Intelligent advertisement putting management method and device, equipment and storage medium

Abstract

The invention discloses an intelligent advertisement putting management method and device, equipment and a storage medium, and relates to the technical field of intelligent recommendation managements.On one hand, browsing data generated by any user side before a user portrait label is newly added through a capture short video platform is subjected to data analysis; extracting the time spent by the user side for preliminarily establishing interests in the related short videos corresponding to the newly added user portrait labels, wherein the time is also the time spent by the user side for establishing interests in the short video contents belonging to the outside of the own big data labels; on the other hand, by analyzing commodity transaction records generated when the user side browses the short video advertisement on the short video platform, the time consumed by the short video advertisement for generating the advertising effect on the user side is analyzed; through matching the two kinds of time obtained through extraction, the effective delivery of the short video advertisement to the user side is improved, and meanwhile, the adhesion degree of the short video platform and the user side is promoted.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent recommendation management, and specifically to an intelligent advertisement placement management method, device, equipment and storage medium. Background Art

[0002] An advertisement placed in the form of a short video is a short video advertisement, which generally appears in various social, short video and news information apps. Through short videos within a few minutes or even just a few seconds, it can be pushed to users at a higher frequency, including style themes such as knowledge sharing, creative advertisements, social hotspots, fashion trends, etc. With the development and popularity of short videos in recent years, short video advertisements have emerged. Compared with the traditional advertisement model, short video advertisements are completely new both in form and content.

[0003] Intelligent placement is an efficient and accurate advertisement placement method in the marketing field. It uses artificial intelligence technology and data analysis methods to accurately display advertisements to the target audience group according to information such as the characteristics, interests and behaviors of users; the intelligent placement system analyzes a large amount of data such as the browsing records, search history, and social media behaviors of users to understand information such as the interests, purchase behaviors and preferences of users. Then, combined with machine learning and artificial intelligence technologies, the system can display personalized advertisements to potential customers at the appropriate time and place. Summary of the Invention

[0004] The purpose of the present invention is to provide an intelligent advertisement placement management method, device, equipment and storage medium to solve the problems raised in the prior art.

[0005] To achieve the above purpose, the present invention provides the following technical solution: An intelligent advertisement placement management method, the method includes:

[0006] Step S1: Monitor the change situation of the user portrait tags generated by the short video platform for each user terminal by collecting the browsing data of each user terminal;

[0007] Step S2: Whenever a new user portrait tag is captured for any user terminal, sort out the short video sequence browsed by any user terminal during the process of generating the new user portrait tag, and classify each short video included in the short video sequence according to the different corresponding user portrait tags;

[0008] Step S3: According to the distribution situation of different categories of short videos presented in the short video sequence, extract several target video sequences, and evaluate the browsing degree value of any user terminal for the target video sequence;

[0009] Step S4: In the short video sequence, analyze the changing trend of the browsing degree values presented by the user side for different target video sequences, and extract the characteristic browsing attribute values of the user side;

[0010] Step S5: Analyze the commodity transaction data generated by the user side through browsing short video advertisements within the short video platform, and analyze the effective duration that can prompt the user side to generate transaction behavior for any short video advertisement;

[0011] Step S6: Through data matching of the characteristic browsing attribute values of the user side and the effective duration of the short video advertisement, perform intelligent delivery on the user side where new user portrait tags are captured.

[0012] Preferably, step S2 includes:

[0013] Step S2-1: When it is monitored that the user portrait tag set Q(T 0 ) generated by a certain user side on the short video platform as of the time stamp T 0 , and the user portrait tag set Q(T e ) generated as of the time stamp T e satisfy Sum(F)=1, F = Q(T e ) - [Q(T e ) ∩ Q(T 0 )], and β ≧ T e - T 0 > 0, extract the short video sequence D browsed by a certain user side within the time range [T 0 , T e ; where β is the average time difference for updating user portrait tags set in advance, and Sum(F) represents the total number of user portrait tags included in the set F;

[0014] That is to say, as of the time stamp T e , the user portrait tag set generated for this user side has one more user portrait tag compared to the user portrait tag set generated for this user side as of the time stamp T 0 . Because the above relationship is satisfied between the two time stamps, it indicates that the time interval between the two time stamps meets the time difference for user portrait tag update;

[0015] Step S2-2: Set the only 1 user portrait tag included in the set F as the target tag, extract the short video corresponding to the target tag from the short video sequence D as the first feature video, and set the other short videos as the second feature videos.

