained in sequence; the tag quantity of each same type tag in several type tag sets corresponding to the same recording period of different recording dates, and the type tag and its corresponding tag quantity are integrated into the period characteristic data of the recording period.

3. The intelligent instruction publishing system of the integrated advertising machine according to claim 2 is characterized in that: The passenger image set is obtained in the following manner: images of persons entering the delivery area are acquired in real time by an image acquisition device, and images of different persons acquired in the same recording period are integrated into the passenger image set.

4. The intelligent instruction publishing system of the integrated advertising machine according to claim 2 is characterized in that: The personnel classification model is obtained through artificial intelligence model training, including: Acquire a number of person images and their corresponding type labels from a database; integrate the person images and their corresponding type labels into a number of sets of training data and test data; The artificial intelligence model is trained using training data, and the trained artificial intelligence model is tested using test data to obtain a personnel classification model whose input is a personnel image and whose output is a corresponding type label; wherein the artificial intelligence model includes a BP neural network model and an RBF neural network model.

5. The intelligent instruction publishing system of the integrated advertising machine according to claim 1 is characterized in that: Generating the time period adaptability score corresponding to each time period according to the advertisement characteristic data and the characteristic data of each time period includes: Extract the product type from the advertisement characteristic data and mark it as ML; obtain each type label from the time period characteristic data and mark it as BLk, where k is the number of the type label; The time period adaptability score SPt of each time period is calculated by the formula SPt=∑(Wtk×max(Pk(ML, BLk))), where k=1, 2, ..., N; N is the total number of type tags; Wtk is the weight coefficient of the type tag numbered k in the recording period numbered t; Pk(ML, BLk) is the degree of matching between the type tag numbered k and the product type.

6. The intelligent instruction publishing system of the integrated advertising machine according to claim 5, characterized in that: The weight coefficient is obtained by: obtaining the number of tags of each type in the period characteristic data of each recording period, and marking it as BStk; The proportional coefficient Btk of the type tag numbered k in the recording period numbered t is calculated by the formula Btk=[ZBt / ∑(ZBt)]×(BStk / ZBt), where t=1, 2, ..., M; M is the total number of recording periods in a recording date; ZBt is the total number of type tags in the recording period numbered t, that is, ZBt=∑(BStk), k=1, 2, ..., N; The weight coefficient Wtk is calculated by the formula Wtk=Btk / ∑(Btk), where k=1, 2,…,N.

7. The intelligent instruction publishing system of the integrated advertising machine according to claim 5, characterized in that: The degree of match between the type label and the product type is obtained in the following way: Get product keywords for product types and personnel keywords for each type of tag; Product keywords and personnel keywords are input into a matching generation model to obtain the matching degree between the product type and the type label of the corresponding personnel; the matching generation model is obtained through artificial intelligence model training.

8. The intelligent instruction publishing system of the integrated advertising machine according to claim 7, characterized in that: The matching degree generation model is obtained through artificial intelligence model training, including: Obtaining from the database a number of product types and their corresponding product keywords, a number of type tags and their corresponding personnel keywords, and a matching score between the product keywords and the personnel keywords; the matching score is an average value of scores given by multiple experts on the degree of relevance between the product keywords and the personnel keywords; Integrate several of the product keywords, personnel keywords and matching scores into several groups of training data and test data; use the training data to train the artificial intelligence model, and use the test data to test the artificial intelligence model, the input obtained is the product keywords and personnel keywords, the output is the artificial intelligence model with the corresponding matching scores, the matching scores are output as the matching degree, and finally a matching degree generation model is obtained; the artificial intelligence model includes a BP neural network model and a RBF neural network model.

9. The intelligent instruction publishing system of the integrated advertising machine according to claim 1, characterized in that: The step of selecting an advertisement delivery time period according to the time period adaptability score includes: The time period adaptability score SPt corresponding to each recording time period is obtained, and the time period adaptability scores are sorted from high to low to obtain a score sorting table; the recording time period ranked first in the sorting table is used as the advertising delivery time period.

10. The intelligent instruction publishing system of the integrated advertising machine according to claim 1, characterized in that: The generating of the parameter adjustment signal according to the environmental information comprises: Extracting light intensity and sound intensity from environmental information; generating a brightness adjustment signal according to the light intensity; generating a volume adjustment signal according to the sound intensity, wherein the parameter adjustment signal includes a brightness adjustment signal and a volume adjustment signal.

11. The intelligent instruction publishing system of the integrated advertising machine according to claim 10, characterized in that: The step of generating a brightness adjustment signal according to the light intensity comprises: The light intensity is marked as GQ; the adjusted brightness PL is calculated by the formula PL = BPL × α1 × GQ / BGQ; where BGQ is the set standard light intensity, BPL is the standard brightness under the set standard light intensity, and α1 is the adjustment coefficient; Determine whether the adjusted brightness is greater than the upper limit of the display brightness; if so, set the adjusted brightness to the upper limit of the display brightness; if not, when the adjusted brightness is less than the lower limit of the display brightness, set the adjusted brightness to the lower limit of the display brightness; A brightness adjustment signal is generated according to the adjusted brightness.

12. The intelligent instruction publishing system of the integrated advertising machine according to claim 10, characterized in that: Generating a volume adjustment signal according to the sound intensity includes: The sound intensity is marked as SQ; the adjusted volume SY is calculated by the formula SY=BSY×α2×SQ / BSQ; wherein BSQ is the set standard sound intensity, BSY is the standard volume under the set standard sound intensity, and α2 is the adjustment coefficient; Determine whether the adjusted volume is greater than the upper limit of the device volume; if so, set the adjusted volume to the upper limit of the device volume; if not, when the adjusted volume is less than the lower limit of the device volume, set the adjusted volume to the lower limit of the device volume; A volume adjustment signal is generated according to the adjusted volume.

13. The intelligent instruction publishing system of the integrated advertising machine according to claim 1, characterized in that: The step of adjusting the parameters of the advertising machine according to the parameter adjustment signal includes: A brightness adjustment signal is obtained, and the screen brightness is adjusted according to the adjustment brightness in the brightness adjustment signal; a volume adjustment signal is obtained, and the device volume is adjusted according to the adjustment volume in the volume adjustment signal.

14. The intelligent instruction publishing system of the integrated advertising machine according to claim 1, characterized in that: The advertisement acquisition module is also used to obtain delivery feedback through the advertisement cloud platform, and adjust the matching degree generation model according to the delivery feedback.

15. The intelligent instruction publishing system of the integrated advertising machine according to claim 14, characterized in that: The adjusting the matching degree generation model according to the delivery feedback includes: Extract the delivery revenue TS from the delivery feedback; obtain the estimated revenue TT; The score correction coefficient PX is calculated by the formula PX = β × arctan ((TS-TT) / TT), where β is the adjustment coefficient and 0<β<1; Obtain the matching score PF between the corresponding product keywords and each personnel keyword in the database; calculate the updated matching score GPF by the formula GPF=(PX+1)×PF, update the corresponding matching score in the database by updating the matching score, and use the product keywords, personnel keywords and the updated matching score to retrain the matching generation model.

Images