Big data information acquisition and analysis system based on cloud computing

Abstract

The invention relates to the field of big data information acquisition and analysis, in particular to a big data information acquisition and analysis system based on cloud computing, which comprises a data acquisition module, a user grading module, a user preference analysis module, an intelligent recommendation module and a database, the user grading module analyzes and obtains a consumption vitality index, a consumption balance metric value and a credit condition weighted value according to the purchase frequency, the credit score, the violation record and the like, and further divides the users into bronze grade to diamond grade; the user preference analysis module obtains a user behavior preference value by calculating an age interaction feature index, a deep browsing proportion value and a local preference feature value; the intelligent recommendation module is used for classifying the video pages according to the priority and providing personalized recommendation contents for the user; according to the method, accurate user analysis and recommendation are realized by integrating multi-source data, the user experience and the platform operation efficiency are effectively improved, and the method has important application value in the field of big data information collection and analysis.

Description

Technical Field

[0001] The present invention relates to the field of big data information collection and analysis, and in particular to a big data information collection and analysis system based on cloud computing. Background Art

[0002] In today's digital age, the popularity of the Internet and the vigorous development of various online platforms have led to an explosive growth in data; massive amounts of user data, credit data, social data, and text data continue to emerge. How to extract valuable information from these complex data and achieve accurate user analysis and personalized services has become a key challenge faced by various industries.

[0003] Relying solely on the purchase records of users on a single platform for analysis will ignore important information such as their interest expressions on social media and browsing preferences on other platforms, making it difficult to gain a comprehensive understanding of user needs. The analysis methods for users are relatively simple and crude, lacking an effective mechanism to comprehensively consider the multi-dimensional behavioral characteristics of users. In terms of user classification, it is usually based on a single indicator such as purchase amount or credit score, which cannot accurately reflect the user's consumption potential, the stability of consumption preferences, and the true status of credit risk; as users' expectations for online service experience continue to increase, the demand for accurate recommendations and customized services has become more urgent; most existing recommendation systems are based on simple rules or general algorithms, which are difficult to meet the diverse needs of users, thus affecting user satisfaction and platform competitiveness.

[0004] In order to solve the above defects, a technical solution is now provided. Summary of the invention

[0005] In order to solve the technical problems raised by the above-mentioned background technology, the present invention provides a big data information collection and analysis system based on cloud computing.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] The present invention is a big data information collection and analysis system based on cloud computing, comprising a data acquisition module, a user classification module, a user preference analysis module, an intelligent recommendation module and a database.

[0008] The data acquisition module connects with various data sources according to the designed multi-source data acquisition interface, obtains different data classification storage, and then sets the acquisition rule interface according to the stored data. The specific process is as follows:

[0009] Build an extensible interface framework and connect with various data source interfaces to collect various types of data, intelligently identify and classify various types of data, including user data, credit data, social data and text data. Identify the age, geographical location and browsing history in user data and store them in the database; identify the user's purchase frequency on the platform, credit score and violation record credit status in the credit data and store them in the database; identify the number of comments, number of link clicks and social copy in the text data and store them in the database;

[0010] According to the display screen, set the collection rule interface for each type of data that needs to be obtained. Specifically: the administrator sets the specified data range, collection time interval and data limit parameters through the display screen; the data range, collection time interval and data limit are divided into six levels and defined with numbers, namely 1, 2, 3, 4, 5 and 6, 6 means the highest level, and the administrator sets the collection rules by clicking the corresponding number point.