[0016] Combined with the real scenario, it is often the case that the short video platform first uses the user portrait tag set Q(T 0)For related short video push to the client for reference, according to the actual browsing data generated by the client, the user portrait tags of the client are adjusted. Specifically, on the basis of the user portrait tag set Q(T 0 ), a new user portrait tag, that is, the target tag, is added.

[0017] Preferably, step S3 includes:

[0018] Step S3-1: Obtain the video duration T of each short video in the short video sequence D L , extract the actual browsing duration T of a certain client for each short video r ; Calculate the browsing degree value β = T r / T L ;

[0019] Step S3-2: Extract the first feature videos presented individually in 1 arrangement and the first feature videos presented continuously in M arrangements in the short video sequence D, and set them as the first target video sequence; Extract the second feature videos presented individually in 1 arrangement and the second feature videos presented continuously in M arrangements in the short video sequence D, and set them as the second target video sequence; where M≥2;

[0020] Step S3-3: Evaluate the browsing index α = [β 1 +β 2 +...+β n / n of a certain client for each first target video sequence or second target video sequence in the short video sequence D, where β 1 , β 2 ,..., β n respectively represent the browsing degree values of a certain client for the 1st, 2nd,..., nth short videos in each first target video sequence or second target video sequence.

[0021] Preferably, step S4 includes:

[0022] Step S4-1: In the order of presenting the first target video sequence and the second target video sequence in the short video sequence D, cumulatively calculate the browsing index for each first target video sequence in turn, and at the same time cumulatively calculate the browsing index for each second target video sequence in turn;

[0023] Step S4-2: When it is captured that the total browsing index δ 1 obtained after cumulatively calculating the browsing index for the 1st to the i-th first target video sequences, and the total browsing index δ 2 obtained after cumulatively calculating the browsing index for the 1st to the j-th second target video sequences satisfy δ 2 -δ 1When <θ, all the first feature videos included within the range of the first to the i-th first target video sequences are extracted and set as the target short videos, and the average actual viewing duration Tg of all the target short videos for a certain client is obtained. The average actual viewing duration Tg is set as the characteristic viewing attribute value of the certain client; where i ≤ R 1 , R 1 is the total number of first target video sequences included in the short video sequence D; j ≤ R 2 , R 2 is the total number of second target video sequences included in the short video sequence D; where θ is the exponential deviation threshold.

[0024] Preferably, step S5 includes:

[0025] Step S5-1: From the back-end of the short video platform, for any short video advertisement, collect the commodity transaction records generated by all clients through browsing any short video advertisement and according to the corresponding redirected commodity links, and set the commodity transaction records as the effect records of any short video advertisement;

[0026] Step S5-2: For each effect record of any short video advertisement, capture the actual viewing duration of any short video advertisement by the corresponding client before each effect record is established within the short video platform, and set the actual viewing duration as the effect duration of any short video advertisement corresponding to each effect record; obtain the average effect duration of any short video advertisement corresponding to all effect records;

[0027] Combined with the actual scenario, before a certain effect record is established within the short video platform by the corresponding client, the actual viewing duration of the corresponding short video advertisement by the client is basically the duration for the client to lock in the interest in the corresponding commodity in the short video advertisement. From the perspective of the short video advertisement itself, the commodity information transmitted to the client within this duration may be exactly the most attractive part of the short video advertisement.

[0028] Preferably, step S6 includes: If the characteristic viewing attribute value X extracted for a certain client with a newly added user portrait label satisfies X ≥ Y with the average effect duration Y corresponding to a certain short video advertisement, feedback to add the certain short video advertisement to the short video sequence to be delivered to a certain user.

