CN114881712B

CN114881712B - Intelligent advertisement putting method, device, equipment and storage medium

Info

Publication number: CN114881712B
Application number: CN202210812199.2A
Authority: CN
Inventors: 易星
Original assignee: Hangzhou Xinhang Interactive Technology Co ltd
Current assignee: Hangzhou Xinhang Interactive Technology Co ltd
Priority date: 2022-07-12
Filing date: 2022-07-12
Publication date: 2022-09-23
Anticipated expiration: 2042-07-12
Also published as: CN114881712A

Abstract

The invention relates to the field of artificial intelligence, and discloses an intelligent advertisement delivery method, an intelligent advertisement delivery device, intelligent advertisement delivery equipment and a storage medium, which are used for improving the accuracy of intelligent advertisement delivery. The method comprises the following steps: if the target user accords with the preset user type, matching a questionnaire template according to the user data and performing path segmentation analysis on the questionnaire template to obtain a front-stage problem path and a rear-stage problem path; generating a front section end node and a front section answer path corresponding to the questionnaire template based on the front section question path, and carrying out real-time analysis on the user real attribute of the target user to obtain a real-time analysis result; carrying out path updating on the rear-section question path according to the real-time analysis result to generate a target end node and a target answer path; inputting the target end node and the target answer path into a questionnaire data processing model to analyze the real requirements of the user, and generating the real requirements of the user; and carrying out intelligent advertisement putting on the target user according to the real requirement of the user.

Description

Intelligent advertisement putting method, device, equipment and storage medium

Technical Field

The invention relates to the field of artificial intelligence, in particular to an intelligent advertisement delivery method, an intelligent advertisement delivery device, intelligent advertisement delivery equipment and a storage medium.

Background

With the rapid development of computer technology, internet media has become an important channel for enterprises to place advertisements due to the characteristics of wide audience area, high propagation efficiency and the like. In order to maximize the advertising revenue of the media, the advertising resources of the advertising provider are put into the media, and the advertising position of the media is realized.

The existing scheme can not realize accurate advertisement putting for users and multidimensional and fine management, so that the accuracy of an intelligent advertisement putting strategy is greatly reduced, the advertisement profit maximization of media can not be realized, the accurate putting requirement of advertisements provided by advertisements can not be met, and the accuracy of the existing scheme is low.

Disclosure of Invention

The invention provides an intelligent advertisement delivery method, an intelligent advertisement delivery device, intelligent advertisement delivery equipment and a storage medium, which are used for improving the accuracy of intelligent advertisement delivery.

The first aspect of the present invention provides a smart advertisement delivery method, including: receiving an intelligent advertisement putting analysis request sent by a client, and acquiring user data of a target user from a preset cloud database according to the intelligent advertisement putting analysis request, wherein the user data comprises user gender and age, user-defined interest types, user online time and single browsing time; performing tagging processing on the user data to obtain a plurality of user tags corresponding to the user data, calculating the similarity between the user tags and a preset user type template to obtain target similarity, and judging whether the target user accords with a preset user type according to the target similarity, wherein the user type comprises: the user is not clear of the preference of the user, the user is not clear of the requirement of the user, and the user actively selects intelligent advertisement delivery; if the target user accords with the preset user type, matching a questionnaire template corresponding to the target user according to the user data, and performing path segmentation analysis on the questionnaire template to obtain a front-stage problem path and a rear-stage problem path; performing questionnaire filling on the questionnaire template based on the front-section question path, generating a front-section end node and a front-section answer path corresponding to the questionnaire template, and performing real-time analysis on the user real attribute of the target user according to the front-section end node and the front-section answer path to obtain a real-time analysis result; carrying out path updating on the rear-section question path according to the real-time analysis result and a preset path updating strategy to generate an updated question path, and carrying out questionnaire filling on the questionnaire template according to the updated question path to obtain a target end node and a target answer path; vector mapping is carried out on the target end node and the target answer path to obtain an initial input vector, the initial input vector is input into a preset questionnaire data processing model to carry out user real demand analysis, and a user real demand corresponding to the target user is generated, wherein the questionnaire data processing model comprises: a feature extraction network, a bidirectional long-time memory network and a prediction network; and matching a target advertisement corresponding to the real demand of the user from a preset advertisement database according to the real demand of the user, and delivering the target advertisement to the client so as to perform intelligent advertisement delivery on the target user.

Optionally, in a first implementation manner of the first aspect of the present invention, the tagging is performed on the user data to obtain a plurality of user tags corresponding to the user data, a similarity between each user tag and a preset user type template is calculated to obtain a target similarity, and whether the target user conforms to a preset user type is determined according to the target similarity, where the user type includes: the method comprises the following steps that a user is not clear of own preference, the user is not clear of own requirement, and the user actively selects intelligent advertisement putting, wherein the steps comprise: labeling the gender and age of the user, the user-defined interest types, the user online time period and the single browsing time length in the user data respectively to obtain a plurality of user labels corresponding to the user data; acquiring a template label corresponding to a preset user type template; calculating the tag hit rate between the user tags and the template tags, and taking the tag hit rate as the similarity between the user tags and a preset user type template to obtain target similarity; judging whether the target user accords with a preset user type according to the target similarity, wherein the user type comprises the following steps: the user is not aware of his preferences, the user is not aware of his needs, and the user actively selects smart advertising.

Optionally, in a second implementation manner of the first aspect of the present invention, if the target user conforms to the preset user type, matching a questionnaire template corresponding to the target user according to the user data, and performing path segmentation analysis on the questionnaire template to obtain a front-stage question path and a rear-stage question path, including: if the target user accords with the preset user type, extracting key information of the user data to obtain key information corresponding to the user data, wherein the key information comprises user gender and age and user-defined interest categories; matching the key information with a plurality of preset questionnaire templates to obtain the information matching degree corresponding to each questionnaire template; sorting the information matching degrees corresponding to the questionnaire templates to obtain target sorting results, and taking the questionnaire template corresponding to the information matching degree with the highest rank in the target sorting results as the questionnaire template corresponding to the target user; and calling a preset template analysis model to perform path segmentation analysis on the questionnaire template to obtain a front-stage problem path and a rear-stage problem path corresponding to the questionnaire template.

Optionally, in a third implementation manner of the first aspect of the present invention, the filling out a questionnaire on the questionnaire template based on the previous-segment question path to generate a previous-segment end node and a previous-segment answer path corresponding to the questionnaire template, and performing real-time analysis on the user real attribute of the target user according to the previous-segment end node and the previous-segment answer path to obtain a real-time analysis result, includes: filling the questionnaire in the questionnaire template based on the front-section question path to obtain a filled questionnaire; analyzing questionnaire information of the filled questionnaire to obtain a front section end node and a front section answer path corresponding to the questionnaire template; analyzing the content of the front section end node and the front section answer path in real time to obtain the content of a front section questionnaire; and calling a preset attribute database to perform real-time analysis on the real attributes of the front-section questionnaire contents to obtain a real-time analysis result.

