CN120146893A

CN120146893A - Marketing scenario intelligent training management method, device and storage medium

Info

Publication number: CN120146893A
Application number: CN202510257667.8A
Authority: CN
Inventors: 吴正
Original assignee: Beijing Zhenghe Siqi Data Technology Co ltd
Current assignee: Beijing Zhenghe Siqi Data Technology Co ltd
Priority date: 2025-03-05
Filing date: 2025-03-05
Publication date: 2025-06-13
Anticipated expiration: 2045-03-05
Also published as: CN120146893B

Abstract

The present invention provides a marketing scenario intelligent training management method, device and storage medium. The present invention relates to the technical field of marketing scenarios. The consumption scenario of the user is determined according to the consumption portrait and the corresponding consumption parameters, and the corresponding marketing scenario is determined according to the consumption scenario of the user, the residence of the user and the corresponding store distribution map, thereby ensuring the accuracy of the marketing scenario; the consumption route of the user is determined based on the current location of the user, the corresponding demand information and the current time, and the corresponding marketing information is matched according to multiple consumption nodes in the user's consumption route and the marketing scenario, and each marketing information is presented to the user in turn. Therefore, a training set is determined according to the content of each marketing information, the level coefficient of each marketing information and the consumption portrait of the user in the city, and the intelligent training of the marketing scenario is triggered according to the training set and the marketing scenario, thereby ensuring the applicability of the marketing scenario in various cities.

Description

Marketing scene intelligent training management method, device and storage medium

Technical Field

The invention relates to the technical field of marketing scenes, in particular to an intelligent training management method, equipment and storage medium for marketing scenes.

Background

Along with the development of science and technology, the marketing scene is gradually applied to life and is suitable for consumption of users in various cities in various popularization levels, at this time, residence places where the users are located and corresponding store distribution diagrams are introduced, interaction is carried out on the residence places where the users are located and the corresponding store distribution diagrams, and corresponding marketing scenes are output, however, the marketing scenes are not compatible with the consumption scenes of the users, the accuracy of the marketing scenes is affected, and the applicability of the marketing scenes in various cities is reduced.

Disclosure of Invention

Based on the foregoing, it is necessary to provide a marketing scenario intelligent training management method, device and storage medium for solving the above technical problems.

A marketing scene intelligent training management method comprises the steps of collecting consumption data sets of users in different cities, determining consumption portraits of the users in the cities according to the consumption data sets, life information of the users and the corresponding cities, determining consumption parameters of the users in the cities based on income levels of the users, corresponding consumption time and favorites of the users in the cities, determining consumption scenes of the users according to the consumption portraits and the corresponding consumption parameters, determining corresponding marketing scenes according to the consumption scenes of the users, residence places of the users and corresponding store distribution diagrams, determining consumption routes of the users in the marketing scenes according to the current positions of the users, corresponding demand information and the current time, matching corresponding marketing information according to a plurality of consumption nodes in the consumption routes of the users and the marketing scenes, sequentially displaying each marketing information to the users, determining grade coefficients of each marketing information according to content of each marketing information, grade coefficients of each marketing information and the number of clicks of the users in the cities, determining a training set according to the content of each marketing information and the consumption training set of the user in the cities, and the intelligent training set according to the triggering of the marketing scenes.

The marketing scene intelligent training management equipment is applied to the marketing scene intelligent training management method, and comprises the following components:

The acquisition module is used for acquiring consumption data sets of users in different cities;

The consumption portrait module is used for determining the consumption portrait of the user in the city according to the consumption data set, the life information of the user and the corresponding city;

The consumption parameter module is used for determining the consumption parameters of the user in the city based on the income level of the user, the corresponding consumption time and the preference of the user in the city;

the marketing scene module is used for determining a consumption scene of the user according to the consumption portrait and the corresponding consumption parameters, and determining a corresponding marketing scene according to the consumption scene of the user, the residence where the user is located and the corresponding store distribution diagram;

The marketing information module is used for determining a consumption route of the user based on the current position of the user, corresponding demand information and current time in the marketing scene, and matching the corresponding marketing information according to a plurality of consumption nodes in the consumption route of the user and the marketing scene, wherein each marketing information is sequentially presented to the user;

And the intelligent training module is used for determining the grade coefficient of each marketing message according to each marketing message, the viewing time of the user and the clicking times of the user, determining a training set according to the content of each marketing message, the grade coefficient of each marketing message and the consumption portrait of the user in the city, and triggering the intelligent training of the marketing scene according to the training set and the marketing scene.

A storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the marketing scenario intelligent training management method described above.

The marketing scene intelligent training management method, the marketing scene intelligent training management device and the storage medium collect consumption data sets of users in different cities, determine consumption figures of the users in the cities according to the consumption data sets, life information of the users and the corresponding cities, determine consumption parameters of the users in the cities based on income levels of the users, corresponding consumption time and favorites of the users in the cities, determine consumption scenes of the users according to the consumption figures and the corresponding consumption parameters, and determine corresponding marketing scenes according to the consumption scenes of the users, residence places of the users and corresponding store distribution diagrams, so that the marketing scene intelligent training management method, the marketing scene intelligent training management device and the marketing scene intelligent training management device are compatible with the overall consideration of the consumption scenes of the users, residence places of the users and the corresponding store distribution diagrams, and ensure the accuracy of the marketing scenes.

Further, in the marketing scene, the consumption route of the user is determined based on the current position of the user, the corresponding demand information and the current time, and corresponding marketing information is matched according to a plurality of consumption nodes in the consumption route of the user and the marketing scene, and each marketing information is sequentially presented to the user, so that marketing information in each aspect is presented according to the consumption route, and subsequent management and control of each marketing information are facilitated.

Therefore, the grade coefficient of each marketing message is determined according to each marketing message, the viewing time of the user and the clicking times of the user, the training set is determined according to the content of each marketing message, the grade coefficient of each marketing message and the consumption portrait of the user in the city, and the intelligent training of the marketing scene is triggered according to the training set and the marketing scene, so that the consideration of the user in each city is compatible, and the applicability of the marketing scene in each city is ensured.

Drawings

FIG. 1 is a schematic illustration of an application scenario of a marketing scenario intelligent training management method in one embodiment;

FIG. 2 is a flow diagram of a marketing scenario intelligent training management method in one embodiment;

FIG. 3 is a block diagram of a marketing scenario intelligent training management device, in one embodiment;

FIG. 4 is an internal block diagram of a marketing scenario intelligent training management device, under one embodiment.

Detailed Description

The present application will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present application more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the application.

Example 1

The marketing scene intelligent training management method provided by the application can be applied to an application environment shown in figure 1. Wherein the terminal 102 communicates with the server 104 via a network. The terminal 102 may be, but not limited to, various personal computers, servers, and marketing scenarios, and the server 104 may be implemented as a stand-alone server or a server cluster composed of a plurality of servers.

Example two

In this embodiment, please refer to fig. 2 to 4, a marketing scene intelligent training management method is applied to a marketing scene intelligent training management scene, where the marketing scene intelligent training management method includes:

step S11, collecting consumption data sets of users in different cities;

step S12, determining the consumption portrait of the user in the city according to the consumption data set, the life information of the user and the corresponding city;

step S13, determining consumption parameters of the user in the city based on the income level of the user, the corresponding consumption time and the preference of the user in the city;

Step S14, determining a consumption scene of the user according to the consumption portrait and the corresponding consumption parameters, and determining a corresponding marketing scene according to the consumption scene of the user, the residence place of the user and the corresponding store distribution diagram;

Step S15, in the marketing scene, determining a consumption route of the user based on the current position of the user, corresponding demand information and current time, and matching corresponding marketing information according to a plurality of consumption nodes in the consumption route of the user and the marketing scene, wherein each marketing information is sequentially presented to the user;

And S16, determining the grade coefficient of each marketing message according to each marketing message, the viewing time of the user and the clicking times of the user, determining a training set according to the content of each marketing message, the grade coefficient of each marketing message and the consumption portrait of the user in the city, and triggering intelligent training of the marketing scene according to the training set and the marketing scene.

In step S11, collecting consumption data sets of users in different cities;

in the implementation process of the invention, the specific steps can be as follows:

S111, marking a payment platform of a user;

s112, exporting past payment data of the user according to a payment platform of the user;

s113, determining a data set of each city according to the past payment data of the user and the matching of the corresponding payment positions;

And S114, determining consumption data of the user in different cities based on the data set of each city and the consumption type of the user.