[0029] To better implement the above method, an intelligent advertisement delivery management device is also proposed. The device includes a platform browsing data collection management module, a short video extraction and classification module, a browsing degree value evaluation management module, a characteristic browsing attribute value calculation module, an effect duration extraction management module, and an advertisement intelligent delivery management module;

[0030] The platform browsing data collection and management module is used to monitor the changes in the user portrait tags generated for each user terminal by collecting the browsing data of each user terminal on the short video platform;

[0031] The short video extraction and classification module is used to, whenever a new user portrait tag is captured for any user terminal, sort out the short video sequence browsed by the user terminal during the process of the new user portrait tag, and classify each short video included in the short video sequence according to the different corresponding user portrait tags;

[0032] The browsing degree value evaluation and management module is used to extract several target video sequences according to the distribution of different categories of short videos in the short video sequence, and evaluate the browsing degree value of the user terminal for any target video sequence;

[0033] The feature browsing attribute value calculation module is used to extract several target video sequences according to the distribution of different categories of short videos in the short video sequence, and evaluate the browsing degree value of the user terminal for any target video sequence;

[0034] The effect duration extraction and management module is used to analyze the change trend of the browsing degree value presented by the user terminal for different target video sequences in the short video sequence, and extract the feature browsing attribute value of the user terminal;

[0035] The intelligent advertising placement management module is used to perform intelligent placement on the user terminals for which new user portrait tags are captured by making data matching between the feature browsing attribute value of the user terminal and the effect duration of the short video advertisement.

[0036] A storage medium stores computer instructions. When the computer instructions are executed by a processor, any of the above intelligent advertising placement management methods can be implemented.

[0037] An intelligent advertising placement management device, the device includes a memory and at least one processor. Computer instructions are stored in the memory. The at least one processor calls the computer instructions in the memory so that the intelligent advertising placement management device executes any of the above intelligent advertising placement management methods.

[0038] Compared with the prior art, the beneficial effects of the present invention are as follows: On the one hand, by capturing the browsing data generated by the short video platform for any user terminal before adding a new user portrait label, analyzing the data, and extracting the time required for the user terminal to initially establish an interest in the relevant short videos corresponding to the new user portrait label. This time is also the time required for the user terminal to establish an interest in short video content outside its own big data labels. On the other hand, by analyzing the commodity transaction records generated by the user terminal through browsing short video advertisements on the short video platform, analyzing the time required for the short video advertisement to generate an advertising effect on the user terminal; by matching the two extracted times, the effective delivery of short video advertisements to the user terminal is improved, and at the same time, the adhesion between the short video platform and the user terminal is promoted. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a schematic flowchart of an intelligent advertising placement management method of the present invention;

[0040] Figure 2 It is a schematic structural diagram of an intelligent advertising placement management device of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0041] Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0042] Embodiment: As Figure 1 - Figure 2 shown, the present invention provides an intelligent advertising placement management method, and the method includes:

[0043] Step S1: Monitor the change situation of the user portrait labels generated by the short video platform for each user terminal by collecting the browsing data of each user terminal;

[0044] Step S2: Whenever a new user portrait label is captured for any user terminal, sort out the short video sequence browsed by the user terminal during the process of adding the new user portrait label, and classify each short video included in the short video sequence according to the different corresponding user portrait labels;

[0045] Among them, step S2 includes:

[0046] Step S2-1: When it is monitored that the set Q(T 0 ) of user portrait labels generated by the short video platform for a certain user terminal as of timestamp T 0 , and the set Q(T e ) of user portrait labels generated as of timestamp T e satisfy Sum(F) = 1, F = Q(T e ) - [Q(T e) ∩ Q(T 0 )], and β ≥ T e -T 0 > 0, extract the short video sequence D browsed by a certain client within the time range [T 0 , T e ; where β is the average interval time difference for updating the user portrait label set in advance, and Sum(F) represents the total number of user portrait labels included in the set F;

[0047] Step S2-2: Set the only 1 user portrait label included in the set F as the target label, extract the short videos corresponding to the target label from the short video sequence D as the first feature videos, and set the other short videos as the second feature videos;

[0048] For example, for the user portrait label set Q(T 0 ) generated by a certain client at the time stamp T 0 ) = {Label 1, Label 2, Label 3, Label 4}, and the user portrait label set Q(T e ) generated at the time stamp T e ) = {Label 1, Label 2, Label 3, Label 4, Label 5}, because compared with the user portrait label set Q(T 0 ) generated at the time stamp T 0 ), a new label: Label 5 is added to the user portrait label set Q(T e ) generated at the time stamp T e );

[0049] Because when the short video platform pushes short videos to the client, it always gives priority to pushing according to the user portrait label corresponding to the client;

