Intelligent advertisement analysis management method and system based on block chain data

Abstract

The invention discloses an intelligent advertisement analysis management method and system based on block chain data, and relates to the technical field of advertisements, and the method comprises the steps: collecting interaction behavior data generated when a user interacts with advertisement content; converting the interactive behavior data into advertisement behavior tags in real time, performing chain association with the front and back advertisement behavior tags through an encrypted hash algorithm, and storing the advertisement behavior tags in a block chain; after a user receives personalized advertisement content, personalized interaction behavior data of the user and the personalized advertisement content are recorded in real time and stored in an advertisement behavior label chain, and the advertisement behavior label chain updates an advertisement putting index according to the personalized interaction behavior data recorded in real time. According to the invention, through real-time conversion and chain storage of the user interaction behavior data, a dynamically updated advertisement behavior label chain is generated, so that the interest preference of the user can be more accurately captured.

Description

Technical Field

[0001] The present invention relates to the field of advertising technology, and particularly to an intelligent advertising analysis and management method and system based on blockchain data. Background Art

[0002] With the rapid development of Internet and big data technologies, the advertising industry has gradually moved towards intelligence and personalization. Traditional advertising delivery methods mainly rely on manual experience and simple statistical analysis, and it is difficult to adapt to the changing interests and behavior patterns of users in real time. In recent years, blockchain technology and artificial intelligence models have been widely used in advertising analysis and management. Through encryption algorithms and data chain structures, the transparency of the advertising delivery process and the efficient processing of data have been achieved.

[0003] However, existing intelligent advertising analysis methods still face some challenges. First, data security and privacy protection are important issues. Existing systems have risks of being tampered with and leaked during data transmission and storage. Second, the degree of personalization of advertising content is not high, and it is difficult to accurately capture users' interest preferences, resulting in unsatisfactory advertising effects. In addition, existing behavior prediction models need to improve real-time performance and accuracy when facing large-scale data. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides an intelligent advertising analysis and management method based on blockchain data to solve the problems of advertising data security and privacy protection, personalized generation of advertising content, and real-time behavior prediction.

[0006] To solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides an intelligent advertising analysis and management method based on blockchain data, which includes collecting interaction behavior data generated when users interact with advertising content;

[0008] Real-time converting the interaction behavior data into advertising behavior tags, and chain associating and storing the advertising behavior tags with the previous and subsequent advertising behavior tags through an encrypted hash algorithm in the blockchain;

[0009] Performing data preprocessing on the advertising behavior tag chain, extracting advertising delivery metrics, and using a behavior sequence analysis model to perform real-time behavior prediction on the advertising delivery metrics, and outputting a dynamically updated user conversion path;

[0010] Based on the dynamically updated user conversion path, a generative adversarial network is called to generate personalized advertising content according to the user's historical interaction behavior data. The version and distribution of the personalized advertising content are recorded through an advertising behavior tag chain, and the personalized advertising content is pushed to the user;

[0011] After the user receives the personalized advertising content, the personalized interaction behavior data between the user and the personalized advertising content is recorded in real time and stored in the advertising behavior tag chain. The advertising behavior tag chain updates the advertising placement metrics according to the real-time recorded personalized interaction behavior data.

[0012] As a preferred solution of the intelligent advertising analysis and management method based on blockchain data according to the present invention, wherein: the interaction behavior data includes interaction operation data, device environment data, interest preference data, behavior sequence data, personalized advertising data and target conversion data.

[0013] As a preferred solution of the intelligent advertising analysis and management method based on blockchain data according to the present invention, wherein: the interaction behavior data is real-time converted into advertising behavior tags, and is chain-linked with the previous and subsequent advertising behavior tags through an encrypted hash algorithm and stored in the blockchain. The specific steps are as follows.

[0014] The interaction behavior data is normalized using a standardization function;

[0015] The normalized interaction behavior data is mapped to a high-dimensional semantic space through a non-linear embedding function to generate a corresponding quantum state probability distribution;

[0016] The density matrix is used to describe the eigenvalues of the quantum state probability distribution and used as semantic features to identify the semantic patterns in the interaction behavior data;

[0017] Cluster analysis and time correlation analysis are performed on the identified semantic patterns to generate advertising behavior tags corresponding to the interaction behavior data;

[0018] The semantic correlation weight between advertising behavior tags is calculated using a quantum state correlation matrix. The expression is:

[0019]

[0020] where, A ij represents the semantic correlation weight between advertising behavior tags i and j, and i and j represent two different advertising behavior tags. represents the complex conjugate quantum state function of advertising behavior tag i at time t, and ψ j (t) represents the quantum state function of advertising behavior tag j at time t. is a Gaussian time weight function, t 0Indicates the reference time point of the current behavior event, α represents the time weight factor, and dt represents the tiny time increment in the integration operation;

[0021] Form a semantic association matrix from the two-dimensional matrix composed of the semantic association weights of all advertising behavior label pairs (i, j);

[0022] Based on the semantic association matrix, construct a set of paths for all possible advertising behavior label pairs

[0023] Construct a chain structure according to the maximum weight path of the semantic association matrix to generate the optimal advertising behavior label chain. The expression is:

[0024]

[0025] Among them, Represents the optimal advertising behavior label chain, Represents finding the maximum weight path from the set of paths for all possible advertising behavior label pairs And Represents summing up the semantic association weights of all advertising behavior label pairs (i, j) in the path set;

[0026] Based on the optimal advertising behavior label chain Generate the encrypted hash value of the optimal advertising behavior label chain. The expression is:

[0027]