In the embodiment of the application, the payment platform of the user is marked, the past payment data of the user is exported according to the payment platform of the user, and the user-based payment platform is further controlled, so that the export of the past payment data of the user is ensured, and the subsequent control of the past payment data of the user is facilitated.

In this case, the payment platform of the user is marked to distinguish which payment platform is used by the user (such as payment treasures, weChat payments, unionpay payments, etc.) so as to facilitate the subsequent data export and processing, typically, the payment platform requires the user to perform identity verification and account binding when registering or using the service of the user, through which the payment platform can acquire the payment information of the user and store the payment information in a database, and in order to mark the payment platform of the user, a specific identifier or tag can be allocated to the user in the database, wherein the identifier is associated with the payment platform of the user.

Further, the present payment data of the user is derived to analyze the payment behavior, consumption habit, etc. of the user, provide basis for subsequent decision or optimization, the payment platform marked in step S111 can be connected to the corresponding payment platform database to inquire and derive the payment data of the user, the payment data generally includes transaction time, transaction amount, transaction type (such as transfer, payment, refund, etc.), transaction counterpart information, etc., and optionally, the payment data is located to the corresponding payment platform database according to the user ID and the payment platform name, all payment records of the user are inquired in the database, and the inquiry result is derived to a designated file format (such as Excel, CSV, PDF, etc.) for subsequent analysis and processing.

Specifically, it is assumed that there is a user a who uses both the payment treasures and the WeChat payments, when the user a registers or uses the services of the two paymate platforms, the payment treasures and the WeChat payments allocate a unique user ID to the user a in respective databases and associate the ID with the payment information of the user, and in order to mark the paymate of the user a, a record may be added to the user a in a unified database, where the record includes the names of the user a, the user IDs of the two paymate platforms, and the corresponding paymate names (e.g. "paytreasures" and "WeChat pays").

It is assumed that we need to derive all payment data of the user a on the payment treasures and the WeChat payments, firstly, we connect to databases of the two payment platforms according to the payment platform names (i.e. "payment treasures" and "WeChat payments") marked by the user a in step S111, then, query all payment records of the user a in the databases, and finally, derive the query result as an Excel file, in which we can clearly see all transaction records of the user a on the payment treasures and the WeChat payments, including transaction time, transaction amount, transaction type, transaction counterpart information, etc.

The method comprises the steps of determining the data set of each city according to the past payment data of the user and the matching of the corresponding payment positions, determining the consumption data of the user in different cities based on the data set of each city and the consumption types of the user, and ensuring the accuracy of the consumption data of the user in different cities by being compatible with the data set of each city and the overall consideration of the consumption types of the user.

At this time, the payment data of the user is classified into each city through the payment data and the payment position information of the user, so as to obtain a data set of each city, firstly, the past payment data of the user needs to be integrated, which generally comprises fields of transaction time, transaction amount, transaction type, transaction position and the like, then, the transaction position information in the payment data is analyzed, which generally relates to a geocoding technology, namely, an address is converted into longitude and latitude coordinates, or the address is matched with a preset city list through an address matching technology, and finally, the payment data is classified into the corresponding city according to the analyzed position information, so as to form a data set of each city.

The method comprises the steps of analyzing consumption habits and preferences of users in different cities through data sets of the cities and consumption type information of the users, firstly dividing the consumption types of the users according to dimensions such as types of commodities and types of services, then screening payment data of the corresponding consumption types from the data sets of the cities, and finally analyzing the screened payment data, wherein the analysis comprises the steps of calculating indexes such as consumption amount and consumption frequency, so that the consumption habits and preferences of the users in different cities are revealed.

In step S12, determining the consumption portrait of the user in the city according to the consumption data set, the life information of the user and the corresponding city;

s121, acquiring a consumption data set;

s122, acquiring life information of the user based on traversal of a life database of the user;

S123, associating the consumption data set, life information of the user and corresponding cities;

S124, determining a first consumption characteristic according to the consumption data set and life information of the user;

S125, determining a second consumption characteristic according to the consumption data set and the corresponding city;

and S126, determining the consumption portrait of the user in the city according to the first consumption feature, the second consumption feature and the stay time of the user in the city.

In the embodiment of the application, the consumption data set is acquired, the life information of the user is acquired based on the traversal of the life database of the user, the traversal of the life database of the user is realized, and the accuracy of the life information of the user is ensured.

At this time, consumption data of the user at different times and different places are collected, including but not limited to transaction amount, transaction time, transaction commodity/service types, transaction merchants and the like, optionally, purchase records of the user on a payment platform, including commodity names, purchase time, prices, payment modes and the like, payment records of the user, recording each transaction of the user, including transaction time, transaction amount, transaction objects and the like, consumption records of the user in a physical store, and the like, can be acquired through a member system, a POS machine and the like, and a consumption data set integrating various data sources is provided, so that enterprises can acquire comprehensive user consumption data quickly.

The method comprises the steps of interfacing API interfaces with third party service providers such as a payment platform and the payment platform, obtaining consumption data of a user in real time, capturing consumption information of the user from a public website, and purchasing an integrated consumption data set from the third party data provider.

The method comprises the steps of collecting life information such as basic information, hobbies and social behaviors of a user so as to further know consumption motivations and behavior modes of the user, at the moment, filling in personal information such as names, ages, sexes, professions and the like when the user registers on a platform, behavior records of the user on social media including posts, comments, praise, concerned people and the like, behavior records of the user on the platform such as browsing records, searching records and clicking records, optionally, accessing social media accounts or behavior logs through authorization of the user to obtain related information, designing questionnaires to collect the basic information and the hobbies of the user, and extracting useful life information from the user behavior logs by using a data mining technology.

Further, the first consumption characteristic is determined according to the consumption data set and the life information of the user, the second consumption characteristic is determined according to the consumption data set and the corresponding city, the multi-dimensional control of the consumption data set, the life information of the user and the corresponding city is compatible, and the accuracy of the first consumption characteristic and the second consumption characteristic is guaranteed.

At this time, the consumption data and life information of the users are associated with specific cities so as to analyze the consumption habits and life styles of the users in different cities, the consumption data sets and life information databases of the users are combined to ensure that the data of each user are complete, corresponding city labels are added for each user according to the consumption records or the geographical position information in the life information of the users, and the integrated data are cleaned to remove repeated, wrong or invalid data and ensure the accuracy and reliability of the data.

The method comprises the steps of analyzing consumption habits, preferences and demands of users to form first consumption characteristics, analyzing consumption data of the users, including consumption amount, consumption frequency, consumption type (such as catering, shopping and entertainment), consumption brands and the like, associating the consumption data of the users with life information of the users, analyzing influences of the life information on the consumption habits, and extracting consumption characteristics of the users such as consumption capacity, consumption preferences, brand loyalty and the like according to analysis results.

The method comprises the steps of analyzing consumption differences and characteristics of users in different cities to form second consumption characteristics, analyzing consumption data of the users in the different cities, including consumption amount, consumption type, consumption frequency and the like, comparing the consumption data of the users in the different cities to find out consumption trends and differences, and extracting the consumption characteristics of the users in specific cities, such as urban consumption hot spots, consumption habit differences and the like, according to analysis results.

Specifically, it is assumed that a certain payment platform wants to analyze consumption differences of users in different cities, firstly, the platform integrates a consumption data set of the users and a life information database of the users, then, corresponding city labels are added for each user according to receiving address information in purchase records of the users, for example, purchase records of the user A show that receiving addresses of the user A are mainly in Beijing city, so that the user A is added with the city label of Beijing, and after data cleaning, the payment platform obtains a complete data set containing user consumption data, life information and city labels.

By analyzing the consumption data of the user A, the payment platform finds that the consumption amount of the user A in shopping is high and mainly purchases high-end brands of clothing and electronic products, meanwhile, the payment platform knows that the user A is a 35-year-old staff and pays attention to personal images and quality life in combination with life information of the user A, and therefore the payment platform extracts first consumption characteristics of the user A, namely high consumption capacity, quality and high-end brands.

