Intelligent cloud customer obtaining system based on big data

Abstract

The invention discloses an intelligent cloud customer acquisition system based on big data, which relates to the technical field of intelligent cloud customer acquisition and comprises a data acquisition module, a processing module, a calculation module and an analysis module. According to the method, various behavior characteristics of the user in the financial field are combined, so that the accurate customer portrait is constructed, the current state of the customer is reflected, the potential demand and the future trend of the customer are reflected, powerful support is provided for subsequent accurate recommendation, and the customer portrait and the financial product or service characteristics are combined, so that the user experience is improved. According to the method, accurate prediction of potential demands of customers is realized, powerful decision support is provided for financial institutions, and finally, a plurality of factors such as potential demand prediction values, market feedback data and user credit scores are comprehensively considered, so that personalized recommendation priority ranking is obtained; the sorting result not only improves the recommendation accuracy and the conversion rate, but also optimizes the resource configuration and improves the operation efficiency of the financial institution.

Description

Technical Field

[0001] The present invention relates to the field of intelligent cloud customer acquisition technology, and specifically to an intelligent cloud customer acquisition system based on big data. Background Art

[0002] With the rapid development of Internet technology, the application of big data in the financial field is becoming increasingly widespread, especially in customer acquisition. Traditional customer acquisition methods mainly rely on advertising, ground promotion activities and other means, which are not only costly but also inefficient. With the rise of big data technology, financial institutions have begun to try to accurately identify potential customers through data analysis and improve customer acquisition efficiency.

[0003] However, the existing customer acquisition system may rely on simple data collection and analysis, lacking in-depth mining and precise construction of customer portraits, resulting in a low match between customer portraits and financial products. In addition, the existing customer acquisition system may lack accurate predictions of customer potential needs, and only uses simple data analysis or statistical models for predictions. The prediction results often deviate greatly from actual needs. In addition, the existing customer acquisition system may adopt a general recommendation algorithm and lack personalized recommendation strategies for the characteristics of the financial field, resulting in unsatisfactory recommendation effects and low conversion rates. Summary of the invention

[0004] The purpose of the present invention is to provide an intelligent cloud customer acquisition system based on big data, which solves the problems raised in the above-mentioned background technology.

[0005] To achieve the above objectives, the present invention provides the following technical solutions: an intelligent cloud customer acquisition system based on big data, comprising:

[0006] Data collection module: obtains user access frequency, user keyword search frequency and usage time on financial apps from financial websites, search engines and financial apps, and stores them in the database;

[0007] Processing module: used to input the user's access frequency, the user's keyword search frequency and the usage time on the financial APP, and clean and organize the input data, and output the number of times user m visits the financial website within time period t, the number of times user m searches for financial keywords in the search engine within time period t, and the usage time of user m on the financial APP within time period t;

[0008] Calculation module: used to input the number of times user m visits financial websites within time period t, the number of times user m searches for financial keywords in search engines within time period t, and the usage time of user m on financial apps within time period t, and output a comprehensive score of user m's behavior within time period t, a potential demand prediction value for user m and financial product p within time period t, and an intelligent recommendation priority for user m and financial product p within time period t;

[0009] Analysis module: used to formulate personalized recommendation strategies for user m and financial product p based on the comprehensive score of user m's behavior in time period t, the potential demand prediction value for user m and financial product p in time period t, and the user m and financial product p in time period t, and give priority to recommending highly matching financial products to high-priority users.

[0010] Optionally, the calculation module includes: a behavior submodule, a demand submodule, a recommendation submodule and an optimization submodule.

[0011] Optionally, the calculation formula of the behavior submodule is as follows:

[0012]

[0013] in:

[0014] UBDX t，m Refers to the comprehensive score of user m’s behavior in time period t;

[0015] UBDA t，m Refers to the number of times user m visits the financial website in time period t, UBDB t，m Refers to the number of times user m searches for financial keywords in the search engine within time period t, UBDC t，m Refers to the usage time and UBDD of user m on financial apps in time period t t，m Refers to the total usage time of user m on all apps in time period t, A1 t Refers to UBDA in time period t t，m The weight of A2 t Refers to UBDB in time period t t，m Weight, A3 t Refers to UBDC in time period t t，m The weight of, log refers to the logarithmic function, t refers to the time period, m refers to the user, A1 t ×UBDA t，m Refers to the contribution of visit frequency to the comprehensive score of user behavior;

[0016] The processing process of the behavior submodule is as follows: the number of times user m visits the financial website in time period t is UBDA t，m, the number of times user m searches for financial keywords in the search engine within time period t UBDB t，m and the usage time of user m on financial apps in time period t UBDC t，m Input to the behavior submodule, and based on the total usage time of user m on all apps in time period t, UBDD t，m , time period t and user m, output the comprehensive score UBDX of user m's behavior in time period t t，m .

