Service strategy selection system

Abstract

The invention discloses a service strategy selection system. The system comprises a server, a work order subsystem, a financial subsystem, a charging subsystem and a data center basic management subsystem, and the development stage of a target data center is determined according to financial data obtained from the financial subsystem; determining an initial weight coefficient of each type of project operation data, and adjusting the initial weight coefficient according to the development stage to obtain a target weight coefficient of each type of project operation data; and determining a service level of each customer project according to the target weight coefficient of each type of project operation data and the project operation data, and determining a service strategy corresponding to each customer project according to the service level of each customer project. The technical problem that in the prior art, due to the fact that different data are dispersed in different systems, the data in all the systems cannot be comprehensively analyzed, and the service strategy matched with the actual requirements of customers and the requirements of all stages of data center profit cannot be determined is solved.

Description

Technical Field

[0001] The present application relates to the technical field of data centers, and in particular to a service strategy selection system. Background Art

[0002] In recent years, with the influx of social capital, the supply of data center rack resources has exceeded the demand. Data center service providers are facing increasingly fierce competition, which is not only reflected in the construction standards and price levels of computer rooms, but also in the differentiated customer service levels. The company's maintenance service personnel are limited, and it is impossible to invest the same scale of service force for all customers. How to reasonably judge the customer service level, select the appropriate service strategy based on this, provide differentiated services, and further maximize the value of the data center through services while ensuring the existing customer service satisfaction is a difficult problem facing IDC (Internet Data Center) service providers.

[0003] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0004] An embodiment of the present application provides a service strategy selection system to at least solve the technical problem that in the prior art, different data are scattered in different systems and it is impossible to conduct a comprehensive analysis of the data in each system, resulting in an inability to determine a service strategy that matches the actual needs of customers and the needs of data centers at different profit stages.

[0005] According to one aspect of an embodiment of the present application, a service strategy selection system is provided, including a server, a work order subsystem, a financial subsystem, a billing subsystem, and a data center basic management subsystem, wherein the server is used to determine the project operation data of each customer project through the work order subsystem, the billing subsystem and the data center basic management subsystem, wherein the project operation data includes at least one of the following: service relationship data, project profit data, project power consumption data, project revenue data, project network data, and project compound growth rate data, the service relationship data includes the failure data of the customer project within a preset statistical period, and the project compound growth rate data includes traffic growth rate data and hardware resource capacity growth rate; the development stage of the target data center is determined based on the financial data obtained from the financial subsystem, wherein the financial data includes the investment data and operating cost data of the data center; the initial weight coefficients of various types of project operation data are determined, and the initial weight coefficients are adjusted according to the development stage to obtain the target weight coefficients of various types of project operation data; the service level of each customer project is determined based on the target weight coefficients of the various types of project operation data and the project operation data, and the service strategy corresponding to each customer project is determined based on the service level of each customer project.

[0006] Optionally, determining the development stage of the target data center based on the financial data obtained from the financial subsystem includes: determining a first preset rack loading rate and a second preset rack loading rate based on the financial data to determine the target data, wherein the second preset rack loading rate is greater than the first preset rack loading rate, the first preset rack loading rate is the rack loading rate when the target data center's revenue and monthly investment depreciation reach a balance point, and the second preset rack loading rate is the loading rate when the target data center reaches a preset revenue target; determining a third preset rack loading rate, wherein the third preset rack loading rate is greater than the second preset rack loading rate; determining the development stage of the target data center based on the first preset rack loading rate, the second preset rack loading rate and the third preset rack loading rate.

[0007] Optionally, determining the development stage of the target data center based on the financial data obtained from the financial subsystem includes: determining a rack occupancy rate value interval corresponding to each development stage based on a first preset rack occupancy rate, a second preset rack occupancy rate and a third preset rack occupancy rate; determining a target rack occupancy rate of the target data center; and determining the development stage of the target data center based on the rack occupancy rate value interval within which the target rack occupancy rate lies.

[0008] Optionally, the development stage includes an initial expansion period, a profit accumulation period, an integrated development period, and a stable service period, wherein the initial expansion period corresponds to a first shelf rate value interval, the profit accumulation period corresponds to a second shelf rate value interval, the integrated development period corresponds to a third shelf rate value interval, and the stable service period corresponds to a fourth shelf rate value interval, wherein the minimum value of the first shelf rate value interval is zero, and the maximum value is a first preset rack shelf rate, the minimum value of the second shelf rate value interval is greater than the first preset rack shelf rate, and the maximum value is the second preset rack shelf rate, the minimum value of the third shelf rate value interval is greater than the second preset rack shelf rate, the maximum value is the third preset rack shelf rate, and the minimum value of the fourth shelf rate value interval is greater than the third preset rack shelf rate.

[0009] Optionally, the initial weight coefficient is adjusted according to the development stage to obtain the target weight coefficient of each type of project operation data, including: determining the core project operation data and non-core project operation data from the various types of project operation data according to the development stage; determining the initial weight coefficient of the non-core project operation data as the target weight coefficient of the non-core project operation data; increasing the initial weight coefficient of the core project operation data, and using the increased result as the target weight coefficient of the core project operation data.