[0011] The user classification module analyzes users based on purchase frequency, credit score and credit status of violation records, obtains consumption vitality index, consumption trade-off metric value and credit status weighted value and divides them into different levels. The specific process is as follows:

[0012] The user classification module is provided with an analysis unit and a classification unit, which obtain the purchase frequency of each user in March, extract the purchase number and purchase amount from the purchase frequency of each user, obtain the total consumption value of each user according to the purchase number and purchase amount, and obtain the consumption trajectory curve of the user with time as the horizontal axis and the total consumption value as the vertical axis; draw a tangent to the curve at the purchase time point, calculate the tangent slope and record it as the amount slope; extract the preset consumption amount from the database, sum up the amount slopes greater than the preset consumption amount to obtain the consumption increase degree; sum up the amount slopes less than the preset consumption amount to obtain the consumption reduction degree; then take the purchase number as the horizontal axis and the purchase interval as the vertical axis to obtain the consumption interval curve, connect two adjacent time intervals to obtain the consumption interval slope, mark the consumption increase degree, consumption reduction degree and consumption interval slope as KQ, KW and KE respectively, and enter them into the formula for calculation after normalization. The consumption vitality index DQ of each user is obtained, where J1, J2 and J3 are weight factors corresponding to the consumption increase degree, consumption decrease degree and consumption interval slope;

[0013] Then obtain the number of returns, return amount, number of discounted product purchases, and discounted product purchase amount of each user, and mark them as HK, HT, HR, and HE respectively. After normalization, substitute them into the formula for calculation Obtain the consumption trade-off metric DP of each user, where J4, J5, J6 and J7 are weight coefficients corresponding to the number of returns, the amount of returns, the number of discounted product purchases and the amount of discounted product purchases;

[0014] Obtain each user's credit score and violation record credit status, identify and extract the user's repayment history, debt level, credit account usage years and recent credit query times, and mark them as SK, SL, SH and SF respectively, normalize them and substitute them into the formula for calculation The weighted value DW of each user's credit status is obtained, where J8, J9, J10 and J11 are weight factors of repayment history, debt level, credit account usage years and recent credit inquiry times respectively;

[0015] The grading unit divides users into five levels according to their consumption vitality index and consumption trade-off metric values. The five levels include bronze, silver, gold, platinum and diamond. Specifically, bronze is in the range of DQ<-1, DP<0.3 and DW<-0.5, which means that the consumption amount of users at this level is growing slowly, and the number of consumption is relatively small and unstable, so they are classified as bronze users; silver is in the range of -1≤DQ<0, 0.3≤DP<0.5 and -0.5≤DW<0.2, which means that the consumption amount of users at this level has an increasing trend. If the consumption amount of users at this level increases significantly but at a slower speed, they are classified as silver users; if the consumption amount of users at this level increases significantly but at a slower speed, they are classified as silver users; if the consumption amount of users at this level increases significantly but at a slower speed, they are classified as gold users; if the consumption amount of users at this level increases significantly but at a slower speed, they are classified as platinum users; if the consumption amount of users at this level increases significantly but at a slower speed, they are classified as platinum users; if the consumption amount of users at this level increases significantly but at a slower speed, they are classified as platinum users; if the consumption amount of users at this level increases significantly but at a slower speed, they are classified as platinum users; if the consumption amount of users at this level increases significantly but at a slower speed, they are classified as platinum users; if the consumption amount of users at this level increases significantly but at a slower speed, they are classified as platinum users.

[0016] The user preference analysis module analyzes the user's age-time interaction characteristic index, depth browsing ratio value, and local preference characteristic value, and obtains the user behavior preference value. The specific process is as follows:

[0017] The age, region and browsing history in the user data are obtained in real time. The time point when each user registers on the platform is marked as the registration time point, and the current time point of the system is marked as the end time point. The registration time point and the end time point are subtracted to obtain the platform registration time. The platform registration time of each user and the corresponding user age are calculated as the interaction item feature to obtain the age-time interaction feature index;

[0018] According to the browsing history of each user, the category of the corresponding video page is obtained, and the categories include news, entertainment, technology and shopping; the dwell time and browsing times of each category of each user are obtained in turn, and the preset dwell time in the database is extracted. If the dwell time of a video page is longer than the preset dwell time, then this time is recorded as the effective dwell time, and the effective dwell time of each category of each user is added to obtain the total effective dwell time, and the total effective dwell time of each category is arranged from large to small to obtain a page dwell time table; then the number of comments and the number of link clicks of the category of each video page are obtained, and the page dwell time table is matched with the number of comments and the number of link clicks of the category, and the standard effective dwell time in the database is extracted. If the effective dwell time of a page in the page dwell table is longer than the standard effective dwell time, and there are comments or link clicks, then the page is marked as a deep dwell page and the total number is counted to obtain the number of deep dwell pages;