[0050] If the short video sequence browsed by the client within the time range [T 0 , T e is {Short video 1, Short video 2, Short video 3, Short video 4, Short video 5, Short video 6, Short video 7}; where, Short video 1 corresponds to Label 1, Short video 2 corresponds to Label 3, Short video 3 corresponds to Label 5, Short video 4 corresponds to Label 2, Short video 5 corresponds to Label 3, Short video 6 corresponds to Label 5, and Short video 7 corresponds to Label 5;

[0051] In summary, it is necessary to set Short video 3, Short video 6, and Short video 7 as the first feature videos, and set the other short videos as the second feature videos;

[0052] Step S3: According to the distribution of different categories of short videos presented in the short video sequence, extract several target video sequences and evaluate the browsing degree value of the client for any target video sequence;

[0053] Among them, step S3 includes:

[0054] Step S3-1: Obtain the video duration T of each short video in the short video sequence D L , extract the actual viewing duration T of each short video by a certain client r ; Calculate the viewing degree value β of each short video by a certain client = T r / T L ;

[0055] Step S3-2: Extract the first feature videos presented in a single arrangement and the first feature videos presented in a continuous arrangement of M in the short video sequence D, and set them as the first target video sequence; Extract the second feature videos presented in a single arrangement and the second feature videos presented in a continuous arrangement of M in the short video sequence D, and set them as the second target video sequence; where M≥2;

[0056] Step S3-3: Evaluate the viewing index α of each first target video sequence or second target video sequence in the short video sequence D by a certain client = [β 1 +β 2 +...+β n / n, where β 1 , β 2 ,..., β n respectively represent the viewing degree values of the 1st, 2nd,..., nth short videos by a certain client in each first target video sequence or second target video sequence.

[0057] Step S4: In the short video sequence, analyze the change trend of the viewing degree values presented by the client for different target video sequences, and extract the characteristic viewing attribute values of the client;

[0058] Among them, step S4 includes:

[0059] Step S4-1: In the order of presenting the first target video sequence and the second target video sequence in the short video sequence D, sequentially accumulate the viewing indexes for each first target video sequence, and simultaneously accumulate the viewing indexes for each second target video sequence;

[0060] Step S4-2: When it is captured that the total viewing index δ 1 obtained after accumulating the viewing indexes for the 1st to the i-th first target video sequences, and the total viewing index δ 2 obtained after accumulating the viewing indexes for the 1st to the j-th second target video sequences satisfy δ 2 -δ 1When <θ, all the first feature videos included within the range of the first to the i-th first target video sequences are extracted and set as the target short videos. The average actual viewing duration Tg of all the target short videos for a certain client is obtained, and the average actual viewing duration Tg is set as the characteristic viewing attribute value of the certain client; where i ≤ R 1 , R 1 is the total number of first target video sequences included in the short video sequence D;

[0061] j ≤ R 2 , R 2 is the total number of second target video sequences included in the short video sequence D; where θ is the exponential deviation threshold;

[0062] Step S5: Analyze the commodity transaction data generated by the client through browsing short video advertisements within the short video platform, and analyze the effective duration of any short video advertisement that can prompt the client to generate a transaction behavior;

[0063] Among them, step S5 includes:

[0064] Step S5-1: From the backend of the short video platform, for any short video advertisement, collect all the commodity transaction records generated by all clients through browsing any short video advertisement and generating corresponding commodity links for jumping, and set the commodity transaction records as the effective records of any short video advertisement;

[0065] Step S5-2: For each effective record of any short video advertisement, capture the actual viewing duration of any short video advertisement by the corresponding client before each effective record is established within the short video platform, and set the actual viewing duration as the effective duration of any short video advertisement corresponding to each effective record; obtain the average effective duration of any short video advertisement corresponding to all effective records;

[0066] Step S6: Through data matching of the characteristic viewing attribute value of the client and the effective duration of the short video advertisement, perform intelligent placement on the clients for which new user portrait tags are captured;

[0067] Among them, step S6 includes: If the characteristic viewing attribute value X extracted for a certain client with a new user portrait tag and the average effective duration Y corresponding to a certain short video advertisement satisfy X ≥ Y, feedback to add the certain short video advertisement to the short video sequence to be placed for the certain user.