[0017] Optionally, the calculation formula of the demand submodule is as follows:

[0018]

[0019] in:

[0020] PSPX t，m，p Refers to the potential demand forecast value for user m and financial product p in time period t;

[0021] PSPA p Refers to the number of different attributes contained in the user profile of financial product p, PSPB t，m，p，i Refers to the matching degree between the i-th dimension of the user m’s portrait and the characteristics of financial product p in time period t, PSPC p Refers to the age span and PSPD of the target user group of financial products p Refers to the actual age of user m in time period t, PSPE p refers to the average age of the target user group of the financial product p, p refers to the financial product, Refers to the sum of the matching degrees of all portrait dimensions of user m and financial product p in time period t;

[0022] The processing process of the demand submodule is as follows: the comprehensive score UBDX of the behavior of user m in time period t is calculated as follows: t，m Input to the demand submodule and based on the number of different attributes PSPA included in the user profile for financial product p p 、The age span of the target user group of financial products PSPC p , the actual age PSPD of user m in time period t p and the average age of the target user group of financial products (PSPE) p Output the potential demand forecast value PSPX for user m and financial product p in time period t t，m，p .

[0023] Optionally, the calculation formula of the recommendation submodule is as follows:

[0024]

[0025]

[0026] in:

[0027] SRPX t，m，p Refers to the intelligent recommendation priority for user m and financial product p in time period t;

[0028] SRPA p Refers to the conversion rate of financial products p to new customers, SRPB p Refers to the sales data of financial product p in time period t, SRPC p Refers to the number of reviews received by financial product p in time period t, SRPD p，t Refers to the favorable rating and SRPE of financial product p in time period t t，m Refers to the credit score, SRPF of user m in time period t p Refers to the minimum credit score, SRPG, among the target user group for financial product p p Refers to the highest credit score among the target user group for financial product p, (SRPE t，m -SRPF p ) / (SRPG p -SRPF p ) refers to the position of the credit score relative to the credit score range of the target users of the financial product, C1 refers to UBDX t，m The weight coefficient, C2 refers to PSPX t，m，p The weight coefficient, C3 refers to SRPB p The weight factor, C4 refers to SRPC p The weight factor, C5 refers to SRPD p，t The weight factor of

[0029] The processing process of the recommendation submodule is as follows: the comprehensive score UBDX of user m in time period t is calculated as follows: t，m and the potential demand forecast value PSPX for user m and financial product p in time period t t，m，p Input to the recommendation submodule, and based on the conversion rate SRPA of financial product p to new customers p 、The sales data of financial product p in time period t SRPB p and the number of reviews SRPC received by financial product p in time period t p Output the intelligent recommendation priority SRPX for user m and financial product p in time period t t，m，p .

[0030] Optionally, the processing process of the optimization submodule is as follows:

[0031] First up: PSPB t，m，p，i，new =PSPB t，m，p，i +α×(SRPX t，m，p -SRPP);

[0032] Second: Set the loop termination condition:

[0033] Termination condition 1: The number of iterations i is 187;

[0034] Termination condition 2: |SRPX t，m，p，new -SRPX t，m，p，old |<0.00013;

[0035] in:

[0036] PSPB t，m，p，i，new Refers to the matching degree between the i-th dimension in the profile of user m in time period t after update and the characteristics of financial product p, SRPP refers to the expected recommendation priority, SRPX t，m，p，new Refers to the intelligent recommendation priority for user m and financial product p in time period t after the update, α refers to the adjustment factor, SRPX t，m，p，old Refers to the intelligent recommendation priority for user m and financial product p in time period t before updating;

[0037] Based on the intelligent recommendation priority SRPX for user m and financial product p in time period t t，m，p , and adjust the matching degree PSPB between the i-th dimension of the user m’s portrait and the characteristics of the financial product p in the time period t by adjusting the factor α t，m，p，i Output the matching degree PSPB between the i-th dimension of the user m’s portrait and the feature of financial product p in time period t after the update t，m，p，i，new , PSPB t，m，p，i，new Replace PSPB in the requirements submodule t，m，p，i , to constantly adjust PSPX t，m，p and SRPX t，m，p , and finally the loop is terminated by termination conditions one and two.

[0038] Optionally, the tools used by the data collection module are: web crawler software, application data analysis tools, user portrait database and credit scoring system.

[0039] Optionally, when the analysis module recommends highly matched financial products to high-priority users, the recommendations are made via SMS, email, and APP push, and the recommendation effects are tracked and evaluated, including user click-through rate and conversion rate. The recommendation strategy is optimized based on the evaluation results.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] 1. The present invention outputs the comprehensive behavior score UBDX of user m in time period t through the behavior submodule t，m ,UBDX t，m The calculation of embodies the massive and diverse characteristics of big data, which can comprehensively and objectively reflect the user's activity and behavioral characteristics in the financial field, provide a data basis for the subsequent precision marketing of financial enterprises, help financial enterprises identify potential users and high-value users, and provide strong support for personalized recommendations and customized services. It helps financial enterprises quickly identify active users and potential users, and provide target user groups for subsequent intelligent recommendations and precision marketing. By analyzing UBDX t，m Financial companies can understand users’ financial behavior patterns and preferences through the changing trends, thereby providing strong support for product design and marketing strategies.