Optionally, in a fourth implementation manner of the first aspect of the present invention, the vector mapping is performed on the target end node and the target answer path to obtain an initial input vector, and the initial input vector is input into a preset questionnaire data processing model to perform user real demand analysis, so as to generate a user real demand corresponding to the target user, where the questionnaire data processing model includes: the characteristic extraction network, the bidirectional long-time and short-time memory network and the prediction network comprise: carrying out vector mapping on the target end node and the target answer path to obtain an initial input vector; inputting the initial input vector into a preset questionnaire data processing model, wherein the questionnaire data processing model comprises: a feature extraction network, a bidirectional long-time memory network and a prediction network; extracting the characteristics of the initial input vector through the characteristic extraction network to obtain a characteristic vector; providing the bidirectional long-time and short-time memory network to perform feature normalization processing on the feature vector to obtain a feature normalization vector; performing feature prediction on the feature normalized vector through the prediction network to obtain a prediction node; and analyzing the prediction node and the end node to obtain an analysis result, and generating a user real demand corresponding to the target user according to the analysis result.

Optionally, in a fifth implementation manner of the first aspect of the present invention, the matching, according to the real user demand, a target advertisement corresponding to the real user demand from a preset advertisement database, and delivering the target advertisement to the client to perform intelligent advertisement delivery for the target user includes: acquiring a plurality of candidate advertisements in a preset advertisement database; performing credibility calculation on the candidate advertisements based on the real user requirements to obtain target credibility corresponding to each candidate advertisement; ranking the candidate advertisements according to the target credibility corresponding to each candidate advertisement, and taking the candidate advertisement with the highest ranking as the advertisement to be delivered corresponding to the target user to obtain the target advertisement; and carrying out advertisement putting display on the target advertisement through the display page of the client so as to carry out intelligent advertisement putting on the target user.

Optionally, in a sixth implementation manner of the first aspect of the present invention, the smart advertisement delivery method further includes: when the target user receives the target advertisement, obtaining feedback information of the target user to the target advertisement in real time, wherein the feedback information comprises: advertisement evaluation scores and advertisement click operation data; analyzing the operation data of the advertisement clicking operation data to obtain a clicking operation analysis result; comprehensively analyzing the feedback information according to the advertisement evaluation score and the click operation analysis result to obtain a target feedback result; and carrying out intelligent advertisement putting on the advertisement selected from the plurality of candidate advertisements for the next time based on the target feedback result.

The invention provides an intelligent advertisement delivery device, which comprises an acquisition module, a display module and a display module, wherein the acquisition module is used for receiving an intelligent advertisement delivery analysis request sent by a client and acquiring user data of a target user from a preset cloud database according to the intelligent advertisement delivery analysis request, and the user data comprises user gender and age, user-defined interest types, user online time and single browsing time; a calculating module, configured to perform tagging processing on the user data to obtain a plurality of user tags corresponding to the user data, calculate a similarity between the user tags and a preset user type template to obtain a target similarity, and determine whether the target user conforms to a preset user type according to the target similarity, where the user type includes: the user is not clear of the preference of the user, the user is not clear of the requirement of the user, and the user actively selects intelligent advertisement delivery; the matching module is used for matching a questionnaire template corresponding to the target user according to the user data and performing path segmentation analysis on the questionnaire template to obtain a front-stage problem path and a rear-stage problem path if the target user conforms to the preset user type; a filling module, configured to fill in a questionnaire on the questionnaire template based on the front-segment question path, generate a front-segment end node and a front-segment answer path corresponding to the questionnaire template, and perform real-time user attribute analysis on the target user according to the front-segment end node and the front-segment answer path, to obtain a real-time analysis result; the updating module is used for updating the path of the rear-section question path according to the real-time analysis result and a preset path updating strategy to generate an updated question path, and filling the questionnaire template according to the updated question path to obtain a target end node and a target answer path; an analysis module, configured to perform vector mapping on the target end node and the target answer path to obtain an initial input vector, input the initial input vector into a preset questionnaire data processing model to perform user real demand analysis, and generate a user real demand corresponding to the target user, where the questionnaire data processing model includes: a feature extraction network, a bidirectional long-time memory network and a prediction network; and the delivery module is used for matching the target advertisement corresponding to the real user demand from a preset advertisement database according to the real user demand and delivering the target advertisement to the client so as to carry out intelligent advertisement delivery on the target user.

Optionally, in a first implementation manner of the second aspect of the present invention, the calculation module is specifically configured to: labeling the gender and age of the user, the user-defined interest types, the user online time period and the single browsing time length in the user data respectively to obtain a plurality of user labels corresponding to the user data; acquiring a template label corresponding to a preset user type template; calculating the tag hit rate between the user tags and the template tags, and taking the tag hit rate as the similarity between the user tags and a preset user type template to obtain target similarity; judging whether the target user accords with a preset user type according to the target similarity, wherein the user type comprises the following steps: the user is not aware of his or her preferences, the user is not aware of his or her needs, and the user actively selects smart advertising.

Optionally, in a second implementation manner of the second aspect of the present invention, the matching module is specifically configured to: if the target user accords with the preset user type, extracting key information of the user data to obtain key information corresponding to the user data, wherein the key information comprises user gender and age and user-defined interest categories; matching the key information with a plurality of preset questionnaire templates to obtain the information matching degree corresponding to each questionnaire template; sorting the information matching degrees corresponding to the questionnaire templates to obtain target sorting results, and taking the questionnaire template corresponding to the information matching degree with the highest rank in the target sorting results as the questionnaire template corresponding to the target user; and calling a preset template analysis model to perform path segmentation analysis on the questionnaire template to obtain a front-stage problem path and a rear-stage problem path corresponding to the questionnaire template.

Optionally, in a third implementation manner of the second aspect of the present invention, the filling module further includes: filling the questionnaire in the questionnaire template based on the front-section question path to obtain a filled questionnaire; analyzing questionnaire information of the filled questionnaire to obtain a front section end node and a front section answer path corresponding to the questionnaire template; performing real-time content analysis on the front section end node and the front section answer path to obtain front section questionnaire content; and calling a preset attribute database to perform real-time analysis on the real attributes of the front-section questionnaire contents to obtain a real-time analysis result.

Optionally, in a fourth implementation manner of the second aspect of the present invention, the analysis module is specifically configured to: carrying out vector mapping on the target end node and the target answer path to obtain an initial input vector; inputting the initial input vector into a preset questionnaire data processing model, wherein the questionnaire data processing model comprises: a feature extraction network, a bidirectional long-time memory network and a prediction network; extracting the characteristics of the initial input vector through the characteristic extraction network to obtain a characteristic vector; providing the bidirectional long-short time memory network to perform feature normalization processing on the feature vector to obtain a feature normalization vector; performing feature prediction on the feature normalized vector through the prediction network to obtain a prediction node; and analyzing the prediction node and the end node to obtain an analysis result, and generating a user real demand corresponding to the target user according to the analysis result.

Optionally, in a fifth implementation manner of the second aspect of the present invention, the releasing module is specifically configured to: acquiring a plurality of candidate advertisements in a preset advertisement database; calculating the credibility of the candidate advertisements based on the real demand of the user to obtain the target credibility corresponding to each candidate advertisement; ranking the candidate advertisements according to the target credibility corresponding to each candidate advertisement, and taking the candidate advertisement with the highest ranking as the advertisement to be delivered corresponding to the target user to obtain the target advertisement; and carrying out advertisement putting display on the target advertisement through the display page of the client so as to carry out intelligent advertisement putting on the target user.

