CN117808535A - Advertisement putting method and system based on user big data analysis - Google Patents
Advertisement putting method and system based on user big data analysis Download PDFInfo
- Publication number
- CN117808535A CN117808535A CN202311852190.5A CN202311852190A CN117808535A CN 117808535 A CN117808535 A CN 117808535A CN 202311852190 A CN202311852190 A CN 202311852190A CN 117808535 A CN117808535 A CN 117808535A
- Authority
- CN
- China
- Prior art keywords
- user
- information
- advertisement
- commodity
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 33
- 238000007405 data analysis Methods 0.000 title claims abstract description 32
- 230000000694 effects Effects 0.000 claims abstract description 78
- 238000006243 chemical reaction Methods 0.000 claims abstract description 26
- 230000006399 behavior Effects 0.000 claims description 58
- 238000004422 calculation algorithm Methods 0.000 claims description 57
- 238000012544 monitoring process Methods 0.000 claims description 48
- 238000012545 processing Methods 0.000 claims description 31
- 238000004458 analytical method Methods 0.000 claims description 25
- 238000012549 training Methods 0.000 claims description 23
- 238000005457 optimization Methods 0.000 claims description 22
- 238000010801 machine learning Methods 0.000 claims description 21
- 238000013135 deep learning Methods 0.000 claims description 20
- 230000008859 change Effects 0.000 claims description 18
- 230000007246 mechanism Effects 0.000 claims description 17
- 238000012360 testing method Methods 0.000 claims description 13
- 230000008569 process Effects 0.000 claims description 12
- 238000002716 delivery method Methods 0.000 claims description 8
- 230000002776 aggregation Effects 0.000 claims description 6
- 238000004220 aggregation Methods 0.000 claims description 6
- 238000004590 computer program Methods 0.000 claims description 6
- 238000012986 modification Methods 0.000 claims description 6
- 230000004048 modification Effects 0.000 claims description 6
- 238000007781 pre-processing Methods 0.000 claims description 6
- 230000035945 sensitivity Effects 0.000 claims description 6
- 238000012216 screening Methods 0.000 claims description 5
- 238000010998 test method Methods 0.000 claims description 5
- 238000012217 deletion Methods 0.000 claims description 4
- 230000037430 deletion Effects 0.000 claims description 4
- 238000002474 experimental method Methods 0.000 claims description 4
- 238000001914 filtration Methods 0.000 claims description 4
- 238000003780 insertion Methods 0.000 claims description 4
- 230000037431 insertion Effects 0.000 claims description 4
- 230000000007 visual effect Effects 0.000 claims description 4
- 238000005192 partition Methods 0.000 claims description 3
- 238000003860 storage Methods 0.000 claims description 3
- 230000008901 benefit Effects 0.000 description 3
- 230000006872 improvement Effects 0.000 description 3
- 239000002699 waste material Substances 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000004140 cleaning Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000005065 mining Methods 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 235000013361 beverage Nutrition 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000019771 cognition Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000010921 in-depth analysis Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 230000026676 system process Effects 0.000 description 1
- 230000008685 targeting Effects 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides an advertisement putting method and system based on user big data analysis. Acquiring user information and advertisement commodity information from a background database in real time; the user information comprises user IP information, geographic position positioning information and user identity information; the advertisement commodity information comprises commodity type information, commodity sales information, commodity browsing information and commodity searching information; and acquiring advertisement commodity information in the area of the user according to the user IP information and the geographic position positioning information, and acquiring user identity information corresponding to the advertisement commodity according to the user behavior data. Through analyzing the big data of the user, the user interest, preference, behavior characteristics and other information can be deeply known, so that accurate target user positioning is realized. Advertisements can be delivered to truly potential interested users, and the advertisement reaching effect and conversion rate are improved.
Description
Technical Field
The invention provides an advertisement putting method and system based on user big data analysis, and belongs to the technical field of big data analysis and advertisement putting.
Background
With the rapid development of the internet, the application of big data technology is becoming more and more widespread. Advertising as an important means of marketing, how to reach target users more precisely is the focus of attention of advertisers and advertising platforms. Traditional advertisement delivery modes are often delivered based on simple user labels, and lack of deep analysis and mining of user behaviors results in poor delivery effects. Therefore, the advertisement putting method based on the user big data analysis is developed to improve the accuracy and the effect of advertisement putting, and is a problem which needs to be solved in the current advertisement industry.
Conventional advertising methods are typically based on simple tags of the user, such as geographic location, age, gender, etc. These tags often lack in-depth analysis and mining of user behavior, resulting in poor delivery. In addition, the traditional advertisement putting method lacks of acquiring and analyzing the identity information of the user, and can not accurately put according to the identity information of the user.
Disclosure of Invention
The invention provides an advertisement putting method and system based on user big data analysis, which are used for solving the problems that the existing advertisement putting mode cannot accurately put according to the characteristics of specific crowd and the putting strategy cannot be timely adjusted according to market feedback and actual conditions:
The invention provides an advertisement putting method based on user big data analysis, which comprises the following steps:
s1: acquiring user information and advertisement commodity information from a background database in real time; the user information comprises user IP information, geographic position positioning information and user identity information; the advertisement commodity information comprises commodity type information, commodity sales information, commodity browsing information and commodity searching information;
s2: acquiring advertisement commodity information in an area where a user is located according to the user IP information and the geographic position positioning information, and acquiring user identity information corresponding to the advertisement commodity according to the user behavior data;
s3: analyzing the advertisement commodity information and the user identity information by combining a machine learning algorithm with a deep learning algorithm, and establishing user portraits of users with different sexes and different ages in the area according to analysis results;
s4: establishing a commodity recommendation model according to the user portrait, and throwing advertisement commodities according to the commodity recommendation model;
s5: and monitoring the advertisement putting effect in real time, and continuously optimizing the recommendation model according to the monitoring effect.
Further, the user information and the advertisement commodity information are obtained from a background database in real time; the user information comprises user IP information, geographic position positioning information and user identity information; the advertisement commodity information comprises commodity type information, commodity sales information, commodity browsing information and commodity searching information; comprising the following steps:
S11: the system is connected with a background database through an API interface, and monitors the user information and the data change of the advertisement commodity information table through a monitoring mechanism provided by the background database;
s12: triggering corresponding events when a data change is detected, wherein the data change comprises insertion, update, modification and deletion;
s13: the data change status information in the database is published to a message queue in the form of a message through Kafka;
s14: processing and analyzing each received message in real time by using a machine learning algorithm;
s15: updating relevant user information and advertisement commodity information in real time according to the data processing result; and storing the results of updating the relevant user information and the advertisement commodity information in real time into a database.