[0042] 2. The present invention outputs the potential demand prediction value PSPX for user m and financial product p in time period t through the demand submodule t，m，p ,PSPX t，m，p The calculation reflects the value characteristics of big data. By mining and analyzing user portrait data, it can accurately predict the user's potential demand for specific financial products, making it easier for financial companies to plan the market in advance and launch products and services that meet user needs to improve user satisfaction and loyalty. t，m，p It can accurately predict the potential demand of users for specific financial products, making it easier for financial companies to plan the market in advance to meet user needs, thereby improving user satisfaction and loyalty. t，m，p The distribution of financial services can help financial enterprises understand the differences in needs of different user groups and provide data support for product differentiation design and personalized recommendations.

[0043] 3. The present invention outputs the intelligent recommendation priority SRPX for user m and financial product p in time period t through the recommendation submodule t，m，p ,PSPX t，m，p The calculation reflects the real-time and high-speed characteristics of big data. It can track and analyze market feedback data in real time and quickly adjust marketing strategies. Through intelligent recommendation priority sorting, financial enterprises can give priority to recommending high-conversion products to high-value users, thereby improving marketing efficiency and conversion rate. p，t It provides priority sorting for intelligent recommendation of financial products, enabling financial enterprises to recommend high-conversion products to high-value users first, thereby improving marketing efficiency and conversion rate. p，t Financial enterprises can evaluate the effectiveness of marketing strategies based on changes in the customer experience, and adjust and optimize recommendation strategies in a timely manner to improve marketing effectiveness and user satisfaction.

[0044] 4. The present invention optimizes the submodule to match the i-th dimension of the portrait with the feature of the financial product p within the time period t. t，m，p，i Perform loop optimization iterations through SRPX t，m，p To adjust PSPB t，m，p，i Make the business logic clearer, easier to understand and implement, and thus more accurately identify financial products that match user needs. This improvement in accuracy helps reduce ineffective marketing and increase conversion rates, which can improve overall customer acquisition efficiency. During the iterative calculation process, the system can continuously optimize the matching algorithm based on user feedback and behavioral data to make the recommended products or services more in line with user expectations and preferences. This personalized recommendation can enhance user satisfaction and loyalty, thereby improving user experience. Through iterative calculation, the system can more accurately predict which user groups are more likely to be converted into actual customers, thereby optimizing resource allocation and investing limited resources in user groups with greater potential. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 This is a flowchart of the method steps of the intelligent cloud customer acquisition system based on big data;

[0046] Figure 2 This is a schematic diagram of the overall structure of the intelligent cloud customer acquisition system based on big data;

[0047] Figure 3 Schematic diagram of the structure of the synthesis module. DETAILED DESCRIPTION

[0048] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0049] This intelligent cloud customer acquisition system based on big data is different from the existing cloud customer acquisition system. The existing cloud customer acquisition system often relies only on simple data collection and analysis, lacks in-depth mining and accurate construction of user information, resulting in a low match between user information portraits and financial products or services. In addition, the existing cloud customer acquisition system lacks accurate prediction of customer potential needs, and only uses simple data analysis or statistical models for prediction. The prediction results often deviate greatly from the actual needs.

[0050] The module of this intelligent cloud customer acquisition system method based on big data combines the user's behavioral characteristics in the financial field, such as visit frequency, transaction behavior, investment preferences, etc., to build an accurate customer portrait. This portrait not only reflects the customer's current status, but also reflects their potential needs and future trends, providing strong support for subsequent accurate recommendations. In addition, this system combines customer portraits with the characteristics of financial products or services to achieve accurate prediction of customer potential needs. This prediction result not only takes into account the customer's current needs, but also combines market trends and competitive situations, providing strong decision-making support for financial institutions.

[0051] Example: See Figures 1 to 3 , this implementation provides an intelligent cloud customer acquisition system based on big data, including:

[0052] Data collection module: obtains user access frequency, user keyword search frequency and usage time on financial apps from financial websites, search engines and financial apps, and stores them in the database;

[0053] Processing module: used to input the user's access frequency, the user's keyword search frequency and the usage time on the financial APP, and clean and organize the input data, and output the number of times user m visits the financial website within time period t, the number of times user m searches for financial keywords in the search engine within time period t, and the usage time of user m on the financial APP within time period t;

[0054] Calculation module: used to input the number of times user m visits financial websites within time period t, the number of times user m searches for financial keywords in search engines within time period t, and the usage time of user m on financial apps within time period t, and output a comprehensive score of user m's behavior within time period t, a potential demand prediction value for user m and financial product p within time period t, and an intelligent recommendation priority for user m and financial product p within time period t;

[0055] Analysis module: used to formulate personalized recommendation strategies for user m and financial product p based on the comprehensive score of user m's behavior in time period t, the potential demand prediction value for user m and financial product p in time period t, and recommend highly matching financial products to high-priority users first;

[0056] The calculation module includes: behavior submodule, demand submodule, recommendation submodule and optimization submodule.

