E-commerce data analysis method and system based on cloud platform, and electronic equipment

Abstract

The invention discloses an e-commerce data analysis method and system based on a cloud platform and electronic equipment, and relates to the technical field of e-commerce data analysis. Performing conjoint analysis on the real-time data and the primary adjustment result to judge whether the server and the storage resources need to be subjected to secondary adjustment or not, and judging whether the server and the storage resources need to be subjected to static adjustment or not after analyzing the states of the server and the storage resources on the current day; and combining the real-time data obtained in the next day with the states of the server and the storage resources in the next day to judge whether the server and the storage resources need to be dynamically adjusted or not. After traffic prediction is carried out through the cloud platform, the elastic expansion capability of the cloud platform can dynamically adjust calculation and storage resources according to service requirements, system collapse in peak periods is avoided, resource investment can be reduced in low-traffic periods, and therefore the cost is optimized.

Description

Technical Field

[0001] The present invention relates to the technical field of e-commerce data analysis, and in particular to an e-commerce data analysis method, system and electronic equipment based on a cloud platform. Background Art

[0002] With the rapid development of the e-commerce industry, e-commerce platforms generate massive amounts of data every day, including user behavior data, transaction data, product information, inventory, logistics information, etc. These data are large in volume, diverse in types, and generated quickly, and need to be analyzed and processed by an analysis system.

[0003] Existing e-commerce platforms usually pre-configure servers and storage resources, which has the following drawbacks:

[0004] 1. Configure a large number of servers and storage resources. E-commerce platforms usually configure servers and storage resources based on estimated peak traffic, which results in a large amount of computing resources being idle during non-peak hours, resulting in resource waste and high operating costs.

[0005] 2. With fewer servers and storage resources, when faced with sudden traffic, the system may crash, freeze, or respond slowly due to excessive server load, which seriously affects the user experience.

[0006] 3. When real-time traffic increases, manual intervention is usually required, which may cause the system to be unable to respond quickly to sudden traffic, or even cause problems with untimely resource expansion. During low-traffic periods, manual adjustment of resources will slow down the response speed.

[0007] Based on this, the present invention proposes an e-commerce data analysis method, system and electronic device based on a cloud platform. After traffic prediction is performed through the cloud platform, the elastic expansion capability of the cloud platform can dynamically adjust computing and storage resources according to business needs to avoid system crashes during peak periods. At the same time, it can also reduce resource investment during low traffic periods, thereby optimizing costs. Summary of the invention

[0008] The purpose of the present invention is to provide an e-commerce data analysis method, system and electronic device based on a cloud platform to address the deficiencies in the background technology.

[0009] In order to achieve the above object, the present invention provides the following technical solution: an e-commerce data analysis method based on a cloud platform, the analysis method comprising the following steps:

[0010] After the analysis system obtains the date of the next day, it analyzes whether the date of the next day is a special date. When the date of the next day is a special date, the cloud platform obtains multiple historical data of the special date and adjusts the server and storage resources in combination with the status of the server and storage resources on that day.

[0011] The analysis system obtains the real-time data of the e-commerce platform the next day, and analyzes the real-time data together with the results of the first adjustment to determine whether it is necessary to make a second adjustment to the server and storage resources. On special dates, the e-commerce platform operates based on the server and storage resources after the first or second adjustment.

[0012] If the date of the next day after the analysis is not a special date, the server and storage resource status of that day is analyzed to determine whether static adjustment of the server and storage resources is required;

[0013] The real-time data obtained on the next day is combined with the status of the server and storage resources on the next day to determine whether it is necessary to dynamically adjust the server and storage resources. On ordinary dates, the e-commerce platform operates based on the initial status, static adjustment or dynamically adjusted server and storage resources.

[0014] In a preferred embodiment, adjusting the server and storage resources once in combination with the server and storage resource status of the day includes the following steps:

[0015] Get multiple data from several years before a specific date, including traffic peak change factors and average page click-through rates;

[0016] Obtain the request timeout frequency of the server and the read and write delay frequency of the memory;

[0017] Comprehensively analyze the traffic peak change factor, average page click rate, request timeout frequency, and read / write delay frequency to obtain a first adjustment coefficient;

[0018] The initial number of servers and the initial number of storage devices are corrected based on the first adjustment coefficient, and the expression is: Where fs c is the corrected number of servers, cs c is the corrected memory quantity, fs h is the initial number of servers, cs h is the initial memory quantity, tj one is the first adjustment coefficient.