Optionally, in a sixth implementation manner of the second aspect of the present invention, the smart advertisement delivery apparatus further includes: an optimization module, configured to obtain feedback information of the target user on the target advertisement in real time when the target user receives the target advertisement, where the feedback information includes: advertisement evaluation scores and advertisement click operation data; analyzing the operation data of the advertisement clicking operation data to obtain a clicking operation analysis result; comprehensively analyzing the feedback information according to the advertisement evaluation score and the click operation analysis result to obtain a target feedback result; and performing intelligent advertisement delivery on advertisements which are selected from the candidate advertisements for the next time based on the target feedback result.

A third aspect of the present invention provides an intelligent advertisement delivery apparatus, including: a memory and at least one processor, the memory having instructions stored therein; the at least one processor invokes the instructions in the memory to cause the smart advertisement delivery device to perform the smart advertisement delivery method described above.

A fourth aspect of the present invention provides a computer-readable storage medium having stored therein instructions, which, when run on a computer, cause the computer to execute the above-mentioned smart advertisement delivery method.

In the technical scheme provided by the invention, user data is analyzed in advance, type division is carried out on target users, a questionnaire template which is most consistent with the target users is matched for the target users to fill, the matching accuracy of the questionnaire template is improved, the contents of the questionnaire are more consistent with the actual requirements of the users, in addition, the questionnaire is divided into a front-section questionnaire and a rear-section questionnaire, the front-section questionnaire is used for carrying out attribute analysis on the users, the results of the attribute analysis are used for carrying out real-time adjustment on the rear-section questionnaire, the questionnaire is more consistent with the target users by carrying out real-time adjustment on the questionnaire, the specific questionnaire data of each target user is constructed, the real requirements of the users fed back by the users in real time are analyzed, the real requirements of the users obtained by analysis are more consistent with the actual requirements of the users, data analysis is carried out aiming at the paths answered by the users in the questionnaire to inquire the target advertisements which are most consistent with the users, therefore, the accuracy of intelligent advertisement putting is improved.

Drawings

FIG. 1 is a schematic diagram of an embodiment of a smart advertisement delivery method according to an embodiment of the present invention;

FIG. 2 is a diagram of another embodiment of a smart advertisement delivery method according to an embodiment of the present invention;

FIG. 3 is a schematic diagram of an embodiment of a smart advertisement delivery device according to an embodiment of the present invention;

FIG. 4 is a schematic diagram of another embodiment of a smart advertisement delivery device according to an embodiment of the present invention;

fig. 5 is a schematic diagram of an embodiment of a smart advertisement delivery device in the embodiment of the present invention.

Detailed Description

The embodiment of the invention provides an intelligent advertisement putting method, an intelligent advertisement putting device, intelligent advertisement putting equipment and a storage medium, which are used for improving the accuracy of intelligent advertisement putting. The terms "first," "second," "third," "fourth," and the like in the description and in the claims, as well as in the drawings, if any, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It will be appreciated that the data so used may be interchanged under appropriate circumstances such that the embodiments described herein may be practiced otherwise than as specifically illustrated or described herein. Moreover, the terms "comprises," "comprising," or "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

For convenience of understanding, a detailed flow of an embodiment of the present invention is described below, and referring to fig. 1, an embodiment of a smart advertisement delivery method in an embodiment of the present invention includes:

101. receiving an intelligent advertisement putting analysis request sent by a client, and acquiring user data of a target user from a preset cloud database according to the intelligent advertisement putting analysis request, wherein the user data comprises user gender and age, user-defined interest types, user online time and single browsing duration;

it is to be understood that the execution subject of the present invention may be an intelligent advertisement delivery device, and may also be a terminal or a server, which is not limited herein. The embodiment of the present invention is described by taking a server as an execution subject.

It should be noted that, an intelligent advertisement delivery analysis request sent by a client is received, the target user browses through an application program in the client, and an intelligent advertisement delivery analysis request is generated by clicking an application program boundary, the intelligent advertisement delivery analysis request is transmitted to a server through the client, the intelligent advertisement delivery analysis request is analyzed through the server, and user data of the target user is obtained from a preset cloud database, the user data input by the target user includes personal gender, age and interest data, user-defined interest types, user online time periods, single browsing time periods and the like.

102. Performing tagging processing on user data to obtain a plurality of user tags corresponding to the user data, calculating the similarity between the user tags and a preset user type template to obtain target similarity, and judging whether a target user accords with a preset user type according to the target similarity, wherein the user type comprises: the user is not clear of the preference of the user, the user is not clear of the requirement of the user, and the user actively selects intelligent advertisement delivery;

specifically, the user data is subjected to tagging processing to obtain a plurality of user tags corresponding to the user data, a certain number of other users having a relationship such as "similar attribute", "similar behavior", and "similar scene" with the target user are determined according to characteristics such as personal attribute, daily behavior, activity scene, and the like of the target user, and the determined users are gathered to form a user group about the target user, the users in the user group are all users with similar characteristics to the target user, the initial label is established to organize and mine characteristics of personal attributes, daily behaviors and activity scenes in user data of the user group from the aspect of object and object connection, projecting user characteristics from the perspective of object and object connections, forming object and object connections that are valid for a user group, thereby creating user tags based on objects and object contacts valid for the user group to which the target user belongs. And then calculating the similarity between the user label and a preset user type template to obtain a target similarity, and judging whether a target user accords with a preset user type according to the target similarity, wherein the user type comprises that the user is not clear of the preference of the user, the user has an undefined demand on the user and the user actively selects intelligent advertisement delivery, and the server identifies the preset user type according to user data input by the target user and judges whether the target user accords with the preset user type.

103. If the target user accords with the preset user type, matching a questionnaire template corresponding to the target user according to the user data, and performing path segmentation analysis on the questionnaire template to obtain a front-stage problem path and a rear-stage problem path;

it should be noted that, if the target user is in accordance with the preset user type, the target user enters questionnaire filling, the server first matches a questionnaire template according to the user data of the target user, the questionnaire template includes a user preference path, is similar to a tree diagram, includes a plurality of question options, and different answers may be directed to different branches. And the target user answers the questions in the questionnaire template after receiving the questionnaire template, the server fills template nodes and throws out new questions until the template nodes reach the end nodes, backtracks the questionnaire template according to the end nodes, and analyzes to obtain the question paths corresponding to the questionnaire template. In this embodiment, the questionnaire template is subjected to path segmentation analysis to obtain a front-stage question path and a rear-stage question path, where the front-stage question path is mainly a preset standard question and is used for analyzing real attributes of a user according to an answer of the user to the front-stage question path and generating the rear-stage question path more consistent with a target user in real time according to the real attributes of the user.