The payment platform further analyzes consumption data of the user A in Beijing and Shanghai, and through comparison, the user A mainly concentrates in a high-end shopping center and a brand exclusive store to purchase clothes, electronic products, luxury goods and the like, while in Shanghai, the user A prefers to visit fashion trend blocks and purchases fashion clothes and accessories, so that the payment platform extracts second consumption characteristics of the user A, namely, quality and brands are emphasized in Beijing, the consumption amount is higher, fashion and trend are emphasized in Shanghai, the consumption frequency is higher, and the consumption amount of a single pen is relatively lower.

Therefore, the consumption portrait of the user in the city is determined according to the first consumption feature, the second consumption feature and the stay time of the user in the city, the integral consideration of the first consumption feature, the second consumption feature and the stay time of the user in the city is compatible, the multi-dimensional control of the first consumption feature, the second consumption feature and the stay time of the user in the city is realized, and the accuracy of the consumption portrait of the user in the city is ensured.

At the moment, the first consumption characteristics (based on the personal life information and the general consumption habit characteristics of the user), the second consumption characteristics (the specific consumption habit of the user in different cities) and the stay time of the user in the city are integrated to comprehensively depict the consumption portrait of the user in the city, and the portrait is helpful for enterprises to deeply understand the consumption behaviors of the user in the city, so that a more accurate marketing strategy is formulated.

The method comprises the steps of integrating a first consumption feature, a second consumption feature and the stay time of a user in the city to form a comprehensive data set, distributing proper weight for each feature according to importance and relativity of the features, for example, for a user staying for a long time, the second consumption feature has higher weight, for a user staying for a short time, the stay time and the first consumption feature are more important, constructing a consumption image of the user in the city by using a data analysis tool or algorithm based on the integrated data set and weight distribution, wherein the consumption image comprises information such as consumption preference, consumption frequency, consumption amount, consumption hot spot and the like of the user, verifying the accuracy of the consumption image by comparing with actual consumption data, and optimizing according to requirements.

Specifically, suppose that a travel service platform wishes to construct a consumption figure for a city that user B is going to-a achievement, a first consumption characteristic of user B shows that user B is a 30 year old young loving food and culture with a high pursuit for a high quality life experience, a second consumption characteristic shows that user B tends to eat at a high-end restaurant, visit cultural attractions, purchase featured souvenirs, and prefer to try street snacks, visit fashion trend blocks while in the sea, and user B plans to stay at the achievement for 5 days.

Based on the information, the travel service platform builds a consumer representation of the user B in a achievement, wherein the consumer preference is that the user B is loved with delicious foods, especially Sichuan dishes and local characteristic snacks, the user B is interested in cultural sceneries and museums, high-quality life experiences such as high-end lodging and customized travel services are favored, the consumer frequency is that the user B is expected to taste the local delicious foods for multiple times due to long residence time (5 days), visit a plurality of cultural sceneries and purchase some characteristic souvenirs, the consumer amount is expected to be relatively high in view of pursuit of the high-quality life by the user B, especially in catering and lodging, the consumer hotspot is that the user B is expected to go to the famous delicious food street (such as in brocade, wide and narrow lane) of the achievement to taste the local snacks, the cultural sceneries such as in the city center or sceneries are visited Wu Houci, the pavilion, and the high-end hotel lodging is selected.

Through the consumer representation, the travel service platform can provide more personalized service recommendations for the user B, such as recommending local restaurants meeting the taste of the user B, customizing travel routes meeting the interests of the user B, and the like, so that the user satisfaction and the platform competitiveness are improved.

In another embodiment of the present application, an example of a consumer representation matching table is described, with the consumer representation matching table example:

In this consumer representation matching table, we determine a consumer representation of the city based on a first consumer characteristic of the user (e.g., food lovers, shopping owners, etc.), a second consumer characteristic (e.g., like to try local feature snacks, prefer high-end shopping malls, etc.), and a stay period of the user in the city (e.g., 3 days, 7 days, etc.).

In addition, the first consumption characteristic is 0.4 (because the first consumption characteristic reflects the long-term consumption habit of the user and has a larger influence on the consumption behavior of the user in the city), the second consumption characteristic is 0.3 (because the second consumption characteristic reflects the specific consumption habit of the user in the city and has a certain influence on the consumption behavior of the user in the city), the stay time is 0.3 (because the stay time determines the nature and time of the consumption of the user in the city), the first consumption characteristic of a certain user is assumed to be a 'food lover' (score is 8 and full score is 10), the second consumption characteristic is 'like to try local feature snack' (score is 7 and full score is 10), and the stay time is 5 days (score is 5 and full score is 10 and is scored according to the length of stay days).

The consumer representation of the user in the achievement is scored as 0.48+0.37+0.3 x 5 = 3.2 + 2.1 + 1.5 = 6.8, and according to the score, we can describe the consumer representation of the user as that the user will tend to taste local distinctive snacks during the achievement, enjoy the delight of the delicates, but will not consume too frequently or try too many different delicates due to the low score of the stay length and certain consumption characteristics.

In step S13, determining a consumption parameter of the user in the city based on the income level of the user, the corresponding consumption time and the preference of the user in the city;

S131, collecting income data of the user based on traversal of a payment platform of the user;

s132, determining the income level of the user according to the income data of the user, the job title of the user in the enterprise and the residence of the user;

S133, determining a plurality of consumption types and corresponding consumption frequencies according to traversal of consumption data of the user in the city;

s134, determining preference of a user in the city according to a plurality of consumption types and cities corresponding to the corresponding consumption frequencies;

s135, interacting the income level of the user, the corresponding consumption time and the preference of the user in the city;

and S136, determining consumption parameters of the user in the city according to the income level of the user, the corresponding consumption time and the favorite interaction of the user in the city.

The embodiment of the application collects the income data of the user based on the traversal of the payment platform of the user, determines the income level of the user according to the income data of the user, the job title of the user in the enterprise and the residence of the user, is compatible with the overall consideration of the income data of the user, the job title of the user in the enterprise and the residence of the user, and ensures the accuracy of the income level of the user.

At this point, the user's paymate is traversed, which involves examining all paymate used by the user, including but not limited to payroll, weChat payments, bank transfer applications, etc., which record all of the user's transaction activities, including revenue, extracting transaction records from these paymate, particularly records related to user revenue, including payroll revenue, bonuses, investment benefits, etc., and the extracted data needs to be consolidated to form a comprehensive view of user revenue, which helps analyze the user's revenue level and revenue source.

Based on the income data collected from the payment platform, the total income, the income source and the income stability of the user are analyzed, the income level of the user can be more accurately estimated by combining the job title of the user in the enterprise, different titles often correspond to different salary ranges, the living cost of the living place and the overall salary level are also important factors for determining the income level of the user, the living cost of the first-line city is higher, the salary level is correspondingly higher, and the second-line or third-line city is lower.

Assuming a user, mr. who uses Payment treasures and WeChat payments as the primary payment tools, by traversing both platforms we can collect Mr.'s following revenue data:

Payment treasures, fixed payroll per month, final annual prizes, investment benefits and the like, weChat payments, occasional part-time revenues, red package revenues and the like. After integrating these data, we can get the total income situation of Mr. who has fixed payroll 10000 yuan each month, final annual prize 30000 yuan, average investment income 2000 yuan each month, and random both part-time income and red package income.

Assuming that he is a project manager of a science and technology company, residing in Beijing, based on his income data and job title, we can perform the following analysis:

Revenue data analysis, namely, the total revenue of Mr. Zhang including fixed payroll, final annual rewards and investment benefits, average total monthly revenue of about 14000 yuan (fixed payroll+investment benefits, final annual rewards being distributed by year), job title consideration, namely, as project manager, the payroll level of Mr. Zhang is higher than that of ordinary staff, in Beijing such first line city, the payroll range of project manager is wider, but in combination with the total income situation, we can infer that he is at medium and upper income level, and living place factor, beijing as first line city, the living cost is higher, however, the total revenue of Mr. Zhang is enough to support his life in Beijing, and has certain saving capacity.

By combining the above factors, the method can determine that the income level of Mr. Zhang is moderate, and the analysis is helpful for enterprises to know the financial condition and the consumption capability of users more accurately, so as to formulate a more proper marketing strategy;

And determining the preference of the user in the city according to the cities corresponding to the consumption types and the corresponding consumption frequencies, and introducing the cities corresponding to the consumption types and the corresponding consumption frequencies to realize multi-dimensional control of the preference of the user in the city.