[0010] Optionally, determining the service level of each customer project based on the target weight coefficient of each type of project operation data and the project operation data includes: processing the various types of project operation data of each customer project and the target weight coefficient of the project operation data through a gradient boosting machine model, and obtaining the service level prediction value of each customer project output by the gradient boosting machine model; determining the service level of each customer project based on the service level prediction value.

[0011] Optionally, the gradient boosting machine model is trained in the following manner: determining a first objective function, wherein the first objective function is composed of a training loss function and a regularization loss function, the training loss function is used to measure the quality of fit of the gradient boosting machine model on the training set, and the regularization is used to measure the complexity of the gradient boosting machine model after fitting; determining a second objective function, wherein the second objective function is used to determine a service level corresponding to a predicted value based on the predicted value output by the gradient boosting machine model; determining a training data set, and target weight values ​​of various types of historical project operation data in the training data set; and training the gradient boosting machine model using the gradient boosting machine algorithm, the first objective function, the second objective function, and the training data set.

[0012] Optionally, processing various types of project operation data and target weight coefficients of project operation data of various customer projects through the gradient boosting machine model includes: determining a first evaluation value of each type of project operation data, and normalizing the first evaluation value to obtain a second evaluation value; inputting the second evaluation value and the target weight coefficient corresponding to the second evaluation value into the gradient boosting machine model.

[0013] Optionally, the fault data includes the cumulative number of faults occurring within a preset statistical period and the number of customer devices affected by the faults, the project power consumption data includes the total power consumption data of the racks that have been opened for the customer project, and the project network data includes the number of communication lines and bandwidth flow corresponding to the customer project.

[0014] Optionally, determining the project operation data of each customer project through the work order subsystem, the financial subsystem and the data center basic management subsystem includes: obtaining fault data from the work order subsystem according to a preset collection cycle, obtaining the number of racks and rack unit price information corresponding to each customer project from the billing subsystem, and obtaining the rack power load information corresponding to each customer project from the data center basic management subsystem; determining the project operation data based on the fault data, the number of racks and rack unit price information, and the rack power load information.

[0015] In an embodiment of the present application, a server, a work order subsystem, a financial subsystem, a billing subsystem, and a data center basic management subsystem are used, wherein the server is used to determine the project operation data of each customer project through the work order subsystem, the billing subsystem and the data center basic management subsystem, wherein the project operation data includes at least one of the following: service relationship data, project profit data, project power consumption data, project revenue data, project network data, and project compound growth rate data; the service relationship data includes the failure data of the customer project within a preset statistical period, and the project compound growth rate data includes the traffic growth rate data and the hardware resource capacity growth rate; the development stage of the target data center is determined based on the financial data obtained from the financial subsystem, wherein the financial data includes the investment data and the operating cost data of the data center; the initial weight coefficients of various types of project operation data are determined, and the weight coefficients are calculated based on the development stage. The initial weight coefficient is adjusted to obtain the target weight coefficient of each type of project operation data; the service level of each customer project is determined based on the target weight coefficient of each type of project operation data and the project operation data, and the service strategy corresponding to each customer project is determined based on the service level of each customer project. By establishing a system for collecting various types of project operation data through interaction of each subsystem, the target weight coefficient of each type of project operation data is determined and calculated based on the data and the corresponding weight, the purpose of accurately determining the service level of the corresponding customer project is achieved, thereby achieving the technical effect of quickly and accurately determining the service strategy corresponding to each customer project, and further solving the technical problem of being unable to determine the service strategy that matches the actual needs of customers and the needs of each stage of data center profitability due to the fact that different data in the prior art are scattered in different systems and the data in each system cannot be comprehensively analyzed. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0017] Figure 1 It is a structural diagram of a service strategy selection system provided according to an embodiment of the present application;

[0018] Figure 2 It is a schematic diagram of a service strategy selection process provided according to an embodiment of the present application;

[0019] Figure 3 It is a schematic diagram of a gradient boosting tree construction principle provided according to an embodiment of the present application;

[0020] Figure 4 It is a schematic diagram of a gradient boosting tree construction principle provided according to an embodiment of the present application;

[0021] Figure 5 It is a structural diagram of a computer terminal (mobile device) provided according to an embodiment of the present application. DETAILED DESCRIPTION

[0022] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0023] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0024] In order to better understand the embodiments of the present application, the technical terms involved in the embodiments of the present application are explained as follows:

[0025] IDC (Internet Data Center) is a combination of traditional data centers and the Internet. In addition to having a safe and reliable computer room environment and professional service management, it can provide various types of network access for customer managed equipment / business systems.

[0026] DCIM (Data Center Infrastructure Management) combines IT (information technology) and equipment management to conduct centralized monitoring and capacity planning of key data center equipment. Through software, hardware, and sensors, DCIM provides an independent management platform to monitor and manage data center IT equipment and infrastructure in real time.