[0068] To better implement the above method, an intelligent advertisement placement management device is also proposed. The device includes a platform browsing data collection management module, a short video extraction and classification module, a browsing degree value evaluation management module, a characteristic browsing attribute value calculation module, an effective duration extraction management module, and an advertisement intelligent placement management module;

[0069] The platform browsing data collection and management module is used to monitor the changes in the user portrait tags generated by the short video platform for each user terminal by collecting the browsing data of each user terminal;

[0070] The short video extraction and classification module is used to, whenever a new user portrait tag is captured for any user terminal, sort out the short video sequence browsed by any user terminal during the process of generating the new user portrait tag, and classify each short video included in the short video sequence according to the different corresponding user portrait tags;

[0071] The browsing degree value evaluation and management module is used to extract a number of target video sequences according to the distribution of different categories of short videos in the short video sequence, and evaluate the browsing degree value of the user terminal for any target video sequence;

[0072] The feature browsing attribute value calculation module is used to extract a number of target video sequences according to the distribution of different categories of short videos in the short video sequence, and evaluate the browsing degree value of the user terminal for any target video sequence;

[0073] The effect duration extraction and management module is used to analyze the change trend of the browsing degree value of the user terminal for different target video sequences in the short video sequence, and extract the feature browsing attribute value of the user terminal;

[0074] The intelligent advertisement placement management module is used to perform intelligent placement on the user terminal that captures the occurrence of a new user portrait tag by making data matching between the feature browsing attribute value of the user terminal and the effect duration of the short video advertisement.

[0075] A storage medium stores computer instructions, and when the computer instructions are executed by a processor, an intelligent advertisement placement management method according to any one of claims 1-6 can be implemented.

[0076] An intelligent advertisement placement management device, the device includes a memory and at least one processor, computer instructions are stored in the memory, and the at least one processor calls the computer instructions in the memory to enable the intelligent advertisement placement management device to execute any of the above intelligent advertisement placement management methods.

[0077] Finally, it should be noted that the above are only the preferred embodiments of the present invention and are not used to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, for those skilled in the art, they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An intelligent advertising delivery management method, characterized in that: The method comprises: Step S1: monitoring the changes in user portrait labels generated by the short video platform for each user terminal by collecting browsing data of each user terminal; Step S2: Whenever a new user portrait tag is captured for any user terminal, the short video sequence browsed by the user terminal during the process of the new user portrait tag being added is sorted, and the short videos contained in the short video sequence are classified according to the corresponding user portrait tags; Step S3: extracting a number of target video sequences according to the distribution of short videos of different categories in the short video sequence, and evaluating the browsing degree value of any target video sequence by the user end; Step S4: in the short video sequence, analyzing the change trend of the browsing degree value presented by the user terminal to different target video sequences, and extracting the characteristic browsing attribute value of the user terminal; Step S5: Analyze the commodity transaction data generated by the user end through browsing the short video advertisements on the short video platform, and analyze the effect duration of any short video advertisement that can prompt the user end to generate transaction behavior; Step S6: By matching the characteristic browsing attribute values ​​of the user end with the effective duration of the short video advertisement, intelligent delivery is performed on the user end that captures the newly added user portrait tag.

2. The intelligent advertising delivery management method according to claim 1, characterized in that: The step S2 comprises: Step S2-1: When the short video platform generates a user portrait tag set Q(T0) for a certain user end at the time of the timestamp T0, and the user portrait tag set Q(T0) generated by the short video platform for a certain user end at the time of the timestamp T0 is detected, e The user portrait label set Q(T e ), satisfying Sum(F)=1, F=Q(T e )-[Q(T e )∩Q(T0)], and β≧T e - When T0>0, extract the time range of the user end [T0, T e ]; where β is the average interval time difference of the preset user portrait tag update, and Sum(F) represents the total number of user portrait tags contained in the set F; Step S2-2: Set the only user portrait tag contained in the set F as the target tag, extract the short video corresponding to the target tag from the short video sequence D as the first feature video, and set the other short videos as the second feature videos.

3. The intelligent advertising delivery management method according to claim 2, characterized in that: The step S3 comprises: Step S3-1: Obtain the video duration T of each short video in the short video sequence D L , extract the actual browsing time T of each short video by the user terminal r ; Calculate the browsing degree value of each short video by the user terminal β = T r / T L ; Step S3-2: extracting a first feature video with a single arrangement and a first feature video with M consecutive arrangements in the short video sequence D as a first target video sequence; extracting a second feature video with a single arrangement and a second feature video with M consecutive arrangements in the short video sequence D as a second target video sequence; wherein M≧2; Step S3-3: Evaluate the browsing index α of the user terminal for each first target video sequence or second target video sequence in the short video sequence D = [β 1 +β 2 +...+β n ] / n, where β 1 , β 2 , ..., β n They respectively represent the browsing degree values ​​of the 1st, 2nd, ..., nth short videos in each first target video sequence or second target video sequence by the certain user terminal.