At this time, the user is comprehensively checked on all consumption data of the city, which includes transaction records of the user on various payment platforms (such as payment treasures and WeChat payments), bank cards and related online and offline shopping platforms, different consumption types such as catering, traffic, entertainment, shopping, housing (rentals or credits), education, medical treatment and the like are identified from the consumption data, and the consumption frequency of the user is calculated for each consumption type, which can be calculated according to the frequency of days, weeks, months or years, depending on the availability of the data and the purpose of analysis.

Based on the consumption types identified in the previous step and the calculated consumption frequency, the investment and preference of the user on various consumption is analyzed, the consumption preference of the user is further determined by combining the consumption environment and cultural characteristics of the city, for example, certain cities are known as food, the catering consumption of the user in the city is particularly frequent, other cities are known in rich shopping or entertainment activities, the preference of the user in the city is determined by combining the consumption types, the frequency and the city characteristics, the enterprise is helped to know the consumption tendency of the user, and the basis is provided for providing personalized services and products.

Specifically, assuming a user, mr. she has recently moved to the Shanghai for living, we have traversed all of her consumption data in the Shanghai in order to understand her consumption habits in the Shanghai.

The restaurant consumption is that Mr. spends 5 times per week in restaurants near the company on weekdays, different restaurants or spot takeouts are tried on weekends, the average frequency of the weekly restaurant consumption is 10 times, the traffic consumption is that she uses subways to and fro every day, the monthly traffic card recharge frequency is 4 times (once per week), the drip-strike is occasionally used, but the frequency is low, the entertainment consumption is that Mr. spends 1-2 times per month in movie theatres, the frequency of attending friends, parties or social activities is 2-3 times per month, the shopping consumption is that her consumption on an online shopping platform is more frequent, records of purchasing daily necessities or clothes are available for several times per week, but the frequency of large-scale shopping (such as electronic products and furniture) is low.

In combination with her consumption type and frequency in Shanghai, and the consumption environment and cultural characteristics in Shanghai, we can conclude that:

The restaurant preference is that Mr. consumes the restaurant with higher frequency and tries the wish of different restaurants strongly, and in consideration of the abundance of the food culture in Shanghai, we can infer that she has a great interest in the food culture in Shanghai.

Entertainment preferences although Mr's movie and social activity frequency is not particularly high, there is a fixed entertainment consumption per month, indicating that she enjoys social and recreational activities, shanghai is an internationalized metropolitan providing a rich entertainment option and Mr's will try different ways of entertainment.

Shopping preference, namely that on-line shopping frequency of Mr. is high, which indicates that she likes a convenient shopping mode, and meanwhile, as the fashion Shanghai, various shopping places are in full view, and although the shopping frequency of her large amount is not high, the fashion trend and new product release are concerned.

The method comprises the steps of determining the user's consumption parameters in the city according to the user's income level, the corresponding consumption time and the user's preference in the city, realizing the user's income level, the corresponding consumption time and the user's preference in the city, ensuring the overall consideration of the user's income level, the corresponding consumption time and the user's preference in the city, and realizing the accurate control of the user's consumption parameters in the city.

In step S14, determining a consumption scene of the user according to the consumption portrait and the corresponding consumption parameters, and determining a corresponding marketing scene according to the consumption scene of the user, the residence where the user is located, and the corresponding store distribution diagram;

s141, acquiring the consumption portrait and corresponding consumption parameters;

s142, matching the corresponding scene weight with the consumption portrait and the corresponding consumption parameters;

S143, determining a consumption scene of the user according to the consumption portrait, the consumption parameters and the corresponding scene weights;

s144, determining the residence place of the user according to the residence data of the user;

S145, determining a corresponding store distribution diagram based on the residence place where the user is located, the surrounding drawings of the residence place and the store marks;

And S146, determining a corresponding marketing scene according to the consumption scene of the user, the residence place of the user and the corresponding store distribution diagram.

In the embodiment of the application, the consumption portrait and the corresponding consumption parameters are acquired, the corresponding scene weights are matched with the consumption portrait and the corresponding consumption parameters, the consumption scene of the user is determined according to the consumption portrait, the consumption parameters and the corresponding scene weights, the consumption portrait, the consumption parameters and the corresponding scene weights are introduced, the whole control is carried out on the consumption portrait, the consumption parameters and the corresponding scene weights, and the accuracy of the consumption scene of the user is ensured.

At this time, the consumer representation and consumer parameters of the user are deeply understood and matched with different consumer scenes, so that a weight is allocated to each scene, the weight reflects the consumer's nature or preference degree under different scenes, firstly, we need to review the consumer representation of the user, which generally comprises the age, sex, income level, occupation, hobbies, consumer habits and other information of the user, secondly, we read specific parameters related to consumer consumption, such as consumption frequency, average consumption amount, consumption hot spot (i.e. the most commonly consumed field or commodity of the user), consumption elasticity (i.e. the sensitivity of the consumption changing with the income level) and the like, and according to the consumer representation and consumption parameters of the user, we match the user with different consumer scenes, the scenes comprise online shopping, offline shopping (such as supermarket, mall, private store and the like), recreation (such as cinema, gymnasium, SPA and the like), travel and the like, finally, we allocate a weight for each matched consumer scene, the allocation of the weight can be based on the historical consumer consumption data, the consumption preference, the competition environment and the weight value is between the consumer's market trend, the weight value and the consumer's preference value is between 0 or the value of the consumer's nature of the value of the number is 0.

The method and the system are characterized in that the consumption portraits, the consumption parameters and the scene weights distributed before are combined to determine the most consumption scene of the user, so that the user can understand the consumption behaviors of the user in different situations more accurately, a basis is provided for the subsequent marketing strategy formulation, firstly, the consumption portraits, the consumption parameters, the scene weights and other information of the user are summarized, secondly, the weight value of each scene is analyzed to find out the scene with the highest weight, the scene is the most important consumption scene of the user, other factors such as time arrangement, geographical position limitation and the like of the user are considered while the main consumption scene is determined, and finally, the most consumption scene of the user is determined by comprehensively considering all the factors.

Specifically, it is assumed that a user is Mr. his consumption portrait shows that he is a middle-aged, high-income, outdoor activity and high-end shopping man, his consumption parameters include an average monthly consumption of 5000 yuan in the high-end market, an average monthly consumption of 2000 yuan in the online shopping platform, and low price sensitivity.

Based on this information we can match mr. Plums with the following consumption scenarios and assign corresponding weights:

The shopping weight of the high-end shopping mall is 0.6, the shopping weight of the on-line shopping mall is 0.3 because Mr. prune prefers shopping in the high-end shopping mall and the consumption amount under the scene is high, the shopping weight of the on-line shopping mall is lower than that of the on-line shopping mall although Mr. prune is also shopping on-line, the shopping weight of the outdoor product store is 0.1, and the shopping in the outdoor product store is occasionally carried out because Mr. prune prefers outdoor activities, but is not the main consumption scene.

Based on his consumption portraits, consumption parameters and scene weights, we can determine that his main consumption scene is a high-end market purchase because he consumes higher in average month at the high-end market and the weight value of the scene is 0.6, which is the highest weight in all scenes, meanwhile we also need to notice that although mr. Prune consumes on the online shopping platform and the weight value is 0.3, his online consumption amount is still lower than that of the high-end market, therefore, we can pay more attention to the scene of the high-end market when making marketing strategies, and personalized shopping experience and preferential activities are provided for mr. Prune.

The corresponding store distribution diagram is determined based on the residence place, the surrounding drawings of the residence place and the store marks, the residence place, the surrounding drawings of the residence place and the store marks of the user are introduced, and the residence place, the surrounding drawings of the residence place and the store marks of the user are integrally considered, so that the accuracy of the store distribution diagram is ensured.

At this time, the accurate residence place is determined according to residence data provided by the user, the data comprise specific information such as household addresses, postal codes, cities, areas or streets of the user, and by analyzing and processing the data, the approximate geographical position of the residence of the user can be clarified, which is important for subsequent market subdivision, marketing strategy establishment and store selection.

Firstly, collecting residence data of a user from various channels such as user registration information, purchase records or questionnaires, then cleaning the collected data, removing repeated, wrong or incomplete information, ensuring the accuracy and reliability of the data, then converting the residence data of the user into specific geographic position information such as longitude and latitude coordinates by using a Geographic Information System (GIS) technology, and finally determining the accurate residence places such as cities, areas and streets where the user is located according to geographic positioning results and administrative division data.