[0019] Obtain the total number of valid stay pages for each user in the database, and obtain the depth browsing ratio value based on the total number of valid stay pages and the number of deep stay pages;

[0020] According to the regional positioning and social copy of each user, the social copy theme of each user corresponding to the regional positioning is identified and classified to obtain several theme categories corresponding to the regional positioning of each user, including food, travel, sports and technology; the number of times each theme category is mentioned in each region is counted to obtain the number of times each theme category is mentioned in each region kAi, and the formula is used to calculate Get each user's local preference feature value fA i , where A represents the region number, i represents the category number of each topic, and n represents the total number of social copy used for analysis in the region;

[0021] The age-related interaction feature index and the depth browsing ratio are marked as KDc and GHa, respectively, and combined with the local preference feature value fA i To perform calculations, use the formula The user behavior preference value GTE is obtained, where R1, R2 and R3 are preset proportional coefficients, g represents the user number, and m1 represents the total number of users; the user behavior preference value is sent to the intelligent recommendation module.

[0022] The intelligent recommendation module provides personalized recommendation content for users based on the real-time recommendation priority. The specific process is as follows:

[0023] According to the user's behavior preference value, the category of the video page is divided into several priorities for recommendation to the user. Specifically: according to the behavior preference value of each user, news items, entertainment items, technology items and shopping items are arranged according to the priority, and marked as the first-level category, the second-level category, the third-level category and the fourth-level category respectively. Any user is selected as the preferred object, the behavior preference value of the corresponding preferred object is obtained and the priority is divided, and the category of the video page of the preferred object is pushed. First, the first-level category is pushed to the user in turn. After the push of the first category is completed, the content of the first category is still retrieved. If there is no first-level category content, the second-level category is pushed to the user. After the push of the second-level category is completed, the content of the second category is still retrieved. If there is no second-level category content, the third category is pushed; and so on, until the push of the categories of all video pages is completed.

[0024] Compared with the prior art, the beneficial effects of the present invention are: by integrating multi-source data, comprehensively considering multi-dimensional factors such as purchase frequency, credit score, credit status of violation records, age-time interaction characteristic indicators, in-depth browsing ratio value and local preference characteristic value, a very accurate and comprehensive user portrait is constructed. This in-depth analysis can accurately divide user levels, such as bronze to diamond, so that enterprises can clearly understand the consumption potential, consumption preference stability and credit risk level of each user, and provide a solid foundation for subsequent differentiated services. For high-level users, exclusive high-end product recommendations and high-quality service experience can be provided, while for low-level users with higher credit risks, targeted risk prevention and control measures can be implemented, effectively improving the risk management and marketing accuracy of enterprises; according to the user behavior preference value, the video page category is prioritized to achieve highly personalized recommendation services. The system no longer relies on traditional general recommendation rules, but deeply explores the individual interest differences and dynamic behavior changes of users to ensure that the recommended content closely matches the actual needs of users; whether it is news, entertainment, technology or shopping information, it can be presented in the order that best suits the interests of users, significantly improving the user's attention and acceptance of recommended content, thereby enhancing the user's participation and stickiness on the platform and promoting business conversion. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. The following drawings are not intentionally scaled to the actual size, and the focus is on illustrating the main purpose of the present invention.

[0026] Figure 1 It is a principle block diagram of the present invention. DETAILED DESCRIPTION

[0027] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work also fall within the scope of protection of the present invention.

[0028] Please refer to Figure 1 As shown, the present invention is a big data information collection and analysis system based on cloud computing, including a data acquisition module, a user classification module, a user preference analysis module, an intelligent recommendation module and a database.