[0019] In a preferred embodiment, the traffic peak change factor, the average page click rate, the request timeout frequency, and the read / write delay frequency are comprehensively analyzed to obtain the first adjustment coefficient, which is expressed as follows:

[0020] In the formula, tj one is the first adjustment coefficient, δ is the traffic peak change factor, ε is the average page click rate, τ is the request timeout frequency, is the read and write delay frequency, e (δ+ε) For historical data, is the state data, α and β are the proportional coefficients of historical data and state data, respectively, and both are greater than 0.

[0021] In a preferred embodiment, the analysis system obtains the real-time data of the e-commerce platform on the next day, and analyzes the real-time data together with the primary adjustment result to determine whether a secondary adjustment of the server and storage resources is required, including the following steps:

[0022] The analysis system obtains the real-time data of the e-commerce platform on the next day. The real-time data includes the traffic growth rate and the user online growth rate, and obtains the correction index based on the traffic growth rate and the user online growth rate;

[0023] The second adjustment coefficient is obtained by correcting the first adjustment coefficient through the correction index, and the expression is: two =tj one *XZ, where tj two is the second adjustment coefficient, tj one is the first adjustment coefficient, XZ is the correction index;

[0024] The correction index is compared with the index threshold, and the index threshold is used to determine whether the corrected number of servers and the corrected number of storage devices need to be corrected by the second adjustment coefficient. If the correction index is less than or equal to the index threshold, it is determined that the corrected number of servers and the corrected number of storage devices do not need to be corrected by the second adjustment coefficient. If the correction index is greater than the index threshold, it is determined that the corrected number of servers and the corrected number of storage devices need to be corrected by the second adjustment coefficient. The expression is: Where fs c is the corrected number of servers, cs c is the corrected memory quantity, fs ct is the number of servers after the second correction, cs ct is the number of memories after secondary correction, tj two is the second adjustment coefficient.

[0025] In a preferred embodiment, obtaining the correction index based on the traffic growth rate and the user online growth rate includes the following steps:

[0026] The traffic growth rate and the user online growth rate are normalized to map their value ranges to [0,1], and the normalized value of the traffic growth rate and the normalized value of the user online growth rate are obtained. The correction index is obtained by summing the normalized value of the traffic growth rate and the normalized value of the user online growth rate.

[0027] In a preferred embodiment, when the date of the next day after analysis is not a special date, the server and storage resource status of the day are analyzed to determine whether static adjustment of the server and storage resources is required, including the following steps:

[0028] When the date of the next day is not a special date, obtain the request timeout frequency of the server and the read and write delay frequency of the memory on that day;

[0029] Normalize the request timeout frequency and the read-write delay frequency to obtain the normalized value of the request timeout frequency and the normalized value of the read-write delay frequency, sum the normalized timeout frequency value and the normalized read-write delay frequency value to obtain the abnormality coefficient, compare the abnormality coefficient with the first abnormality threshold, if the abnormality coefficient is less than or equal to the first abnormality threshold, determine that there is no need to statically adjust the server and storage resources, if the abnormality coefficient is greater than the first abnormality threshold, determine that there is a need to statically adjust the server and storage resources, and statically adjust the server and storage resources according to the static adjustment table.

[0030] In a preferred embodiment, the real-time data acquired on the next day is combined with the server and storage resource status on the next day to determine whether it is necessary to dynamically adjust the server and storage resources, including the following steps:

[0031] Get the real-time data for the next day, including the traffic growth rate and the user online growth rate, the server request timeout frequency and the storage read and write delay frequency for the next day;

[0032] The traffic growth rate, user online growth rate, request timeout frequency, and read / write delay frequency are calculated using a linear regression analysis algorithm to obtain the dynamic adjustment coefficient. The expression is: In the formula, tj dynamic is the dynamic adjustment coefficient, R is the influencing factor, R is 1.513, θ1, θ2, θ3, θ4 are the traffic growth rate, user online growth rate, request timeout frequency, and read and write delay frequency, respectively, γ1, γ2, γ3, γ4 are regression coefficients, and the regression coefficients are greater than 0;

[0033] The dynamic adjustment coefficient is compared with the dynamic adjustment threshold. The dynamic adjustment threshold is used to predict the overall load change trend of the e-commerce platform. If the dynamic adjustment coefficient is less than or equal to the dynamic adjustment threshold, it is predicted that the e-commerce platform will show a low load trend change. If the dynamic adjustment coefficient is greater than the dynamic adjustment threshold, it is predicted that the e-commerce platform will show a high load trend change.