Further, each received message is processed and analyzed in real time by using a machine learning algorithm; comprising the following steps:
s141: dividing the received message data set into small data blocks according to a certain rule, and distributing the data blocks to available computing nodes;
s142: each Map task acquires a data block, and for each message, the Map module converts the data block into a key value pair form, wherein the key represents the characteristic of the message, the characteristic comprises a keyword and a time stamp, and the value represents the message or related messages;
S143: the Map module outputs the generated intermediate result to the distributed file system in the form of key value pairs;
s144: the buffer module partitions and sorts the intermediate results output by the Map stage according to keys, and sends the data of the same key to the same Reduce task node;
s145: each Reduce module receives partitioned intermediate result data from the Shuffle module;
s146: the Reduce module processes each received key value pair, wherein the processing comprises aggregation, screening and statistics, and a final output result is obtained;
s147: and the Reduce module outputs the processed result to the database.
Further, the user identity information comprises user registration information, purchase preference and user behavior data; the user registration information comprises user gender, age, user name and mobile phone number; the purchase preference comprises a preference class, a brand preference, a price sensitivity, a special requirement media influence and a regional culture difference; the user behavior data includes user browsing behavior, clicking behavior, user purchase history, and searching behavior.
Further, the advertisement commodity information and the user identity information are analyzed through a machine learning algorithm and a deep learning algorithm, and user figures of users with different sexes and different ages in the area are established according to analysis results; comprising the following steps:
S31: acquiring a data set of advertisement commodity information and user identity information through a database; and preprocessing the data set;
s32: extracting features from the user identity information, wherein the features comprise key features and main features, the key features comprise user gender and age, and the main features comprise purchase preference and user behavior data;
s33: associating the advertisement commodity information with the user identity information, and connecting the two data sets through common attributes to form an integral data set; the common attributes include commodity IDs and user IDs;
s34: dividing the integral data set into a training set and a testing set; the training set is used for training a model, and the testing set is used for evaluating the performance of the model;
s35: training and parameter tuning are carried out on the model through a training set, and the trained model is evaluated and verified through a testing set;
and evaluating the performance and generalization capability of the model by calculating indexes such as accuracy, recall rate, F1 value and the like.
S36: according to the trained model, predicting and classifying new user data, dividing the users into different age groups with different sexes through the characteristics, and establishing user portraits;
S37: and visually presenting the user portrait result through a visual tool.
Further, a commodity recommendation model is established according to the user portrait, and advertisement commodities are put in according to the commodity recommendation model; comprising the following steps:
calculating the correlation and similarity between the user and the advertisement commodity through a collaborative filtering algorithm, and matching the user image data with the advertisement commodity characteristics;
generating a candidate advertisement commodity list for each user according to the matching degree between the user portrait and the advertisement commodity; the candidate good list may be ranked according to predicted user preference for the advertised good to determine the most appropriate advertised good.
And setting an advertisement delivery mechanism and a strategy according to the budget, the delivery channel and the marketing target of the advertisement.
Further, the advertisement putting effect is monitored in real time, and the recommendation model is continuously optimized according to the monitoring effect; comprising the following steps:
determining key indexes for evaluating advertisement putting effect; the key indexes comprise click rate, conversion rate and return on investment rate;
collecting related data in the advertisement putting process in real time through a data monitoring module, wherein the related data comprises advertisement exposure rate, click quantity and conversion events;
Based on the collected data, analyzing the data through a deep learning algorithm to obtain advertisement putting condition information and generating a monitoring report;
according to the monitoring result, adjusting and optimizing the selection and processing mode of the advertisement commodity characteristics, adjusting the parameters of the recommendation model, and evaluating and trying different recommendation algorithms or improving the existing algorithm;
comparing different optimization strategies through an A/B test method; randomly dividing a part of users into an experiment group and a control group, respectively applying different models or strategies to carry out advertisement delivery, and evaluating the effect of optimizing the strategy by comparing index changes of the two groups of data;
establishing a continuous monitoring mechanism, periodically checking and evaluating advertisement putting effect, and performing iterative optimization according to feedback data; and according to the changes of different periods and requirements, timely adjusting a recommendation model and an advertisement putting strategy.
The invention provides an advertisement putting system based on user big data analysis, which comprises:
and a data acquisition module: acquiring user information and advertisement commodity information from a background database in real time; the user information comprises user IP information, geographic position positioning information and user identity information; the advertisement commodity information comprises commodity type information, commodity sales information, commodity browsing information and commodity searching information;
The regional information determining module: acquiring advertisement commodity information in an area where a user is located according to the user IP information and the geographic position positioning information, and acquiring user identity information corresponding to the advertisement commodity according to the user behavior data;
a user portrait creation module: analyzing the advertisement commodity information and the user identity information by combining a machine learning algorithm with a deep learning algorithm, and establishing user portraits of users with different sexes and different ages in the area according to analysis results;
an advertisement commodity recommending module: establishing a commodity recommendation model according to the user portrait, and throwing advertisement commodities according to the commodity recommendation model;
model optimization module: and monitoring the advertisement putting effect in real time, and continuously optimizing the recommendation model according to the monitoring effect.
The invention provides electronic equipment, which comprises a memory, a processor and a computer program stored in the memory and capable of running on the memory, wherein the processor executes the program to realize any advertisement delivery method based on user big data analysis.
The invention provides a non-transitory computer readable storage medium, on which a computer program is stored, the program being executed by a processor to implement the advertisement delivery method based on user big data analysis as described in any one of the above.
The invention has the beneficial effects that: through analyzing the big data of the user, the user interest, preference, behavior characteristics and other information can be deeply known, so that accurate target user positioning is realized. The advertisements can be put into users with real potential interests, so that the advertisement reaching effect and conversion rate are improved; by further analyzing the user big data, the user groups interested in the advertisement can be identified, so that the advertisement content, the delivery channel and the time are optimized. Thus, the click rate, conversion rate and return on investment rate of the advertisement can be improved, and the effect and benefit of advertisement putting are improved; through analysis of the big data of the user, the personalized requirements and the preferences of the user can be known, and accordingly advertising content and recommendation strategies can be adjusted according to the personalized characteristics of the user. The personalized advertisement can attract the attention of the user, and the participation degree and the purchase willingness of the user are increased; based on analysis of big data of users, advertisement putting effect can be monitored in real time, and optimization adjustment is carried out according to the monitoring result. Thus, problems and improvement spaces can be quickly found, and the advertising effect and the delivery benefit are improved; through analysis based on the big data of the user, more personalized and highly relevant advertisement content and recommendation can be provided for the user; the interference of the user on the advertisement can be reduced, and the satisfaction degree and the brand cognition degree of the user are improved; through accurate target user positioning and optimizing the delivery strategy, the waste of advertisement resources can be avoided, and unnecessary advertisement delivery cost is reduced. Meanwhile, the click rate and the conversion rate of the advertisement are improved, and a better advertisement effect can be obtained with less throwing cost.