A distribution diagram of surrounding stores of a user is drawn according to the residence of the user, geographical drawings of the residence and marked store information, and the diagram helps us intuitively know the business environment of the region where the user is located, including key information such as the types, the numbers and the distribution positions of the stores, the relative distance between the stores and the residence of the user and the like.

Firstly, geographical drawings of a residence of a user are required to be collected, which generally comprises geographical information such as streets, buildings, public facilities and the like, meanwhile, marked store information such as names, types, addresses and the like of stores are required to be collected, then, the collected geographical drawings and the store information are integrated to ensure that the corresponding relation between the geographical drawings and the store information is accurate, then, GIS technology or professional mapping software is utilized to map the integrated data into a store distribution diagram, the store types, the quantity and the distribution positions of the periphery of the residence of the user are clearly displayed in the map, and finally, the distribution diagram is optimized and adjusted according to requirements, such as adding labels, labeling distances or providing other relevant information, so that the subsequent analysis and marketing requirements are better met.

Specifically, the house data of the king women of the user are assumed to comprise the house address of the user, namely' a certain district of the Santun street in the King Korea of Beijing, through data cleaning and geographic positioning, the residence of the king women can be determined to be the Santun street in the King Korea of Beijing, and the information is very important for subsequent analysis of the consumption habits of the king women, formulation of targeted marketing strategies and popularization of nearby stores.

After determining that the residence area of her is the Santun street in the Yankee area of Beijing, the geographic drawing and marked store information of the area are collected, by integrating the data, a distribution map of Wang Nvshi peripheral stores is drawn, the distribution map shows the distribution situation of various stores in the Santun street, such as a high-end market, a fashion brand store, a food restaurant, a coffee shop and the like, the business environment of the area where the king women are located can be intuitively known through the map, powerful support is provided for subsequent marketing strategy formulation, and for example, the nearby fashion brand or high-end market can be selected for directional popularization and marketing according to the consumption image and preference of the king women.

Therefore, the corresponding marketing scene is determined according to the consumption scene of the user, the residence place of the user and the corresponding store distribution diagram, the consumption scene of the user, the residence place of the user and the corresponding store distribution diagram are controlled in a multi-dimensional mode, and the accuracy of the marketing scene is guaranteed.

The consumption scene refers to a specific environment and situation of a consumer when the consumer purchases and uses a product or service, different consumption scenes can excite different demands of the consumer so as to influence the purchase decision of the consumer, daily life habits, interests and potential demands of a target user group are analyzed, the characteristics of the product or service are combined to determine the applicable consumption scene, and the positioning of the consumption scene is verified and optimized through market research, user interviews and other modes.

The residence of the user is one of important factors to be considered when the marketing strategy is formulated, enterprises can more specifically conduct marketing and resource allocation by knowing the regional distribution of the user, geographic position data of the user including cities, areas, streets and the like are collected and analyzed, the user data are visualized by means of Geographic Information Systems (GIS) and the like to form a user distribution diagram, and key areas and potential markets are identified according to the user distribution diagram.

The store distribution diagram shows the physical store layout situation of enterprises in different regions, the enterprises can know the market coverage situation and potential market blank point of the enterprises through analyzing the store distribution diagram, store information of the enterprises including store positions, areas, operation conditions and the like is collected and arranged, the store information is visualized by map software or GIS tools to form the store distribution diagram, and the rationality and the potential market blank point of the store distribution diagram are analyzed by combining with the user distribution diagram.

After the consumption scene, residence and store distribution diagram of the user are determined, the enterprise needs to combine the information to determine the corresponding marketing scene, the marketing scene refers to a marketing activity of the enterprise for a specific user group in specific time and space, the consumption scene, residence and store distribution diagram of the user are comprehensively analyzed to identify potential marketing opportunities, specific marketing activity schemes including activity topics, time, places, participation modes and the like are formulated according to the marketing opportunities, and activity propaganda and popularization are carried out in an online and offline combined mode to attract the participation of the user.

Taking an intelligent home company as an example, a new entity storefront is arranged in a large shopping center of a city center of a first-line city, a plurality of large residential communities are arranged around the shopping center, a user has high requirements on intelligent home products, in combination with consumption scenes such as home entertainment and home office of the user, the user plans to hold a scene of 'intelligent home experience day' during the operation of the new entity storefront, and the activity content comprises intelligent product display, field experience, expert explanation, preferential promotion and the like.

The intelligent household product is widely focused and accepted during the activities, good marketing effects are achieved, the consumption scene, residence and store distribution diagram of the users are deeply analyzed, the enterprises can determine the corresponding marketing scenes and formulate the corresponding marketing strategies and activity schemes by combining specific product characteristics and market demands, and the intelligent household product is beneficial to improving brand awareness and market share of the enterprises and can better meet the demands and expectations of the users.

Specifically, a marketing scene matching table is designed to correlate the consumption scene, residence, store distribution and corresponding marketing scene of the user, and the following is a simplified example of the marketing scene matching table:

the method comprises the steps of user ID, consumption scene, residence place, store distribution, marketing scene and targeted marketing activities, wherein the consumption scene is generated by the user in daily life, the residence place is a specific geographic position of the user, the store distribution is the number and distribution of stores nearby the residence place of the user, and the marketing scene is determined according to the consumption scene and residence place of the user and the store distribution.

In step S15, in the marketing scenario, determining a consumption route of the user based on the current position of the user, the corresponding demand information and the current time, and matching the corresponding marketing information according to a plurality of consumption nodes in the consumption route of the user and the marketing scenario, wherein each marketing information is sequentially presented to the user;

s151, in the marketing scene, collecting demand information of a user;

s152, determining the current position of the user based on the position data of the user;

s153, determining a consumption route of the user according to the current position of the user, the corresponding demand information and the current time;

s154, determining a plurality of consumption nodes in the consumption route of the user based on the consumption route of the user, the store distribution diagram and the consumption budget of the user;

s155, associating a plurality of consumption nodes in the consumption route of the user and marketing scenes;

and S156, matching corresponding marketing information based on a plurality of consumption nodes in the consumption route of the user and the marketing scene, and sequentially presenting the marketing information to the user.

In the embodiment of the application, in the marketing scene, the demand information of the user is collected, the current position of the user is determined based on the position data of the user, the consumption route of the user is determined according to the current position of the user, the corresponding demand information and the current time, the integral consideration of the current position of the user, the corresponding demand information and the current time is compatible, and the accuracy of the consumption route of the user is ensured.

At this time, in the marketing scenario, the requirement information of the user is collected, and optionally, the requirement information of the user in a specific marketing scenario is obtained by means of questionnaire, interview of the user, online behavior tracking (such as browsing records, searching keywords and the like) or intelligent device collection (such as using habit of an intelligent home system), and the requirement information comprises preference, price sensitivity, purchasing frequency, specific function requirement and the like of the user on products.

Location data including GPS data, wiFi hotspot information, mobile device signals, etc. needs to be collected from the user's devices (e.g., smartphones, car navigation systems, etc.), the collected location data needs to be processed and analyzed to determine the accurate location of the user, and privacy protection regulations must be strictly complied with to ensure the security and privacy of the user's data when processing the user's location data.

And determining a consumption route of the user according to the current position of the user, the corresponding demand information and the current time, wherein the current position of the user, the corresponding demand information and the current time are introduced, and the demand information comprises demands explicitly expressed by the user (such as finding nearby restaurants, shopping centers and the like) and potential demands obtained through data analysis (such as favorite commodity types, consumption habits and the like of the user).

The current time has a great influence on the consumption route, for example, the user tends to find a restaurant more at lunch time and a entertainment place more at night, and the current position, the demand information and the current time of the user are combined so as to be convenient for the user to plan an optimal consumption route.

The navigation application is used for collecting GPS data through a smart phone of a user and uploading the GPS data to a server in real time, the server receives the data and then determines the current position of the user by using a map matching technology and displays the current position on a map, so that the user can clearly know the position of the user and information such as surrounding roads, buildings and the like, the user is assumed to open the application in lunch time and express the requirement of searching for nearby restaurants, the application firstly screens out a nearby restaurant list according to the current position of the user and the lunch time, then utilizes a path searching technology to plan an optimal route to the nearest restaurant for the user, meanwhile, the application can also consider traffic conditions (such as congestion conditions, road construction and the like) to provide more accurate arrival time estimation for the user, the current position of the user can be determined based on the position data of the user, and an optimal consumption route can be planned for the user according to the requirement information and the current time of the user, convenience of the user is improved, and more marketing opportunities are provided for a user.