[0034] When it is predicted that the overall e-commerce platform will show a high load trend, and the initial number of servers and the initial number of memories are not statically adjusted, the expression for adjustment through the dynamic adjustment coefficient is: Where fs c is the corrected number of servers, cs c is the corrected memory quantity, fs h is the initial number of servers, cs h is the initial memory quantity, tj dynamic is the dynamic adjustment coefficient;

[0035] When it is predicted that the e-commerce platform as a whole will show a high load trend, and the initial number of servers and the initial number of memories are statically adjusted, the expression adjusted by the dynamic adjustment coefficient is: Where fs c is the corrected number of servers, cs c is the corrected memory quantity, fs jh is the number of servers after static adjustment, cs jh is the amount of memory after static adjustment, tj dynamic is the dynamic adjustment coefficient.

[0036] In a preferred embodiment, after the analysis system obtains the date of the next day, analyzing whether the date of the next day is a special date includes the following steps:

[0037] The analysis system obtains the date of the next day and obtains all special dates from the special date library of the e-commerce platform;

[0038] Use the acquired date of the next day to match the date in the special date library, and determine whether it is a special date by directly comparing the date;

[0039] If the date of the next day matches a date or date range in the special date library, the analysis system marks the type of the date;

[0040] Based on the result output of the date matching, it is judged whether the next day is a special date and its type. If it is a special date, the information of the special date is fed back. If the next day is not a special date, the analysis system returns the corresponding status.

[0041] E-commerce data analysis system based on cloud platform, including date analysis module, special date adjustment module and general date adjustment module;

[0042] Date analysis module: after obtaining the date of the next day, analyze whether the date of the next day is a special date;

[0043] Special date adjustment module: When the date of the next day is analyzed to be a special date, after obtaining multiple historical data of the special date based on the cloud platform, the server and storage resources are adjusted once in combination with the server and storage resource status of the day. The real-time data of the e-commerce platform is obtained the next day, and the real-time data is combined with the result of the first adjustment to determine whether the server and storage resources need to be adjusted again;

[0044] Ordinary date adjustment module: when the date of the next day is not a special date, the server and storage resource status of the day are analyzed to determine whether static adjustment of the server and storage resources is required. The real-time data obtained on the next day is combined with the server and storage resource status of the next day to determine whether dynamic adjustment of the server and storage resources is required.

[0045] An electronic device for e-commerce data analysis based on a cloud platform comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the e-commerce data analysis method based on a cloud platform are implemented.

[0046] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0047] The present invention analyzes whether the date of the next day is a special date. When the date of the next day is analyzed to be a special date, the server and storage resources are adjusted once in combination with the server and storage resource status of the day, and the real-time data is combined with the result of the first adjustment to determine whether the server and storage resources need to be adjusted twice. When the date of the next day is not a special date, the server and storage resource status of the day are analyzed to determine whether the server and storage resources need to be statically adjusted, and the real-time data obtained the next day is combined with the server and storage resource status of the next day to determine whether the server and storage resources need to be dynamically adjusted. After the traffic forecast is performed through the cloud platform, the elastic expansion capability of the cloud platform can dynamically adjust the computing and storage resources according to business needs, avoiding system crashes during peak periods, and also reducing resource investment during low traffic periods, thereby optimizing costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0049] Figure 1 The present invention is a flow chart of the method. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are 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.

[0051] Example 1: Please refer to Figure 1 As shown, the e-commerce data analysis method based on the cloud platform described in this embodiment includes the following steps:

[0052] After the analysis system obtains the date of the next day, it analyzes whether the date of the next day is a special date. When the analysis shows that the date of the next day is a special date, after obtaining multiple historical data of the special date based on the cloud platform, the server and storage resources are adjusted once in combination with the status of the server and storage resources on that day. The analysis system obtains real-time data of the e-commerce platform on the next day, and jointly analyzes the real-time data with the result of the first adjustment to determine whether a second adjustment of the server and storage resources is required. On a special date, the e-commerce platform operates based on the servers and storage resources after the first adjustment or the second adjustment. When the analysis shows that the date of the next day is not a special date, the status of the server and storage resources on that day is analyzed to determine whether static adjustment of the servers and storage resources is required. The real-time data obtained on the next day is combined with the status of the server and storage resources on the next day to determine whether dynamic adjustment of the servers and storage resources is required. On an ordinary date, the e-commerce platform operates based on the initial state, static adjustment or dynamic adjustment of the servers and storage resources.