[0028] Among them, H k Represents the encrypted hash value of the k-th optimal advertising behavior label chain. k represents the index variable of the optimal advertising behavior label chain. SHA3 is the secure hash algorithm, and ρ k (t) represents the quantum state density function of the k-th advertising behavior label chain at time t, Represents the Gaussian decay factor, and β represents the time weight parameter;

[0029] When a new advertising behavior label is added to the advertising behavior label chain, recalculate the semantic association weights between the new advertising behavior label and all advertising behavior labels in the optimal advertising behavior label chain, update the optimal path, and calculate the encrypted hash value of the new optimal advertising behavior label chain;

[0030] Store the semantic association weight of the new advertising behavior label and the encrypted hash value of the updated optimal advertising behavior label chain as new label information in the blockchain. Each advertising behavior label chain serves as a block;

[0031] Establish the semantic index of each advertising behavior label chain and use the quantum Bloom filter to optimize the query efficiency of the advertising behavior label chain. The expression is:

[0032]

[0033] Among them, Q(x) represents the result of the quantum Bloom filter, m represents the number of hash functions in the Bloom filter, z represents the index of the hash function in the Bloom filter, and b z represents the Boolean value of the z-th hash function, and γ z represents the weight parameter of the z-th dimension, and x represents the input query value. represents the symbol conversion term, represents the weight decay function;

[0034] When the advertising behavior label chain is updated, synchronously update the semantic indexes of the corresponding blockchain and the advertising behavior label chain.

[0035] As a preferred solution of the intelligent advertising analysis and management method based on blockchain data according to the present invention, wherein: performing data preprocessing on the advertising behavior label chain, extracting advertising placement metrics, and using a behavior sequence analysis model to perform real-time behavior prediction on the advertising placement metrics, and outputting a dynamically updated user conversion path. The specific steps are as follows.

[0036] Clean and standardize the advertising behavior label chain, extract the time series, click-through rate, and conversion rate of user interaction behaviors, and generate a behavior feature matrix containing advertising placement metrics based on semantic association weights;

[0037] Organize the advertising placement metrics in the behavior feature matrix into a sliding window sequence in chronological order, and at the same time calculate the time interval features between the advertising placement metrics to form a behavior sequence containing short-term and long-term dependencies;

[0038] Input the behavior sequence into a behavior sequence analysis model based on LSTM, combine the historical behavior feature matrix and the current behavior feature matrix, and predict the next advertising behavior label of the user and its conversion probability;

[0039] Screen the advertising behavior labels according to the conversion probability and insert them into the current short-term behavior sequence, and at the same time record the conversion probability, and output a dynamically updated user conversion path.

[0040] As a preferred solution of the intelligent advertising analysis and management method based on blockchain data according to the present invention, wherein: based on the dynamically updated user conversion path, call a generative adversarial network to generate personalized advertising content according to the user's historical interaction behavior data. The specific steps are as follows.

[0041] Based on the user conversion path, extract the behavior feature matrix and the advertising context, and combine them to generate a joint feature matrix;

[0042] Based on the semantic association weights between advertising behavior tags, the joint feature matrix is ​​weighted to generate a high-dimensional feature vector;

[0043] Input the high-dimensional feature vector into the bidirectional generative adversarial network, the generator generates personalized advertising content, and the discriminator verifies the authenticity of the personalized advertising content and its matching degree with the interactive behavior data;

[0044] Through behavioral path analysis, verify whether personalized advertising content matches the user conversion path and optimize the relevance and logic of the content.

[0045] As a preferred solution of the intelligent advertising analysis and management method based on blockchain data described in the present invention, the specific steps are as follows: the version and distribution of personalized advertising content are recorded through the advertising behavior tag chain, and the personalized advertising content is pushed to the user.

[0046] Based on the generated personalized advertising content, the semantic features described by the eigenvalues ​​of the quantum state density matrix are extracted, and a unique version identifier is generated using an encrypted hash algorithm, which is then associated and stored with the advertising behavior tag chain;

[0047] Generate a distribution record for each advertising content, store the distribution record as a block in the blockchain, and generate an advertising distribution record chain;

[0048] Combine the user's advertising behavior tag chain and advertising distribution record chain to dynamically match the advertising content version with the highest conversion probability;

[0049] The selected advertising content version is pushed to the user through the content distribution interface.

[0050] As a preferred solution of the intelligent advertising analysis and management method based on blockchain data described in the present invention, after the user receives the personalized advertising content, the personalized interaction behavior data between the user and the personalized advertising content is recorded in real time and stored in the advertising behavior tag chain. The advertising behavior tag chain updates the advertising delivery index according to the personalized interaction behavior data recorded in real time. The specific steps are as follows:

[0051] Monitor click, view and conversion data through native application integrated SDK, and collect personalized interaction behavior data in timestamp format;

[0052] Perform cryptographic hashing on the personalized interaction behavior data, update the advertising behavior tag, and semantically associate it with the historical advertising behavior tag chain and store it in the blockchain;

[0053] Click, view and conversion data are extracted from the updated advertising behavior tag chain and integrated with the behavior feature matrix to dynamically correct the real-time values ​​of advertising delivery indicators.