Further, a plurality of consumption nodes in the consumption route of the user are determined based on the consumption route of the user, the store distribution diagram and the consumption budget of the user, interaction of the consumption route of the user, the store distribution diagram and the consumption budget of the user is introduced, division of the plurality of consumption nodes in the consumption route is achieved, the overall consideration of the consumption route of the user, the store distribution diagram and the consumption budget of the user is compatible, and accuracy of the plurality of consumption nodes in the consumption route of the user is guaranteed.

For a consumption route, one or more feasible consumption routes are planned based on the starting position, the target position and the stopover point of the user and combined with map data and traffic information, and the time period of travel of the user, such as early peak, afternoon break, evening shopping and the like, and the business hours of all consumption nodes are considered, so that the nodes on the route are ensured to be within the available time of the user.

Aiming at the store distribution diagram, the planned consumption route is overlapped with the store distribution diagram to identify the store position around the route, and store types meeting the user requirements are screened out according to the consumption preference (such as catering, shopping, entertainment and the like) of the user.

For the consumption budget, the consumption budget of the user is known, including the whole budget and the single budget (such as catering budget, shopping budget and the like), and the store meeting the user budget is screened out according to the price information (such as people average consumption, commodity price range and the like) of the store.

Aiming at the consumption nodes, the priority ranking is carried out on the screened stores by combining the factors such as the preference of the user, the scoring of the stores, the preferential activities and the like, and a plurality of consumption nodes of the user in the consumption route are determined according to the route planning, the time nodes, the budget limit and the priority ranking.

Specifically, assuming a user plans to conduct a day of shopping and entertainment on the weekend, the starting point is at home, the ending point is at movie theatre, and the consumption budget is 500 yuan, the following is a specific example illustration:

The user starts from home, plans to go to shopping mall, goes to food street and finally arrives at movie theatre, plans an optimal consumption route in consideration of traffic condition and store business hours, screens out stores such as fashion clothing, electronic products and cosmetic, screens out various restaurants and snack spreads around the shopping mall, sets the whole budget as 500 yuan, wherein the shopping budget is 300 yuan, the catering budget is 200 yuan, and screens out stores meeting the budget, such as restaurants with average consumption of less than 100 yuan, according to the price information of the stores.

In shopping center, the user selects a fashion clothing store and an electronic product store with high score and preferential activities, in food street, the user selects a restaurant with good public praise and moderate price, and finally, the user goes to movie theatre to watch the movie, thus completing shopping and entertainment activities in one day.

In another embodiment of the application, a consumption node matching table is constructed, the consumption node matching table is used for recording potential consumption nodes in a user consumption route and performing preliminary screening according to factors such as consumption budget, store type, position and the like of the user, and the following is an example of the consumption node matching table:

In this consumption node match table:

The consumption node number is used to uniquely identify each consumption node, the store type indicates the type of store, such as a cafe, a bookstore, etc., the location information describes the location of the store relative to the user's consumption route, the estimated consumption amount is the amount the user expects to consume at the store, whether the budget is met is a boolean value indicating whether the consumption node is within the user's consumption budget, and the user preference indicates the user's preference for the store type, which may be high, medium, low, etc.

Through the preliminary screening, we can exclude consumer nodes that do not meet the budget or that have low user preference, such as the "high-end restaurants" in the consumer node matching table.

Based on the multiple consumption nodes in the consumption route of the user and the marketing scene, the corresponding marketing information is matched, and each marketing information is sequentially presented to the user, so that the matching of the multiple consumption nodes in the consumption route of the user and the marketing scene is realized, the presentation of the marketing information is ensured, and the accuracy of the marketing information is improved.

At this time, the position and type of each node are defined according to the consumption route of the user and a plurality of consumption nodes (such as shops, restaurants and movie theatres) determined before, a marketing scene related to each consumption node is defined for each consumption node, the marketing scene is based on various factors such as holidays (such as christmas promotions), events (such as new products to be marketed), user behaviors (such as first shopping offers) and the like, each consumption node is related to a corresponding marketing scene, data analysis is usually involved, so that the consumption behaviors and historical preferences of the user in a specific scene are known, and the relationship between the marketing scene and the consumption nodes is dynamically adjusted according to real-time feedback and position change of the user, so that timeliness and relativity of marketing activities are ensured.

The method comprises the steps of establishing a database containing various marketing information, wherein the information comprises coupons, discounts, gifts, activity details and the like, matching marketing information related to each consumption node from a marketing information base according to consumption routes and related marketing scenes of users, formulating a presentation strategy of the marketing information, considering presentation time points (such as before the users arrive at the nodes, when the users consume the marketing information in the nodes and the like), modes (such as short messages, APP pushing, field presentation and the like) and frequencies (such as one-time presentation or multiple reminding), carrying out personalized adjustment on the presented marketing information according to consumption histories, preferences and real-time feedback of the users so as to improve the participation degree and satisfaction degree of the users, monitoring the presentation effect of the marketing information, and collecting feedback data of the users for optimizing future marketing activities.

Specifically, suppose a user plans to go from home, go to shopping mall to shop, go to restaurant to eat, go to movie theatre to watch movies, in which case the shopping mall node is associated with a "winter promotion" marketing scenario, because multiple merchants are promoting winter clothing and household items.

The restaurant node is associated with a "food saver" marketing scenario in that the restaurant participates in a local food saver activity providing featured dishes and preferential packages, and the theater node is associated with a "new film showing" marketing scenario in that there is a new film of interest to the user being shown.

Assume that the user has planned a route for consumption from the shopping mall to the restaurant to the movie theater. When a user approaches a shopping center, pushing a piece of coupon information of 'winter promotion' through an APP to encourage the user to consume in the shopping center, before the user uses a meal, sending a piece of detailed information of 'food festival' activity, including special dishes and preferential packages, through a short message to attract the user to a designated restaurant for dining, and when the user arrives at a movie theater, displaying a piece of preferential information of 'new film showing' on a self-service ticket vending machine of the movie theater, such as buying ticket to send popcorn or small gifts, thereby improving the viewing experience of the user.

In another embodiment of the present application, to clearly illustrate how marketing information is matched based on a user's consumption node and marketing scenario, we can construct a marketing information matching table, the following being an example of a marketing information matching table:

In the marketing information matching table, a consumption node number corresponds to a store selected by a user in a consumption route, a store type describes a specific type of each consumption node, a marketing scene is a marketing theme set according to the store type, seasons, holidays and other factors, and marketing information is preferential or promotion information specifically provided for the user.

In step S16, determining the grade coefficient of each marketing message according to each marketing message, the viewing time of the user and the clicking times of the user, determining a training set according to the content of each marketing message, the grade coefficient of each marketing message and the consumption portrait of the user in the city, and triggering the intelligent training of the marketing scene according to the training set and the marketing scene;

s161, acquiring each marketing message, wherein each marketing message is presented to a user based on the mobile device;

s162, collecting the view time length of the user and the clicking times of the user based on the presentation of each marketing message;

S163, performing multiple interactions on each marketing message, the viewing time of the user and the clicking times of the user, and determining the grade coefficient of each marketing message according to the multiple interactions of each marketing message, the viewing time of the user and the clicking times of the user;

S164, determining the content of each marketing message according to each marketing message traversal;

S165, interacting the content of each marketing message, the grade coefficient of each marketing message and the consumption portrait of the user in the city, and determining a training set according to the content of each marketing message, the grade coefficient of each marketing message and the interaction of the consumption portrait of the user in the city;

And S166, determining a plurality of training dimensions based on the matching of the training set and the marketing scene, marking corresponding training data in the plurality of training dimensions, and triggering intelligent training of the marketing scene according to the plurality of training dimensions, the corresponding training data and the corresponding marketing scene to output an optimized marketing scene so as to realize the optimal management of the marketing scene.

In the specific implementation process of the method, each marketing message is acquired and presented to the user based on the mobile equipment, the viewing time length of the user and the clicking times of the user are acquired based on the presentation of each marketing message, the viewing time length of the user and the clicking times of the user are subjected to multiple interactions, the grade coefficient of each marketing message is determined according to the multiple interactions of each marketing message, the viewing time length of the user and the clicking times of the user, the multiple interactions of each marketing message, the viewing time length of the user and the clicking times of the user are realized, and the accuracy of the grade coefficient of each marketing message is ensured.