[0053] This application analyzes whether the date of the next day is a special date. When the date of the next day is a special date, the server and storage resources are adjusted once in combination with the status of the server and storage resources on that day, and the real-time data is combined with the result of the first adjustment to determine whether the server and storage resources need to be adjusted twice. When the date of the next day is not a special date, the server and storage resource status of that day is analyzed to determine whether the server and storage resources need to be statically adjusted, and the real-time data obtained the next day is combined with the server and storage resource status of the next day to determine whether the server and storage resources need to be dynamically adjusted. After traffic prediction is performed through the cloud platform, the elastic expansion capability of the cloud platform can dynamically adjust computing and storage resources according to business needs to avoid system crashes during peak periods, and can also reduce resource investment during low traffic periods, thereby optimizing costs.

[0054] An electronic device for e-commerce data analysis based on a cloud platform comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein: when the processor executes the computer program, steps of an e-commerce data analysis method based on a cloud platform are implemented.

[0055] Embodiment 2: After the analysis system obtains the date of the next day, it analyzes whether the date of the next day is a special date, including the following steps:

[0056] The analysis system obtains the date of the next day and obtains all special dates that may affect the business from the special date library of the e-commerce platform. The special dates in the special date library are generally divided into the following categories:

[0057] Fixed holidays, shopping festivals, platform-defined promotion days, seasonal events

[0058] Use the acquired date of the next day to match the date in the special date library. You can determine whether it is a special date by directly comparing the date.

[0059] Once the next day's date matches a date or date range in the special date library, the system needs to mark the type of the date;

[0060] If the next day's date matches multiple special dates, the system needs to prioritize these dates based on historical user traffic.

[0061] According to the result output of the date matching, it is judged whether the next day is a special date and its type. If it is a special date, the specific information of the date is fed back. If the next day is not a special date, the system should return the corresponding status, such as "ordinary date".

[0062] When the date of the next day is a special date, after obtaining multiple historical data of the special date based on the cloud platform, the server and storage resources are adjusted once in combination with the server and storage resource status of the day, including the following steps:

[0063] Get multiple data from several years before a specific date, including traffic peak change factors and average page click-through rates;

[0064] Obtain the request timeout frequency of the server and the read and write delay frequency of the memory;

[0065] The first adjustment coefficient is obtained by comprehensively analyzing the traffic peak change factor, average page click rate, request timeout frequency, and read / write delay frequency. The expression is:

[0066] In the formula, tj one is the first adjustment coefficient, δ is the traffic peak change factor, ε is the average page click rate, τ is the request timeout frequency, is the read and write delay frequency, e (δ+ε) For historical data, is the state data, α and β are the proportional coefficients of historical data and state data, respectively, and both are greater than 0;

[0067] The larger the first adjustment coefficient is, the greater the historical traffic peak change factor and the average page click rate on a special date are under comprehensive analysis, and the server request timeout frequency and the memory read and write delay frequency on the day before the special date are also greater. Therefore, the number of servers and memories needs to be increased. The initial number of servers and the initial number of memories are corrected based on the first adjustment coefficient. The expression is: Where fs c is the corrected number of servers, cs c is the corrected memory quantity, fs h is the initial number of servers, cs h is the initial memory quantity, tj one is the first adjustment coefficient;

[0068] It should be noted that since the calculated values ​​of the corrected number of servers and the corrected number of memories may contain decimals, in this application, when the corrected number of servers or the corrected number of memories contains decimals, the decimals are rounded to the nearest digit. For example, if the corrected number of servers is 2.12, then the corrected number of servers is rounded to 3, and so on.

[0069] The analysis system obtains the real-time data of the e-commerce platform the next day, and analyzes the real-time data together with the results of the first adjustment to determine whether a second adjustment of the server and storage resources is required, including the following steps:

[0070] The analysis system obtains real-time data of the e-commerce platform one day later. The real-time data includes the traffic growth rate and the user online growth rate.

[0071] The traffic growth rate and the online user growth rate are normalized so that the value range of the traffic growth rate and the online user growth rate is mapped to [0,1], and the normalized traffic growth rate and the normalized online user growth rate are obtained. The normalized traffic growth rate and the normalized online user growth rate are summed to obtain a correction index. The larger the correction index, the greater the increase in traffic and users of the e-commerce platform in a short period of time.