[0054] Second aspect, the present invention provides an intelligent advertising analysis and management system based on blockchain data, including a data collection module, a storage module, a behavior prediction module, a push module, and an interaction record module;

[0055] The data collection module is used to collect interaction behavior data generated when a user interacts with advertising content;

[0056] The storage module is used to convert the interaction behavior data into advertising behavior tags in real time, and perform chain association with the previous and subsequent advertising behavior tags through an encrypted hash algorithm and store them in the blockchain;

[0057] The behavior prediction module is used to perform data preprocessing on the advertising behavior tag chain, extract advertising placement metrics, and use a behavior sequence analysis model to perform real-time behavior prediction on the advertising placement metrics, and output a dynamically updated user conversion path;

[0058] The push module is used to, based on the dynamically updated user conversion path, call a generative adversarial network, generate personalized advertising content according to the user's historical interaction behavior data, record the version and distribution status of the personalized advertising content through the advertising behavior tag chain, and push the personalized advertising content to the user;

[0059] The interaction record module is used to, after the user receives the personalized advertising content, record the personalized interaction behavior data of the user with the personalized advertising content in real time and store it in the advertising behavior tag chain, and the advertising behavior tag chain updates the advertising placement metrics according to the real-time recorded personalized interaction behavior data.

[0060] Third aspect, the present invention provides a computer device, including a memory and a processor, where the memory stores a computer program, and: when the computer program is executed by the processor, any step of the intelligent advertising analysis and management method based on blockchain data as described in the first aspect of the present invention is implemented.

[0061] Fourth aspect, the present invention provides a computer-readable storage medium, on which a computer program is stored, and: when the computer program is executed by the processor, any step of the intelligent advertising analysis and management method based on blockchain data as described in the first aspect of the present invention is implemented.

[0062] The beneficial effects of the present invention are as follows: By introducing blockchain technology, the advertising data has immutability and high security during the transmission and storage processes, greatly improving data security and privacy protection. Through the real-time conversion and chained storage of user interaction behavior data, a dynamically updated advertising behavior tag chain is generated, enabling more accurate capture of users' interest preferences. By adopting generative adversarial network technology, personalized advertising content is generated based on users' historical behavior data, not only improving the relevance and attractiveness of the advertisements, but also being able to record the personalized interaction behavior data in real time after the users receive the advertising content, further optimizing the advertising placement metrics and user conversion path. In summary, the present invention has significant beneficial effects in terms of data security, personalized advertising generation, and real-time behavior prediction, and can effectively improve the accuracy and effectiveness of advertising placement. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for use in the description of the embodiments. Obviously, the following-described drawings are only some embodiments of the present invention, and those of ordinary skill in the art can obtain other drawings based on these drawings without creative efforts.

[0064] Figure 1 It is a flowchart of the intelligent advertising analysis and management method based on blockchain data in Embodiment 1.

[0065] Figure 2 It is a module diagram of the intelligent advertising analysis and management system based on blockchain data in Embodiment 1. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0066] In order to make the above-mentioned objects, features, and advantages of the present invention more obvious and understandable, the following will make a detailed description of the specific embodiments of the present invention with reference to the drawings in the specification.

[0067] In the following description, many specific details are set forth to fully understand the present invention, but the present invention can also be implemented in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0068] Secondly, the so-called "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that can be included in at least one implementation manner of the present invention. The "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it a separate or selectively exclusive embodiment from other embodiments.

[0069] Embodiment 1, refer to Figure 1 and Figure 2, which is the first embodiment of the present invention. This embodiment provides an intelligent advertising analysis and management method based on blockchain data, including the following steps:

[0070] S1. Collect the interaction behavior data generated when users interact with advertising content.

[0071] Furthermore, the interaction behavior data includes interaction operation data, device environment data, interest preference data, behavior sequence data, personalized advertising data, and target conversion data.

[0072] Specifically, the integrated SDK monitors and collects the interaction operation data of users with advertising content in real time, which includes operations such as clicks, swipes, and long presses. At the same time, it also collects device environment data, such as device type, operating system, browser version, etc. Then, according to the operation records of users, the interest preference data of users is extracted, and by analyzing the advertising content browsed by users and their stay time, etc., the interest fields of users are inferred. Subsequently, the behavior sequence data of users is generated. Through time series analysis, the advertising interaction behaviors of users at different time points are recorded and organized into behavior sequence data. Next, the interest preference data and behavior sequence data of users are combined to generate personalized advertising data for identifying and matching advertising content suitable for users. Finally, the target conversion data of users is tracked, and it is recorded whether users have completed specific target behaviors, such as purchases, registrations, etc., after advertising interactions to evaluate the actual effect of the advertising.

[0073] S2. Real-time convert the interaction behavior data into advertising behavior tags, and perform chained association with the previous and subsequent advertising behavior tags through an encryption hash algorithm and store them in the blockchain.

[0074] Furthermore, a standardization function is used to normalize the interaction behavior data;

[0075] Among them, the normalization process helps to improve the comparability of data and lay a foundation for subsequent processing steps.

[0076] Map the normalized interaction behavior data to a high-dimensional semantic space through a non-linear embedding function to generate a corresponding quantum state probability distribution;

[0077] Among them, generating the quantum state probability distribution is to capture the semantic features in the data.

[0078] Preferably, the role of the non-linear embedding function is to capture the complex semantic relationships in the data, making the originally linearly inseparable data linearly separable in the high-dimensional space.

[0079] Use the density matrix to describe the eigenvalues of the quantum state probability distribution and use them as semantic features to identify the semantic patterns in the interaction behavior data;

[0080] Among them, the semantic features include, but are not limited to, the semantic feature strength described by the eigenvalues of the quantum state density matrix, and other semantic feature information related to the interaction behavior data.

[0081] It should be noted that the density matrix can not only accurately describe the semantic features in the data, but also reflect the probability nature of the data distribution. By performing cluster analysis and time correlation analysis on the semantic patterns of the identifiers, advertisement behavior labels corresponding to the interaction behavior data can be generated.

[0082] Perform cluster analysis and time correlation analysis on the semantic patterns of the identifiers to generate advertisement behavior labels corresponding to the interaction behavior data;

[0083] Preferably, cluster analysis helps to discover potential patterns and structures in the data, while time correlation analysis can reveal the time-dependent relationships in the behavior data.