At this time, each marketing message is acquired, each marketing message is presented to the user based on the mobile device, the mobile device and the interaction of the user are further controlled, meanwhile, the enterprise displays marketing content to the user on various marketing channels (such as social media, search engines, emails, APP pushing and the like) including advertisements, coupons, activity information and the like, the time spent by the user browsing or viewing a certain marketing message reflects the interest degree and the attention degree of the user on the information, the number of times the user clicks a certain marketing message generally indicates the interest of the user on the information or the willingness of the user to act, in actual operation, the enterprise can acquire the data through a data analysis tool or platform, and for example, the clicking action and browsing duration of the user can be recorded through embedding tracking codes in the webpage or the APP.

Multiple interactions of each marketing message, the viewing time length of the user and the clicking times of the user refer to comprehensive analysis of three variables of the marketing message, the viewing time length of the user and the clicking times, and the mutual influence and the correlation among the three variables are considered, for example, the clicking times of a certain marketing message are high, but the viewing time length indicates that the message attracts the user to click, but the content is not attractive enough, otherwise, if the viewing time length is long, but the clicking times are small, the message content is attractive, but the factor for promoting the action of the user is lacking, a grade coefficient is allocated to each marketing message based on the result of the multiple interaction analysis and is used for evaluating the attractive force and the effect of the message, the grade coefficient can be a numerical value or grade classification (such as A, B, C and the like) depending on the evaluation standard and the requirement of an enterprise, alternatively, the enterprise can use a data analysis model or algorithm for multiple interaction analysis and grade coefficient determination in actual operation, for example, a machine learning algorithm can be used for training a large amount of historical data to identify attractive force and effect modes of different marketing messages, and grade coefficients are allocated for new information according to the attractive force and effect modes.

Specifically, it is assumed that a website of an electronic commerce shows an advertisement of a new mobile phone on a first page, the website records that the time for viewing the advertisement by a user A is 30 seconds, clicking for 2 times, and the user B only views for 5 seconds without clicking, the data provide a basis for subsequent analysis, the website of the electronic commerce uses a data analysis model to analyze based on the data of the user A and the data of the user B and more similar user data, and the data analysis model finds that marketing information with the viewing time between 15 seconds and 30 seconds and the clicking times of more than 1 usually has higher conversion rate, so that the data analysis model is distributed with a higher ranking coefficient (such as A level) for the advertisement viewed and clicked by the user A, representing that the information has higher attraction and effect, and the data analysis model is distributed with a lower ranking coefficient (such as C level) for the information which is only briefly viewed and not clicked by the user B, and the information is required to be optimized or adjusted.

The method comprises the steps of obtaining marketing information of a user, determining the content of each marketing information according to the traversal of each marketing information, interacting the content of each marketing information, the grade coefficient of each marketing information and the consumption portrait of the user in the city, determining a training set according to the content of each marketing information, the grade coefficient of each marketing information and the interaction of the consumption portrait of the user in the city, realizing the content of each marketing information, the grade coefficient of each marketing information and the interaction of the consumption portrait of the user in the city, and ensuring the richness of the training set;

at this time, all available marketing messages are traversed, namely each message is checked one by one, the purpose of the traversal is to know the content of each marketing message in detail, the content comprises text description, pictures, videos, linked goods or service details and the like, and the determination of the content is critical to the subsequent steps, because the matching degree of the marketing messages and the user images is directly influenced, and the final marketing strategy is formulated.

And carrying out interaction analysis on the marketing information content determined in the last step and the grade coefficient of the marketing information (reflecting the attraction and effect of the information) and the consumption portrait of the user in the city (reflecting the consumption habit, preference, purchasing power and other information of the user), wherein the purpose of the interaction analysis is to find out the association between the marketing information and the portrait of the user and which information is more likely to attract the user of a specific type.

Based on this interactive analysis, the system or analyst builds a training set, which is a data set comprising a plurality of data points, each data point comprising the content of the marketing message, the ranking coefficients, and the user portrayal features that may match the message, which will be used for subsequent machine learning model training or strategy formulation.

Specifically, it is assumed that an e-commerce platform is preparing a marketing campaign for a summer promotion, in step S164, an analyst traverses all marketing messages related to the summer promotion, such as discount information of sun cream, recommendation of summer clothing, advertisement of beach vacation packages, etc., and for each message, the analyst records specific content, such as discount strength of sun cream, style description of summer clothing, price of beach vacation packages, contained service, etc.

In the example of a summer promotion on an e-commerce platform, an analyst may interact with each marketing message (e.g., the discount information for sunscreens) with its ranking factor (e.g., based on user clicks and evaluation of purchase data) and the consumer image of the user in the city (e.g., whether the user frequently purchases sunscreens, whether he prefers high-end brands, whether he is frequently involved in a promotional campaign, etc.), by which analysis the analyst may find the discount information for sunscreens particularly effective for a population of users who frequently purchase sunscreens and are price sensitive.

Based on this finding, the analyst would construct a training set containing a plurality of data points resembling "sunscreens discount information + high ranking factor + user portraits that frequently purchase sunscreens and are price sensitive"; this training set can then be used to train a machine learning model that can predict which marketing messages are most likely to attract which types of users, thereby helping the e-commerce platform to formulate more accurate marketing strategies;

In another embodiment of the application, a relationship matching table is established to record the correspondence between the content of the marketing message, the rating coefficient and the consumer representation of the user, and the relationship matching table can be used to construct a training set. Relationship matching table example:

in this relationship matching table, each row represents a training data point, for example, the first data represents a marketing message for summer clothing discounts, with a ranking factor of A, for a population of users who like summer clothing and consume a medium level, this data point being included in training set 1.

The intelligent training of the marketing scene is triggered according to the training dimensions, the corresponding training data and the corresponding marketing scene, so that the optimized marketing scene is output to realize the optimized management of the marketing scene, the interaction of the training dimensions, the corresponding training data and the corresponding marketing scene is realized, the accurate triggering of the intelligent training of the marketing scene is ensured, and the optimized management of the marketing scene is further controlled.

At this time, the training set (including data such as consumer consumption portraits and marketing information) is matched with a specific marketing scene, so that a plurality of key training dimensions are identified, wherein the dimensions may include user interests, consumption behaviors, product characteristics, marketing channels and the like, and factors which have important influence on the marketing effect are extracted as training dimensions by analyzing the data in the training set and combining the specific characteristics of the marketing scene.

After the training dimensions are determined, the data in each dimension need to be marked so that the data can be accurately identified and utilized by subsequent intelligent training, labels or codes are distributed to the data in each training dimension, accuracy and traceability of the data are guaranteed, for example, the preference degree of different types of products can be coded and marked for users in the dimension of interest of the users, further, a machine learning or deep learning algorithm is utilized to simulate and optimally train a marketing scene by combining a plurality of training dimensions and corresponding training data, a proper machine learning model is selected, the marked training data is input into the model to be trained, and the prediction and adaptation capability of the model to the marketing scene are improved by adjusting model parameters and an optimization algorithm.

After intelligent training, the model outputs optimized marketing scene schemes, the schemes aim at improving marketing effect, meeting user requirements and reducing marketing cost, and the intelligent training results are applied to actual marketing scenes, and the schemes are continuously adjusted and optimized in the modes of A/B test, user feedback and the like, so that the optimal management of the marketing scenes is finally realized.

Specifically, assume we have an online retail platform that wishes to optimize the context of its email marketing campaign, the following are specific examples of operations according to the steps described above:

A plurality of training dimensions are determined based on matching of the training set and the marketing scene, wherein the training set comprises data of purchase history, browsing behaviors, interest and hobbies of a user, the marketing scene comprises an email marketing activity aiming at promoting new products on a platform, and the training dimensions comprise the steps of determining the following training dimensions of user purchase preference (such as product type and price sensitivity), browsing behavior characteristics (such as page stay time and click rate) and interest and hobbies (such as sports and fashion) according to characteristics of the training set and the marketing scene.

Coding and marking the purchase preference of the user, such as marking the user who purchases sports products as 'sports preference', quantifying the browsing behavior characteristics, such as calculating the average page stay time and click rate of each user, and classifying and marking the interest and hobbies, such as marking the user who likes fashion content as 'fashion fan'.