[0072] Therefore, the first adjustment coefficient is corrected by the correction index to obtain the second adjustment coefficient, and the expression is: two =tj one *XZ, where tj two is the second adjustment coefficient, tj oneis the first adjustment coefficient, XZ is the correction index;

[0073] It should be noted that after obtaining the second adjustment coefficient, the revised number of servers and the revised number of storage devices cannot be directly modified by the second adjustment coefficient;

[0074] The correction index is compared with the index threshold, and the index threshold is used to determine whether the corrected number of servers and the corrected number of storage devices need to be corrected by the second adjustment coefficient. If the correction index is less than or equal to the index threshold, it is determined that the corrected number of servers and the corrected number of storage devices do not need to be corrected by the second adjustment coefficient. If the correction index is greater than the index threshold, it is determined that the corrected number of servers and the corrected number of storage devices need to be corrected by the second adjustment coefficient. The expression is: Where fs c is the corrected number of servers, cs c is the corrected memory quantity, fs ct is the number of servers after the second correction, cs ct is the number of memories after secondary correction, tj two is the second adjustment coefficient.

[0075] If the date of the next day after the analysis is not a special date, the server and storage resource status of the day are analyzed to determine whether static adjustment of the server and storage resources is required, including the following steps:

[0076] When the date of the next day is not a special date, obtain the request timeout frequency of the server and the read and write delay frequency of the memory on that day;

[0077] The request timeout frequency and the read / write delay frequency are normalized to obtain the normalized value of the request timeout frequency and the normalized value of the read / write delay frequency. The normalized value of the timeout frequency and the normalized value of the read / write delay frequency are summed to obtain the abnormality coefficient. The larger the abnormality coefficient is, the worse the status of the server and the storage on that day is. The abnormality coefficient is compared with the first abnormality threshold. If the abnormality coefficient is less than or equal to the first abnormality threshold, it is determined that there is no need to perform static adjustment on the server and storage resources. If the abnormality coefficient is greater than the first abnormality threshold, it is determined that there is a need to perform static adjustment on the server and storage resources, and the server and storage resources are statically adjusted according to the static adjustment table. The static adjustment table is shown in Table 1:

[0078] Table 1 Static adjustment table

[0079]

[0080] In Table 1, yc is the abnormal coefficient, yz1 is the first abnormal threshold, yz2 is the second abnormal threshold, yz3 is the third abnormal threshold, yz4 is the fourth abnormal threshold, yz5 is the fifth abnormal threshold, and yz n-1 is the n-1th abnormal threshold, yz n is the nth abnormal threshold;

[0081] It can be seen from Table 1 that at different gradients, the increase in the initial number of servers and the initial number of storage devices is different. The larger the abnormal coefficient, the greater the increase in the initial number of servers and the initial number of storage devices.

[0082] Combining the real-time data obtained on the next day with the server and storage resource status on the next day to determine whether it is necessary to dynamically adjust the server and storage resources includes the following steps:

[0083] Obtain real-time data for the next day, including the traffic growth rate and the user online growth rate, the server request timeout frequency and the memory read and write delay frequency (obtained in real time) for the next day;

[0084] The traffic growth rate, user online growth rate, request timeout frequency, and read / write delay frequency are calculated using a linear regression analysis algorithm to obtain the dynamic adjustment coefficient. The expression is: In the formula, ij dynamic is the dynamic adjustment coefficient, R is the influencing factor, R is 1.513, θ1, θ2, θ3, θ4 are the traffic growth rate, user online growth rate, request timeout frequency, and read and write delay frequency, respectively, γ1, γ2, γ3, γ4 are regression coefficients, and the regression coefficients are greater than 0;

[0085] In this application, since the dynamic adjustment coefficient is obtained in real time, the overall load change trend of the e-commerce platform is predicted by the dynamic adjustment coefficient;

[0086] The dynamic adjustment coefficient is compared with the dynamic adjustment threshold. The dynamic adjustment threshold is used to predict the overall load change trend of the e-commerce platform. If the dynamic adjustment coefficient is less than or equal to the dynamic adjustment threshold, it is predicted that the e-commerce platform will show a low load trend change. If the dynamic adjustment coefficient is greater than the dynamic adjustment threshold, it is predicted that the e-commerce platform will show a high load trend change.