[0084] Calculate the semantic association weight between advertisement behavior labels using the quantum state correlation matrix, and the expression is:

[0085]

[0086] Among them, A ij represents the semantic association weight between advertisement behavior labels i and j, where i and j represent two different advertisement behavior labels, represents the complex conjugate quantum state function of advertisement behavior label i at time t, and ψ j (t) represents the quantum state function of advertisement behavior label j at time t, is the Gaussian time weight function, t 0 represents the reference time point of the current behavior event, α represents the time weight factor, and dt represents the infinitesimal time increment in the integration operation;

[0087] It should be noted that ψ j (t) represents the quantum state function of advertisement behavior label j at time t. This description stems from the concept in quantum mechanics and is mainly used to capture the dynamic semantic features of the advertisement behavior label. Specifically, the quantum state function describes the state of advertisement behavior label j at time t, and this state can be regarded as a probability distribution in the high-dimensional semantic space. The "quantum state function" here does not refer to the real quantum state in physics, but a mathematical model used to represent and analyze the complex semantic relationships in the advertisement behavior data.

[0088] Specifically, the quantum state function can be regarded as a probability wave function, which provides the state information of advertisement behavior label j at time t. Through this representation method, the changes of advertisement behavior labels at different time points can be accurately modeled and analyzed.

[0089] Preferably, a quantum state function is used to describe advertising behavior tags, which can not only capture the complex semantic relationships in interaction behavior data, but also accurately model and analyze these relationships through a mathematical model, so as to optimize the advertising placement strategy and improve the advertising effect.

[0090] Form a semantic association matrix from the two-dimensional matrix composed of the semantic association weights of all advertising behavior tag pairs (i, j);

[0091] It should be noted that the process of using the quantum state association matrix to calculate the semantic association weights between advertising behavior tags can construct a two-dimensional semantic association matrix between advertising behavior tags. By constructing a chain structure through the maximum weight path of the semantic association matrix, an optimal advertising behavior tag chain is generated, thereby optimizing the advertising placement strategy.

[0092] Based on the semantic association matrix, construct a set of paths for all possible advertising behavior tag pairs

[0093] Among them, each advertising behavior tag path is composed of advertising behavior tag pairs, and its weight is represented by the semantic association weight between advertising behavior tag pairs.

[0094] Specifically, based on the semantic association matrix, first traverse all advertising behavior tag pairs, and generate a potential advertising behavior tag path for each pair of advertising behavior tags. Each advertising behavior tag path is composed of two tags i and tag j, and the weight of the advertising behavior tag path is the semantic association weight A between tag i and tag j ij . Then, use the combination method of all advertising behavior tag pairs to construct a set of advertising behavior tag pair paths

[0095] Construct a chain structure according to the maximum weight path of the semantic association matrix to generate an optimal advertising behavior tag chain, and the expression is:

[0096]

[0097] Among them, represents the optimal advertising behavior tag chain, represents finding the maximum weight path from the set of paths of all possible advertising behavior tag pairs , represents summing the semantic association weights of all advertising behavior tag pairs (i, j) in the path set;

[0098] Preferably, by calculating the semantic association weights between advertising behavior tags, the most semantically relevant tag path is found, ensuring the accuracy and relevance of advertising placement, thereby maximizing the advertising effect. By constructing an optimal advertising behavior tag chain, the advertising placement strategy can be dynamically optimized, improving the user conversion rate and the effectiveness of advertising placement, ensuring that the advertising content always highly matches the user needs, achieving personalized recommendation and precision marketing, effectively enhancing the ROI (return on investment) of advertising placement, and enhancing the user experience.

[0099] Based on the optimal advertising behavior tag chain , the encrypted hash value of the optimal advertising behavior tag chain is generated, and the expression is:

[0100]

[0101] where H k represents the encrypted hash value of the k-th optimal advertising behavior tag chain, k represents the index variable of the optimal advertising behavior tag chain, SHA3 is a secure hash algorithm, and ρ k (t) represents the quantum state density function of the k-th advertising behavior tag chain at time t, represents the Gaussian decay factor, and β represents the time weight parameter;

[0102] It should be noted that by generating the encrypted hash value of the optimal advertising behavior tag chain in this way, the security and immutability of the advertising behavior tag chain are ensured. For newly added advertising behavior tags, it is necessary to recalculate their semantic association weights with all tags in the existing tag chain, update the optimal path, and generate a new encrypted hash value to ensure the dynamic optimization of the tag chain.

[0103] When a newly added advertising behavior tag is added to the advertising behavior tag chain, recalculate the semantic association weights between the new advertising behavior tag and all advertising behavior tags in the optimal advertising behavior tag chain, update the optimal path, and calculate the encrypted hash value of the new optimal advertising behavior tag chain;

[0104] Among them, the newly added advertising behavior tags refer to the new behavior tags generated during the advertising placement process. These tags are created based on the latest advertising interaction behavior data of users. They reflect the behavior characteristics of users in the latest advertising interaction, including clicks, swipes, views, conversions, etc. The introduction of these newly added tags can provide more accurate user behavior analysis, thereby optimizing the advertising placement strategy and improving the advertising effect. Whenever new interaction behavior data is generated, corresponding newly added advertising behavior tags will be generated, and these tags need to be semantically associated and analyzed with the existing tags to ensure the dynamic optimization and update of the advertising behavior tag chain.