[0057] In this embodiment: This system realizes data-driven precision marketing and personalized recommendations by collecting and analyzing massive behavioral data and personal credit data of users in the financial field. With the help of the real-time processing capability of big data technology, this system can track and analyze market feedback data in real time, quickly adjust marketing strategies to improve marketing effectiveness, and provide users with personalized financial product and service recommendations by mining and analyzing user portrait data, thereby improving user satisfaction and loyalty, and realizing intelligent recommendation priority sorting and marketing strategy optimization, thereby improving the scientificity and accuracy of decision-making. The subscript t in the present invention indicates that the time period helps enterprises to analyze and predict user behavior data in different time periods and formulate more accurate marketing strategies, and u indicates that users are helpful. It helps enterprises identify the characteristics and behavior patterns of different users, realize personalized marketing and precise recommendations, and p indicates that the product helps enterprises analyze the market demand and user preferences of different products, optimize product design and marketing strategies, and the existence of these indicators enables the formula to more accurately describe and calculate the relationship between users, products and time periods, providing more accurate and reliable data support for intelligent cloud customer acquisition. Multiple groups of sub-modules are combined to cover multiple aspects such as user behavior analysis, potential demand prediction and intelligent recommendation priority sorting, forming a complete system of intelligent cloud customer acquisition process. By comprehensively collecting and analyzing user behavior data in the financial field, it can more accurately understand user needs and preferences, and provide strong support for subsequent intelligent recommendations and precise marketing. UBDX t，m It can quantify the user's financial behavior activity in a specific period of time and provide basic data for potential demand forecasting and intelligent recommendation. t，m，p Through machine learning algorithms, the matching degree between user profiles and financial products is quantified and calculated, which can accurately predict the user's potential demand for specific financial products / services. t，m，pTaking into account multiple factors such as potential demand forecast value, market feedback data and personal credit data, the system can derive the priority ranking of intelligent recommendations and improve the accuracy and conversion rate of marketing. With the help of big data technology and cloud computing platform, the formulas in multiple groups of sub-modules can quickly process and analyze massive data and improve customer acquisition efficiency. Through the automated and intelligent recommendation system, it can track and analyze user behavior and market feedback in real time, quickly adjust marketing strategies, and improve marketing effects. Through accurate user portraits and behavior analysis, it can quickly identify potential users and high-value users and improve customer acquisition efficiency. Through the automated recommendation system and real-time market feedback analysis, it can quickly adjust marketing strategies and improve marketing efficiency. Through personalized recommendations and customized services, it can meet the personalized needs of users and improve user experience and satisfaction. Through accurate demand prediction and priority ranking, this system can provide users with financial products / services that are more in line with their needs and preferences. Through intelligent recommendation systems and real-time data analysis, this system can optimize the allocation of marketing resources and improve resource utilization efficiency and conversion rate. Through accurate user portraits and demand predictions, this system can provide enterprises with more accurate market insights and decision support, helping enterprises to formulate more scientific and reasonable marketing strategies and product plans.

[0058] See also Figures 1 to 3 , the behavior submodule processing process is as follows:

[0059]

[0060] in:

[0061] UBDX t，m Refers to the comprehensive score of user m’s behavior in time period t, UBDX t，m Comprehensively reflects the user's behavioral characteristics;

[0062] UBDA t，m Refers to the number of times user m visits the financial website in time period t, UBDA t，m That is, the number of times a user visits a financial website within a certain period of time (such as a month), which reflects the user's attention to financial information;

[0063] UBDB t，m Refers to the number of times user m searches for financial keywords in search engines within time period t, UBDB t，m That is, the number of times users search for financial-related keywords in search engines, reflecting the users’ active willingness to acquire financial knowledge;

[0064] UBDC t，m Refers to the usage time of user m on financial apps in time period t, UBDC t，mThat is, the length of time users spend on financial apps directly reflects their actual use of financial services;

[0065] UBDD t，m Refers to the total usage time of user m on all apps in time period t, UBDD t，m That is, the total time users spend on all apps is used as the denominator to calculate the proportion of app usage time, so as to more accurately assess the degree of user investment in financial apps;

[0066] A1 t Refers to UBDA in time period t t，m The weight of A2 t Refers to UBDB in time period t t，m Weight, A3 t Refers to UBDC in time period t t，m The weight of, log refers to the logarithmic function, t refers to the time period, m refers to the user, A1 t ×UBDA t，m Refers to the contribution of visit frequency to the comprehensive score of user behavior;

[0067] The processing process of the behavior submodule is as follows: the number of times user m visits the financial website in time period t is UBDA t，m , the number of times user m searches for financial keywords in the search engine within time period t UBDB t，m and the usage time of user m on financial apps in time period t UBDC t，m Input to the behavior submodule, and based on the total usage time of user m on all apps in time period t, UBDD t，m , time period t and user m, output the comprehensive score UBDX of user m's behavior in time period t t，m .

[0068] In this example: UBDX t，m The calculation of embodies the massive and diverse characteristics of big data, and can comprehensively and objectively reflect the user's activity and behavioral characteristics in the financial field, provide a data basis for the subsequent precision marketing of financial enterprises, help financial enterprises identify potential users and high-value users, and provide strong support for personalized recommendations and customized services. t，m As a comprehensive score of financial users’ behaviors in a specific period of time, UBDX t，m It can help financial enterprises quickly identify active users and potential users, and provide target user groups for subsequent intelligent recommendations and precision marketing. t，m Financial enterprises can understand the changing trends of users’ financial behavior patterns and preferences, and provide strong support for product design and marketing strategies.t，m It can be used as a quantitative indicator of user activity to screen potential users and high-value users. It can also be combined with user portrait data to analyze user behavior patterns and preferences, providing strong support for product design and marketing strategies.