Selecting a proper machine learning model (such as logistic regression, random forest and the like) for training, inputting marked training data into the model for training, adjusting model parameters to optimize the prediction capability of the E-mail marketing activity effect, intelligently training, outputting an optimized E-mail marketing activity scheme by the model, for example, aiming at a user group with 'sports preference' and 'long page residence time', sending an E-mail containing the latest sports product popularization information, collecting data through A/B test and user feedback, adjusting and optimizing the optimized marketing scene, for example, timely adjusting popularization strategies or replacing popularization products when a certain type of users have bad product popularization response to a specific type of products.

Example III

In this embodiment, as shown in fig. 3, there is provided a marketing scenario intelligent training management device, including:

Example IV

In this embodiment, a marketing scenario intelligent training management device is provided, an internal structure diagram of the marketing scenario intelligent training management device may be shown in fig. 4, the marketing scenario intelligent training management device comprises a processor, a memory, a network interface, a display screen and an input device which are connected through a system bus, wherein the processor of the marketing scenario intelligent training management device is used for providing computing and control capabilities, the memory of the marketing scenario intelligent training management device comprises a nonvolatile storage medium and an internal memory, the nonvolatile storage medium stores an operating system and a computer program, the nonvolatile storage medium is provided with a database for storing user behavior data and user images, the internal memory is used for providing an environment for the operation of the operating system and the computer program in the nonvolatile storage medium, a network interface of the marketing scenario intelligent training management device is used for communicating with other devices deployed with application software, the computer program is executed by the processor to realize a marketing scenario intelligent training management method, the display screen of the marketing scenario intelligent training management device may be a liquid crystal display screen or an electronic ink display screen, the input device of the marketing scenario intelligent training management device may be a touch screen, a touch pad, or the like.

The above examples illustrate only a few embodiments of the application, which are described in detail and are not to be construed as limiting the scope of the application, it being understood that variations and modifications can be made by those skilled in the art without departing from the spirit of the application, which is within the scope of the application, which is therefore subject to the appended claims.

Claims

1. The intelligent training management method for the marketing scene is characterized by comprising the following steps of:

collecting consumption data sets of users in different cities;

determining the consumption portrait of the user in the city according to the consumption data set, the life information of the user and the corresponding city;

Determining consumption parameters of the user in the city based on the income level of the user, the corresponding consumption time and the preference of the user in the city;

Determining a consumption scene of the user according to the consumption portrait and the corresponding consumption parameters, and determining a corresponding marketing scene according to the consumption scene of the user, the residence place of the user and the corresponding store distribution diagram;

In the marketing scene, determining a consumption route of a user based on the current position of the user, corresponding demand information and current time, and matching corresponding marketing information according to a plurality of consumption nodes in the consumption route of the user and the marketing scene, wherein each marketing information is sequentially presented to the user;

and determining the grade coefficient of each marketing message according to each marketing message, the viewing time of the user and the clicking times of the user, determining a training set according to the content of each marketing message, the grade coefficient of each marketing message and the consumption portrait of the user in the city, and triggering the intelligent training of the marketing scene according to the training set and the marketing scene.

2. The marketing scenario intelligent training and management method of claim 1, wherein the collecting the consumption data sets of users in different cities comprises:

marking a payment platform of a user;

exporting past payment data of the user according to a payment platform of the user;

determining a data set of each city according to the past payment data of the user and the matching of the corresponding payment positions;

Consumption data of users in different cities is determined based on the data sets of the respective cities and the consumption types of the users.

3. The marketing scenario intelligent training and management method of claim 2, wherein the determining the consumption portraits of the users in the city according to the consumption data set, the life information of the users and the corresponding cities comprises:

acquiring a consumption data set;

collecting life information of a user based on traversal of a life database of the user;

Associating the consumption data set, life information of the user and corresponding cities;

Determining a first consumption characteristic according to the consumption data set and the life information of the user;

Determining a second consumption feature according to the consumption data set and the corresponding city;

and determining the consumption portrait of the user in the city according to the first consumption characteristic, the second consumption characteristic and the stay time of the user in the city.

4. The marketing scenario intelligent training and management method of claim 3, wherein the determining the consumption parameters of the user in the city based on the user's revenue level, the corresponding consumption time, and the user's preference in the city comprises:

Collecting revenue data for the user based on the traversal of the user's paymate;

determining the income level of the user according to the income data of the user, the job title of the user in the enterprise and the residence of the user;

determining a plurality of consumption categories and corresponding consumption frequencies according to traversal of consumption data of the user in the city;

Determining preference of a user in a city according to a plurality of consumption types and cities corresponding to the corresponding consumption frequencies;

interacting the income level of the user, the corresponding consumption time and the preference of the user in the city;

The consumption parameters of the user in the city are determined according to the income level of the user, the corresponding consumption time and the favorite interactions of the user in the city.

5. The method for intelligent training management of a marketing scenario according to claim 1, wherein determining a consumption scenario of a user according to the consumption portrait and the corresponding consumption parameters, and determining a corresponding marketing scenario according to the consumption scenario of the user, a residence where the user is located, and a corresponding store distribution map, comprises:

Acquiring the consumption portrait and corresponding consumption parameters;

Matching the corresponding scene weight to the consumption portrait and the corresponding consumption parameters;

Determining a consumption scene of the user according to the consumption portrait, the consumption parameters and the corresponding scene weights;

determining the residence place of the user according to the residence data of the user;

Determining a corresponding store distribution diagram based on the residence place where the user is located, the surrounding drawings of the residence place and the store marks;

and determining a corresponding marketing scene according to the consumption scene of the user, the residence place of the user and the corresponding store distribution diagram.

6. The method for intelligent training management of a marketing scenario according to claim 1, wherein in the marketing scenario, a consumption route of a user is determined based on a current location of the user, corresponding demand information and a current time, and a plurality of consumption nodes in the consumption route of the user and the marketing scenario match corresponding marketing information, and each marketing information is sequentially presented to the user, comprising:

in the marketing scene, collecting demand information of a user;

determining a current location of the user based on the location data of the user;

determining a consumption route of the user according to the current position of the user, the corresponding demand information and the current time;

determining a plurality of consumption nodes in the user's consumption route based on the user's consumption route, the store profile, and the user's consumption budget;

Associating a plurality of consumption nodes in a consumption route of a user with a marketing scenario;

based on a plurality of consumption nodes in the consumption route of the user and the marketing scene matching corresponding marketing information, each marketing information is sequentially presented to the user.

7. The method for intelligent training management of a marketing scenario according to claim 1, wherein determining the rating coefficient of each marketing message according to each marketing message, the viewing time of the user and the number of clicks of the user, determining the training set according to the content of each marketing message, the rating coefficient of each marketing message and the consumption portrayal of the user in the city, and triggering the intelligent training of the marketing scenario according to the training set and the marketing scenario comprises:

Acquiring each marketing message, wherein each marketing message is presented to a user based on mobile equipment;

collecting the checking time length of the user and the clicking times of the user based on the presentation of each marketing message;

And determining the grade coefficient of each marketing message according to the multiple interactions of each marketing message, the viewing time of the user and the clicking times of the user.

8. The method for intelligent training management of a marketing scenario according to claim 7, wherein the determining the rating coefficient of each marketing message according to each marketing message, the viewing time of the user and the number of clicks of the user, determining the training set according to the content of each marketing message, the rating coefficient of each marketing message and the consumption portrayal of the user in the city, and triggering the intelligent training of the marketing scenario according to the training set and the marketing scenario, further comprises:

Determining the content of each marketing message according to each marketing message traversal;

The method comprises the steps of carrying out interaction on the content of each marketing message, the grade coefficient of each marketing message and the consumption portrait of a user in the city, and determining a training set according to the content of each marketing message, the grade coefficient of each marketing message and the interaction of the consumption portrait of the user in the city;

And triggering intelligent training of the marketing scene according to the training dimensions, the corresponding training data and the corresponding marketing scene to output an optimized marketing scene so as to realize optimal management of the marketing scene.

9. Marketing scene intelligent training management equipment, characterized in that it is applied to the marketing scene intelligent training management method according to any one of claims 1 to 8, comprising:

10. A storage medium having stored thereon a computer program, which when executed by a processor, implements the steps of the marketing scenario intelligent training management method of any one of claims 1 to 8.