[0105] It should be noted that when a newly added advertising behavior tag is added to the advertising behavior tag chain, it is necessary to recalculate its semantic association weights with all tags in the existing tag chain. This is because the newly added advertising behavior tag may change the original semantic association structure and affect the maximum weight path of the advertising behavior tag chain. The specific steps are as follows: First, calculate the semantic association weights between the new advertising behavior tag and each advertising behavior tag in the maximum weight path. Then, based on these new weights, update the maximum weight path of the advertising behavior tag chain to ensure that the new path reflects the latest association relationship. Finally, perform an encrypted hash processing on the updated maximum weight path to generate a new encrypted hash value to ensure the security and immutability of the data.

[0106] Store the semantic association weights of the newly added advertising behavior tag and the encrypted hash value of the updated optimal advertising behavior tag chain as new tag information in the blockchain, with each advertising behavior tag chain as a block;

[0107] Preferably, taking each advertising behavior tag chain as a block can ensure the integrity, transparency, and immutability of the advertising behavior data, thereby improving the security and credibility of the data.

[0108] Establish a semantic index for each advertising behavior tag chain and use a quantum Bloom filter to optimize the query efficiency of the advertising behavior tag chain. The expression is:

[0109]

[0110] where Q(x) represents the result of the quantum Bloom filter, m represents the number of hash functions in the Bloom filter, z represents the index of the hash function in the Bloom filter, b z represents the Boolean value of the z-th hash function, γ z represents the weight parameter of the z-th dimension, x represents the input query value, represents the symbol conversion term, represents the weight decay function;

[0111] It should be noted that by introducing multi-dimensional weight parameters, the query accuracy and efficiency are significantly improved. When the advertising behavior tag chain is updated, synchronously update the corresponding blockchain and the semantic index of the advertising behavior tag chain to ensure real-time performance and consistency.

[0112] When the advertising behavior tag chain is updated, synchronously update the corresponding blockchain and the semantic index of the advertising behavior tag chain to ensure the consistency and integrity of all relevant data, thereby improving the query efficiency and the accuracy of data management.

[0113] S3. Perform data preprocessing on the advertising behavior label chain, extract advertising placement metrics, use a behavior sequence analysis model to perform real-time behavior prediction on the advertising placement metrics, and output a dynamically updated user conversion path.

[0114] Furthermore, clean and standardize the advertising behavior label chain, extract the time series, click-through rate, and conversion rate of user interaction behaviors, and generate a behavior feature matrix containing advertising placement metrics based on semantic association weights;

[0115] Among them, the semantic association weight refers to calculating the semantic association weight between advertising behavior labels through a quantum state correlation matrix. The semantic association weight reflects the semantic association degree between different advertising behavior labels and helps to construct the advertising behavior label chain.

[0116] Organize the advertising placement metrics in the behavior feature matrix into a sliding window sequence in chronological order, and at the same time calculate the time interval features between the advertising placement metrics to form a behavior sequence containing short-term and long-term dependencies;

[0117] Specifically, divide the time period into continuous and overlapping windows, each window contains a set of advertising placement metrics arranged in chronological order, and then calculate the time interval features between the advertising placement metrics within these windows to ensure that the data can reflect short-term and long-term behavior dependencies. Through this method, not only local behavior sequences containing short-term dependencies are obtained, but also long-term dependency features across time periods are captured to form a comprehensive behavior sequence for further analysis and prediction.

[0118] Input the behavior sequence into a behavior sequence analysis model based on LSTM, combine the historical behavior feature matrix and the current behavior feature matrix, and predict the next advertising behavior label of the user and its conversion probability;

[0119] Screen the advertising behavior labels according to the conversion probability and insert them into the current short-term behavior sequence, and at the same time record the conversion probability, and output a dynamically updated user conversion path.

[0120] Among them, the user conversion path refers to the user behavior path generated based on the predicted user advertising behavior labels and their conversion probabilities. The user conversion path reflects the behavior trajectory of the user from seeing the advertisement to completing the conversion.

[0121] S4. Based on the dynamically updated user conversion path, call a generative adversarial network to generate personalized advertising content according to the user's historical interaction behavior data, record the version and distribution of the personalized advertising content through the advertising behavior label chain, and push the personalized advertising content to the user.

[0122] Furthermore, based on the user conversion path, extract the behavior feature matrix and advertising context, and combine them to generate a joint feature matrix;

[0123] Specifically, extract the user behavior feature matrix, including the interaction behavior data between the user and the advertisement. Then, obtain the advertisement context information, such as the attributes of the advertisement content, display location, and target audience, etc. Then, combine these user behavior features with the advertisement context information, and generate a joint feature matrix through weighted processing. This joint feature matrix contains multi-dimensional features of the user's historical behavior and advertisement content, providing a data basis for further personalized advertisement generation and delivery. In this way, it is possible to effectively integrate user behavior and advertisement content information, and implement a more accurate and personalized advertisement delivery strategy.

[0124] Based on the semantic association weights between advertisement behavior tags, perform weighted processing on the joint feature matrix to generate a high-dimensional feature vector;

[0125] Input the high-dimensional feature vector into a bidirectional generative adversarial network. The generator generates personalized advertisement content (including advertisement text, images, and videos), and the discriminator verifies the authenticity of the personalized advertisement content and its matching degree with the interaction behavior data;

[0126] Through behavior path analysis (combining elements such as user behavior sequences and conversion probabilities), verify whether the personalized advertisement content matches the user's conversion path, and optimize the relevance and logic of the content.

[0127] Specifically, extract the advertisement interaction data of each user at different time points from the behavior feature matrix. After sorting by timestamp, use a fixed-length window to slide step by step to create a sequence, ensuring that the data within each window represents a continuous time period.