[0069] See also Figures 1 to 3 , the requirements submodule processing process is as follows:

[0070]

[0071] in:

[0072] PSPX t，m，p Refers to the potential demand forecast value for user m and financial product p in time period t, PSPX t，m，p Comprehensively reflects the user's potential needs and the degree of matching with the product;

[0073] PSPA p Refers to the number of different attributes contained in the user profile of financial product p, PSPA p That is, the number of different attributes included in the user profile, such as age, gender, occupation, income, etc. For example, for a financial product, the profile dimensions may include age, income, investment preferences, etc. The number of profile dimensions is the total number of these dimensions;

[0074] PSPB t，m，p，i Refers to the matching degree between the i-th dimension of the user m’s portrait and the characteristics of financial product p in time period t, PSPB t，m，p，i That is, the matching degree between the i-th dimension in the user portrait and the financial product p, with a value range of 0 to 1, indicating the degree of fit between the dimension and the product or service;

[0075] PSPC p Refers to the age span of the target user group of a financial product, PSPC p That is, the age span of the target user group, i.e., the difference between the maximum age and the minimum age, is used to calculate the relative difference between the user age and the target age;

[0076] PSPD p Refers to the actual age of user m in time period t;

[0077] PSPE p Refers to the average age of the target user group of financial products, PSPE p That is, the average age of financial product p, which is used to evaluate the degree of match between the user age and the target user group of the product;

[0078] p refers to financial products;

[0079] Refers to the sum of the matching degrees of all portrait dimensions of user m and financial product p in time period t. That is, each portrait dimension has a matching degree, and the sum of these matching degrees reflects the overall matching degree between the user and the product / service;

[0080] The processing process of the demand submodule is as follows: the comprehensive score of user m’s behavior in time period t is UBDX t，m Input to the demand submodule and based on the number of different attributes PSPA included in the user profile for financial product p p 、The age span of the target user group of financial products PSPC p , the actual age PSPD of user m in time period t p and the average age of the target user group of financial products (PSPE) p Output the potential demand forecast value PSPX for user m and financial product p in time period t t，m，p .

[0081] In this embodiment: the PSPX t，m，p The calculation of reflects the value characteristics of big data. By mining and analyzing user portrait data, it can accurately predict the user's potential demand for specific financial products, which helps financial companies to plan the market in advance, launch products and services that meet user needs, and improve user satisfaction and loyalty. t，m，p It can accurately predict users' potential demand for specific financial products, help financial companies to plan the market in advance to meet user needs, and thus improve user satisfaction and loyalty. t，m，p The distribution of PSPX can help financial companies understand the differences in needs of different user groups and provide data support for product differentiation design and personalized recommendations. t，m，p It can be used as a basis for predicting users' potential demand for specific financial products, and can also be combined with market trends and competitive situations to provide data support for product differentiation design and personalized recommendations.

[0082] See also Figures 1 to 3 , the recommended submodule processing process is as follows:

[0083]

[0084] in:

[0085] SRPX t，m，p Refers to the intelligent recommendation priority for user m and financial product p in time period t;

[0086] SRPA p Refers to the conversion rate of financial product p to new customers, SRPA pThat is, the conversion rate of the product to new customers in historical data, ranging from 0 to 1, indicating the product's ability to attract new customers;

[0087] SRPB p Refers to the sales data of financial product p in time period t, SRPB p That is, the sales data of the product directly reflects the market acceptance of the product;

[0088] SRPC p Refers to the number of reviews received by financial product p in time period t, SRPC p That is, the number of reviews a product has received, which reflects the level of attention the product has received and the level of user engagement;

[0089] SRPD p，t Refers to the favorable rating of financial product p in time period t, SRPD p，t That is, the product's praise rate, that is, the proportion of the number of favorable comments to the total number of reviews, reflects the product's quality and user satisfaction;

[0090] SRPE t，m Refers to the credit score of user m in time period t, SRPEt,m is the credit score of the user, such as the score obtained from the credit report, which is used to evaluate the credit status of the user;

[0091] SRPF p Refers to the minimum credit score among the target user group for financial product p;

[0092] SRPG p Refers to the highest credit score among the target user group for financial product p, SRPG p That is, the lowest and highest credit scores among the target user group of the financial product, which are used to calculate the relative position of the user's credit score to the credit score range;

[0093] (SRPE t，m -SRPF p ) / (SRPG p -SRPF p ) refers to the position of the credit score relative to the credit score range of the target users of the financial product, C1 refers to UBDX t，m The weight coefficient, C2 refers to PSPX t，m，p The weight coefficient, C3 refers to SRPB p The weight factor, C4 refers to SRPC p The weight factor, C5 refers to SRPD p，t The weight factor of

[0094] The processing process of the recommendation submodule is as follows: the comprehensive score UBDX of user m’s behavior in time period t is calculated as t，m and the potential demand forecast value PSPX for user m and financial product p in time period t t，m，p Input to the recommendation submodule, and based on the conversion rate SRPA of financial product p to new customers p 、The sales data of financial product p in time period t SRPB p and the number of reviews SRPC received by financial product p in time period t p Output the intelligent recommendation priority SRPX for user m and financial product p in time period t t，m，p .