[0027] Gradient Boosting Machine (GBM), a popular ensemble learning algorithm, improves the performance of a model by combining multiple weak prediction models (usually decision trees). As a machine learning tool, the Gradient Boosting Machine has shown excellent performance in solving complex prediction problems. Its core concept is based on the gradient descent method, which gradually builds a series of weak learners to iteratively optimize model predictions, and finally converges into a strong prediction model.

[0028] In the related technologies, different business data are scattered in different subsystems, which makes it difficult to comprehensively consider. In addition, the service strategy is based on people's subjective experience, which is not accurate enough. In addition, the strategy formulation takes a long time and cannot be adjusted dynamically and instantly. Specifically, the related technologies have the following problems:

[0029] 1. The evaluation standards are confusing and the business data is scattered, making it difficult to accurately determine the service strategy. At the same time, the scattered data is not conducive to determining the current development status, which affects the accuracy of the service strategy. Since data centers are asset-heavy industries, different service strategies need to be established during the entire life cycle of data centers from the completion of construction and production (sales rate 0%) to the basic completion of sales (sales rate of 90%). Existing technologies cannot accurately determine the current stage of development.

[0030] 2. Manual judgment of service strategies is time-consuming and lacks objectivity, and it is impossible to dynamically adjust service strategies in a timely manner when business conditions change.

[0031] In order to solve the above problems, relevant solutions are provided in the embodiments of the present application, which are described in detail below.

[0032] According to an embodiment of the present application, a system embodiment of a service policy selection system is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0033] The service strategy selection system provided in the embodiment of the present application is as follows: Figure 1As shown, it includes a server 10, a work order subsystem 20, a financial subsystem 30, a billing subsystem 40, and a data center basic management subsystem 50, wherein the server 10 is used to determine the project operation data of each customer project through the work order subsystem 20, the billing subsystem 40 and the data center basic management subsystem 50, wherein the project operation data includes at least one of the following: service relationship data, project profit data, project power consumption data, project revenue data, project network data, and project compound growth rate data, the service relationship data includes the failure data of the customer project within a preset statistical period, and the project compound growth rate data includes the traffic growth rate data and the hardware resource capacity growth rate; the development stage of the target data center is determined based on the financial data obtained from the financial subsystem 30, wherein the financial data includes the investment data and the operating cost data of the data center; the initial weight coefficients of various types of project operation data are determined, and the initial weight coefficients are adjusted according to the development stage to obtain the target weight coefficients of various types of project operation data; the service level of each customer project is determined based on the target weight coefficients of the various types of project operation data and the project operation data, and the service strategy corresponding to each customer project is determined based on the service level of each customer project.

[0034] Optionally, the server 10 opens interfaces with the billing subsystem 40, the data center basic management subsystem 50, the work order subsystem 20, and the financial subsystem 30, performs information synchronization at a set frequency (such as daily / weekly / preferably monthly), and stores the relevant information of each user in the database. Including:

[0035] Billing subsystem 40: obtains the number of racks activated, power consumption of a single rack, and unit price of racks for each customer in the data center; the number of lines activated (number of lines and bandwidth), and unit price of lines.

[0036] Data center basic management subsystem 50: obtains power load data of each customer rack in the data center.

[0037] Work order subsystem 20: Obtain the impact of data center failures / abnormal events on customers, including the impact scope, duration, and affected rack scope of the failure / event.

[0038] Financial subsystem 30: obtains data center investment data and operating cost data.

[0039] In some optional embodiments of the present application, the service strategy selection process is as follows: Figure 2 As shown: First, extract and quantify the parameters of each dimension (i.e., the operation data of each project), apply the trained model to generate the customer level, and determine the service strategy corresponding to each customer project based on the service level of each customer project.

[0040] In some optional embodiments of the present application, the project operation data is standardized, and a data center business development stage model is created. The system automatically determines which development stage the current data center is in based on the project operation data, and uses the customer data of each stage of the commissioned computer room to train the gradient boosting machine model. According to the development stage of the data center to be evaluated, the core project is determined, and the customer service levels of the data center are output by inputting each parameter item of the customer into the trained gradient boosting machine model. It makes up for the lack of existing data center user service level evaluation technology, so that the service upgrader changes from passive to active, and triggers customer service and marketing strategy analysis in combination with the change of customer service level indicators, reducing the risk of customer business shrinkage or even complete loss. Among them, the project operation data is obtained from each subsystem according to the interface, including: S (service relationship, that is, the frequency, number, and scope of customer impact of failure), P (profit, that is, customer project profit margin index), P (power, that is, the total IT load power consumption of the customer project), R (income, that is, the annual income of the customer project), I (network, that is, the Internet export rate of the customer project, the number of customer dedicated lines), I (compound growth rate of business, that is, the total IT capacity growth rate of the customer, the total traffic growth rate) and other data.