[0128] Among them, during the construction of each sliding window, calculate the exact time difference between adjacent advertisement delivery metrics, and save it as a time interval feature together with the original data within the window. This not only retains the user's activity pattern within each time period but also captures the temporal relationship between activities. By adjusting the size and step length of the sliding window, it is possible to effectively identify short-term change trends (such as recent interactions) and long-term behavior patterns (such as trends over weeks or months) in user behavior, thus forming a comprehensive behavior sequence that can reflect both immediate responses and historical preferences.

[0129] Furthermore, extract semantic features described by the eigenvalues of the quantum state density matrix based on the generated personalized advertisement content, generate a unique version identifier using an encrypted hash algorithm, and store it in association with the advertisement behavior tag chain;

[0130] Generate a distribution record for each advertisement content, store the distribution record as a block in the blockchain, and generate an advertisement distribution record chain;

[0131] Specifically, capture the detailed information of the advertisement, including but not limited to the advertisement ID, target audience characteristics, distribution timestamp, and expected coverage, and associate this metadata with the identifier of the advertisement content itself; then, create a data structure containing all the above information as the distribution record of the advertisement content, and at the same time calculate the cryptographic hash value of this record to ensure its integrity and immutability, and then encapsulate this distribution record together with its hash value into a block format, ready to be stored in the blockchain; finally, after verifying that the new block meets the requirements of the blockchain protocol, link it to the end of the existing blockchain, so that the newly added distribution record becomes a part of the advertisement distribution record chain, thus completing the process from generating the distribution record to forming a continuous and immutable advertisement distribution record chain.

[0132] Combine the user's advertisement behavior tag chain and the advertisement distribution record chain to dynamically match the advertisement content version with the highest conversion probability.

[0133] Specifically, by analyzing the user's historical advertisement behavior tag chain, extract specific interaction operation data, device environment data, interest preference data, behavior sequence data, personalized advertisement data, and target conversion data; then, compare these specific data characteristics with the advertisement content versions and their distribution effects in the existing advertisement distribution record chain, and evaluate the conversion probabilities of different versions of advertisements for users with similar interaction patterns; then, use machine learning algorithms to calculate the potential response degrees of the current user to each advertisement version in real time, and predict which version is most likely to prompt the user to complete the expected conversion behavior; finally, select the advertisement content version with the highest conversion probability to ensure that it is the option that best matches the user's current behavior path and personal preferences, thus achieving precise push.

[0134] Push the selected advertisement content version to the user through the content distribution interface.

[0135] S5. After the user receives the personalized advertisement content, record the personalized interaction behavior data of the user with the personalized advertisement content in real time and store it in the advertisement behavior tag chain, and the advertisement behavior tag chain updates the advertisement placement metrics according to the real-time recorded personalized interaction behavior data.

[0136] Monitor click, view, and conversion data through the SDK (Software Development Kit) integrated with the local application, and collect personalized interaction behavior data in timestamp format.

[0137] Among them, the SDK is used to monitor the interaction between the user and the advertisement, such as clicks, views, etc., and record conversion data.

[0138] Preferably, the SDK integrated through the local application can accurately capture every interaction of the user with the advertisement. This includes not only direct operations (such as clicks), but also indirect responses (such as dwell time). Marking these data with timestamps can ensure the accuracy and reliability of subsequent analysis.

[0139] Perform encrypted hashing on the personalized interaction behavior data, update the advertisement behavior tags, and semantically associate them with the historical advertisement behavior tag chain and store them in the blockchain;

[0140] It should be noted that performing encrypted hashing on the collected data, on the one hand, ensures the security of user data, and on the other hand, simplifies the data structure for efficient storage. The updated advertisement behavior tags reflect the latest changes in user interests. When they are semantically associated with the historical tag chain, a continuous and evolving user behavior trajectory is formed. This process enhances the memory ability of the system and helps to understand user needs more deeply.

[0141] Extract click, view, and conversion data from the updated advertisement behavior tag chain, and fuse them with the behavior feature matrix to dynamically correct the real-time values of the advertisement placement metrics.

[0142] It should be noted that by fusing the latest user interaction data (such as clicks, views, and conversions) with the historical behavior feature matrix, subtle changes in user behavior patterns can be captured. This real-time data integration ensures that the advertisement placement strategy is always based on the latest and most accurate information, thereby improving the relevance and effectiveness of the advertisement content.

[0143] Preferably, by real-time fusing the latest user interaction data with the historical behavior feature matrix, this process can capture subtle changes in user behavior patterns, which can instantaneously respond to changes in market trends and shifts in consumer preferences, ensuring that the advertisement reaches the target audience at the most appropriate time, thereby maximizing the return on investment of the advertisement.

[0144] This embodiment also provides an intelligent advertising analysis and management system based on blockchain data, including: a data collection module, a storage module, a behavior prediction module, a push module, and an interaction record module; the data collection module is used to collect interaction behavior data generated when users interact with advertising content; the storage module is used to convert the interaction behavior data into advertising behavior tags in real time, and perform chained association with the previous and subsequent advertising behavior tags through an encrypted hash algorithm and store them in the blockchain; the behavior prediction module is used to perform data preprocessing on the advertising behavior tag chain, extract advertising placement metrics, and use a behavior sequence analysis model to perform real-time behavior prediction on the advertising placement metrics, and output a dynamically updated user conversion path; the push module is used to, based on the dynamically updated user conversion path, call a generative adversarial network, generate personalized advertising content according to the user's historical interaction behavior data, record the version and distribution status of the personalized advertising content through the advertising behavior tag chain, and push the personalized advertising content to the user; the interaction record module is used to, after the user receives the personalized advertising content, record the personalized interaction behavior data of the user and the personalized advertising content in real time and store it in the advertising behavior tag chain, and the advertising behavior tag chain updates the advertising placement metrics according to the real-time recorded personalized interaction behavior data.