[0087] When it is predicted that the overall e-commerce platform will show a high load trend, and the initial number of servers and the initial number of memories are not statically adjusted, the expression for adjustment through the dynamic adjustment coefficient is: Where fs c is the corrected number of servers, cs c is the corrected memory quantity, fs h is the initial number of servers, csh is the initial memory quantity, tj dynamic is the dynamic adjustment coefficient;

[0088] When it is predicted that the e-commerce platform as a whole will show a high load trend, and the initial number of servers and the initial number of memories are statically adjusted, the expression adjusted by the dynamic adjustment coefficient is: Where fs c is the corrected number of servers, cs c is the corrected memory quantity, fs jh is the number of servers after static adjustment, cs jh is the amount of memory after static adjustment, tj dynamic is the dynamic adjustment coefficient.

[0089] In this application,

[0090] The calculation logic of the traffic peak change factor is: obtain the traffic peaks of special dates in previous years (for example, the previous five years), establish a data set with the traffic peaks of special dates in previous years, calculate the average traffic peak and the traffic peak standard deviation in the data set (the calculation method belongs to the prior art and will not be repeated in this application), and divide the average traffic peak by the traffic peak standard deviation to obtain the traffic peak change factor. The larger the traffic peak change factor, the larger the traffic peak on special dates in previous years and the smaller the fluctuation, that is, the overall traffic peak on special dates in previous years is larger and more stable.

[0091] The calculation logic of the average page click rate is as follows: after obtaining the page click rates of special dates in previous years (for example, the previous five years), the average page click rate is calculated. The larger the average page click rate, the larger the overall historical page click rate of special dates. The calculation logic of the page click rate is as follows: obtain the number of pages in the e-commerce platform, obtain the number of clicks on each page, sum the number of clicks on all pages to obtain the total number of clicks, obtain the number of impressions of each page, sum the number of impressions of all pages to obtain the total number of impressions, and obtain the page click rate by dividing the total number of clicks by the total number of impressions.

[0092] The calculation logic of the request timeout frequency is as follows: obtain the number of request timeouts and the total number of requests of the server in a certain period of time, and divide the number of request timeouts by the total number of requests to obtain the request timeout frequency. The greater the request timeout frequency, the worse the server's response ability.

[0093] The calculation logic of the read and write delay frequency is: obtain the total number of read and write operations and the number of read and write delays of the memory in a certain period of time, and obtain the read and write delay frequency by dividing the number of read and write delays by the total number of read and write operations. The larger the proportion of delays in the memory, the higher the read and write delay frequency.

[0094] The calculation logic of the traffic growth rate is: subtract the traffic of the previous period from the current period to obtain the traffic difference, subtract the current period from the previous period to obtain the monitoring duration, and obtain the traffic growth rate by dividing the traffic difference by the previous monitoring duration.

[0095] The calculation logic of the user online growth rate is as follows: subtract the number of online users in the previous period from the number of online users in the current period to obtain the user difference, subtract the current period from the previous period to obtain the monitoring duration, and obtain the user online growth rate by dividing the user difference by the previous monitoring duration.

[0096] Embodiment 3: The e-commerce data analysis system based on the cloud platform described in this embodiment includes a date analysis module, a special date adjustment module, and a general date adjustment module;

[0097] Date analysis module: after obtaining the date of the next day, analyze whether the date of the next day is a special date, and send the analysis results to the special date adjustment module and the ordinary date adjustment module;

[0098] Special date adjustment module: when the date of the next day is analyzed to be a special date, after obtaining multiple historical data of the special date based on the cloud platform, the server and storage resources are adjusted once in combination with the server and storage resource status of the day, and the real-time data of the e-commerce platform is obtained the next day. The real-time data is combined with the result of the first adjustment to determine whether the server and storage resources need to be adjusted again. On the special date, the e-commerce platform operates based on the server and storage resources adjusted once or twice;

[0099] Ordinary date adjustment module: when the date of the next day is not a special date, the server and storage resource status of the day are analyzed to determine whether static adjustment of the server and storage resources is required. The real-time data obtained on the next day is combined with the server and storage resource status of the next day to determine whether dynamic adjustment of the server and storage resources is required. On ordinary dates, the e-commerce platform operates based on the initial state, static adjustment or dynamically adjusted server and storage resources.