[0029] The data acquisition module connects with various data sources according to the designed multi-source data acquisition interface, obtains different data classification storage, and then sets the acquisition rule interface according to the stored data. The specific process is as follows:

[0030] Build an extensible interface framework and connect with various data source interfaces to collect various types of data, intelligently identify and classify various types of data, including user data, credit data, social data and text data. Identify the age, geographical location and browsing history in user data and store them in the database; identify the user's purchase frequency on the platform, credit score and violation record credit status in the credit data and store them in the database; identify the number of comments, number of link clicks and social copy in the text data and store them in the database;

[0031] According to the display screen, set the collection rule interface for each type of data that needs to be obtained. Specifically: the administrator sets the specified data range, collection time interval and data limit parameters through the display screen; the data range, collection time interval and data limit are divided into six levels and defined with numbers, namely 1, 2, 3, 4, 5 and 6, 6 means the highest level, and the administrator sets the collection rules by clicking the corresponding number point.

[0032] The user classification module analyzes users based on purchase frequency, credit score and credit status of violation records, obtains consumption vitality index, consumption trade-off metric value and credit status weighted value and divides them into different levels. The specific process is as follows:

[0033] The user classification module is provided with an analysis unit and a classification unit, which obtain the purchase frequency of each user in March, extract the purchase number and purchase amount from the purchase frequency of each user, obtain the total consumption value of each user according to the purchase number and purchase amount, and obtain the consumption trajectory curve of the user with time as the horizontal axis and the total consumption value as the vertical axis; draw a tangent to the curve at the purchase time point, calculate the tangent slope and record it as the amount slope; extract the preset consumption amount from the database, sum up the amount slopes greater than the preset consumption amount to obtain the consumption increase degree; sum up the amount slopes less than the preset consumption amount to obtain the consumption reduction degree; then take the purchase number as the horizontal axis and the purchase interval as the vertical axis to obtain the consumption interval curve, connect two adjacent time intervals to obtain the consumption interval slope, mark the consumption increase degree, consumption reduction degree and consumption interval slope as KQ, KW and KE respectively, and enter them into the formula for calculation after normalization. The consumption vitality index DQ of each user is obtained, where J1, J2 and J3 are weight factors corresponding to the consumption increase degree, consumption decrease degree and consumption interval slope;

[0034] Then obtain the number of returns, return amount, number of discounted product purchases, and discounted product purchase amount of each user, and mark them as HK, HT, HR, and HE respectively. After normalization, substitute them into the formula for calculation Obtain the consumption trade-off metric DP of each user, where J4, J5, J6 and J7 are weight coefficients corresponding to the number of returns, the amount of returns, the number of discounted product purchases and the amount of discounted product purchases;

[0035] Obtain each user's credit score and violation record credit status, identify and extract the user's repayment history, debt level, credit account usage years and recent credit query times, and mark them as SK, SL, SH and SF respectively, normalize them and substitute them into the formula for calculation The weighted value DW of each user's credit status is obtained, where J8, J9, J10 and J11 are weight factors of repayment history, debt level, credit account usage years and recent credit inquiry times respectively;

[0036] The grading unit divides users into five levels according to their consumption vitality index and consumption trade-off metric values. The five levels include bronze, silver, gold, platinum and diamond. Specifically, bronze is in the range of DQ<-1, DP<0.3 and DW<-0.5, which means that the consumption amount of users at this level is growing slowly, and the number of consumption is relatively small and unstable, so they are classified as bronze users; silver is in the range of -1≤DQ<0, 0.3≤DP<0.5 and -0.5≤DW<0.2, which means that the consumption amount of users at this level has an increasing trend. If the consumption amount of users at this level increases significantly but at a slower speed, they are classified as silver users; if the consumption amount of users at this level increases significantly but at a slower speed, they are classified as silver users; if the consumption amount of users at this level increases significantly but at a slower speed, they are classified as gold users; if the consumption amount of users at this level increases significantly but at a slower speed, they are classified as platinum users; if the consumption amount of users at this level increases significantly but at a slower speed, they are classified as platinum users; if the consumption amount of users at this level increases significantly but at a slower speed, they are classified as platinum users; if the consumption amount of users at this level increases significantly but at a slower speed, they are classified as platinum users; if the consumption amount of users at this level increases significantly but at a slower speed, they are classified as platinum users; if the consumption amount of users at this level increases significantly but at a slower speed, they are classified as platinum users.