[0041] In some embodiments of the present application, determining the development stage of the target data center based on financial data obtained from the financial subsystem includes: determining the target data based on the financial data, determining a first preset rack loading rate and a second preset rack loading rate, wherein the second preset rack loading rate is greater than the first preset rack loading rate, the first preset rack loading rate is the rack loading rate when the target data center's revenue and monthly investment depreciation reach a balance point, and the second preset rack loading rate is the loading rate when the target data center reaches a preset revenue target; determining a third preset rack loading rate, wherein the third preset rack loading rate is greater than the second preset rack loading rate; determining the development stage of the target data center based on the first preset rack loading rate, the second preset rack loading rate and the third preset rack loading rate.

[0042] As an optional implementation, the first preset rack occupancy rate is the percentage of racks when revenue and monthly depreciation of investment reach a balance point, the second preset rack occupancy rate is the percentage of racks when the data center reaches a set profit target (such as 10%), and the third preset rack occupancy rate is the occupancy rate that reaches an excellent level in the industry.

[0043] After the data center definition is completed, the service strategy selection system determines the development stage of the data center based on the data center rack rate (actual number of operating racks / total number of built racks) obtained from the data center basic management subsystem 50.

[0044] The determination of the first preset rack loading rate is automatically completed by the service strategy selection system. The investment cost of the data center includes land, buildings, external power supply, chillers, low-voltage power distribution, precision air conditioners, UPS, batteries, cabinets, bridges, and fire protection. Assume that the investment depreciation period corresponding to each specialty is:

[0045] major Depreciation period Investment amount (unit: yuan) land term1 invest1 Buildings term2 invest2 External power term3 invest3 Chiller term4 invest4 Low voltage power distribution term5 invest5 Precision air conditioning term6 invest6 ups term7 invest7 Battery term8 invest8 Cabinet term9 invest9 Bridge term10 invest10 Fire protection major term11 invest11

[0046] The unit price of the sold racks is set as price (unit: yuan / year), and the target price of the racks to be sold is set as list (unit: yuan / year). Then the first preset rack shelf rate is obtained according to the following formula:

[0047]

[0048] Among them, regarding the determination of the second preset rack loading rate, the second preset rack loading rate is set to Y, the profit rate of the racks sold is set to promar, the actual annual operating cost is recorded as incuco (unit: yuan), the target profit rate of the racks to be sold is set to protar, and the estimated operating cost is recorded as tarco (unit: yuan), then Y is obtained according to the following formula:

[0049]

[0050] In some embodiments of the present application, determining the development stage of a target data center based on financial data obtained from a financial subsystem includes: determining a rack occupancy rate value interval corresponding to each development stage based on a first preset rack occupancy rate, a second preset rack occupancy rate, and a third preset rack occupancy rate; determining a target rack occupancy rate of the target data center; and determining the development stage of the target data center based on the rack occupancy rate value interval within which the target rack occupancy rate lies.

[0051] In some embodiments of the present application, the development stage includes an initial expansion period, a profit accumulation period, an integrated development period, and a stable service period, wherein the initial expansion period corresponds to a first shelf rate value interval, the profit accumulation period corresponds to a second shelf rate value interval, the integrated development period corresponds to a third shelf rate value interval, and the stable service period corresponds to a fourth shelf rate value interval, wherein the minimum value of the first shelf rate value interval is zero, and the maximum value is a first preset rack shelf rate, the minimum value of the second shelf rate value interval is greater than the first preset rack shelf rate, and the maximum value is the second preset rack shelf rate, the minimum value of the third shelf rate value interval is greater than the second preset rack shelf rate, and the maximum value is the third preset rack shelf rate, and the minimum value of the fourth shelf rate value interval is greater than the third preset rack shelf rate.

[0052] In some optional embodiments of the present application, the development of a data center is defined as four stages, namely, the initial expansion stage, the profit accumulation stage, the integrated development stage, and the stable service stage. The goals of each stage, the corresponding shelf rate percentage interval start (inclusive) and end boundary values ​​and definitions are shown in the following table:

[0053]

[0054] Wherein, X% is the first preset rack loading rate, Y% is the second preset rack loading rate, and 90% is an optional value of the third preset rack loading rate.

[0055] Optionally, data centers are usually invested and constructed in batches, so for a certain data center, with the additional investment, the first preset rack occupancy rate will change multiple times, causing the data center to switch, oscillate, and fluctuate between the initial expansion period and the profit accumulation period.

[0056] As an optional implementation method, a data center has a total of 4,000 cabinets, 20% of which have been completed. The average investment cost of a single cabinet is 240,000 yuan, the monthly investment depreciation is 3,500 yuan / rack, and the monthly depreciation of the initial investment is 2.8 million yuan. Currently, 6 customers have put 360 cabinets on the shelves, with an average monthly rent of 8,000 yuan per cabinet, and a monthly rental income of 2.88 million yuan. The initial expansion period boundary X% is calculated to be 43.75%. At present, it is actually judged that the data center has completed the initial expansion period and entered the profit accumulation period. When judging the customer service level, the profit pr is recorded as the core project operation data.