104. Performing questionnaire filling on a questionnaire template based on the front-section question path to generate a front-section end node and a front-section answer path corresponding to the questionnaire template, and performing real-time user attribute analysis on a target user according to the front-section end node and the front-section answer path to obtain a real-time analysis result;

and the target user fills in the problems in the questionnaire template, and the server fills in the template nodes and throws out new problems until the problem nodes or the end nodes of the front section end nodes are filled. And the server predicts the real attribute of a target user through a deep learning network model according to the final node and the intermediate user answer path, and then generates a real-time analysis result corresponding to the target user according to the predicted user requirement. In this embodiment, each question in the previous-stage answer path and the answer fed back by the target user are set with a corresponding feature factor, a feature value corresponding to each feature factor can be determined based on the previous-stage end node and the previous-stage answer path, a value obtained after discretization and normalization of the feature values can be used as the feature factor corresponding to the previous-stage answer path, discretization or normalization is performed, an array formed by the feature values corresponding to each feature factor is used as the target feature, and the real attribute of the user is analyzed in real time for the target user according to the target feature, so that a real-time analysis result is obtained.

105. Carrying out path updating on the rear-section question path according to the real-time analysis result and a preset path updating strategy to generate an updated question path, and carrying out questionnaire filling on a questionnaire template according to the updated question path to obtain a target end node and a target answer path;

specifically, the real-time analysis result is used for indicating the user real attribute of the target user, and the path update is performed on the subsequent problem path according to the real-time analysis result and the preset path update strategy to generate an updated problem path, where the preset path update strategy specifically includes: firstly, a path updating rule base is established, all realization paths related to the target user are listed through the path updating rule base, the path updating rule base is directly inquired in the process of updating the path, and the rule is exported to form an updating problem path. Listing possible paths for path updating according to methods such as a preset path search method or a rule base, namely all candidate problem paths, counting the problem correlation degree in each candidate problem path, selecting an optimal path from a plurality of candidate problem paths by combining data such as the user online time period and the single browsing time period of the target user, generating an updated problem path, and finally filling a questionnaire to a questionnaire template according to the updated problem path to obtain a target end node and a target answer path.

106. Vector mapping is carried out on a target end node and a target answer path to obtain an initial input vector, the initial input vector is input into a preset questionnaire data processing model to carry out user real demand analysis, and user real demands corresponding to a target user are generated, wherein the questionnaire data processing model comprises: a feature extraction network, a bidirectional long-time memory network and a prediction network;

specifically, vector mapping is performed on a target end node and a target answer path to obtain an initial input vector, specifically, vector mapping is performed through a word vector layer, the word vector layer comprises a plurality of coding units, the target end node and the target answer path are converted into vectors through the coding units in the word vector layer, each coding unit outputs a target characteristic value, a plurality of target characteristic values corresponding to the coding units are finally obtained, and then the target characteristic values are converted into the vectors to obtain the initial input vector. Vector conversion is carried out on the target end node and the target answer path to obtain an initial input vector, and the recognition rate of the neural network can be increased. The bidirectional long-short time memory network comprises a forward long-short time memory network and a reverse long-short time memory network, the forward long-short time memory network comprises a plurality of LSTM units, each LSTM unit outputs a sequence element, a forward hidden state sequence is output according to the sequence element, the reverse long-short time memory network comprises a plurality of LSTM units, each LSTM unit outputs a sequence element, and a reverse hidden state sequence is output according to the sequence element. It should be noted that the prediction network is configured to perform information tagging on the target feature sequence, where the information tagging may be performed according to different side hierarchies or levels, and in this embodiment, information classification is performed according to sentence levels, and each sentence is used as a minimum unit to tag the target feature sequence to obtain a target sentence level sequence, so as to generate a user real requirement corresponding to a target user according to the target sentence level sequence.

107. And matching the target advertisement corresponding to the real demand of the user from a preset advertisement database according to the real demand of the user, and delivering the target advertisement to the client so as to perform intelligent advertisement delivery on the target user.

Specifically, the server calculates the reliability according to the real requirements of the users to obtain the target reliability, the server sorts the target reliability, selects the maximum value of the candidate advertisements corresponding to the target reliability to obtain the maximum value of the candidate advertisements, and the server takes the candidate advertisements corresponding to the maximum value as the target advertisements and intelligently puts the target advertisements to the target users according to the target advertisements. Specifically, after the intelligent advertisement delivery is finished, the server presets a time period in the application program browsed by the user, and sends questionnaire information to the target user again to obtain feedback information of the target user on the target advertisement; the server analyzes the feedback information to obtain a feedback result, and the feedback result reflects the satisfaction degree of the target user on intelligent advertisement putting of the application program; and the server optimizes the intelligent advertisement putting process based on the feedback result.

Optionally, when the target user receives the target advertisement, the server obtains feedback information of the target user on the target advertisement in real time, where the feedback information includes: advertisement evaluation scores and advertisement click operation data; performing operation data analysis on the advertisement click operation data to obtain a click operation analysis result; comprehensively analyzing the feedback information according to the advertisement evaluation score and the click operation analysis result to obtain a target feedback result; and selecting the next advertisement from the candidate advertisements to carry out intelligent advertisement putting based on the target feedback result.

In the embodiment of the invention, the user data is analyzed in advance, the types of the target users are divided, the questionnaire template which is most matched with the target users is matched for the target users to fill, the matching accuracy of the questionnaire template is improved, the contents of the questionnaire are more accordant with the actual requirements of the users, the questionnaire is divided into a front-stage questionnaire and a rear-stage questionnaire, the front-stage questionnaire is used for carrying out attribute analysis on the users, the results of the attribute analysis are used for carrying out real-time adjustment on the rear-stage questionnaire, the questionnaire is adjusted in real time to enable the questionnaire to be more accordant with the target users, the specific questionnaire data of each target user is constructed, the real requirements of the users fed back by the users in real time are analyzed, the real requirements of the users obtained by analysis are more accordant with the actual requirements of the users, data analysis is carried out on the paths answered by the users in the questionnaire to inquire the target advertisements which are most accordant with the users, therefore, the accuracy of intelligent advertisement putting is improved.

Referring to fig. 2, another embodiment of the intelligent advertisement delivery method according to the embodiment of the present invention includes:

201. receiving an intelligent advertisement putting analysis request sent by a client, and acquiring user data of a target user from a preset cloud database according to the intelligent advertisement putting analysis request, wherein the user data comprises user gender and age, user-defined interest types, user online time and single browsing duration;

specifically, the server queries user data of a target user from a preset application program, wherein the user data comprises user gender, user age and user interest; the server generates a user portrait corresponding to the target user based on the user gender, the user age and the user interest; the server calculates the similarity between the user portrait and a preset user type and compares the similarity with a preset target value; and if the similarity is greater than or equal to the preset target value, the server determines that the user conforms to the preset user type. Further, if the similarity is smaller than a preset target value, determining that the user does not conform to a preset user type, generating a user portrait corresponding to the user by the server based on the user gender, the user age and the user interest, and generating a triple corresponding to the target user by the server, wherein the triple comprises the gender, the age and the interest; the server calculates whether the user portrait accords with the similarity corresponding to the preset user type or not, and compares the similarity with a preset target value; and if the similarity is smaller than the preset target value, the server determines that the target user does not conform to the three types of users.