[0145] This embodiment also provides a computer device applicable to the case of an intelligent advertising analysis and management method based on blockchain data, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the intelligent advertising analysis and management method based on blockchain data as proposed in the above embodiment.

[0146] This computer device can be a terminal, and this computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. Among them, the processor of this computer device is used to provide computing and control capabilities. The memory of this computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of this computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a carrier network, NFC (Near Field Communication), or other technologies. The display screen of this computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of this computer device can be a touch layer covering the display screen, or a button, a trackball, or a touchpad provided on the outer shell of the computer device, or an external keyboard, touchpad, or mouse, etc.

[0147] This embodiment also provides a storage medium, on which a computer program is stored. When the program is executed by a processor, it implements the intelligent advertising analysis and management method based on blockchain data as proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, abbreviated as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, abbreviated as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, abbreviated as EPROM), programmable read-only memory (Programmable Red-Only Memory, abbreviated as PROM), read-only memory (Read-Only Memory, abbreviated as ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0148] In summary, by introducing blockchain technology, the advertising data in the present invention has the characteristics of immutability and high security during the transmission and storage processes, greatly improving the data security and privacy protection. Through the real-time conversion and chained storage of user interaction behavior data, a dynamically updated advertising behavior tag chain is generated, enabling more accurate capture of users' interest preferences. By adopting the generative adversarial network technology, personalized advertising content is generated based on users' historical behavior data, which not only improves the relevance and attractiveness of the advertisements, but also can record the personalized interaction behavior data in real time after the users receive the advertising content, further optimizing the advertising placement metrics and user conversion path. In conclusion, the present invention has significant beneficial effects in terms of data security, personalized advertising generation, and real-time behavior prediction, and can effectively improve the accuracy and effectiveness of advertising placement.

[0149] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

Claims

1. An intelligent advertising analysis and management method based on blockchain data, characterized by: include, Collect interactive behavior data generated when users interact with advertising content; The interactive behavior data is converted into advertising behavior tags in real time, and linked with the previous and next advertising behavior tags through the encrypted hash algorithm and stored in the blockchain; Preprocess the data of the advertising behavior tag chain, extract advertising delivery indicators, use the behavior sequence analysis model to make real-time behavior predictions on advertising delivery indicators, and output dynamically updated user conversion paths; Based on the dynamically updated user conversion path, the generative adversarial network is called to generate personalized advertising content according to the user's historical interactive behavior data. The version and distribution of personalized advertising content are recorded through the advertising behavior label chain, and personalized advertising content is pushed to users. After the user receives the personalized advertising content, the personalized interaction behavior data between the user and the personalized advertising content will be recorded in real time and stored in the advertising behavior tag chain. The advertising behavior tag chain updates the advertising delivery indicators based on the personalized interaction behavior data recorded in real time.

2. The intelligent advertising analysis and management method based on blockchain data according to claim 1, characterized in that: The interactive behavior data includes interactive operation data, device environment data, interest preference data, behavior sequence data, personalized advertising data and target conversion data.

3. The intelligent advertising analysis and management method based on blockchain data as claimed in claim 2, characterized in that: The interactive behavior data is converted into advertising behavior tags in real time, and linked with the previous and next advertising behavior tags through an encrypted hash algorithm and stored in the blockchain. The specific steps are as follows: The interaction behavior data is normalized using a standardization function; The normalized interaction behavior data is mapped to a high-dimensional semantic space through a nonlinear embedding function to generate the corresponding quantum state probability distribution; Use density matrix to describe the eigenvalues ​​of quantum state probability distribution and use it as semantic features to identify semantic patterns in interactive behavior data. Perform cluster analysis and temporal correlation analysis on the semantic patterns of the logos to generate advertising behavior labels corresponding to the interactive behavior data; The quantum state association matrix is ​​used to calculate the semantic association weights between advertising behavior tags. The expression is: Among them, A ij represents the semantic association weight between advertising behavior labels i and j, where i and j represent two different advertising behavior labels. represents the complex conjugate quantum state function of advertising behavior label i at time t, ψ j (t) represents the quantum state function of the advertising behavior tag j at time t, is a Gaussian time weight function, t0 represents the reference time point of the current behavior event, α represents the time weight factor, and dt represents the small time increment in the integration operation; A two-dimensional matrix consisting of the semantic association weights of all advertising behavior label pairs (i, j) is formed to form a semantic association matrix; Based on the semantic association matrix, construct a set of all possible advertising behavior label pairs. A chain structure is constructed based on the maximum weight path of the semantic association matrix to generate the optimal advertising behavior label chain. The expression is: in, represents the optimal advertising behavior label chain, Represents the path set from all possible advertising behavior labels Find the maximum weight path in It means summing up the semantic association weights of all advertising behavior label pairs (i, j) in the path set; In the optimal advertising behavior tag chain Based on this, the encrypted hash value of the optimal advertising behavior label chain is generated, and the expression is: Among them, H k represents the encrypted hash value of the kth optimal advertising behavior label chain, k represents the index variable of the optimal advertising behavior label chain, SHA3 is the secure hash algorithm, and ρ k (t) represents the quantum state density function of the kth advertising behavior tag chain at time t, represents the Gaussian attenuation factor, β represents the time weight parameter; When a new advertising behavior label is added to the advertising behavior label chain, the semantic association weights of the new advertising behavior label and all advertising behavior labels in the optimal advertising behavior label chain are recalculated, the optimal path is updated, and the encrypted hash value of the new optimal advertising behavior label chain is calculated; The semantic association weight of the newly added advertising behavior label and the encrypted hash value of the updated optimal advertising behavior label chain are stored in the blockchain as the newly added label information, and each advertising behavior label chain is regarded as a block; Establish a semantic index for each advertising behavior label chain, and use quantum Bloom filter to optimize the query efficiency of advertising behavior label chain. The expression is: Where Q(x) represents the result of the quantum Bloom filter, m represents the number of hash functions in the Bloom filter, z represents the index of the hash function in the Bloom filter, and b z represents the Boolean value of the zth hash function, γ z represents the weight parameter of the zth dimension, x represents the input query value, represents a sign conversion term, represents the weight decay function; When the advertising behavior tag chain is updated, the semantic index of the corresponding blockchain and the advertising behavior tag chain is updated synchronously.