[0057] In some embodiments of the present application, the initial weight coefficient is adjusted according to the development stage to obtain the target weight coefficient of each type of project operation data, including: determining the core project operation data and non-core project operation data from the various types of project operation data according to the development stage; determining the initial weight coefficient of the non-core project operation data as the target weight coefficient of the non-core project operation data; increasing the initial weight coefficient of the core project operation data, and using the increased result as the target weight coefficient of the core project operation data.

[0058] In some optional embodiments of the present application, according to the data center development stage determined in step 4, the core project operation data is set, and the weight of the core project operation data is * 2. The remaining non-core project operation data are normally valued, and the weight is not set.

[0059] Optionally, in the initial expansion period, revenue re is used as the core project operation data, in the profit accumulation period, profit pr is used as the core project operation data, in the integrated development period, network in is used as the core project operation data, and in the stable service period, service relationship se is used as the core project operation data.

[0060] In some embodiments of the present application, determining the service level of each customer project based on the target weight coefficient of each type of project operation data and the project operation data includes: processing the various types of project operation data of each customer project and the target weight coefficient of the project operation data through a gradient boosting machine model, and obtaining the service level prediction value of each customer project output by the gradient boosting machine model; determining the service level of each customer project based on the service level prediction value.

[0061] In some optional embodiments of the present application, determining the service level of each customer project includes inputting customer data into a trained gradient boosting model to obtain a predicted service level for each customer. Based on the model calculation results, differentiated service strategies are formulated for different service levels to improve customer satisfaction and business stability.

[0062] Optionally, the gradient model will be updated and iterated, and the model will be updated periodically or triggered by events. When the first step of data collection is performed periodically, if the following changes occur in the data center, the fourth step will be successfully performed to determine the current development stage of the data center: there are new customers, the number of racks opened / bandwidth changes by more than 5% compared with the last model operation, new failures / abnormal events occur in the data center, and the cost data of the data center increases by more than 5% compared with the last model operation. After completing the re-determination of the development stage of the data center, update the model training, update the customer service level dimension parameters that are adapted to the current stage, and update the calculation and adapt the service strategy.

[0063] In some embodiments of the present application, the gradient boosting machine model is trained in the following manner: determining a first objective function, wherein the first objective function is composed of a training loss function and a regularization loss function, the training loss function is used to measure the quality of the gradient boosting machine model's fit on the training set, and the regularization is used to measure the complexity of the gradient boosting machine model after fitting; determining a second objective function, wherein the second objective function is used to determine a service level corresponding to a predicted value based on a predicted value output by the gradient boosting machine model; determining a training data set, and target weight values ​​for various types of historical project operation data in the training data set; and training the gradient boosting machine model using the gradient boosting machine algorithm, the first objective function, the second objective function, and the training data set.

[0064] Optionally, when determining the first objective function, a gradient boosting tree is used for classification. The gradient boosting tree is a supervised learning method that automatically learns the weights of each dimension through the gradient boosting algorithm. During the model training process, the gradient boosting machine will automatically adjust the weight of each feature (i.e., dimension) according to its influence on the target variable (customer service level). Finally, the customer service level is output. A clear objective function is required as the optimization target during training. The following design of the first objective function is divided into two parts:

[0065] Obj(Θ)=L(Θ)+Ω(Θ)

[0066] Among them, L(Θ) represents the training loss, which is used to measure the quality of the model fitting on the training set, and Ω(Θ) represents regularization, which is used to measure the complexity of the model after fitting.

[0067] Optionally, define the customer service level as high, medium and low. Construct a second objective function for judging the level, and in the subsequent process, use the return value to mark the data.

[0068] Optionally, determining the training data set includes constructing training sets for each stage based on historical information of data center customers that have completed four stages.

[0069] In some optional embodiments of the present application, during the initial expansion phase of a data center, some customer historical data is quantified as follows:

[0070] ZWA customers have high service levels of se: 0.8, pr: 0.15, po: 0.6, re: 0.4, in: 0.4, ir: 0.3;

[0071] YY customer se: 0.2, pr: 0.12, po: 0.4, re: 0.3, in: 0.5, ir: 0.6 service level;

[0072] Optionally, use the gradient boosting algorithm to train the model. During the training process, the model will learn how to predict the target variable, that is, the customer service level, based on the input features. After training, the data set is reorganized and merged based on the calculation results.

[0073] Optionally, the gradient boosting tree construction principle is as follows Figure 3 , Figure 4 As shown:

[0074] 1. Single regression tree construction:

[0075] A single regression tree is used to classify attributes, adding points to positively impacted classification groups and deducting points from negatively impacted classification groups. According to the customer group, information such as service relationship, profit, electricity, network, and service growth rate can be input to perform conditional classification based on a certain attribute. For example, whether the standardized profit is greater than 0.8 is used as a classification for the profit attribute to distinguish whether the customer belongs to the integrated development stage group. If it meets the requirements (such as A and C), the comprehensive score will be assessed at +0.3 points. If it does not meet the requirements, the comprehensive score will be -0.1 points (such as B, D, and E).