202. Performing labeling processing on user data to obtain a plurality of user labels corresponding to the user data, calculating the similarity between the user labels and a preset user type template to obtain target similarity, and judging whether a target user accords with a preset user type according to the target similarity, wherein the user type comprises: the user is not clear of the preference of the user, the user is not clear of the requirement of the user, and the user actively selects intelligent advertisement delivery;

specifically, labeling processing is respectively carried out on the gender and age of the user, the user-defined interest types, the user online time period and the single browsing time length in the user data to obtain a plurality of user labels corresponding to the user data; acquiring a template label corresponding to a preset user type template; calculating the tag hit rate between a plurality of user tags and template tags, and taking the tag hit rate as the similarity between the user tags and a preset user type template to obtain target similarity; judging whether the target user accords with a preset user type according to the target similarity, wherein the user type comprises the following steps: the user is not aware of his or her preferences, the user is not aware of his or her needs, and the user actively selects smart advertising. Further, if the user accords with a preset user type, keyword extraction is carried out on the user data to obtain key information corresponding to the user data; the server matches the key information with a plurality of preset questionnaire templates to obtain the key information and the keyword hit rate corresponding to each questionnaire template; the server ranks the keyword hit rate corresponding to each questionnaire template, and takes the questionnaire template with the highest keyword hit rate as the questionnaire template corresponding to the target user; and the server calls a preset template analysis model to perform path analysis on the questionnaire template to obtain a problem path corresponding to the questionnaire template. If the target user accords with the preset user type, the server extracts keywords from the user data to obtain target keywords, the server extracts the keywords by firstly converting the user data into text data, then extracts the text data through a preset OCR character recognition model to obtain standard text data, and finally removes repeated contents from the standard text data to obtain the target keywords; the server respectively matches the target keywords with a plurality of preset questionnaire templates to obtain keyword hit rates corresponding to the questionnaire templates, wherein the keyword hit rates are the number of keyword hits; and the server ranks the keyword hit rate corresponding to each questionnaire template, and takes the questionnaire template with the highest keyword hit rate as the questionnaire template corresponding to the target user.

Obtaining a communication rate and a recommended click rate based on the problem path; specifically, the server optimizes the template according to indexes such as a traffic rate and a recommended click rate obtained through on-line service data statistical analysis, wherein the traffic rate indicates whether a user can walk to a final node of the template. Filling information into the questionnaire template based on the click through rate and the recommended click rate to obtain a filled questionnaire template; one template can be used as a node of another template, namely a sub-template. The templates may have inclusion relationship, for example, a universal template may include several types of templates as sub-templates, and the weight of the template node may be adjusted according to the actual filling condition of the user. Performing data extraction on the filled questionnaire template to obtain an end node and a user answer path corresponding to the questionnaire template; specifically, the server fills information into the questionnaire template based on the circulation rate and the recommended click rate to obtain the filled questionnaire template; the server extracts data of the filled questionnaire template to obtain an end node and a user answer path corresponding to the questionnaire template; the server generates the real demand of the user based on the end node and the user answer path.

203. If the target user accords with the preset user type, matching a questionnaire template corresponding to the target user according to the user data, and performing path segmentation analysis on the questionnaire template to obtain a front-stage problem path and a rear-stage problem path;

specifically, if the target user accords with a preset user type, extracting key information of user data to obtain key information corresponding to the user data, wherein the key information comprises user gender and age and user-defined interest categories; matching the key information with a plurality of preset questionnaire templates to obtain the information matching degree corresponding to each questionnaire template; sorting the information matching degrees corresponding to the questionnaire templates to obtain target sorting results, and taking the questionnaire template corresponding to the information matching degree with the highest rank in the target sorting results as the questionnaire template corresponding to the target user; and calling a preset template analysis model to perform path segmentation analysis on the questionnaire template to obtain a front-stage problem path and a rear-stage problem path corresponding to the questionnaire template. The questionnaire template is subjected to path segmentation analysis to obtain a front-stage question path and a rear-stage question path, wherein the front-stage question path is mainly a preset standard question and has the functions of analyzing the real attribute of a user according to the answer of the user to the front-stage question path and generating the rear-stage question path which is more consistent with a target user in real time according to the real attribute of the user.

204. Performing questionnaire filling on a questionnaire template based on the front-section question path to generate a front-section end node and a front-section answer path corresponding to the questionnaire template, and performing real-time user attribute analysis on a target user according to the front-section end node and the front-section answer path to obtain a real-time analysis result;

specifically, questionnaire filling is carried out on the questionnaire template based on the front-section question path to obtain a filled questionnaire; analyzing questionnaire information of the filled questionnaire to obtain a front section end node and a front section answer path corresponding to the questionnaire template; analyzing the real-time content of the front section end node and the front section answer path to obtain the content of the front section questionnaire; and calling a preset attribute database to perform real attribute real-time analysis on the content of the front-stage questionnaire to obtain a real-time analysis result. The method comprises the steps of setting corresponding characteristic factors for each question in a front-section answer path and an answer fed back by a target user, determining characteristic values corresponding to the characteristic factors based on a front-section end node and the front-section answer path, discretizing or normalizing the characteristic values to obtain characteristic factors corresponding to the front-section answer path, taking an array formed by the characteristic values corresponding to the characteristic factors as target characteristics, and analyzing real attributes of the target user in real time according to the target characteristics to obtain real-time analysis results.

205. Carrying out path updating on the rear-section question path according to the real-time analysis result and a preset path updating strategy to generate an updated question path, and carrying out questionnaire filling on a questionnaire template according to the updated question path to obtain a target end node and a target answer path;

206. carrying out vector mapping on a target end node and a target answer path to obtain an initial input vector, inputting the initial input vector into a preset questionnaire data processing model to carry out user real demand analysis, and generating a user real demand corresponding to a target user, wherein the questionnaire data processing model comprises the following steps: a feature extraction network, a bidirectional long-time memory network and a prediction network;

specifically, vector mapping is carried out on a target end node and a target answer path to obtain an initial input vector; inputting the initial input vector into a preset questionnaire data processing model, wherein the questionnaire data processing model comprises: a feature extraction network, a bidirectional long-time memory network and a prediction network; extracting the characteristics of the initial input vector through a characteristic extraction network to obtain a characteristic vector; providing a bidirectional long-time and short-time memory network to perform feature normalization processing on the feature vector to obtain a feature normalization vector; performing feature prediction on the feature normalized vector through a prediction network to obtain a prediction node; and analyzing the prediction node and the end node to obtain an analysis result, and generating a user real demand corresponding to the target user according to the analysis result. The server performs feature extraction on the initial word vector through a forward long-short-time memory network in a bidirectional long-short-time memory network layer to obtain a forward hidden state sequence; the server extracts the characteristics of the initial word vectors through a reverse long-short-time memory network in the bidirectional long-short-time memory network layer to obtain a reverse hidden state sequence; and the server splices the forward hidden state sequence and the reverse hidden state sequence to obtain a target characteristic sequence. It should be noted that the forward hidden state sequence is a linear sequence that captures the maximum feature value of each forward dimension of the initial word vector through a linear channel of a forward long-short-term memory network; the reverse hidden state sequence is a linear sequence which captures the maximum characteristic value of each reverse dimension of the initial word vector through a linear channel of a reverse long-time memory network. The forward long-short time memory network comprises a plurality of LSTM units, each LSTM unit outputs a sequence element, a forward hidden state sequence is output according to the sequence element, the reverse long-short time memory network comprises a plurality of LSTM units, each LSTM unit outputs a sequence element, and a reverse hidden state sequence is output according to the sequence elements, for example: when the forward hidden state sequence is [1,2,1] and the reverse hidden state sequence is [2,3,4], the server splices the forward hidden state sequence and the reverse hidden state sequence to obtain a target feature sequence of [1,2,1,2,3,4 ].