Drawings
FIG. 1 is a step diagram of an advertisement delivery method based on user big data analysis according to the present invention;
FIG. 2 is a block diagram of an advertisement delivery system based on user big data analysis according to the present invention.
Detailed Description
In order that the above-recited objects, features and advantages of the present invention will be more clearly understood, a more particular description of the invention will be rendered by reference to specific embodiments thereof which are illustrated in the appended drawings. It should be noted that, in the case of no conflict, the embodiments of the present application and the features in the embodiments may be combined with each other.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, and the described embodiments are merely some, rather than all, embodiments of the present invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.
In one embodiment of the present invention, as shown in fig. 1, an advertisement delivery method based on user big data analysis, the method includes:
s1: acquiring user information and advertisement commodity information from a background database in real time; the user information comprises user IP information, geographic position positioning information and user identity information; the advertisement commodity information comprises commodity type information, commodity sales information, commodity browsing information and commodity searching information;
s2: acquiring advertisement commodity information in an area where a user is located according to the user IP information and the geographic position positioning information, and acquiring user identity information corresponding to the advertisement commodity according to the user behavior data;
s3: analyzing the advertisement commodity information and the user identity information by combining a machine learning algorithm with a deep learning algorithm, and establishing user portraits of users with different sexes and different ages in the area according to analysis results;
s4: establishing a commodity recommendation model according to the user portrait, and throwing advertisement commodities according to the commodity recommendation model;
s5: and monitoring the advertisement putting effect in real time, and continuously optimizing the recommendation model according to the monitoring effect.
The working principle of the technical scheme is as follows: : acquiring IP information, geographic position positioning information and user identity information of a user, and type information, sales information, browsing information and search information of advertisement commodities in real time through a background database; determining the area where the user is located according to the IP information and the geographic position positioning information of the user, and acquiring advertisement commodity information in the area from a database; acquiring the association information of the user and the advertisement commodity, namely the user identity information corresponding to the advertisement commodity according to the behavior data of the user, such as browsing records, searching records and the like; analyzing the advertisement commodity information and the user identity information by using a machine learning algorithm and a deep learning algorithm to know interests, preferences and consumption behaviors of users of different sexes and different age groups, and establishing user portraits; according to the user portrait, a commodity recommendation model is established, and the model can recommend advertisement commodities according to the characteristics and the historical behaviors of the user; and putting the advertisement commodity according to the commodity recommendation model. Through accurate positioning and personalized recommendation, the advertisements are put into a target user group, and the conversion rate and the user participation of the advertisements are improved; and monitoring the advertisement putting effect in real time, wherein the advertisement putting effect comprises indexes such as click rate, conversion rate and the like. And carrying out data analysis according to the monitoring result, evaluating the effect of the advertisement, and continuously optimizing the recommendation model.
The technical scheme has the effects that: the user information and the advertisement commodity information are obtained in real time through the background database, so that timeliness and accuracy of data can be ensured, and a basis is provided for subsequent analysis and recommendation; through the user IP information and the geographic position positioning information, the advertisement commodity can be accurately put into the users in the corresponding areas, and the coverage rate and the accuracy of the advertisement are improved; the advertisement commodity information and the user identity information are analyzed by combining a deep learning algorithm, so that the behavior mode and preference of the user can be mined, and the user needs and interests can be further known; through analysis results, user portraits of users with different sexes and different ages can be established, advertisers are helped to better know target user groups, and accordingly advertisement putting strategies are optimized; according to the user portraits, a commodity recommendation model is established, advertisement commodities suitable for users can be accurately recommended according to interests and preferences of the users, and the click rate and conversion rate of advertisements are improved; problems and optimization space can be found in time by monitoring advertisement putting effect in real time, and advertisement effect and ROI (return on investment) are improved; and continuously optimizing the recommendation model according to the monitoring effect, adjusting the recommendation algorithm and parameters, and improving the effect and the accuracy of advertisement delivery.
In one embodiment of the invention, the user information and the advertisement commodity information are acquired from a background database in real time; the user information comprises user IP information, geographic position positioning information and user identity information; the advertisement commodity information comprises commodity type information, commodity sales information, commodity browsing information and commodity searching information; comprising the following steps:
s11: the system is connected with a background database through an API interface, and monitors the user information and the data change of the advertisement commodity information table through a monitoring mechanism provided by the background database;
s12: triggering corresponding events when a data change is detected, wherein the data change comprises insertion, update, modification and deletion;
s13: the data change status information in the database is published to a message queue in the form of a message through Kafka;
s14: processing and analyzing each received message in real time by using a machine learning algorithm;
s15: updating relevant user information and advertisement commodity information in real time according to the data processing result; and storing the results of updating the relevant user information and the advertisement commodity information in real time into a database.
The working principle of the technical scheme is as follows: the system is connected with a background database through an API interface, and monitors the data change of a user information table and an advertisement commodity information table in real time by utilizing a monitoring mechanism provided by the background database; when detecting the data change, the system triggers corresponding events including operations such as insertion, updating, modification and deletion of the data; using Kafka as a message queue, and publishing the data change information in the database into the message queue in a message form; kafka has the characteristics of high throughput and low delay, and can effectively process a large number of messages; after receiving the data change information in the message queue, the system processes and analyzes each message in real time by using a machine learning algorithm. This may include steps of data cleaning, feature extraction, model training, etc. to obtain more valuable information; and according to the data processing result, the system updates the related user information and advertisement commodity information in real time, and stores the updated result into a background database for subsequent inquiry and use.
The technical scheme has the effects that: by acquiring user information in real time, including IP information, geographic location information, and user identity information, the system may perform personalized advertisement pushing based on these information. For example, according to the characteristics and preferences of the region where the user is located, advertisement commodities relevant to the local area are displayed to the user, so that the click rate and conversion rate of the advertisements are improved, and sales of merchants and advertisers are increased; by monitoring the data change of the advertisement commodity information table, the system can acquire the sales information, browsing information and searching information of the commodity in real time. The real-time updated information can help merchants to better know market demands and user behaviors, optimize commodity pricing strategies, sales promotion activities and recommendation strategies, and improve sales performance and user satisfaction; by processing and analyzing each message received in real-time using machine learning algorithms, the system can mine potential patterns and trends that are hidden in the mass data. The method is beneficial to predicting the behaviors and interests of users and is capable of providing insight into market trends in advance, so that more accurate marketing strategies are formulated for merchants, market risks are reduced, and competitiveness is improved; by employing a combination of message queues and a background database, the system is able to process and store large amounts of data in a high throughput and low latency manner. This enables the system to cope with highly concurrent data access and update and ensures reliability and consistency of the data; due to the instantaneity of the technical scheme, a merchant can acquire user feedback and market data in real time and adjust advertisement pushing and commodity strategies in time so as to adapt to continuously changing market demands and user preferences and improve the accuracy and the agility of business decisions.