[0037] The user preference analysis module analyzes the user's age-time interaction characteristic index, depth browsing ratio value, and local preference characteristic value, and obtains the user behavior preference value. The specific process is as follows:

[0038] The age, region and browsing history in the user data are obtained in real time. The time point when each user registers on the platform is marked as the registration time point, and the current time point of the system is marked as the end time point. The registration time point and the end time point are subtracted to obtain the platform registration time. The platform registration time of each user and the corresponding user age are calculated as the interaction item feature to obtain the age-time interaction feature index;

[0039] According to the browsing history of each user, the category of the corresponding video page is obtained, and the categories include news, entertainment, technology and shopping; the dwell time and number of views of each category of each user are obtained in turn, and the preset dwell time in the database is extracted. If the dwell time of a certain video page is longer than the preset dwell time, then this time is recorded as the effective dwell time, and the effective dwell time of each category of each user is added to obtain the total effective dwell time, and the total effective dwell time of each category is arranged from large to small to obtain a page dwell time table; then the number of comments and the number of link clicks of the category of each video page are obtained, and the page dwell time table is matched with the number of comments and the number of link clicks of the category, and the standard effective dwell time in the database is extracted. If the effective dwell time of a page in the page dwell table is longer than the standard effective dwell time, and there are comments or link clicks, then the page is marked as a deep dwell page and the total number is counted to obtain the number of deep dwell pages;

[0040] Obtain the total number of valid stay pages for each user in the database, and obtain the depth browsing ratio value based on the total number of valid stay pages and the number of deep stay pages;

[0041] According to the regional positioning and social copy of each user, the social copy theme of each user corresponding to the regional positioning is identified and classified to obtain several theme categories corresponding to the regional positioning of each user, including food, travel, sports and technology; the number of times each theme category is mentioned in each region is counted to obtain the number of times each theme category is mentioned in each region kAi, and the formula is used to calculate Get each user's local preference feature value fA i , where A represents the region number, i represents the category number of each topic, and n represents the total number of social copy used for analysis in the region;

[0042] The age-related interaction feature index and the depth browsing ratio are marked as KDc and GHa, respectively, and combined with the local preference feature value fA i To perform calculations, use the formula The user behavior preference value GTE is obtained, where R1, R2 and R3 are preset proportional coefficients, g represents the user number, and m1 represents the total number of users; the user behavior preference value is sent to the intelligent recommendation module.

[0043] The intelligent recommendation module provides personalized recommendation content for users based on the real-time recommendation priority. The specific process is as follows:

[0044] According to the user's behavior preference value, the category of the video page is divided into several priorities for recommendation to the user. Specifically: according to the behavior preference value of each user, news items, entertainment items, technology items and shopping items are arranged according to the priority, and marked as the first-level category, the second-level category, the third-level category and the fourth-level category respectively. Any user is selected as the preferred object, the behavior preference value of the corresponding preferred object is obtained and the priority is divided, and the category of the video page of the preferred object is pushed. First, the first-level category is pushed to the user in turn. After the push of the first category is completed, the content of the first category is still retrieved. If there is no first-level category content, the second-level category is pushed to the user. After the push of the second-level category is completed, the content of the second category is still retrieved. If there is no second-level category content, the third category is pushed; and so on, until the push of the categories of all video pages is completed.

[0045] The above is an explanation of the present invention and should not be considered as a limitation thereof. Although several exemplary embodiments of the present invention have been described, it will be readily appreciated by those skilled in the art that many modifications may be made to the exemplary embodiments without departing from the novel teachings and advantages of the present invention. Therefore, all such modifications are intended to be included within the scope of the present invention as defined in the claims. It should be understood that the above is an explanation of the present invention and should not be considered as being limited to the specific embodiments disclosed, and modifications to the disclosed embodiments and other embodiments are intended to be included within the scope of the appended claims. The present invention is defined by the claims and their equivalents.