Optionally, the questionnaire data processing model in this embodiment may also be a convolutional neural network model, where the convolutional neural network model inputs the end node and the user answer path into a preset questionnaire data processing model; the server performs convolution operation on the questionnaire data and the user answer path through a convolution neural network in the questionnaire data processing model to obtain a feature vector corresponding to the questionnaire data; the server performs feature normalization processing on the feature vectors to obtain prediction nodes; the server analyzes the prediction node and the end node to obtain an analysis result; and the server generates the user real demand corresponding to the target user based on the analysis result. It should be noted that the preset questionnaire data processing model includes an input layer, a convolutional neural network, and an output layer, where the input layer: a one-hot vector encoding layer (one-hot vector); hiding the layer: convolution functions, i.e., linear units; an output layer: the dimensions are the same as those of the input layer, and are used for logistic regression. The server performs one-hot vector coding on the target filling information through an input layer to obtain a low-dimensional vector, wherein the low-dimensional vector comprises the following components: [0,0,0,1,0,1,0,0], the server performs feature abstraction operation on the low-dimensional vector through a convolutional neural network layer to obtain an abstract feature value, and the abstract feature value is also a feature vector; and the server performs logistic regression operation on the abstract characteristic values through the output layer to obtain the prediction nodes, wherein the logistic regression operation is softmax regression operation, and the prediction nodes can be online prediction results.

207. Acquiring a plurality of candidate advertisements in a preset advertisement database;

208. performing credibility calculation on a plurality of candidate advertisements based on the real requirements of the user to obtain the target credibility corresponding to each candidate advertisement;

209. ranking the candidate advertisements according to the target credibility corresponding to each candidate advertisement, and taking the candidate advertisement with the highest ranking as the advertisement to be delivered corresponding to the target user to obtain the target advertisement;

210. and carrying out advertisement putting and displaying on the target advertisement through a display page of the client so as to carry out intelligent advertisement putting on the target user.

Specifically, the server acquires a plurality of candidate advertisements in a preset advertisement database; the server carries out reliability calculation on a plurality of candidate advertisements based on the real requirements of the user to obtain the target reliability corresponding to each candidate advertisement; the server ranks the candidate advertisements according to the target credibility corresponding to each candidate advertisement, and takes the candidate advertisement with the highest rank as the advertisement to be delivered corresponding to the target user to obtain the target advertisement; and the server displays the target advertisement through the display page of the application program so as to intelligently place the advertisement to the target user. The server ranks a preset candidate advertisement list based on target credibility to obtain ranking, defines the preference degree of the user to measure the preference index of the user to the candidate advertisement based on the target user and considers the user preference and the matching degree of the candidate advertisement, generates the preference index corresponding to the target user, calculates the matching degree of the advertisement in all the candidate advertisements to the target user, and recommends the target advertisement according to the highest target credibility and the preference index.

In the above, the intelligent advertisement delivery method in the embodiment of the present invention is described, and in the following, referring to fig. 3, an embodiment of the intelligent advertisement delivery device in the embodiment of the present invention includes:

the acquisition module 301 is configured to receive an intelligent advertisement delivery analysis request sent by a client, and acquire user data of a target user from a preset cloud database according to the intelligent advertisement delivery analysis request, where the user data includes user gender and age, user-defined interest categories, a user online time period, and a single browsing time period;

a calculating module 302, configured to perform tagging processing on the user data to obtain multiple user tags corresponding to the user data, calculate a similarity between the user tag and a preset user type template to obtain a target similarity, and determine whether the target user meets a preset user type according to the target similarity, where the user type includes: the user is not clear of the preference of the user, the user is not clear of the requirement of the user, and the user actively selects intelligent advertisement delivery;

a matching module 303, configured to, if the target user meets the preset user type, match a questionnaire template corresponding to the target user according to the user data, and perform path segmentation analysis on the questionnaire template to obtain a front-stage question path and a rear-stage question path;

a filling module 304, configured to fill a questionnaire in the questionnaire template based on the front-segment question path, generate a front-segment end node and a front-segment answer path corresponding to the questionnaire template, and perform real-time user attribute analysis on the target user according to the front-segment end node and the front-segment answer path, to obtain a real-time analysis result;

an updating module 305, configured to perform path updating on the back-stage question path according to the real-time analysis result and a preset path updating policy, generate an updated question path, and perform questionnaire filling on the questionnaire template according to the updated question path, so as to obtain a target end node and a target answer path;

an analysis module 306, configured to perform vector mapping on the target end node and the target answer path to obtain an initial input vector, and input the initial input vector into a preset questionnaire data processing model to perform user real demand analysis, so as to generate a user real demand corresponding to the target user, where the questionnaire data processing model includes: a feature extraction network, a bidirectional long-time memory network and a prediction network;

and the delivering module 307 is configured to match a target advertisement corresponding to the user real demand from a preset advertisement database according to the user real demand, and deliver the target advertisement to the client, so as to perform intelligent advertisement delivery for the target user.

Referring to fig. 4, another embodiment of the intelligent advertisement delivery apparatus according to the embodiment of the present invention includes:

a calculating module 302, configured to perform tagging processing on the user data to obtain a plurality of user tags corresponding to the user data, calculate a similarity between the user tags and a preset user type template to obtain a target similarity, and determine whether the target user conforms to a preset user type according to the target similarity, where the user type includes: the user is not clear of the preference of the user, the user is not clear of the requirement of the user, and the user actively selects intelligent advertisement delivery;

a matching module 303, configured to match a questionnaire template corresponding to the target user according to the user data if the target user meets the preset user type, and perform path segmentation analysis on the questionnaire template to obtain a front-stage question path and a rear-stage question path;

a filling module 304, configured to perform questionnaire filling on the questionnaire template based on the front-segment question path, generate a front-segment end node and a front-segment answer path corresponding to the questionnaire template, and perform real-time user attribute analysis on the target user according to the front-segment end node and the front-segment answer path, so as to obtain a real-time analysis result;

Optionally, the calculating module 302 is specifically configured to: labeling the gender and age of the user, the user-defined interest types, the user online time period and the single browsing time length in the user data respectively to obtain a plurality of user labels corresponding to the user data; acquiring a template label corresponding to a preset user type template; calculating the tag hit rate between the user tags and the template tags, and taking the tag hit rate as the similarity between the user tags and a preset user type template to obtain target similarity; judging whether the target user accords with a preset user type according to the target similarity, wherein the user type comprises the following steps: the user is not aware of his or her preferences, the user is not aware of his or her needs, and the user actively selects smart advertising.

Optionally, the matching module 303 is specifically configured to: if the target user accords with the preset user type, extracting key information of the user data to obtain key information corresponding to the user data, wherein the key information comprises user gender and age and user-defined interest categories; matching the key information with a plurality of preset questionnaire templates to obtain the information matching degree corresponding to each questionnaire template; sorting the information matching degrees corresponding to the questionnaire templates to obtain target sorting results, and taking the questionnaire template corresponding to the information matching degree with the highest rank in the target sorting results as the questionnaire template corresponding to the target user; and calling a preset template analysis model to perform path segmentation analysis on the questionnaire template to obtain a front-stage problem path and a rear-stage problem path corresponding to the questionnaire template.