4. The intelligent advertising delivery management method according to claim 3, characterized in that: The step S4 comprises: Step S4-1: accumulating browsing indexes of the first target video sequences and the second target video sequences in the short video sequence D in sequence, and accumulating browsing indexes of the second target video sequences in sequence; Step S4-2: When the total browsing index δ1 obtained by accumulating the browsing indexes of the 1st to the i-th first target video sequences and the total browsing index δ2 obtained by accumulating the browsing indexes of the 1st to the j-th second target video sequences satisfy δ2-δ1<θ, all the first feature videos contained in the range of the 1st to the i-th first target video sequences are extracted and set as target short videos, and the average actual browsing time Tg of all target short videos by the user terminal is obtained, and the average actual browsing time Tg is set as the characteristic browsing attribute value of the user terminal; wherein, i≦R1, R1 is the total number of first target video sequences included in the short video sequence D; j≦R2, R2 is the total number of second target video sequences included in the short video sequence D; wherein θ is the index deviation threshold.

5. The intelligent advertising delivery management method according to claim 4, characterized in that: The step S5 comprises: Step S5-1: From the backend of the short video platform, for any short video advertisement, collect the commodity transaction records generated by all users browsing the arbitrary short video advertisement and following the corresponding redirected commodity links, and set the commodity transaction records as the effect records of the arbitrary short video advertisement; Step S5-2: For each effect record of any short video advertisement, capture the actual browsing time of the corresponding user terminal for the arbitrary short video advertisement before the corresponding user terminal establishes each effect record in the short video platform, and set the actual browsing time as the effect time corresponding to each effect record of any short video advertisement; obtain the average effect time of all effect records corresponding to any short video advertisement.

6. The intelligent advertising delivery management method according to claim 4, characterized in that: The step S6 includes: if the characteristic browsing attribute value X extracted from a certain user terminal with a newly added user portrait tag satisfies X≧Y with the average effect duration Y corresponding to a certain short video advertisement, feedback is given to add the certain short video advertisement into the short video sequence to be delivered to the certain user.

7. An intelligent advertising delivery management device, used to execute an intelligent advertising delivery management method according to any one of claims 1 to 5, characterized in that: The device includes a platform browsing data collection management module, a short video extraction and classification module, a browsing degree value evaluation management module, a feature browsing attribute value calculation module, an effect duration extraction management module, and an advertising intelligent delivery management module; The platform browsing data collection and management module is used to monitor the changes in the user portrait labels generated by the short video platform for each user terminal by collecting browsing data of each user terminal; The short video extraction and classification module is used to sort out the short video sequence browsed by any user terminal in the process of capturing the occurrence of a new user portrait tag, and classify the short videos contained in the short video sequence according to the corresponding user portrait tags. The browsing degree value evaluation management module is used to extract a number of target video sequences according to the distribution of short videos of different categories in the short video sequence, and evaluate the browsing degree value of any target video sequence by the user terminal; The characteristic browsing attribute value calculation module is used to extract a number of target video sequences according to the distribution of short videos of different categories in the short video sequence, and evaluate the browsing degree value of any target video sequence by the user end; The effect duration extraction management module is used to analyze the change trend of the browsing degree value presented by the user terminal to different target video sequences in the short video sequence, and extract the characteristic browsing attribute value of the user terminal; The advertising intelligent delivery management module is used to perform intelligent delivery to the user terminal that has captured the newly added user portrait tag by matching the characteristic browsing attribute value of the user terminal with the effective duration of the short video advertisement.

8. A storage medium, characterized in that: The storage medium stores computer instructions, and when the computer instructions are executed by the processor, the intelligent advertising delivery management method described in any one of claims 1 to 6 can be implemented.

9. An intelligent advertising delivery management device, characterized in that: The device includes a memory and at least one processor, wherein the memory stores computer instructions, and the at least one processor calls the computer instructions in the memory to enable the intelligent advertisement delivery management device to execute an intelligent advertisement delivery management method as described in any one of claims 1-6.

Images