[0095] In this embodiment: this SRPD p，t The calculation reflects the real-time and high-speed characteristics of big data. It can track and analyze market feedback data in real time and quickly adjust marketing strategies. Through intelligent recommendation priority sorting, financial companies can give priority to recommending high-conversion products to high-value users, thereby improving marketing efficiency and conversion rate. p，t It provides priority sorting for intelligent recommendation of financial products, enabling financial enterprises to recommend high-conversion products to high-value users first, thereby improving marketing efficiency and conversion rate. p，t Financial enterprises can evaluate the effectiveness of marketing strategies, adjust and optimize recommendation strategies in a timely manner, and improve marketing effectiveness and user satisfaction. p，t The use of can be used as the basis for intelligent recommendation priority sorting, and high-conversion rate products can be recommended to high-value users first. It can also be combined with user feedback and marketing effect evaluation to continuously optimize recommendation strategies and marketing strategies and improve marketing efficiency and conversion rate.

[0096] See also Figures 1 to 3 , the optimization submodule processing process is as follows:

[0097] First up: PSPB t，m，p，i，new =PSPB t，m，p，i +α×(SRPX t，m，p -SRPP);

[0098] Second: Set the loop termination condition:

[0099] Termination condition 1: The number of iterations i is 187;

[0100] Termination condition 2: |SRPX t，m，p，new -SRPX t，m，p，old |<0.00013;

[0101] in:

[0102] PSPBt，m，p，i，new Refers to the matching degree between the i-th dimension in the profile of user m in time period t after update and the characteristics of financial product p, SRPP refers to the expected recommendation priority, SRPX t，m，p，new Refers to the intelligent recommendation priority for user m and financial product p in time period t after the update, α refers to the adjustment factor, SRPX t，m，p，old Refers to the intelligent recommendation priority for user m and financial product p in time period t before updating;

[0103] Based on the intelligent recommendation priority SRPX for user m and financial product p in time period t t，m，p , and adjust the matching degree PSPB between the i-th dimension of the user m’s portrait and the characteristics of the financial product p in the time period t by adjusting the factor α t，m，p，i Output the matching degree PSPB between the i-th dimension of the user m’s portrait and the feature of financial product p in time period t after the update t，m，p，i，new , PSPB t，m，p，i，new Replace PSPB in the requirements submodule t，m，p，i , to constantly adjust PSPX t，m，p and SRPX t，m，p , and finally the loop is terminated by termination conditions one and two.

[0104] In this embodiment: Through this iterative form system can continuously according to SRPX t，m，p To adjust the matching degree, so as to more accurately identify financial products or services that match user needs. This improvement in accuracy helps to reduce ineffective marketing and increase conversion rate, thereby improving overall customer acquisition efficiency. During the iterative calculation process, the system will continuously optimize the matching algorithm based on user feedback and behavioral data, so that the recommended products or services are more in line with user expectations and preferences. This personalized recommendation can enhance user satisfaction and loyalty, thereby improving user experience. In the big data intelligent cloud customer acquisition system, resources, such as marketing budgets and human resources, are limited. Through iterative calculation, the system can more accurately predict which user groups are more likely to be converted into actual customers, thereby optimizing resource allocation and investing limited resources in more potential user groups. During the iterative calculation process, the system will continuously learn and optimize the algorithm to improve the system's computing efficiency and accuracy. This performance improvement helps the system better cope with the needs of large-scale data processing and real-time recommendations. Through the iterative form, the system is completely based on big data for analysis and decision-making, which can more accurately capture market trends and user needs. The overall system can provide customized product and service recommendations based on the user's personalized needs and behavioral data, and the system can easily expand processing capabilities and algorithm complexity to adapt to the growing amount of data and user needs.

[0105] SRPXt，m，p Directly reflects the user's potential demand and preference for financial products or services, and adjusts PSPB through iteration t，m，p，i , the system can optimize the recommendation strategy more directly according to user needs, PSPB t，m，p，i It is an important indicator to measure the matching degree between users and financial products or services. t，m，p To adjust the matching degree PSPB t，m，p，i , making the business logic clearer and easier to understand and implement. During the iteration process, the system can compare SRPX t，m，p This measurable effect helps the system continuously optimize algorithms and recommendation strategies. t，m，p It is calculated based on the user's potential needs and preferences, and the matching degree PSPB t，m，p，i It is an important indicator to measure the degree of match between users and financial products or services. The two are closely linked in logic. t，m，p The change will directly affect the calculation result of the matching degree. In the intelligent cloud customer acquisition system, SRPX t，m，p and matching degree each have different functions and roles. SRPX t，m，p It is mainly used to reflect the user's potential needs and preferences, while matching is used to measure the degree of matching between the user and the product. Therefore, both play an irreplaceable role in the system.