Claims

1. A big data information collection and analysis system based on cloud computing, including a data acquisition module, a user classification module, an intelligent recommendation module and a database, characterized in that: It also includes a user preference analysis module; The user classification module is equipped with an analysis unit and a classification unit, which analyzes users according to purchase frequency, credit score and credit status of violation records, obtains consumption vitality index, consumption trade-off measurement value and credit status weighted value and divides them into different levels; the intelligent recommendation module intelligently provides personalized recommendation content for users according to real-time recommendation priority; The user preference analysis module analyzes the user's age-time interaction feature index and in-depth browsing ratio value, specifically: obtaining the age, region and browsing history in the user data in real time, marking the time point when each user registers on the platform as the registration time point, and ending the system's current time point as the end time point, subtracting the registration time point from the end time point to obtain the platform registration time, and calculating the interaction item feature of each user's platform registration time and the corresponding user age to obtain the age-time interaction feature index; According to the browsing history of each user, the category to which the corresponding video page belongs is obtained; the dwell time and the number of views of each category of each user are obtained in turn, and the preset dwell time in the database is extracted. If the dwell time of a video page is longer than the preset dwell time, this time is recorded as the effective dwell time, and the effective dwell time of each category of each user is added to obtain the total effective dwell time, and the total effective dwell time of each category is arranged from large to small to obtain a page dwell time table; then the number of comments and the number of link clicks of the category to which each video page belongs are obtained, and the page dwell time table is matched with the number of comments and the number of link clicks of the category to which it belongs, and the standard effective dwell time in the database is extracted. If the effective dwell time of a page in the page dwell table is longer than the standard effective dwell time, and there are comments or link clicks, the page is marked as a deep dwell page and the total number is counted to obtain the number of deep dwell pages; the total number of effective dwell pages of each user in the database is obtained, and the deep browsing ratio value is obtained according to the total number of effective dwell pages and the number of deep dwell pages.

2. The cloud computing-based big data information collection and analysis system according to claim 1 is characterized in that: The user preference analysis module obtains the local preference feature value, specifically: According to the regional positioning and social copy of each user, the social copy theme of each user corresponding to the regional positioning is identified and classified to obtain several theme categories corresponding to the regional positioning of each user, including food, travel, sports and technology; the number of times each theme category is mentioned in each region is counted to obtain the number of times each theme category is mentioned in each region kAi, and the formula is used to calculate Get each user's local preference feature value fA i , where A represents the region number, i represents the category number of each topic, and n represents the total number of social copy used for analysis in the region.

3. The cloud computing-based big data information collection and analysis system according to claim 1, characterized in that: The user preference analysis module obtains the following by comprehensive calculation of the obtained data: The age-related interaction feature index and the depth browsing ratio are marked as KDc and GHa, respectively, and combined with the local preference feature value fA i To perform calculations, use the formula The user behavior preference value GTE is obtained, where R1, R2 and R3 are preset proportional coefficients, g represents the user number, and m1 represents the total number of users; the user behavior preference value is sent to the intelligent recommendation module.