In one embodiment of the invention, each received message is processed and analyzed in real time by using a machine learning algorithm; comprising the following steps:
s141: dividing the received message data set into small data blocks according to a certain rule, and distributing the data blocks to available computing nodes;
s142: each Map task acquires a data block, and for each message, the Map module converts the data block into a key value pair form, wherein the key represents the characteristic of the message, the characteristic comprises a keyword and a time stamp, and the value represents the message or related messages;
s143: the Map module outputs the generated intermediate result to the distributed file system in the form of key value pairs;
s144: the buffer module partitions and sorts the intermediate results output by the Map stage according to keys, and sends the data of the same key to the same Reduce task node;
s145: each Reduce module receives partitioned intermediate result data from the Shuffle module;
s146: the Reduce module processes each received key value pair, wherein the processing comprises aggregation, screening and statistics, and a final output result is obtained;
s147: and the Reduce module outputs the processed result to the database.
The working principle of the technical scheme is as follows: the received message data set will be divided into small data blocks according to certain rules and these data blocks will be distributed over the available computing nodes. This is done to break down large-scale datasets into smaller tasks for parallel processing and to increase computational efficiency; the Map task on each compute node will take a block of data and for each message the Map module will convert it into the form of a key-value pair. Keys represent features of the message such as keywords and time stamps, while values represent the message itself or other information related thereto. The Map module outputs the generated intermediate result to the distributed file system in the form of key value pairs so as to facilitate subsequent processing; the Shuffle module performs partitioning and sorting operations on intermediate results output by the Map stage according to the keys. The method can send the data with the same key to the same Reduce task node so as to facilitate subsequent processing and aggregation operation; each Reduce task node receives partitioned intermediate result data from the Shuffle stage. The Reduce module processes each received key value pair, including aggregation, screening, statistics and other operations, so as to obtain a final output result; the Reduce module outputs the processed results to a database for subsequent data analysis and use.
The technical scheme has the effects that: the large-scale message data set is divided into small data blocks and distributed to a plurality of computing nodes for parallel processing, so that the processing speed and the computing efficiency can be improved. Each computing node can independently process the distributed data blocks, so that the computing load of a single node is reduced, and the processing speed of the whole system is increased; real-time processing and analysis of each received message may be implemented. And the Map stage converts each message and extracts the characteristics, and the Reduce stage aggregates, screens and counts the messages to finally obtain a real-time output result. This is very important for application scenarios requiring fast response and timely decisions; the distributed computing and MapReduce framework has good expandability, and computing nodes can be flexibly increased according to requirements so as to adapt to the ever-increasing message data volume and processing requirements. By adding the computing nodes, the system can smoothly expand the processing capacity, and ensure efficient and reliable processing service; because the data is divided and distributed to a plurality of computing nodes for processing, even if one computing node fails, the whole system can still normally operate, and the work of other nodes is not influenced. The distributed fault tolerance ensures high availability and stability of the system; through the MapReduce framework, the system can perform multidimensional processing and analysis on the received message data, including operations such as feature extraction, aggregation, screening, statistics and the like. These processing and analysis operations may help users understand and mine potential information and patterns in message data, supporting deeper data analysis and application development.
In one embodiment of the present invention, the user identity information includes user registration information, purchase preference, and user behavior data; the user registration information comprises user gender, age, user name and mobile phone number; the purchase preference comprises a preference class, a brand preference, a price sensitivity, a special requirement media influence and a regional culture difference; the user behavior data includes user browsing behavior, clicking behavior, user purchase history, and searching behavior.
The working principle of the technical scheme is as follows: the user identity information comprises user registration information, purchase preference and user behavior data; the user registration information comprises user gender, age, user name and mobile phone number; the purchase preference comprises a preference class, a brand preference, a price sensitivity, a special requirement media influence and a regional culture difference; the user behavior data includes user browsing behavior, clicking behavior, user purchase history, and searching behavior.
Specific:
class preference: users may have a preference for certain specific product categories, such as fashion apparel, electronics, household items, food and beverage, and the like. They are more inclined to purchase products of interest or frequent use.
Brand preference: users may feel or trust certain specific brands that they tend to purchase products of those brands. Factors such as brand image, quality, public praise, etc. may affect the user's purchasing decisions.
Price sensitivity: the sensitivity of the user to price is also an aspect of purchase preference. Some users prefer to purchase lower priced products, while some users pay more attention to the quality and value of the products, willing to pay a higher price for high quality products.
Special requirements are: the user may have special requirements, such as requirements for a particular function, material, specification or design. They will be more concerned about whether the product meets their own specific needs and choose to purchase according to the needs.
Buying habit: the purchasing habits and shopping preferences of the user are also part of the purchasing preferences. For example, some users prefer online shopping and some prefer to go to physical store for purchase; some users prefer to purchase premium promotional products and some users pay more attention to purchasing full value products.
Media impact and social factors: the user may be affected by media, advertising, friends, or social circles, generating interests and likes for certain goods or brands, creating purchase preferences.
Regional culture difference: the user may have differences in demand and purchase preference for products in different geographical and cultural contexts. Factors such as regional unique cultures, customs and habits can influence the purchasing decisions and preferences of the user.
The technical scheme has the effects that: through user registration information and purchase preference analysis, the user's preferences, interests and needs can be known. Based on this information, the system may provide personalized recommendation services that recommend products or services to the user that meet their category preferences, brand preferences, and special needs. Personalized recommendation can improve user satisfaction, promote user purchasing decision and increase sales; the users may be subdivided into different groups or types by comprehensive analysis of user registration information, purchase preferences, and user behavior data. For example, users are divided according to the dimensions of gender, age, regional culture difference and the like, the preferences and demands of different groups are known, and marketing activities and service strategies are developed in a targeted manner. The market accuracy can be improved by the client subdivision, the resource utilization is optimized, and the user viscosity is enhanced; by analyzing user behavior data such as browsing behavior, clicking behavior and purchase history, the behavior trace and purchase habit of the user can be known. The system can conduct personalized guidance and pushing according to the user behavior data, and provide services and products which are more in line with the interests and demands of users. This helps to promote user retention and loyalty, increasing the user's repurchase and long-term value; by analyzing the user registration information, purchase preferences, and user behavior data, a large amount of market feedback and user opinion can be obtained. Such data may be used in market research to learn user satisfaction with the product or service, needs, and improvement suggestions. Based on these feedback and opinion, enterprises can perform product optimization and service improvement, improving product competitiveness and user experience.