Optionally, the filling module 304 further includes: filling the questionnaire in the questionnaire template based on the front-section question path to obtain a filled questionnaire; analyzing questionnaire information of the filled questionnaire to obtain a front section end node and a front section answer path corresponding to the questionnaire template; performing real-time content analysis on the front section end node and the front section answer path to obtain front section questionnaire content; and calling a preset attribute database to perform real-time analysis on the real attributes of the front-section questionnaire contents to obtain a real-time analysis result.

Optionally, the analysis module 306 is specifically configured to: carrying out vector mapping on the target end node and the target answer path to obtain an initial input vector; inputting the initial input vector into a preset questionnaire data processing model, wherein the questionnaire data processing model comprises: a feature extraction network, a bidirectional long-time memory network and a prediction network; extracting the characteristics of the initial input vector through the characteristic extraction network to obtain a characteristic vector; providing the bidirectional long-time and short-time memory network to perform feature normalization processing on the feature vector to obtain a feature normalization vector; performing feature prediction on the feature normalized vector through the prediction network to obtain a prediction node; and analyzing the prediction node and the end node to obtain an analysis result, and generating a user real demand corresponding to the target user according to the analysis result.

Optionally, the releasing module 307 is specifically configured to: acquiring a plurality of candidate advertisements in a preset advertisement database; performing credibility calculation on the candidate advertisements based on the real user requirements to obtain target credibility corresponding to each candidate advertisement; ranking the candidate advertisements according to the target credibility corresponding to each candidate advertisement, and taking the candidate advertisement with the highest ranking as the advertisement to be delivered corresponding to the target user to obtain the target advertisement; and carrying out advertisement putting and displaying on the target advertisement through the display page of the client so as to carry out intelligent advertisement putting on the target user.

Optionally, the smart advertisement delivery device further includes:

an optimization module 308, configured to obtain feedback information of the target user on the target advertisement in real time when the target user receives the target advertisement, where the feedback information includes: advertisement evaluation scores and advertisement click operation data; performing operation data analysis on the advertisement click operation data to obtain a click operation analysis result; comprehensively analyzing the feedback information according to the advertisement evaluation score and the click operation analysis result to obtain a target feedback result; and performing intelligent advertisement delivery on advertisements which are selected from the candidate advertisements for the next time based on the target feedback result.

Fig. 3 and fig. 4 describe the intelligent advertisement delivery apparatus in the embodiment of the present invention in detail from the perspective of the modular functional entity, and the intelligent advertisement delivery apparatus in the embodiment of the present invention is described in detail from the perspective of hardware processing.

Fig. 5 is a schematic structural diagram of a smart advertisement delivery device 500 according to an embodiment of the present invention, where the smart advertisement delivery device 500 may have a relatively large difference due to different configurations or performances, and may include one or more processors (CPUs) 510 (e.g., one or more processors) and a memory 520, and one or more storage media 530 (e.g., one or more mass storage devices) storing applications 533 or data 532. Memory 520 and storage media 530 may be, among other things, transient or persistent storage. The program stored on storage medium 530 may include one or more modules (not shown), each of which may include a sequence of instructions operating on smart advertisement delivery device 500. Still further, processor 510 may be configured to communicate with storage medium 530 to execute a series of instruction operations in storage medium 530 on smart advertisement delivery device 500.

Smart advertisement delivery device 500 may also include one or more power supplies 540, one or more wired or wireless network interfaces 550, one or more input-output interfaces 560, and/or one or more operating systems 531, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, and so forth. Those skilled in the art will appreciate that the smart advertisement delivery device configuration shown in fig. 5 does not constitute a limitation of the smart advertisement delivery device and may include more or fewer components than shown, or some components in combination, or a different arrangement of components.

The invention further provides an intelligent advertisement delivery device, which includes a memory and a processor, where the memory stores computer readable instructions, and when the computer readable instructions are executed by the processor, the processor executes the steps of the intelligent advertisement delivery method in the foregoing embodiments.

The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium, or a volatile computer-readable storage medium, having stored therein instructions, which, when executed on a computer, cause the computer to perform the steps of the intelligent advertisement delivery method.

Further, the computer-readable storage medium may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function, and the like; the storage data area may store data created according to the use of the blockchain node, and the like.

The block chain is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, a consensus mechanism, an encryption algorithm and the like. A block chain (Blockchain), which is essentially a decentralized database, is a series of data blocks associated by using a cryptographic method, and each data block contains information of a batch of network transactions, so as to verify the validity (anti-counterfeiting) of the information and generate a next block. The blockchain may include a blockchain underlying platform, a platform product service layer, an application service layer, and the like.

It can be clearly understood by those skilled in the art that, for convenience and simplicity of description, the specific working processes of the above-described systems, apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.

The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention, which is substantially or partly contributed by the prior art, or all or part of the technical solution may be embodied in a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a portable hard disk, a read-only memory (ROM), a Random Access Memory (RAM), a magnetic disk, an optical disk, or other various media capable of storing program codes.

The above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims

1. A smart advertisement delivery method, characterized in that the smart advertisement delivery method comprises:

receiving an intelligent advertisement putting analysis request sent by a client, and acquiring user data of a target user from a preset cloud database according to the intelligent advertisement putting analysis request, wherein the user data comprises user gender and age, user-defined interest types, user online time and single browsing duration;

performing tagging processing on the user data to obtain a plurality of user tags corresponding to the user data, calculating the similarity between the user tags and a preset user type template to obtain target similarity, and judging whether the target user accords with a preset user type according to the target similarity, wherein the user type comprises: the user is not clear of the preference of the user, the user is not clear of the requirement of the user, and the user actively selects intelligent advertisement delivery;

if the target user accords with the preset user type, matching a questionnaire template corresponding to the target user according to the user data, and performing path segmentation analysis on the questionnaire template to obtain a front-stage problem path and a rear-stage problem path;

performing questionnaire filling on the questionnaire template based on the front-section question path, generating a front-section end node and a front-section answer path corresponding to the questionnaire template, and performing real-time analysis on the user real attribute of the target user according to the front-section end node and the front-section answer path to obtain a real-time analysis result;

performing path updating on the rear-section question path according to the real-time analysis result and a preset path updating strategy to generate an updated question path, and performing questionnaire filling on the questionnaire template according to the updated question path to obtain a target end node and a target answer path;

vector mapping is carried out on the target end node and the target answer path to obtain an initial input vector, the initial input vector is input into a preset questionnaire data processing model to carry out user real demand analysis, and a user real demand corresponding to the target user is generated, wherein the questionnaire data processing model comprises: a feature extraction network, a bidirectional long-time memory network and a prediction network;

and matching a target advertisement corresponding to the real user demand from a preset advertisement database according to the real user demand, and delivering the target advertisement to the client so as to perform intelligent advertisement delivery on the target user.