4. The intelligent advertising analysis and management method based on blockchain data as claimed in claim 3, characterized in that: The data preprocessing of the advertising behavior tag chain, extracting advertising delivery indicators, using the behavior sequence analysis model to perform real-time behavior prediction on the advertising delivery indicators, and outputting dynamically updated user conversion paths are specifically performed as follows: Clean and standardize the advertising behavior tag chain, extract the time series, click-through rate and conversion rate of user interaction behavior, and generate a behavioral feature matrix containing advertising delivery indicators based on semantic association weights; The advertising delivery indicators in the behavior feature matrix are organized into a sliding window sequence in chronological order, and the time interval characteristics between the advertising delivery indicators are calculated to form a behavior sequence containing short-term and long-term dependencies; The behavior sequence is input into the LSTM-based behavior sequence analysis model, and the historical behavior feature matrix and the current behavior feature matrix are combined to predict the user's next advertising behavior label and its conversion probability; Filter advertising behavior tags according to conversion probability and insert them into the current short-term behavior sequence. Meanwhile, record the conversion probability and output dynamically updated user conversion path.

5. The intelligent advertising analysis and management method based on blockchain data as claimed in claim 4, characterized in that: Based on the dynamically updated user conversion path, the generative adversarial network is called to generate personalized advertising content according to the user's historical interactive behavior data. The specific steps are as follows: Based on the user conversion path, the behavior feature matrix and the advertising context are extracted and combined to generate a joint feature matrix; Based on the semantic association weights between advertising behavior tags, the joint feature matrix is ​​weighted to generate a high-dimensional feature vector; Input the high-dimensional feature vector into the bidirectional generative adversarial network, the generator generates personalized advertising content, and the discriminator verifies the authenticity of the personalized advertising content and its matching degree with the interactive behavior data; Through behavioral path analysis, verify whether personalized advertising content matches the user conversion path and optimize the relevance and logic of the content.

6. The intelligent advertising analysis and management method based on blockchain data as claimed in claim 5, characterized in that: The specific steps of recording the version and distribution of personalized advertising content through the advertising behavior tag chain and pushing personalized advertising content to users are as follows: Based on the generated personalized advertising content, the semantic features described by the eigenvalues ​​of the quantum state density matrix are extracted, and a unique version identifier is generated using an encrypted hash algorithm, which is then associated and stored with the advertising behavior tag chain; Generate a distribution record for each advertising content, store the distribution record as a block in the blockchain, and generate an advertising distribution record chain; Combine the user's advertising behavior tag chain and advertising distribution record chain to dynamically match the advertising content version with the highest conversion probability; The selected advertising content version is pushed to the user through the content distribution interface.

7. The intelligent advertising analysis and management method based on blockchain data according to claim 6, characterized in that: After the user receives the personalized advertising content, the personalized interaction behavior data between the user and the personalized advertising content will be recorded in real time and stored in the advertising behavior tag chain. The advertising behavior tag chain updates the advertising delivery index according to the personalized interaction behavior data recorded in real time. The specific steps are as follows: Monitor click, view and conversion data through native application integrated SDK, and collect personalized interaction behavior data in timestamp format; Perform cryptographic hashing on the personalized interaction behavior data, update the advertising behavior tag, and semantically associate it with the historical advertising behavior tag chain and store it in the blockchain; Click, view and conversion data are extracted from the updated advertising behavior tag chain and integrated with the behavior feature matrix to dynamically correct the real-time values ​​of advertising delivery indicators.

8. An intelligent advertising analysis and management system based on blockchain data, based on the intelligent advertising analysis and management method based on blockchain data according to any one of claims 1 to 7, characterized in that: Including data collection module, storage module, behavior prediction module, push module and interaction recording module; The data collection module is used to collect interaction behavior data generated when users interact with advertising content; The storage module is used to convert the interactive behavior data into advertising behavior tags in real time, and link them with the previous and next advertising behavior tags through an encrypted hash algorithm and store them in the blockchain; The behavior prediction module is used to perform data preprocessing on the advertising behavior tag chain, extract advertising delivery indicators, use the behavior sequence analysis model to perform real-time behavior prediction on the advertising delivery indicators, and output a dynamically updated user conversion path; The push module is used to call the generative adversarial network based on the dynamically updated user conversion path, generate personalized advertising content according to the user's historical interactive behavior data, record the version and distribution of the personalized advertising content through the advertising behavior tag chain, and push the personalized advertising content to the user; The interaction recording module is used to record the personalized interaction behavior data between the user and the personalized advertising content in real time after the user receives the personalized advertising content and store it in the advertising behavior tag chain. The advertising behavior tag chain updates the advertising delivery index according to the personalized interaction behavior data recorded in real time.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the smart advertising analysis and management method based on blockchain data are implemented in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the smart advertising analysis and management method based on blockchain data according to any one of claims 1 to 7 are implemented.

Images