[0100] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0101] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0102] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. An e-commerce data analysis method based on a cloud platform, characterized by: The analytical method comprises the following steps: After the analysis system obtains the date of the next day, it analyzes whether the date of the next day is a special date. When the date of the next day is a special date, the cloud platform obtains multiple historical data of the special date and adjusts the server and storage resources in combination with the status of the server and storage resources on that day. The analysis system obtains the real-time data of the e-commerce platform the next day, and analyzes the real-time data together with the results of the first adjustment to determine whether it is necessary to make a second adjustment to the server and storage resources. On special dates, the e-commerce platform operates based on the server and storage resources after the first or second adjustment. If the date of the next day after the analysis is not a special date, the server and storage resource status of that day is analyzed to determine whether static adjustment of the server and storage resources is required; The real-time data obtained on the next day is combined with the status of the server and storage resources on the next day to determine whether it is necessary to dynamically adjust the server and storage resources. On ordinary dates, the e-commerce platform operates based on the initial status, static adjustment or dynamically adjusted server and storage resources.

2. The e-commerce data analysis method based on the cloud platform according to claim 1, characterized in that: Adjust the server and storage resources based on the server and storage resource status of the day, including the following steps: Get multiple data from several years before a specific date, including traffic peak change factors and average page click-through rates; Obtain the server's request timeout frequency and the memory's read and write delay frequency; Comprehensively analyzing the traffic peak variation factor, the average page click rate, the request timeout frequency, and the read / write delay frequency to obtain a first adjustment coefficient; The initial number of servers and the initial number of storage devices are corrected based on the first adjustment coefficient, and the expression is: Where fs c is the corrected number of servers, cs c is the corrected memory quantity, fs h is the initial number of servers, cs h is the initial memory quantity, tj one is the first adjustment coefficient.

3. The e-commerce data analysis method based on the cloud platform according to claim 2 is characterized in that: The first adjustment coefficient is obtained by comprehensively analyzing the traffic peak change factor, average page click rate, request timeout frequency, and read / write delay frequency. The expression is: In the formula, tj one is the first adjustment coefficient, δ is the traffic peak change factor, ε is the average page click rate, τ is the request timeout frequency, is the read and write delay frequency, e (δ+ε) For historical data, is the state data, α and β are the proportional coefficients of historical data and state data, respectively, and both are greater than 0.

4. The e-commerce data analysis method based on the cloud platform according to claim 3 is characterized by: The analysis system obtains the real-time data of the e-commerce platform the next day, and analyzes the real-time data together with the results of the first adjustment to determine whether a second adjustment of the server and storage resources is required, including the following steps: The analysis system obtains the real-time data of the e-commerce platform on the next day. The real-time data includes the traffic growth rate and the user online growth rate, and obtains the correction index based on the traffic growth rate and the user online growth rate; The second adjustment coefficient is obtained by correcting the first adjustment coefficient through the correction index, and the expression is: two =tj one *XZ, where tj two is the second adjustment coefficient, tj one is the first adjustment coefficient, XZ is the correction index; The correction index is compared with the index threshold, and the index threshold is used to determine whether the corrected number of servers and the corrected number of storage devices need to be corrected by the second adjustment coefficient. If the correction index is less than or equal to the index threshold, it is determined that the corrected number of servers and the corrected number of storage devices do not need to be corrected by the second adjustment coefficient. If the correction index is greater than the index threshold, it is determined that the corrected number of servers and the corrected number of storage devices need to be corrected by the second adjustment coefficient. The expression is: Where fs c is the corrected number of servers, cs C is the corrected memory quantity, fs ct is the number of servers after the second correction, cs ct is the number of memories after secondary correction, tj two is the second adjustment coefficient.

5. The e-commerce data analysis method based on the cloud platform according to claim 4 is characterized in that: Obtaining a correction index based on the traffic growth rate and the user online growth rate includes the following steps: The traffic growth rate and the user online growth rate are normalized to map their value ranges to [0, 1], and the normalized value of the traffic growth rate and the normalized value of the user online growth rate are obtained. The correction index is obtained by summing the normalized value of the traffic growth rate and the normalized value of the user online growth rate.