2. The intelligent advertisement delivery method according to claim 1, wherein the tagging is performed on the user data to obtain a plurality of user tags corresponding to the user data, the similarity between the user tags and a preset user type template is calculated to obtain a target similarity, and whether the target user conforms to a preset user type is determined according to the target similarity, wherein the user type includes: the method comprises the following steps that a user is not clear of own preference, the user is not clear of own requirement, and the user actively selects intelligent advertisement putting, wherein the steps comprise:

labeling the gender and age of the user, the user-defined interest types, the user online time period and the single browsing time length in the user data respectively to obtain a plurality of user labels corresponding to the user data;

acquiring a template label corresponding to a preset user type template;

calculating the tag hit rate between the user tags and the template tags, and taking the tag hit rate as the similarity between the user tags and a preset user type template to obtain target similarity;

judging whether the target user accords with a preset user type according to the target similarity, wherein the user type comprises the following steps: the user is not aware of his preferences, the user is not aware of his needs, and the user actively selects smart advertising.

3. The smart advertisement delivery method according to claim 1, wherein if the target user is in accordance with the preset user type, matching a questionnaire template corresponding to the target user according to the user data, and performing path segmentation analysis on the questionnaire template to obtain a front-stage question path and a rear-stage question path, includes:

if the target user accords with the preset user type, extracting key information of the user data to obtain key information corresponding to the user data, wherein the key information comprises user gender and age and user-defined interest categories;

matching the key information with a plurality of preset questionnaire templates to obtain the information matching degree corresponding to each questionnaire template;

sorting the information matching degrees corresponding to the questionnaire templates to obtain target sorting results, and taking the questionnaire template corresponding to the information matching degree with the highest rank in the target sorting results as the questionnaire template corresponding to the target user;

and calling a preset template analysis model to perform path segmentation analysis on the questionnaire template to obtain a front-stage problem path and a rear-stage problem path corresponding to the questionnaire template.

4. The intelligent advertisement delivery method according to claim 1, wherein the questionnaire template is filled based on the previous-stage question path, a previous-stage end node and a previous-stage answer path corresponding to the questionnaire template are generated, and real-time analysis of user attributes is performed on the target user according to the previous-stage end node and the previous-stage answer path, so as to obtain a real-time analysis result, and the method comprises:

filling the questionnaire in the questionnaire template based on the front-section question path to obtain a filled questionnaire;

analyzing questionnaire information of the filled questionnaire to obtain a front section end node and a front section answer path corresponding to the questionnaire template;

analyzing the content of the front section end node and the front section answer path in real time to obtain the content of a front section questionnaire;

and calling a preset attribute database to perform real-time analysis on the real attributes of the front-section questionnaire contents to obtain a real-time analysis result.

5. The intelligent advertisement delivery method according to claim 3, wherein the vector mapping is performed on the target end node and the target answer path to obtain an initial input vector, and the initial input vector is input into a preset questionnaire data processing model to perform user real demand analysis, so as to generate a user real demand corresponding to the target user, wherein the questionnaire data processing model includes: the characteristic extraction network, the bidirectional long-time and short-time memory network and the prediction network comprise:

carrying out vector mapping on the target end node and the target answer path to obtain an initial input vector;

inputting the initial input vector into a preset questionnaire data processing model, wherein the questionnaire data processing model comprises: a feature extraction network, a bidirectional long-time memory network and a prediction network;

extracting the characteristics of the initial input vector through the characteristic extraction network to obtain a characteristic vector;

providing the bidirectional long-time and short-time memory network to perform feature normalization processing on the feature vector to obtain a feature normalization vector;

performing feature prediction on the feature normalized vector through the prediction network to obtain a prediction node;

and analyzing the prediction node and the end node to obtain an analysis result, and generating a user real demand corresponding to the target user according to the analysis result.

6. The intelligent advertisement delivery method according to claim 2, wherein the matching of the target advertisement corresponding to the real user demand from a preset advertisement database according to the real user demand and the delivery of the target advertisement to the client for intelligent advertisement delivery to the target user comprises:

acquiring a plurality of candidate advertisements in a preset advertisement database;

performing credibility calculation on the candidate advertisements based on the real user requirements to obtain target credibility corresponding to each candidate advertisement;

ranking the candidate advertisements according to the target credibility corresponding to each candidate advertisement, and taking the candidate advertisement with the highest ranking as the advertisement to be delivered corresponding to the target user to obtain the target advertisement;

and carrying out advertisement putting and displaying on the target advertisement through the display page of the client so as to carry out intelligent advertisement putting on the target user.

7. The smart advertisement delivery method according to claim 6, further comprising:

when the target user receives the target advertisement, obtaining feedback information of the target user to the target advertisement in real time, wherein the feedback information comprises: advertisement evaluation scores and advertisement click operation data;

performing operation data analysis on the advertisement click operation data to obtain a click operation analysis result;

comprehensively analyzing the feedback information according to the advertisement evaluation score and the click operation analysis result to obtain a target feedback result;

and selecting the next advertisement from the candidate advertisements for intelligent advertisement delivery based on the target feedback result.

8. The utility model provides an intelligent advertisement puts in device which characterized in that, intelligent advertisement puts in device includes:

the system comprises an acquisition module, a display module and a display module, wherein the acquisition module is used for receiving an intelligent advertisement putting analysis request sent by a client and acquiring user data of a target user from a preset cloud database according to the intelligent advertisement putting analysis request, and the user data comprises user gender and age, user-defined interest types, user online time and single browsing time;

a calculating module, configured to perform tagging processing on the user data to obtain a plurality of user tags corresponding to the user data, calculate a similarity between the user tags and a preset user type template to obtain a target similarity, and determine whether the target user conforms to a preset user type according to the target similarity, where the user type includes: the user is not clear of the preference of the user, the user is not clear of the requirement of the user, and the user actively selects intelligent advertisement delivery;

the matching module is used for matching a questionnaire template corresponding to the target user according to the user data and carrying out path segmentation analysis on the questionnaire template to obtain a front-stage question path and a rear-stage question path if the target user accords with the preset user type;

a filling module, configured to perform questionnaire filling on the questionnaire template based on the front-segment question path, generate a front-segment end node and a front-segment answer path corresponding to the questionnaire template, and perform real-time user attribute analysis on the target user according to the front-segment end node and the front-segment answer path, so as to obtain a real-time analysis result;

the updating module is used for updating the path of the rear-section question path according to the real-time analysis result and a preset path updating strategy to generate an updated question path, and filling the questionnaire template according to the updated question path to obtain a target end node and a target answer path;

an analysis module, configured to perform vector mapping on the target end node and the target answer path to obtain an initial input vector, input the initial input vector into a preset questionnaire data processing model to perform user real demand analysis, and generate a user real demand corresponding to the target user, where the questionnaire data processing model includes: a feature extraction network, a bidirectional long-time memory network and a prediction network;

and the delivery module is used for matching the target advertisement corresponding to the real user demand from a preset advertisement database according to the real user demand and delivering the target advertisement to the client so as to carry out intelligent advertisement delivery on the target user.

9. An intelligent advertisement delivery device, comprising: a memory and at least one processor, the memory having instructions stored therein;

the at least one processor invokes the instructions in the memory to cause the smart advertisement delivery device to perform the smart advertisement delivery method of any of claims 1-7.

10. A computer-readable storage medium having instructions stored thereon, wherein the instructions, when executed by a processor, implement a smart advertisement delivery method as recited in any one of claims 1-7.