According to one embodiment of the invention, the advertisement commodity information and the user identity information are analyzed through a machine learning algorithm and a deep learning algorithm, and user figures of users with different sexes and different ages in the area are established according to analysis results; comprising the following steps:
s31: acquiring a data set of advertisement commodity information and user identity information through a database; and preprocessing the data set;
s32: extracting features from the user identity information, wherein the features comprise key features and main features, the key features comprise user gender and age, and the main features comprise purchase preference and user behavior data;
s33: associating the advertisement commodity information with the user identity information, and connecting the two data sets through common attributes to form an integral data set; the common attributes include commodity IDs and user IDs;
s34: dividing the integral data set into a training set and a testing set; the training set is used for training a model, and the testing set is used for evaluating the performance of the model;
s35: training and parameter tuning are carried out on the model through a training set, and the trained model is evaluated and verified through a testing set; evaluating the performance and generalization capability of the model by calculating indexes such as accuracy, recall rate, F1 value and the like;
S36: according to the trained model, predicting and classifying new user data, dividing the users into different age groups with different sexes through the characteristics, and establishing user portraits;
s37: and visually presenting the user portrait result through a visual tool.
The working principle of the technical scheme is as follows: and acquiring a data set of the advertisement commodity information and the user identity information from the database, and preprocessing. Preprocessing comprises data cleaning, missing value processing, feature standardization and the like so as to ensure the quality and consistency of data; key features are extracted from the user identity information, including user gender and age, and key features such as purchase preferences and user behavior data. These features will be input to model training; associating the advertisement commodity information with the user identity information through common attributes (such as commodity ID and user ID), and connecting the two data sets into a whole data set; the overall dataset is divided into a training set and a testing set. The training set is used for training the model and optimizing parameters, and the testing set is used for evaluating the performance and generalization capability of the model; the model is trained and parameter tuned by using the training set. The machine learning algorithm and the deep learning algorithm will learn the complex relationship between the advertised commodity information and the user identity information and build a model. The trained model is evaluated and validated using the test set. And calculating indexes such as accuracy, recall rate, F1 value and the like to evaluate the performance and generalization capability of the model. And predicting and classifying the new user data according to the trained model. Dividing the users into different gender and age groups according to the characteristics of the users, such as gender and age, and establishing user portraits; and finally, visually presenting the user image result through a visualization tool. Thus, the enterprise can intuitively know the characteristics of users with different sexes and ages to formulate more accurate marketing strategies and personalized services.
The technical scheme has the effects that: the quality and consistency of the data can be improved by preprocessing the acquired advertisement commodity information and user identity information data, so that the accuracy of subsequent model training and prediction is improved; extracting key features and main features from the user identity information, including gender, age, purchase preference, user behavior data and the like, wherein the features are helpful for better understanding and describing the characteristics and behaviors of the user; through associating advertisement commodity information and user identity information through common attributes, the two data sets can be connected into an integral data set, and basic data is provided for subsequent model training and classification; and training and parameter tuning are carried out on the whole data set through a machine learning algorithm and a deep learning algorithm, a model is built, and the performance and generalization capability of the model are evaluated through a test set. Thus, the accuracy and the reliability of the model can be improved; and predicting and classifying new user data according to the trained model, dividing the users into different gender and age groups, and establishing user portraits. Thus, the characteristics and the requirements of the user group can be better understood; through visual means to present the user image results visually, the enterprise can know the characteristics and behavior preferences of users of different sexes and ages more intuitively. This provides a basis for formulating accurate marketing strategies and personalized services. In summary, the technical scheme can help enterprises to deeply understand the user groups, and formulate more accurate marketing strategies and personalized services according to the characteristics of users of different sexes and age groups, so that the marketing effect and the user satisfaction of the enterprises are improved.
According to one embodiment of the invention, a commodity recommendation model is established according to the user portrait, and advertisement commodities are put in according to the commodity recommendation model; comprising the following steps:
calculating the correlation and similarity between the user and the advertisement commodity through a collaborative filtering algorithm, and matching the user image data with the advertisement commodity characteristics;
generating a candidate advertisement commodity list for each user according to the matching degree between the user portrait and the advertisement commodity; the candidate good list may be ranked according to predicted user preference for the advertised good to determine the most appropriate advertised good.
And according to the budget, the delivery channel and the marketing target of the advertisement, an advertisement delivery mechanism and a strategy are formulated, and the budget in the next advertisement delivery period is distributed. Which is a kind ofIn, budgeting for the next advertising periodAnd (3) distributing, wherein the calculation formula is as follows:
；
wherein,representing advertisement campaign +.>In->Time period of individual advertisement delivery period->The number of advertisement presentation opportunities received satisfying the rules of delivery, G representing the total budget of advertisement campaign a set by the advertiser during a complete advertisement delivery period,/for a complete advertisement delivery period>And representing the number of time periods in the advertisement delivery period, wherein V is a time window set by the demand side platform, and V is more than or equal to 1, and data used for limiting the recently occurring V advertisement delivery periods is used as a basis for prediction and statistics.
The working principle of the technical scheme is as follows: and calculating the correlation and similarity between the user and the advertisement commodity through a collaborative filtering algorithm to recommend the commodity, and matching the user portrait data with the characteristics of the advertisement commodity, for example, comparing and matching the characteristics of the user such as gender, age, purchase preference and the like with the attribute of the advertisement commodity so as to judge the preference degree of the user on the advertisement commodity. Generating a candidate advertisement commodity list: and generating a candidate advertisement commodity list for each user according to the matching degree. The advertisement goods in the list are ordered according to the predicted preference degree of the goods by the user so as to determine the most suitable advertisement goods. Advertisement delivery mechanism and strategy: and setting an advertisement delivery mechanism and a strategy according to the budget, the delivery channel and the marketing target of the advertisement. Wherein, when the advertisement delivery mechanism and strategy are prepared, the following aspects need to be considered:
budget allocation: the amount of delivery for each advertising channel or media is determined based on the budget of the advertising campaign. This may be assessed and analyzed by factors such as audience size, coverage, advertisement pricing, etc. for different channels. For example: the budget is 10 ten thousand yuan people's coins. A budget of 40% is allocated to internet advertisements, 30% to television advertisements, 20% to radio advertisements, 10% to outdoor advertisements.
The delivery period: and determining the optimal delivery time period according to the data analysis of the behavior habits and the advertising effects of the target audience. For example, if the target user prefers to browse shopping websites at night, the exposure of advertisements may be increased at night. For example, through analysis of user behavior data, it was found that 8 to 10 pm is a period of time when the target audience is using the smart device. Thus, it is decided to increase the frequency and exposure of advertisement delivery at 8 to 10 pm.
Advertisement channel selection: and selecting a proper advertisement channel according to the characteristics of the target audience and the advertisement target. Different channels have different characteristics and audience groups. For example, young people may be reached through a social media platform, while a wider population of users may be covered by television advertising. For example, considering that the target is very young, advertisements are selected to be placed on social media platforms (e.g., micro-letter, micro-blog) and the advertisement content is pushed in cooperation with some scientific-type blogs. In addition, it is also decided to place advertisements in the technological programming of television stations to cover a wider population of users.