4. The cloud computing-based big data information collection and analysis system according to claim 1, characterized in that: The user classification module analyzes users based on purchase frequency, credit score and credit status of violation records, obtains consumption vitality index, consumption trade-off metric value and credit status weighted value and divides them into different levels. The specific process is as follows: Get the purchase frequency of each user in March, extract the number of purchases and the purchase amount from the purchase frequency of each user, get the total consumption value of each user based on the number of purchases and the purchase amount, and get the consumption trajectory curve of the user with time as the horizontal axis and the total consumption value as the vertical axis; Draw a tangent to the curve at the purchase time point, calculate the tangent slope and record it as the amount slope; extract the preset consumption amount from the database, sum the slopes of the amounts greater than the preset consumption amount to obtain the consumption increase; sum the slopes of the amounts less than the preset consumption amount to obtain the consumption decrease; then use the number of purchases as the horizontal axis and the purchase interval as the vertical axis to obtain a consumption interval curve, connect two adjacent time intervals to obtain the consumption interval slope, mark the consumption increase degree, consumption decrease degree and consumption interval slope as KQ, KW and KE respectively, and enter them into the formula for calculation after normalization. The consumption vitality index DQ of each user is obtained, where J1, J2 and J3 are weight factors corresponding to the consumption increase degree, consumption decrease degree and consumption interval slope; Then obtain the number of returns, return amount, number of discounted product purchases, and discounted product purchase amount of each user, and mark them as HK, HT, HR, and HE respectively. After normalization, substitute them into the formula for calculation Obtain the consumption trade-off metric DP of each user, where J4, J5, J6 and J7 are weight coefficients corresponding to the number of returns, the amount of returns, the number of discounted product purchases and the amount of discounted product purchases; Obtain each user's credit score and violation record credit status, identify and extract the user's repayment history, debt level, credit account usage years and recent credit query times, and mark them as SK, SL, SH and SF respectively, normalize them and substitute them into the formula for calculation The weighted value DW of each user's credit status is obtained, where J8, J9, J10 and J11 are weight factors of repayment history, debt level, credit account usage years and recent credit inquiry times, respectively.

5. The cloud computing-based big data information collection and analysis system according to claim 1, characterized in that: The grading unit divides the users into five levels according to the consumption vitality index and consumption trade-off metric value of each user, specifically: The five levels include bronze, silver, gold, platinum and diamond. Specifically, bronze users are classified as bronze users if their DQ is in the range of -1, DP is in the range of 0.3 and DW is in the range of -0.5; silver users are classified as silver users if their -1≤DQ<0, 0.3≤DP<0.5 and -0.5≤DW<0.2; gold users are classified as gold users if their 0≤DQ<1, 0.5≤DP<0.7 and 0.2≤DW<0.8; platinum users are classified as platinum users if their 1≤DQ<2, 0.7≤DP<0.9 and 0.8≤DW<1.5; and diamond users are classified as platinum users if their DQ is in the range of 2, DP≥0.9 and DW≥1.

5.

6. The cloud computing-based big data information collection and analysis system according to claim 1, characterized in that: The intelligent recommendation module provides personalized recommendation content for users based on the real-time recommendation priority. The specific process is as follows: According to the user behavior preference value, the category of the video page is divided into several priorities for recommendation to the user. Specifically: according to the behavior preference value of each user, the news items, entertainment items, technology items and shopping items are arranged according to the priority, and are marked as the first-level category, the second-level category, the third-level category and the fourth-level category respectively. Any user is selected as the preferred object, the behavior preference value of the corresponding preferred object is obtained and the priority is divided, and the category of the video page of the preferred object is pushed. First, the first-level category is pushed to the user in turn. After the push of the first category is completed, the content of the first category is still retrieved. If there is no first-level category content, the second-level category is pushed to the user. After the push of the second-level category is completed, the content of the second category is still retrieved. If there is no second-level category content, the third category is pushed; and so on, until the push of the categories of all video pages is completed.

7. The cloud computing-based big data information collection and analysis system according to claim 1, characterized in that: The data acquisition module connects with various data sources according to the designed multi-source data acquisition interface, obtains different data classification storage, and then sets the acquisition rule interface according to the stored data. The specific process is as follows: Build an extensible interface framework and connect with various data source interfaces to collect various types of data, intelligently identify and classify various types of data, including user data, credit data, social data and text data, and store the age, geographical location and browsing history in the user data in the database after identification; The purchase frequency, credit score and violation record credit status of users on the platform in the credit data are identified and stored in the database after classification; the number of comments, number of link clicks and social copywriting in the text data are identified and stored in the database after classification.

8. The cloud computing-based big data information collection and analysis system according to claim 1, characterized in that: Set the collection rule interface for each type of data to be acquired according to the display screen, specifically: The administrator sets the specified data range, collection time interval and data limit parameters through the display screen; the data range, collection time interval and data limit are divided into six levels and defined with numbers, namely 1, 2, 3, 4, 5 and 6, and the administrator sets the collection rules by clicking the corresponding number points.

Images