[0076] 2. Tree integration: For example, for the same group of customers, multiple classifications of different attributes are performed, and the final scores are calculated and summarized for evaluation. For example, if customer A scores 0.3, 0.6, and -0.4 in profit, total electricity consumption, and service relationship index, respectively, the final comprehensive score is 0.5, and the same applies to other customers to be classified.

[0077] In some optional embodiments of the present application, the training set at a certain stage is as follows:

[0078] The core item pr of the A1 training set currently accounts for 30%;

[0079] WW customer se: 20, pr: 15, po: 60, re: 40, in: 40, ir: 30 score 80 points service level A;

[0080] YY customer se: 10, pr: 12, po: 40, re: 30, in: 50, ir: 60 score 70 points service level B;

[0081] Through regression algorithm, se, po, re, in, ir weights, ww calculation score close to 80 points, YY customer score close to 70 points;

[0082] The classification results of the model are obtained through the gradient boosting algorithm: WW A grade YY B grade.

[0083] Applying the above weights to the new data value center, the new data center is at the same stage, and the core item is still pr accounting for 30%.

[0084] All customers know se, pr, po, re, in, ir.

[0085] Apply the weights obtained from the training set to calculate:

[0086] ZZ customer 68 points, grade B;

[0087] AA customers: 90 points, grade A;

[0088] The trained gradient boosting machine model is used to calculate ZZ B level AA A level, verifying that the weights are correct.

[0089] In some embodiments of the present application, processing various types of project operation data and target weight coefficients of project operation data of various customer projects through a gradient boosting machine model includes: determining a first evaluation value of each type of project operation data, and normalizing the first evaluation value to obtain a second evaluation value; inputting the second evaluation value and the target weight coefficient corresponding to the second evaluation value into the gradient boosting machine model.

[0090] As an optional implementation, the first evaluation value of the project operation data is determined by:

[0091] (1) S (service relationship server): the service relationship is recorded as se. Because the power supply availability of data centers is usually 99.99%-99.999%, once a failure occurs, it is intolerable for customers. Therefore, the number of customer devices n and the cumulative number of occurrences m affected by the failure of the computer room facilities during the phase time obtained from the work order subsystem 20 (consistent with the set information synchronization frequency) need to be processed in a nonlinear manner. The processing of n requires that its growth has a larger growth when it is smaller, but the growth rate gradually slows down, because once a failure occurs, when the number reaches a certain level, it will have an impact on the integrity of the business, and the customer's sensitivity to changes in the number of devices will decrease. The processing of m requires that its growth has a larger growth when it is smaller, but the growth rate gradually increases, because each failure will cause strong dissatisfaction among customers, and after accumulating to a certain level of emotion, it will explode. Taking the above factors into consideration, n and m are processed to obtain the final data se:

[0092] se=log(n+1)*(exp(m)-1)

[0093] (2) P (profit): denoted as pr. The development of an enterprise is closely related to profitability. Before implementing a customer project, the data center service provider will conduct a preliminary assessment of the profit level of the customer project. The profit margin data in the financial subsystem 30 is percentage data. The profit margin data is obtained from the financial subsystem 30 and directly assigned to pr.

[0094] (3) P (Power): denoted as po. The total power consumption data value (in kW) of the racks activated by the customer directly reflects the customer's resource usage. This data can be directly read from the data center basic management subsystem 50;

[0095] (4) R (revenue): denoted as re. The monthly rental income of customers is the direct value of customers to the service provider. The unit is ten thousand yuan and is directly obtained from the financial subsystem 30.

[0096] (5) I (Internet): denoted as in. The Internet is divided into two parts: the number of lines ic and the bandwidth flow ib. The more lines there are, the more subsidiaries / related business companies the data center provides services to. For service providers, its significance is far greater than the level of customer business flow. ic is a single digit or double digit. If ib is in mb, its number is between single digit and six digits. First, normalize ib:

[0097] ib=(ib-ibmin) / (ibmax-ibmin)

[0098] By using ib=ib*10, ib is processed to the same level as ic, and the importance difference between ic and ib is reflected by weight setting:

[0099] in=0.7*ic+0.3*ib

[0100] (6) I (Compound Growth Rate of Business Increase Rate): denoted as ir. The month-on-month (preferably monthly) growth / decrease rate ir is calculated for the customer managed device quantity data obtained from the data center basic management subsystem 50, reflecting the customer's positioning of the importance of the data center business.

[0101] Optionally, the second evaluation value of the project operation data is determined by:

[0102] The above attributes (se, pr, po, re, in, ir) are denoted as K i , where k ij It is recorded as the jth item in the i attribute (j represents the quantitative unnormalized score of different customers). All attributes are transformed as follows to obtain the new attribute k′ ij , normalized:

[0103]

[0104] In some embodiments of the present application, the fault data includes the cumulative number of faults occurring within a preset statistical period and the number of customer devices affected by the faults, the project power consumption data includes the total power consumption data of the racks that have been opened for the customer project, and the project network data includes the number of communication lines and bandwidth flow corresponding to the customer project.