[0106] In the specific implementation process, multiple sub-modules in this method are used to form an intelligent cloud customer acquisition system based on big data. The number of times user m visits the financial website within time period t is calculated as UBDA. t，m , the number of times user m searches for financial keywords in the search engine within time period t UBDB t，m and the usage time of user m on financial apps in time period t UBDC t，m Input to the behavior submodule, output the comprehensive behavior score UBDX of user m in time period t t，m ,UBDX t，m The calculation can quantify the user's financial behavior activity in a specific period of time, providing basic data for potential demand forecasting and intelligent recommendation;

[0107] By calculating the comprehensive score UBDX of user m’s behavior in time period t t，m Input to the demand submodule, output the potential demand forecast value PSPX for user m and financial product p in time period t t，m，p ,PSPX t，m，p The calculation can quantify the matching degree between user profile and financial products, and accurately predict the user's potential demand for specific financial products;

[0108] By calculating the comprehensive score UBDX of user m’s behavior in time period t t，m and the potential demand forecast value PSPX for user m and financial product p in time period t t，m，p Input to the recommendation submodule, output the intelligent recommendation priority SRPX for user m and financial product p in time period t t，m，p , SRPX t，m，p The calculation takes into account multiple factors such as potential demand forecast value, market feedback data and personal credit data, and can derive the priority ranking of intelligent recommendations to improve the accuracy and conversion rate of marketing;

[0109] Based on SRPX t，m，p To iterate the matching degree PSPB between the i-th dimension of the user m’s portrait and the feature of financial product p in time period t t，m，p，i , through SRPX t，m，p To adjust the matching degree PSPB t，m，p，i It makes the business logic clearer, easier to understand and implement, so as to more accurately identify financial products that match user needs. This improvement in accuracy helps to reduce ineffective marketing and increase conversion rates, thereby improving overall customer acquisition efficiency. During the iterative calculation process, the system will continuously optimize the matching algorithm based on user feedback and behavioral data to make the recommended products or services more in line with user expectations and preferences. This personalized recommendation can enhance user satisfaction and loyalty, thereby improving user experience. Through iterative calculation, the system can more accurately predict which user groups are more likely to be converted into actual customers, thereby optimizing resource allocation and investing limited resources in user groups with greater potential.

[0110] This allows the various sub-modules to cooperate with each other in calculations, and to perform overall cycles and iterations, so that the overall system has the effect of automatic optimization and updating, and thus better adaptability.

[0111] See also Figure 1 , Figure 2 and Figure 3 The tools used in the data collection module are: web crawler software, application data analysis tools, user portrait database and credit scoring system; when recommending highly matching financial products to high-priority users in the analysis module, recommendations are made through SMS, email and APP push, and the recommendation effect is tracked and evaluated, including user click-through rate and conversion rate. According to the evaluation results, the recommendation strategy is optimized.

[0112] In this embodiment: the user's access frequency, search keywords, click behavior and other network behavior data are captured from financial websites and search engines through web crawler software, the user's usage time, browsing history, transaction behavior and other data on financial apps are collected through application data analysis tools, the user's age, gender, income, investment preference, credit history and other personal information and portrait data are obtained through the user portrait database, the user's credit status is evaluated through the credit scoring system, and credit scoring data is provided.

[0113] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent cloud customer acquisition system based on big data, characterized by: include: Data collection module: obtains user access frequency, user keyword search frequency and usage time on financial apps from financial websites, search engines and financial apps, and stores them in the database; Processing module: used to input the user's access frequency, the user's keyword search frequency and the usage time on the financial APP, and clean and organize the input data, and output the number of times user m visits the financial website within time period t, the number of times user m searches for financial keywords in the search engine within time period t, and the usage time of user m on the financial APP within time period t; Calculation module: used to input the number of times user m visits financial websites within time period t, the number of times user m searches for financial keywords in search engines within time period t, and the usage time of user m on financial apps within time period t, and output a comprehensive score of user m's behavior within time period t, a potential demand prediction value for user m and financial product p within time period t, and an intelligent recommendation priority for user m and financial product p within time period t; Analysis module: used to formulate personalized recommendation strategies for user m and financial product p based on the comprehensive score of user m's behavior in time period t, the potential demand prediction value for user m and financial product p in time period t, and the user m and financial product p in time period t, and give priority to recommending highly matching financial products to high-priority users.

2. The intelligent cloud customer acquisition system based on big data according to claim 1 is characterized by: The calculation module includes: a behavior submodule, a demand submodule, a recommendation submodule and an optimization submodule.

3. The intelligent cloud customer acquisition system based on big data according to claim 2 is characterized in that: The calculation formula of the behavior submodule is as follows: in: UBDX t，m Refers to the comprehensive score of user m’s behavior in time period t; UBDA t，m Refers to the number of times user m visits the financial website in time period t, UBDB t，m Refers to the number of times user m searches for financial keywords in the search engine within time period t, UBDC t，m Refers to the usage time and UBDD of user m on financial apps in time period t t，m Refers to the total usage time of user m on all apps in time period t, A1 t Refers to UBDA in time period t t，m The weight of A2 t Refers to UBDB in time period t t，m Weight, A3 t Refers to UBDC in time period t t，m The weight of, log refers to the logarithmic function, t refers to the time period, m refers to the user, A1 t ×UBDA t，m Refers to the contribution of visit frequency to the comprehensive score of user behavior; The processing process of the behavior submodule is as follows: the number of times user m visits the financial website in time period t is UBDA t，m , the number of times user m searches for financial keywords in the search engine within time period t UBDB t，m and the usage time of user m on financial apps in time period t UBDC t，m Input to the behavior submodule, and based on the total usage time of user m on all apps in time period t, UBDD t，m , time period t and user m, output the comprehensive score UBDX of user m's behavior in time period t t，m .