Positioning and orientation: according to the characteristics of the user portrait and the advertisement commodity, accurate positioning and orientation are performed. And displaying the advertisement to the user meeting the target audience condition by utilizing the targeting function provided by the advertisement platform. For example, advertisements are targeted to male users between 18 and 35 years of age based on user portraits and advertising merchandise characteristics, with users purchasing electronic products and paying attention to technological information interests. In addition, advertisements are also targeted to users in one or two cities according to the geographic location of the users.
The technical scheme has the effects that: by establishing user portraits and commodity recommendation models, personalized advertisement commodity recommendation can be provided for each user according to the interests and characteristics of the user. Therefore, the interest and purchase willingness of the user to the advertisement can be increased, and the conversion rate of the advertisement is improved; by calculating the correlation and similarity between the user and the advertisement commodity and matching the user portrait and the advertisement commodity characteristics, the advertisement commodity related to the user requirement can be more accurately recommended and put in. Thus, the pertinence and the accuracy of the advertisement can be improved, and the noise of advertisement pushing and the dislike degree of a user are reduced; by generating a candidate advertisement commodity list for each user and sorting the commodities according to the preference degree of the users, the most suitable advertisement commodities can be selected for delivery. Thus, the click rate, conversion rate and return rate of the advertisement can be improved, and the effect and commercial value of advertisement putting are effectively improved; personalized advertisement recommendation can provide commodity information which better meets the interests and demands of users, and better shopping experience is provided for the users. The user sees more goods which accord with the preference of the user in the advertisement, so that the information overload and advertisement interference can be reduced, and the acceptance and acceptance of the user to the advertisement can be enhanced; and setting an advertisement delivery mechanism and a strategy according to the budget, the delivery channel and the marketing target of the advertisement. Advertisement resources can be reasonably distributed, proper delivery channels and opportunities can be selected, and the advertisement delivery effect and cost effectiveness can be improved. Through the formula, more accurate advertisement delivery can be realized, the exposure and conversion effects of advertisements are improved, and meanwhile, the waste of budget is reduced, so that the marketing target of advertisers is achieved. Meanwhile, through prediction and statistics according to the data of the latest V periods, the change trend of the user behavior and the advertising effect can be reflected more accurately, and therefore budget allocation is adjusted. The advertiser can make decisions in time, and the flexibility and the efficiency of advertisement delivery are improved; the allocation mechanism in the formula assigns the budget weighted according to the number of advertisement presentation opportunities so that the budget is more prone to presentation opportunities for more periods. The method can maximize the exposure of the advertisement, improve the click rate and conversion rate of the advertisement and achieve better advertisement effect; male (Male) Of the formula (I)The number of advertisement presentation opportunities satisfying the delivery rule is reflected. By counting and predicting opportunities meeting the delivery rules, the delivery quality and accuracy of advertisements can be better controlled, and inefficient delivery and budget waste are avoided.
According to one embodiment of the invention, the advertisement putting effect is monitored in real time, and the recommendation model is continuously optimized according to the monitoring effect; comprising the following steps:
determining key indexes for evaluating advertisement putting effect; the key indexes comprise click rate, conversion rate and return on investment rate;
collecting related data in the advertisement putting process in real time through a data monitoring module, wherein the related data comprises advertisement exposure rate, click quantity and conversion events;
based on the collected data, analyzing the data through a deep learning algorithm to obtain advertisement putting condition information and generating a monitoring report;
according to the monitoring result, adjusting and optimizing the selection and processing mode of the advertisement commodity characteristics, adjusting the parameters of the recommendation model, and evaluating and trying different recommendation algorithms or improving the existing algorithm;
comparing different optimization strategies through an A/B test method; randomly dividing a part of users into an experiment group and a control group, respectively applying different models or strategies to carry out advertisement delivery, and evaluating the effect of optimizing the strategy by comparing index changes of the two groups of data;
Establishing a continuous monitoring mechanism, periodically checking and evaluating advertisement putting effect, and performing iterative optimization according to feedback data; and according to the changes of different periods and requirements, timely adjusting a recommendation model and an advertisement putting strategy.
The working principle of the technical scheme is as follows: determining key indexes for evaluating advertisement putting effects, such as Click Through Rate (CTR), conversion rate (CVR) and Return On Investment (ROI), according to the targets and requirements of advertisement putting; these metrics can quantify the effectiveness and commercial value of the advertisement for evaluating and comparing the effectiveness of different optimization strategies; in the advertisement putting process, relevant data including advertisement exposure rate, click quantity, conversion event and the like are collected in real time through a data monitoring module. The data can provide real-time condition and effect information of advertisement delivery, and provide basis for subsequent analysis and optimization; analyzing and processing the collected data by applying a deep learning algorithm to obtain detailed information of advertisement putting conditions; by analyzing the relationship between the user behavior and the advertisement characteristics, potential rules and trends are found, and a monitoring report is generated for reference; according to the monitoring result, the parameters of the recommendation model are adjusted according to the selection and processing mode of the advertisement commodity characteristics, and different recommendation algorithms are evaluated and tried or the existing algorithm is improved. Training and optimizing the model by utilizing the historical data and the real-time data, and improving the accuracy and effect of advertisement recommendation; and (3) randomly dividing a part of users into an experiment group and a control group by an A/B test method, and respectively applying different models or strategies to carry out advertisement delivery. And evaluating the effect of the optimization strategy by comparing index changes of the two groups of data. Selecting an optimal scheme according to a test result, and performing subsequent optimization and adjustment; and establishing a continuous monitoring mechanism, periodically checking and evaluating the advertisement putting effect, and performing iterative optimization according to the feedback data. The recommendation model and advertisement placement strategies continue to be improved over time and demand to maintain effectiveness and competitiveness of advertisement placement.
The technical scheme has the effects that: by collecting and analyzing the data in the advertisement putting process in real time, indexes such as the exposure rate, the click quantity and the conversion event of the advertisement can be known in time, so that optimization measures can be taken in time; the effect of the advertisement can be objectively evaluated by determining key indexes such as click rate, conversion rate, return on investment and the like, and the key indexes are used as the basis of optimization; the advertisement putting data is analyzed by utilizing a deep learning algorithm, so that the behavior mode and the preference of a user can be identified, the advertisement is accurately put in a targeted mode, and the conversion rate and the return on investment are improved; based on the collected data, a monitoring report is generated, detailed advertisement putting conditions and effect analysis are provided for advertisers, the advertisers are helped to know advertisement effects and optimize advertisement strategies; according to the monitoring result, the selection and processing modes of the commercial features of the advertisement can be adjusted and optimized, and the attraction and click rate of the advertisement are improved; according to the monitoring result, evaluating and attempting different recommendation algorithms or improving the existing algorithm, and adjusting parameters of a recommendation model to improve the advertising effect; comparing the effects of different optimization strategies through an A/B test method, helping to determine the optimal advertisement delivery scheme and providing experimental data support decisions; and establishing a continuous monitoring mechanism, periodically checking and evaluating the advertisement putting effect, and performing iterative optimization according to the feedback data to keep the competitiveness and effect of advertisement putting.