[0105] In some embodiments of the present application, determining the project operation data of each customer project through the work order subsystem, the financial subsystem and the data center basic management subsystem includes: obtaining fault data from the work order subsystem according to a preset collection cycle, obtaining the number of racks and rack unit price information corresponding to each customer project from the billing subsystem, and obtaining the rack power load information corresponding to each customer project from the data center basic management subsystem; determining the project operation data based on the fault data, the number of racks and rack unit price information, and the rack power load information.

[0106] Through the above steps, the purpose of accurately determining the service level of the corresponding customer project can be achieved, thereby realizing the technical effect of quickly and accurately determining the service strategy corresponding to each customer project, and further solving the technical problem that it is impossible to determine the service strategy that matches the actual needs of customers and the needs of each stage of data center profitability due to the fact that different data in the prior art are scattered in different systems and it is impossible to conduct a comprehensive analysis of the data in each system. Specifically, the present application has the following advantages.

[0107] 1. This system connects the various business subsystems, integrates data for comprehensive evaluation, and proposes a service level evaluation dimensional model for data center customers from six dimensions: S (server), P (profit), P (power), R (revenue), I (internet), and I (compound growth rate of business). It effectively connects the data of the marketing business system and the background maintenance system to form an effective customer service level evaluation system.

[0108] 2. This system determines the weights of various assessments according to the different stages of data center development, accurately assesses the service level, clarifies the judgment methods and business development centers of each stage, and thus transforms them into the core elements of the dimensional model of each stage. It is conducive to combining the perspective of healthy development of the data center and the objective and quantitative evaluation of the model to accurately and quickly guide data center suppliers to focus on service forces in different development periods, give priority to ensuring the most powerful customer service level for data center development, and through differentiated service models, while saving corporate service costs, provide customers with comprehensive service perception.

[0109] 3. The system dynamically updates the gradient booster model based on the operating status, makes timely service level predictions, and accurately customizes differentiated service strategies that match actual customer needs and data center profit needs.

[0110] The system embodiment provided in the embodiments of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Figure 5 FIG. 1 shows a hardware structure block diagram of a computer terminal (or mobile device) for implementing a service strategy selection system. Figure 5 As shown, the computer terminal 50 (or mobile device 50) may include one or more (502a, 502b, ..., 502n are used to illustrate) processors 502 (the processor 502 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 504 for storing data, and a transmission module 506 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the BUS bus), a network interface, a power supply and / or a camera. It can be understood by those skilled in the art that Figure 5 The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 5 More or fewer components as shown, or with Figure 5 Different configurations shown.

[0111] It should be noted that the one or more processors 502 and / or other data processing circuits described above may generally be referred to herein as "data processing circuits". The data processing circuits may be embodied in whole or in part as software, hardware, firmware, or any other combination thereof. In addition, the data processing circuit may be a single independent processing module, or may be incorporated in whole or in part into any of the other components in the computer terminal 50 (or mobile device). As described in the embodiments of the present application, the data processing circuit acts as a processor control (e.g., selection of a variable resistor terminal path connected to an interface).

[0112] The memory 504 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the service policy selection system in the embodiment of the present application. The processor 502 executes various functional applications and data processing by running the software programs and modules stored in the memory 504, that is, realizing the above-mentioned service policy selection system. The memory 504 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 504 may further include a memory remotely arranged relative to the processor 502, and these remote memories may be connected to the computer terminal 50 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0113] The transmission device 506 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the computer terminal 50. In one example, the transmission device 506 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 506 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0114] The display may be, for example, a touch screen liquid crystal display (LCD) that enables a user to interact with a user interface of the computer terminal 50 (or mobile device).

[0115] In the above embodiments of the present application, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0116] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0117] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0118] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0119] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the relevant technology or all or part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk, etc. Various media that can store program codes.

[0120] The above is only a preferred implementation of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A service strategy selection system, characterized in that: Including servers, work order subsystem, financial subsystem, billing subsystem, data center basic management subsystem, among which, The server is used to determine the project operation data of each customer project through the work order subsystem, the billing subsystem and the data center basic management subsystem, wherein the project operation data includes at least one of the following: service relationship data, project profit data, project power consumption data, project revenue data, project network data, and project compound growth rate data. The service relationship data includes the fault data of the customer project within a preset statistical period, and the project compound growth rate data includes traffic growth rate data and hardware resource capacity growth rate; Determining the development stage of the target data center based on the financial data obtained from the financial subsystem, wherein the financial data includes investment data and operating cost data of the data center; Determine the initial weight coefficients of various types of project operation data, and adjust the initial weight coefficients according to the development stage to obtain target weight coefficients of various types of project operation data; The service level of each of the customer projects is determined based on the target weight coefficient of each type of project operation data and the project operation data, and the service strategy corresponding to each of the customer projects is determined based on the service level of each of the customer projects.