6. The e-commerce data analysis method based on the cloud platform according to claim 1, characterized in that: If the date of the next day after the analysis is not a special date, the server and storage resource status of the day are analyzed to determine whether static adjustment of the server and storage resources is required, including the following steps: When the date of the next day is not a special date, obtain the request timeout frequency of the server and the read and write delay frequency of the memory on that day; Normalize the request timeout frequency and the read-write delay frequency to obtain the normalized value of the request timeout frequency and the normalized value of the read-write delay frequency, sum the normalized timeout frequency value and the normalized read-write delay frequency value to obtain the abnormality coefficient, compare the abnormality coefficient with the first abnormality threshold, if the abnormality coefficient is less than or equal to the first abnormality threshold, determine that there is no need to statically adjust the server and storage resources, if the abnormality coefficient is greater than the first abnormality threshold, determine that there is a need to statically adjust the server and storage resources, and statically adjust the server and storage resources according to the static adjustment table.

7. The e-commerce data analysis method based on the cloud platform according to claim 6 is characterized by: Combining the real-time data obtained on the next day with the server and storage resource status on the next day to determine whether it is necessary to dynamically adjust the server and storage resources includes the following steps: Get the real-time data for the next day, including the traffic growth rate and the user online growth rate, the server request timeout frequency and the storage read and write delay frequency for the next day; The traffic growth rate, user online growth rate, request timeout frequency, and read / write delay frequency are calculated using a linear regression analysis algorithm to obtain the dynamic adjustment coefficient. The expression is: In the formula, tj dynamic is the dynamic adjustment coefficient, R is the influencing factor, R is 1.513, θ1, θ2, θ3, θ4 are the traffic growth rate, user online growth rate, request timeout frequency, and read and write delay frequency, γ1, γ2, γ3, γ4 are regression coefficients, and the regression coefficient is greater than θ; The dynamic adjustment coefficient is compared with the dynamic adjustment threshold. The dynamic adjustment threshold is used to predict the overall load change trend of the e-commerce platform. If the dynamic adjustment coefficient is less than or equal to the dynamic adjustment threshold, it is predicted that the e-commerce platform will show a low load trend change. If the dynamic adjustment coefficient is greater than the dynamic adjustment threshold, it is predicted that the e-commerce platform will show a high load trend change. When it is predicted that the overall e-commerce platform will show a high load trend, and the initial number of servers and the initial number of memories are not statically adjusted, the expression for adjustment through the dynamic adjustment coefficient is: Where fs c is the corrected number of servers, cs c is the corrected memory quantity, fs h is the initial number of servers, cs h is the initial memory quantity, tj dynamic is the dynamic adjustment coefficient; When it is predicted that the e-commerce platform as a whole will show a high load trend, and the initial number of servers and the initial number of memories are statically adjusted, the expression adjusted by the dynamic adjustment coefficient is: Where fs c is the corrected number of servers, cs c is the corrected memory quantity, fs jh is the number of servers after static adjustment, cs jh is the amount of memory after static adjustment, tj dynamic is the dynamic adjustment coefficient.

8. The e-commerce data analysis method based on the cloud platform according to claim 7, characterized in that: After the analysis system obtains the date of the next day, it analyzes whether the date of the next day is a special date, including the following steps: The analysis system obtains the date of the next day and obtains all special dates from the special date library of the e-commerce platform; Use the acquired date of the next day to match the date in the special date library, and determine whether it is a special date by directly comparing the date; If the date of the next day matches a date or date range in the special date library, the analysis system marks the type of the date; Based on the result output of the date matching, it is judged whether the next day is a special date and its type. If it is a special date, the information of the special date is fed back. If the next day is not a special date, the analysis system returns the corresponding status.

9. An e-commerce data analysis system based on a cloud platform, used to implement the e-commerce data analysis method based on a cloud platform according to any one of claims 1 to 8, characterized in that: Including date analysis module, special date adjustment module and general date adjustment module; Date analysis module: after obtaining the date of the next day, analyze whether the date of the next day is a special date; Special date adjustment module: When the date of the next day is analyzed to be a special date, after obtaining multiple historical data of the special date based on the cloud platform, the server and storage resources are adjusted once in combination with the server and storage resource status of the day. The real-time data of the e-commerce platform is obtained the next day, and the real-time data is combined with the result of the first adjustment to determine whether the server and storage resources need to be adjusted again; Ordinary date adjustment module: when the date of the next day is not a special date, the server and storage resource status of the day are analyzed to determine whether static adjustment of the server and storage resources is required. The real-time data obtained on the next day is combined with the server and storage resource status of the next day to determine whether dynamic adjustment of the server and storage resources is required.

10. An electronic device for e-commerce data analysis based on a cloud platform, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the cloud platform-based e-commerce data analysis method are implemented.

Images