In one embodiment of the present invention, as shown in FIG. 2, an advertisement delivery system based on user big data analysis, the system comprising:
and a data acquisition module: acquiring user information and advertisement commodity information from a background database in real time; the user information comprises user IP information, geographic position positioning information and user identity information; the advertisement commodity information comprises commodity type information, commodity sales information, commodity browsing information and commodity searching information;
the regional information determining module: acquiring advertisement commodity information in an area where a user is located according to the user IP information and the geographic position positioning information, and acquiring user identity information corresponding to the advertisement commodity according to the user behavior data;
a user portrait creation module: analyzing the advertisement commodity information and the user identity information by combining a machine learning algorithm with a deep learning algorithm, and establishing user portraits of users with different sexes and different ages in the area according to analysis results;
an advertisement commodity recommending module: establishing a commodity recommendation model according to the user portrait, and throwing advertisement commodities according to the commodity recommendation model;
model optimization module: and monitoring the advertisement putting effect in real time, and continuously optimizing the recommendation model according to the monitoring effect.
The working principle of the technical scheme is as follows: : acquiring IP information, geographic position positioning information and user identity information of a user, and type information, sales information, browsing information and search information of advertisement commodities in real time through a background database; determining the area where the user is located according to the IP information and the geographic position positioning information of the user, and acquiring advertisement commodity information in the area from a database; acquiring the association information of the user and the advertisement commodity, namely the user identity information corresponding to the advertisement commodity according to the behavior data of the user, such as browsing records, searching records and the like; analyzing the advertisement commodity information and the user identity information by using a machine learning algorithm and a deep learning algorithm to know interests, preferences and consumption behaviors of users of different sexes and different age groups, and establishing user portraits; according to the user portrait, a commodity recommendation model is established, and the model can recommend advertisement commodities according to the characteristics and the historical behaviors of the user; and putting the advertisement commodity according to the commodity recommendation model. Through accurate positioning and personalized recommendation, the advertisements are put into a target user group, and the conversion rate and the user participation of the advertisements are improved; and monitoring the advertisement putting effect in real time, wherein the advertisement putting effect comprises indexes such as click rate, conversion rate and the like. And carrying out data analysis according to the monitoring result, evaluating the effect of the advertisement, and continuously optimizing the recommendation model.
The technical scheme has the effects that: the user information and the advertisement commodity information are obtained in real time through the background database, so that timeliness and accuracy of data can be ensured, and a basis is provided for subsequent analysis and recommendation; through the user IP information and the geographic position positioning information, the advertisement commodity can be accurately put into the users in the corresponding areas, and the coverage rate and the accuracy of the advertisement are improved; the advertisement commodity information and the user identity information are analyzed by combining a deep learning algorithm, so that the behavior mode and preference of the user can be mined, and the user needs and interests can be further known; through analysis results, user portraits of users with different sexes and different ages can be established, advertisers are helped to better know target user groups, and accordingly advertisement putting strategies are optimized; according to the user portraits, a commodity recommendation model is established, advertisement commodities suitable for users can be accurately recommended according to interests and preferences of the users, and the click rate and conversion rate of advertisements are improved; problems and optimization space can be found in time by monitoring advertisement putting effect in real time, and advertisement effect and ROI (return on investment) are improved; and continuously optimizing the recommendation model according to the monitoring effect, adjusting the recommendation algorithm and parameters, and improving the effect and the accuracy of advertisement delivery.
An embodiment of the invention provides an electronic device, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the memory, wherein the processor executes the program to realize the advertisement delivery method based on the big data analysis of the user.
In one embodiment of the invention, a non-transitory computer readable storage medium has stored thereon a computer program that is executed by a processor to implement the method of advertising based on user big data analysis as described in any of the above.
It will be apparent to those skilled in the art that various modifications and variations can be made to the present invention without departing from the spirit or scope of the invention. Thus, it is intended that the present invention also include such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.
Claims (10)
1. An advertisement delivery method based on user big data analysis, characterized in that the method comprises the following steps:
acquiring user information and advertisement commodity information from a background database in real time; the user information comprises user IP information, geographic position positioning information and user identity information; the advertisement commodity information comprises commodity type information, commodity sales information, commodity browsing information and commodity searching information;
Acquiring advertisement commodity information in an area where a user is located according to the user IP information and the geographic position positioning information, and acquiring user identity information corresponding to the advertisement commodity according to user behavior data;
analyzing the advertisement commodity information and the user identity information by combining a machine learning algorithm with a deep learning algorithm, and establishing user portraits of users with different sexes and different ages in the area according to analysis results;
establishing a commodity recommendation model according to the user portrait, and throwing advertisement commodities according to the commodity recommendation model;
and monitoring the advertisement putting effect in real time, and continuously optimizing the recommendation model according to the monitoring effect.
2. The advertising method based on user big data analysis according to claim 1, wherein the user information and the advertisement commodity information are obtained from a background database in real time; the user information comprises user IP information, geographic position positioning information and user identity information; the advertisement commodity information comprises commodity type information, commodity sales information, commodity browsing information and commodity searching information; comprising the following steps:
the system is connected with a background database through an API interface, and monitors the user information and the data change of the advertisement commodity information table through a monitoring mechanism provided by the background database;
Triggering corresponding events when a data change is detected, wherein the data change comprises insertion, update, modification and deletion;
the data change status information in the database is published to a message queue in the form of a message through Kafka;
processing and analyzing each received message in real time by using a machine learning algorithm;
updating relevant user information and advertisement commodity information in real time according to the data processing result; and storing the results of updating the relevant user information and the advertisement commodity information in real time into a database.
3. The advertising method based on user big data analysis according to claim 2, wherein each received message is processed and analyzed in real time by using a machine learning algorithm; comprising the following steps:
dividing the received message data set into small data blocks according to a certain rule, and distributing the data blocks to available computing nodes;
each Map task acquires a data block, and for each message, the Map module converts the data block into a key value pair form;
the Map module outputs the generated intermediate result to the distributed file system in the form of key value pairs;
The buffer module partitions and sorts the intermediate results output by the Map stage according to keys, and sends the data of the same key to the same Reduce task node;
each Reduce module receives partitioned intermediate result data from the Shuffle module;
the Reduce module processes each received key value pair, wherein the processing comprises aggregation, screening and statistics, and a final output result is obtained;
and the Reduce module outputs the processed result to the database.