2. The service strategy selection system according to claim 1, characterized in that: Determining the development stage of the target data center based on the financial data obtained from the financial subsystem includes: Determine the target data according to the financial data to determine a first preset rack loading rate and a second preset rack loading rate, wherein the second preset rack loading rate is greater than the first preset rack loading rate, the first preset rack loading rate is the rack loading rate when the target data center's income and monthly investment depreciation reach a balance point, and the second preset rack loading rate is the loading rate when the target data center reaches a preset revenue target; Determine a third preset rack loading rate, wherein the third preset rack loading rate is greater than the second preset rack loading rate; The development stage of the target data center is determined according to the first preset rack occupancy rate, the second preset rack occupancy rate and the third preset rack occupancy rate.

3. The service strategy selection system according to claim 2, characterized in that: Determining the development stage of the target data center based on the financial data obtained from the financial subsystem includes: Determine the rack loading rate value interval corresponding to each development stage according to the first preset rack loading rate, the second preset rack loading rate and the third preset rack loading rate; Determining a target rack occupancy rate of the target data center; The development stage of the target data center is determined according to the rack occupancy rate value interval in which the target rack occupancy rate is located.

4. The service strategy selection system according to claim 3, characterized in that: The development stages include the initial expansion period, profit accumulation period, integration development period, and stable service period, among which: The initial expansion period corresponds to the first shelf-loading rate value interval, the profit accumulation period corresponds to the second shelf-loading rate value interval, the integrated development period corresponds to the third shelf-loading rate value interval, and the stable service period corresponds to the fourth shelf-loading rate value interval, wherein the minimum value of the first shelf-loading rate value interval is zero, and the maximum value is the first preset rack shelf-loading rate, the minimum value of the second shelf-loading rate value interval is greater than the first preset rack shelf-loading rate, and the maximum value is the second preset rack shelf-loading rate, the minimum value of the third shelf-loading rate value interval is greater than the second preset rack shelf-loading rate, and the maximum value is the third preset rack shelf-loading rate, and the minimum value of the fourth shelf-loading rate value interval is greater than the third preset rack shelf-loading rate.

5. The service strategy selection system according to claim 1, characterized in that: The initial weight coefficient is adjusted according to the development stage to obtain the target weight coefficients of various project operation data, including: Determining core project operation data and non-core project operation data from each type of project operation data according to the development stage; Determining an initial weight coefficient of the non-core project operation data as a target weight coefficient of the non-core project operation data; The initial weight coefficient of the core project operation data is increased, and the increase result is used as the target weight coefficient of the core project operation data.

6. The service strategy selection system according to claim 1, characterized in that: Determining the service level of each of the customer projects based on the target weight coefficient of each type of project operation data and the project operation data includes: Processing various types of project operation data of each of the customer projects and target weight coefficients of the project operation data through a gradient boosting machine model, and obtaining a service level prediction value of each of the customer projects output by the gradient boosting machine model; The service level of each of the customer projects is determined according to the service level prediction value.

7. The service strategy selection system according to claim 6, characterized in that: The gradient boosting machine model is trained in the following way: Determine a first objective function, wherein the first objective function is composed of a training loss function and a regularization loss function, the training loss function is used to measure the quality of the gradient boosting machine model fitting on the training set, and the regularization is used to measure the complexity of the gradient boosting machine model after fitting; Determining a second objective function, wherein the second objective function is used to determine a service level corresponding to the predicted value according to the predicted value output by the gradient boosting machine model; Determine a training data set and target weight values ​​of various types of historical project operation data in the training data set; The gradient boosting machine model is trained using a gradient boosting machine algorithm, the first objective function, the second objective function and the training data set.

8. The service strategy selection system according to claim 1, characterized in that: The various types of project operation data of each of the client projects and the target weight coefficients of the project operation data are processed by the gradient boosting machine model, including: Determine a first evaluation value of each type of project operation data, and normalize the first evaluation value to obtain a second evaluation value; The second evaluation value and the target weight coefficient corresponding to the second evaluation value are input into the gradient boosting machine model.

9. The service strategy selection system according to claim 1, characterized in that: The fault data includes the cumulative number of faults occurring within the preset statistical period and the number of customer devices affected by the faults; the project power consumption data includes the total power consumption data of the racks that have been opened for the customer project; and the project network data includes the number of communication lines and bandwidth flow corresponding to the customer project.

10. The service strategy selection system according to claim 1, characterized in that: The project operation data of each customer project determined by the work order subsystem, the financial subsystem and the data center basic management subsystem include: Acquire the fault data from the work order subsystem according to a preset collection cycle, acquire the rack quantity and rack unit price information corresponding to each customer project from the billing subsystem, and acquire the rack power load information corresponding to each customer project from the data center basic management subsystem; The project operation data is determined according to the fault data, the number of racks and the rack unit price information, and the rack power load information.

Images