4. The intelligent cloud customer acquisition system based on big data according to claim 3 is characterized by: The calculation formula of the demand submodule is as follows: in: PSPX t，m，p Refers to the potential demand forecast value for user m and financial product p in time period t; PSPA p Refers to the number of different attributes contained in the user profile of financial product p, PSPB t，m，p，i Refers to the matching degree between the i-th dimension of the user m’s portrait and the characteristics of financial product p in time period t, PSPC p Refers to the age span and PSPD of the target user group of financial products p Refers to the actual age of user m in time period t, PSPE p refers to the average age of the target user group of the financial product p, p refers to the financial product, Refers to the sum of the matching degrees of all portrait dimensions of user m and financial product p in time period t; The processing process of the demand submodule is as follows: the comprehensive score UBDX of the behavior of user m in time period t is calculated as follows: t，m Input to the demand submodule and based on the number of different attributes PSPA included in the user profile for financial product p p 、The age span of the target user group of financial products PSPC p , the actual age PSPD of user m in time period t p and the average age of the target user group of financial products (PSPE) p Output the potential demand forecast value PSPX for user m and financial product p in time period t t，m，p .

5. The intelligent cloud customer acquisition system based on big data according to claim 4 is characterized in that: The calculation formula of the recommended submodule is as follows: in: SRPX t，m，p Refers to the intelligent recommendation priority for user m and financial product p in time period t; SRPA p Refers to the conversion rate of financial products p to new customers, SRPB p Refers to the sales data of financial product p in time period t, SRPC p Refers to the number of reviews received by financial product p in time period t, SRPD p，t Refers to the favorable rating and SRPE of financial product p in time period t t，m Refers to the credit score and SRPF of user m in time period t p Refers to the minimum credit score, SRPG, among the target user group for financial product p p Refers to the highest credit score among the target user group for financial product p, (SRPE t，m -SRPF p ) / (SRPG p -SRPF p ) refers to the position of the credit score relative to the credit score range of the target users of the financial product, C1 refers to UBDX t，m The weight coefficient, C2 refers to PSPX t，m，p The weight coefficient, C3 refers to SRPB p The weight factor, C4 refers to SRPC p The weight factor, C5 refers to SRPD p，t The weight factor of The processing process of the recommendation submodule is as follows: the comprehensive score UBDX of user m in time period t is calculated as follows: t，m and the potential demand prediction value PSPX for user m and financial product p in time period t t，m，p Input to the recommendation submodule, and based on the conversion rate SRPA of financial product p to new customers p 、The sales data of financial product p in time period t SRPB p and the number of reviews SRPC received by financial product p in time period t p Output the intelligent recommendation priority SRPX for user m and financial product p in time period t t，m，p .

6. The intelligent cloud customer acquisition system based on big data according to claim 5 is characterized in that: The processing process of the optimization submodule is as follows: First up: PSPB t，m，p，i，new =PSPB t，m，p，i +α×(SRPX t，m，p -SRPP); Second: Set the loop termination condition: Termination condition 1: The number of iterations i is 187; Termination condition 2: |SRPX t，m，p，new -SRPX t，m，p，old |<0.00013; in: PSPB t，m，p，i，new Refers to the matching degree between the i-th dimension in the profile of user m in time period t after update and the characteristics of financial product p, SRPP refers to the expected recommendation priority, SRPX t，m，p，new Refers to the intelligent recommendation priority for user m and financial product p in time period t after the update, α refers to the adjustment factor, SRPX t，m，p，old Refers to the intelligent recommendation priority for user m and financial product p in time period t before updating; Based on the intelligent recommendation priority SRPX for user m and financial product p in time period t t，m，p , and adjust the matching degree PSPB between the i-th dimension of the user m’s portrait and the characteristics of the financial product p in the time period t by adjusting the factor α t，m，p，i Output the matching degree PSPB between the i-th dimension of the user m’s portrait and the feature of financial product p in time period t after the update t，m，p，i，new , PSPB t，m，p，i，new Replace PSPB in the requirements submodule t，m，p，i , to constantly adjust PSPX t，m，p and SRPX t，m，p , and finally the loop is terminated by termination conditions one and two.

7. The intelligent cloud customer acquisition system based on big data according to claim 1 is characterized by: The tools used by the data collection module are: web crawler software, application data analysis tools, user portrait database and credit scoring system.

8. The intelligent cloud customer acquisition system based on big data according to claim 1 is characterized by: When the analysis module recommends highly matched financial products to high-priority users, the recommendation is made through SMS, email and APP push, and the recommendation effect is tracked and evaluated, including user click-through rate and conversion rate. The recommendation strategy is optimized based on the evaluation results.

Images