4. The advertising method based on user big data analysis according to claim 1, wherein the user identity information includes user registration information, purchase preference, and user behavior data; the user registration information comprises user gender, age, user name and mobile phone number; the purchase preference comprises a preference class, a brand preference, a price sensitivity, a special requirement media influence and a regional culture difference; the user behavior data includes user browsing behavior, clicking behavior, user purchase history, and searching behavior.
5. The advertisement putting method based on user big data analysis according to claim 1 or 4, wherein the advertisement commodity information and the user identity information are analyzed by combining a machine learning algorithm with a deep learning algorithm, and user figures of users with different sexes and different ages in the area are established according to analysis results; comprising the following steps:
Acquiring a data set of advertisement commodity information and user identity information through a database; and preprocessing the data set;
extracting features from the user identity information, wherein the features comprise key features and main features, the key features comprise user gender and age, and the main features comprise purchase preference and user behavior data;
associating the advertisement commodity information with the user identity information, and connecting the two data sets through common attributes to form an integral data set;
dividing the integral data set into a training set and a testing set;
training and parameter tuning are carried out on the model through a training set, and the trained model is evaluated and verified through a testing set;
according to the trained model, predicting and classifying new user data, dividing the users into different age groups with different sexes through the characteristics, and establishing user portraits;
and visually presenting the user portrait result through a visual tool.
6. The advertising method based on user big data analysis according to claim 1, wherein the commodity recommendation model is established according to the user portrait, and advertising commodities are delivered according to the commodity recommendation model; comprising the following steps:
Calculating the correlation and similarity between the user and the advertisement commodity through a collaborative filtering algorithm, and matching the user image data with the advertisement commodity characteristics;
generating a candidate advertisement commodity list for each user according to the matching degree between the user portrait and the advertisement commodity;
and setting an advertisement delivery mechanism and a strategy according to the budget, the delivery channel and the marketing target of the advertisement.
7. The advertising method based on user big data analysis according to claim 1, wherein the advertising effect is monitored in real time, and the recommendation model is continuously optimized according to the monitored effect; comprising the following steps:
determining key indexes for evaluating advertisement putting effect; the key indexes comprise click rate, conversion rate and return on investment rate;
collecting related data in the advertisement putting process in real time through a data monitoring module, wherein the related data comprises advertisement exposure rate, click quantity and conversion events;
based on the collected data, analyzing the data through a deep learning algorithm to obtain advertisement putting condition information and generating a monitoring report;
according to the monitoring result, adjusting and optimizing the selection and processing mode of the advertisement commodity characteristics, adjusting the parameters of the recommendation model, and evaluating and trying different recommendation algorithms or improving the existing algorithm;
Comparing different optimization strategies through an A/B test method; randomly dividing a part of users into an experiment group and a control group, respectively applying different models or strategies to carry out advertisement delivery, and evaluating the effect of optimizing the strategy by comparing index changes of the two groups of data;
establishing a continuous monitoring mechanism, periodically checking and evaluating advertisement putting effect, and performing iterative optimization according to feedback data; and according to the changes of different periods and requirements, timely adjusting a recommendation model and an advertisement putting strategy.
8. An advertising system based on user big data analysis, the system comprising:
and a data acquisition module: acquiring user information and advertisement commodity information from a background database in real time; the user information comprises user IP information, geographic position positioning information and user identity information; the advertisement commodity information comprises commodity type information, commodity sales information, commodity browsing information and commodity searching information;
the regional information determining module: acquiring advertisement commodity information in an area where a user is located according to the user IP information and the geographic position positioning information, and acquiring user identity information corresponding to the advertisement commodity according to user behavior data;
A user portrait creation module: analyzing the advertisement commodity information and the user identity information by combining a machine learning algorithm with a deep learning algorithm, and establishing user portraits of users with different sexes and different ages in the area according to analysis results;
an advertisement commodity recommending module: establishing a commodity recommendation model according to the user portrait, and throwing advertisement commodities according to the commodity recommendation model;
model optimization module: and monitoring the advertisement putting effect in real time, and continuously optimizing the recommendation model according to the monitoring effect.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the memory, the processor executing the program to implement the user big data analysis based advertisement delivery method of any of claims 1-7.
10. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the program is executed by a processor to implement the user big data analysis based advertisement delivery method of any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311852190.5A CN117808535A (en) | 2023-12-29 | 2023-12-29 | Advertisement putting method and system based on user big data analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311852190.5A CN117808535A (en) | 2023-12-29 | 2023-12-29 | Advertisement putting method and system based on user big data analysis |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117808535A true CN117808535A (en) | 2024-04-02 |
Family
ID=90426437
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311852190.5A Pending CN117808535A (en) | 2023-12-29 | 2023-12-29 | Advertisement putting method and system based on user big data analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117808535A (en) |
-
2023
- 2023-12-29 CN CN202311852190.5A patent/CN117808535A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| France et al. | Marketing analytics: Methods, practice, implementation, and links to other fields | |
| Gensler et al. | Listen to your customers: Insights into brand image using online consumer-generated product reviews | |
| CN108476334B (en) | Cross-screen optimization of advertisement placement | |
| Fan et al. | Demystifying big data analytics for business intelligence through the lens of marketing mix | |
| JP5662446B2 (en) | A learning system for using competitive evaluation models for real-time advertising bidding | |
| US8543518B2 (en) | Deducing shadow user profiles for ad campaigns | |
| US20110082824A1 (en) | Method for selecting an optimal classification protocol for classifying one or more targets | |
| US20100205024A1 (en) | System and method for applying in-depth data mining tools for participating websites | |
| Dam et al. | Marketing intelligence from data mining perspective: A literature review | |
| CN115860833A (en) | Television advertisement slot targeting based on consumer online behavior | |
| CN103971265A (en) | User Terminal And Method And System For Providing Advertisement | |
| Müllensiefen et al. | Using clustering of rankings to explain brand preferences with personality and socio-demographic variables | |
| Zheng et al. | A scalable purchase intention prediction system using extreme gradient boosting machines with browsing content entropy | |
| Ziemba et al. | Dynamic decision support in the internet marketing management | |
| Zhan et al. | Identifying market structure to monitor product competition using a consumer-behavior-based intelligence model | |
| Zhang et al. | Mining target users for mobile advertising based on telecom big data | |
| KR102238438B1 (en) | System for providing commercial product transaction service using price standardization | |
| Allaymoun et al. | Business intelligence model to analyze social network advertising | |
| JP2020166785A (en) | Television program evaluation system | |
| Boonjing et al. | Data mining for customers' positive reaction to advertising in social media | |
| US20150142782A1 (en) | Method for associating metadata with images | |
| CN117035826A (en) | Digital marketing method and marketing platform | |
| Zhang et al. | Discovering consumers’ purchase intentions based on mobile search behaviors | |
| CN117808535A (en) | Advertisement putting method and system based on user big data analysis | |
| Lin | Big Data Analysis of Personalized Recommendation in E-Commerce |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |