Cloud disk space excitation method based on multi-dimensional behavior perception and electronic equipment

Abstract

The invention discloses a cloud disk space excitation method based on multi-dimensional behavior perception and electronic equipment, and belongs to the technical field of computers. The method comprises the steps that a data acquisition layer monitors user behavior data in real time, and the user behavior data comprises a storage demand, behavior activeness, a storage extension request and a life cycle state; a state quantification layer performs state quantification on user behaviors based on the user behavior data to obtain a plurality of state quantification formulas corresponding to the user behavior data; the intelligent decision engine generates an excitation strategy based on a plurality of state quantization formulas; and the feedback closed loop layer dynamically updates strategy parameters based on instant reward feedback to realize continuous optimization of the system. According to the method and the device, user requirements can be more accurately matched, so that a more personalized and differentiated excitation scheme is provided, the utilization efficiency of storage resources is improved, the satisfaction degree of users is enhanced, and the digital operation competitiveness and the service quality of cloud resource service providers are promoted.

Description

Technical Field

[0001] This application belongs to the field of computer technology, and particularly relates to a cloud disk space incentive method and an electronic device based on multi-dimensional behavior perception. Background Art

[0002] With the vigorous development and popularization of cloud storage technology, individual and enterprise users increasingly rely on cloud disks as the core tools for data storage and management. However, in the face of the growing demand for data storage, the limited cloud disk space has become a challenge that users must face. To address this challenge, users have to frequently purchase additional storage space, which undoubtedly increases their usage costs. To attract and retain users, cloud disk service providers are constantly exploring and innovating operation strategies. Among them, the space incentive distribution system has become a widely used marketing method due to its high efficiency.

[0003] In the traditional cloud disk space incentive distribution system, although it has improved the activity to a certain extent by motivating users' registration, login, and purchase behaviors, the system often distributes incentive coupons equally to all users based on preset rules, ignoring the differences between users and their actual storage needs, and not considering the temporal characteristics of users' behaviors when distributing space incentives. For example, the activity decays over time, resulting in untimely incentives, and not tracking users' feedback in a timely manner after distributing incentive coupons to users. In addition, the resource allocation may also be one-size-fits-all, causing waste. For example, issuing coupons to saturated users not only fails to accurately match the personalized needs of users but may also cause waste of incentive resources.

[0004] In view of the above problems, the cloud disk space incentive method and electronic device based on multi-dimensional behavior perception in this application are proposed. Summary of the Invention

[0005] To solve the deficiencies of the existing technology described above, this application provides a cloud disk space incentive method and an electronic device based on multi-dimensional behavior perception, which solves the problems in the existing technology such as single dimension of behavior perception, rigid incentive strategies, inaccurate resource allocation, and inability to track users' feedback on the incentive method in a timely manner.

[0006] The technical effects to be achieved by this application are realized through the following solutions: In a first aspect, this application provides a cloud disk space incentive method based on multi-dimensional behavior perception, and the method includes: The data collection layer monitors user behavior data in real time, where the user behavior data includes storage requirements, behavior activity, storage expansion requests, and life cycle status; The status quantification layer quantifies the user behavior based on the user behavior data to obtain a plurality of status quantification formulas corresponding to the user behavior data; The intelligent decision-making engine generates incentive policies based on multiple state quantification formulas; The feedback closed-loop layer dynamically updates the policy parameters based on immediate reward feedback to achieve continuous optimization of the system.

[0007] In some embodiments, the state quantification formula includes a storage requirement index calculation formula, and the storage requirement index calculation formula is as follows: , where, , sigmod represents the normalization function, S used represents the storage usage, S total represents the total storage of the package, Trend(H30) represents the linear regression trend slope based on 30-day historical data, represents the proportion of video files in the total files, and α, β, and γ all represent weight coefficients.

[0008] In some embodiments, the state quantification formula further includes a behavior activity calculation formula, and the behavior activity calculation formula is as follows: , where, , represents the file size, i represents the i th file, i has a value range of 1 to n , n is a positive integer, both represent weight coefficients, represents the session density, = number of consecutive operations / total number of operations.

[0009] In some embodiments, the state quantification formula further includes a storage expansion request intensity calculation formula, and the storage expansion request intensity calculation formula is as follows: , where,

[0010] where, S used represents the storage usage, S total represents the total storage of the package, represents the explicit request intensity, if , represents the implicit demand score, represents the average number of uploaded files per day.

[0011] In some embodiments, the state quantification formula further includes a life cycle state calculation formula, and the life cycle state calculation formula is as follows: , wherein, represents , H represents the number of operations of the core functions of users in the past 7 days, P represents the payment status of users, P = 0 means unpaid, P = 1 means paid, S represents the storage utilization rate of users, , 0, 1, 2, and 3 respectively represent new users, growing users, mature users, and silent users.

[0012] In some embodiments, the incentive policy includes a coupon type decision function, and the coupon type decision function is as follows: , wherein, represents a set of different coupon types, represents , , represents a four-dimensional real number space, represents the historical average reward, parameters of , , represents a reward calculation model, c represents a specific coupon, represents the optimal c obtained by maximizing the reward.

[0013] In some embodiments, the incentive policy further includes a reward calculation model, and the reward calculation model is as follows: , wherein, represents , = 1 represents , represents , , represents a one-dimensional real number space, both represent .

[0014] In some embodiments, the feedback closed-loop layer dynamically updates the policy parameters based on immediate reward feedback, including updating the parameters in real time according to the following formula , and : , wherein, , wherein, represents the actual reward, represents the learning rate, represents the parameter of the new coupon type feature vector, represents the new historical average reward, represents the number of selections of new various coupons, represents the derivative symbol...

[0015] In a second aspect, the present application provides an electronic device, which includes: a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the method described in any one of the foregoing is implemented.

[0016] In a third aspect, the present application provides a computer-readable storage medium, which stores one or more programs, and the one or more programs can be executed by one or more processors to implement the method described in any one of the foregoing.

[0017] Through the cloud disk space incentive method and electronic device based on multi-dimensional behavior perception provided by the present application, this method can more accurately match user needs, thereby providing more personalized and differentiated incentive programs, improving the utilization efficiency of storage resources and enhancing user satisfaction, and assisting the digital operation competitiveness and service quality of cloud resource service providers. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present application or the existing technical solutions, the following will briefly introduce the drawings required for the description of the embodiments or the existing technical solutions. Obviously, the drawings in the following description are only some embodiments recorded in the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0019] Figure 1 is a flowchart of the cloud disk space incentive method based on multi-dimensional behavior perception in an embodiment of the present application; Figure 2 is a schematic block diagram of an electronic device in an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] To make the purpose, technical solutions, and advantages of the present application clearer, the following will clearly and completely describe the technical solutions of the present application in conjunction with specific embodiments and the corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present application.

[0021] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in one or more embodiments of this application should have the ordinary meanings understood by those of ordinary skill in the art to which this application belongs. The terms "first", "second" and similar words used in one or more embodiments of this application do not denote any order, quantity or importance, but are only used to distinguish different components. Words such as "including" or "comprising" mean that the elements or objects appearing before this word cover the elements or objects listed after this word and their equivalents, without excluding other elements or objects. Words such as "connected" or "coupled" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right", etc. are only used to represent relative positional relationships, and when the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0022] This application proposes a cloud disk space incentive system based on multi-dimensional behavior perception and collaborative optimization, and constructs a closed-loop technical system of "dynamic perception-intelligent decision-making-elastic incentive". Its characteristics are as follows:

[0023] ① Innovatively design a state quantification model that integrates storage demand characteristics, active behavior characteristics, storage expansion request characteristics, and life cycle characteristics, and realize the accurate characterization of the user's storage state by adjusting the characteristic weight ratio; ② Incorporate historical reward weights, dynamic exploration items, and user state matching degrees into the decision-making, and establish an adaptive matching rule between the user state and the incentive strategy; ③ Develop a dynamic weight allocation strategy with storage resource utilization constraints, and combine the feature space iteration mechanism optimized by policy gradients to form the continuous self-optimization ability of the incentive strategy. Through the technical integration of multi-dimensional feature collaborative analysis, dynamic decision-making algorithm optimization, and resource constraint feedback adjustment, this system effectively solves technical problems such as rigid incentive strategies and serious resource mismatches in traditional solutions, and realizes the collaborative optimization of storage resource allocation efficiency and user incentive effects.

[0024] The following will detail various non-limiting embodiments of this application with reference to the accompanying drawings.

[0025] First, with reference to Figure 1 , the cloud disk space incentive method based on multi-dimensional behavior perception of this application will be described in detail.

[0026] This application provides a cloud disk space incentive method based on multi-dimensional behavior perception. The method includes: S1: The data acquisition layer monitors user behavior data in real time, where the user behavior data includes storage demand, behavior activity, storage expansion request, and life cycle status; S2: The state quantization layer quantizes the user behavior based on the user behavior data, obtaining multiple state quantization formulas corresponding to the user behavior data; S3: The intelligent decision-making engine generates an incentive policy based on the multiple state quantization formulas; S4: The feedback closed-loop layer dynamically updates the policy parameters based on the immediate reward feedback to achieve continuous optimization of the system.

[0027] This application proposes a cloud disk space incentive system based on multi-dimensional state perception and dynamic optimization mechanism, which is implemented through the following technical modules: (1) Data collection layer: Real-time monitor the user storage usage data and behavior operation sequences (such as upload / delete frequency, access thermodynamic characteristics).

[0028] (2) State quantization layer: Construct a four-dimensional feature vector St(u) that integrates the storage demand index (Dt), behavior activity (At), storage expansion request intensity (Rt), and user life cycle (ϕ), and dynamically adjust the weight matrix ω through reinforcement learning.

[0029] (3) Intelligent decision-making engine: Design an improved UCB algorithm, embed the cosine similarity calculation of the user state vector and the coupon type feature vector (Wc), and generate a personalized incentive policy.

[0030] (4) Feedback closed-loop layer (or feedback closed-loop mechanism): Dynamically update the policy parameters (Qc, Wc), etc. based on the immediate reward feedback (such as redemption rate, change in storage utilization rate) to achieve continuous optimization of the system.

[0031] Exemplarily, the data collection layer uses the multi-source heterogeneous data fusion technology to capture and process the following four types of core user behavior characteristics in real time, providing high-quality input for the state quantization layer: Storage Demand: Mainly collect the real-time storage usage of the user's cloud disk (accuracy 0.01GB), the maximum usage space of the user package, the file type distribution characteristics (such as classifying and counting the number of files and occupied space by video, document, image, etc.), and the historical storage growth curve (such as recording the storage volume change in the past 30 days) Activity Level: Collect the activity frequency of the user on the platform, such as the time, file size, and type information of the user's file upload, download, delete, and share operations Storage Expansion Request: Use the storage demand data and behavior activity data for discriminant calculation.

[0032] Lifecycle Status (Account Status): Integrate the user registration duration, active frequency (such as the number of logins in the past 7 days, usage of core functions (such as file upload, download, sharing, AI function usage, etc.)), membership level, and package type data to construct a four-state classification model for new users, growing users, mature users, and inactive users.

[0033] In some embodiments, the status quantification formula includes a storage requirement index calculation formula, and the storage requirement index calculation formula is as follows: , where, , sigmod represents the normalization function, S used represents the storage usage, S total represents the total storage of the package, Trend(H30) represents the linear regression trend slope based on 30-day historical data, represents the proportion of video files in the total files, and α, β, and γ all represent weight coefficients. Exemplarily, α, β, and γ can be set according to business goals or can be optimized and allocated according to the Shapley value. Specifically, for example, they can be set as α = 0.5, β = 0.3, and γ = 0.2.

[0034] In some embodiments, the status quantification formula further includes a behavior activity calculation formula, and the behavior activity calculation formula is as follows: , where, , represents the file size, i represents the i th file, i has a value range from 1 to n , n is a positive integer, both represent weight coefficients, represents the session density, = number of consecutive operations / total number of operations.

[0035] Exemplarily, the operation type weights can be determined according to business goals. For example, if the operation types include upload, share, download, and delete, the corresponding operation type weights can be set as follows: Upload = 1.0: Directly increases the storage usage, the core driving factor; Share = 0.8: Indirectly stimulates the storage demand (caching is required for others to access); Download = 0.6: Reflects the content reuse value but may be accompanied by a deletion risk; Delete = 0.3: Negative operation (reduces the storage demand), with the lowest weight.

[0036] Specifically, it can be set that ; the weights and parameter values here are all exemplary and can be set according to the actual situation.

[0037] Exemplarily, the standard for continuous operations is, for example, that the operation interval is less than 10 minutes. The selection of the 10-minute threshold is mainly through user behavior analysis, the distribution of user operation time intervals, and the highest storage for users with a high proportion of continuous operations within 10 minutes. The relevance decreases when it exceeds 10 minutes.

[0038] The above activity calculation formula gives higher weights to large file operations and high-frequency continuous operations.

[0039] In some embodiments, the state quantization formula further includes a storage expansion request intensity calculation formula, and the storage expansion request intensity calculation formula is as follows: , where

[0040] where , represents the explicit request intensity. If , represents the implicit demand score, represents the average daily number of uploaded files. In the implicit demand score, the storage pressure is the core driver, and the activity gain is used as an auxiliary for decision-making. One gain is obtained for every number of uploaded files greater than 50.

[0041] In some embodiments, the state quantization formula further includes a life cycle state calculation formula, and the life cycle state calculation formula is as follows: , where represents , H represents the number of operations of the user's core functions in the past 7 days, P represents the user's payment status, P = 0 means unpaid, P = 1 means paid, S represents the user's storage utilization rate, , 0, 1, 2, and 3 respectively represent new users, growth stage, maturity stage, and silent users.

[0042] Exemplarily, uploading 1 file + using the AI function 2 times means H = 3.

[0043] Explanation of priority and mutual exclusion. Priority order: silent users > new users > growth stage > maturity stage (if multiple states are satisfied simultaneously, they are overwritten according to the priority).

[0044] Threshold adjustment: The threshold can be optimized according to the data distribution of different services. For example, if the proportion of paying users is relatively low, the constraint of the paying situation can be weakened in the user threshold during the growth period.

[0045] In some embodiments, the incentive strategy includes a coupon type decision function, and the coupon type decision function is as follows: , where, represents a set of different coupon types, represents , , represents a four-dimensional real number space, represents the historical average reward, parameters of, , , represents a reward calculation model, c represents a specific coupon, represents the optimal c obtained by maximizing the reward.

[0046] Exemplarily, when the starting phase N < 1, the logarithmic term is disabled and the coupon type is directly randomly selected (exploring all possibilities).

[0047] In some embodiments, the incentive strategy further includes a reward calculation model, and the reward calculation model is as follows: , where, ∆U represents the user's usage behavior of the coupon, ∆U ∈ {-1, 1}, ∆U = 1 means received, ∆U = -1 means not received, ∆S represents the change rate of the storage space usage, ∆S ∈ R, R represents a one-dimensional real number space, μ1, μ2 = 1 both represent weight coefficients, and μ1 + μ2 = 1. Exemplarily, it can be set according to the business goal, or the most suitable parameter distribution can be explored using an AB test, or it can be updated using the gradient descent method. In this application, according to the business goal, it is expected that the user receives and uses the space coupon. For example, μ1 = 0.5, μ2 = 0.5.

[0048] In some embodiments, the feedback closed-loop layer dynamically updates the policy parameters based on the immediate reward feedback, including updating the parameters in real time according to the following formula , and : , where, , where, represents the actual reward, represents the learning rate, Represents the parameter vector of the new coupon type feature, Represents the new historical average reward, Represents the number of selections of the new various coupons, Represents the derivative symbol., The rate of change with behavior.

[0049] Through the analysis of business data, it is found that the user status The stronger (such as high activity), the parameter The more significant the impact on behavior (click, space growth, etc.). Therefore, it is assumed that the behavior and the user status Are linearly related. This assumption is based on the actual business on the one hand, and on the other hand, the time-consuming for single parameter update is reduced from millisecond level (non-linear model) to microsecond level, meeting the high-concurrency scenario (such as processing 100,000 requests per second).

[0050] At this time , Among them, The rate of change with The rate of change.

[0051] This application can assume that C is {C1, C2, C3}.

[0052] 1. Expanded coupon (c1) Content: Give 20GB of extra space with a validity period of 7 days.

[0053] Applicable scenario: The short-term storage demand of active users surges (such as photography enthusiasts backing up photos).

[0054] 2. Discount coupon (c2) Content: Enjoy a 20% discount on the upgraded package, limited to use within 7 days.

[0055] Applicable scenario: Long-term renewal users enhance their willingness to pay.

[0056] 3. Trial experience coupon (c3) Content: Free experience of the premium cloud disk function (such as VIP viewing in the screening hall) for 7 days.

[0057] Applicable scenario: New users or low-active users cultivate usage habits Assume the user 1 status vector:

[0058] Coupon initial parameters: ; ; ; ; First Iteration: In the cold start phase, N=0, UCB exploration is disabled (to avoid imaginary numbers), and the system randomly selects a coupon type. Suppose we choose a trial coupon for experience c3.

[0059] Reward calculation: Assume that the user receives the coupon and the storage utilization rate increases by 10%.

[0060] rc3=0.5×1+0.5×0.1=0.55; parameter update: Historical Reward Updates: ; Eigenvector update: ; Count Updates: ; For user 2 state vector:

[0061] Calculate the scores of various coupons: ; ; ; Result: Discount coupon c2 is selected.

[0062] Reward calculation: Assume that the user has not claimed the coupon and the storage usage rate remains unchanged.

[0063] rc2=-0.5; Parameter update: Historical Reward Updates:

[0064] Eigenvector update: ; Count Updates: .

[0065] And so on.

[0066] Among them, the conditions for stopping iteration of the decision model can comprehensively consider business goals and mathematical convergence. The update rate is lower than the threshold for multiple consecutive rounds (such as 0.1%). For business goals, consider achieving a preset target for the coupon redemption rate.

[0067] The cloud disk space incentive method based on multi-dimensional behavior perception in this application has the following advantages: 1. Innovatively design the user behavior state vector, integrating the state quantification model of storage demand characteristics, active behavior characteristics, storage expansion request characteristics and life cycle characteristics to accurately characterize the user storage status.

[0068] 2. The dynamic multi-factor decision function incorporates historical reward weights, dynamic exploration terms, and user state matching degrees into the decision-making process, enabling the reasonable quantification of the long-term returns of various coupon types and the balance between cold start exploration and utilization during the stable period.

[0069] The business objective-driven dual-factor reward model combines click-through rate and storage efficiency, which can be dynamically adjusted to adapt to business priorities, and the reward model fully matches the core business objectives. It should be noted that the method of one or more embodiments of the present application can be executed by a single device, such as a computer or a server. The method of this embodiment can also be applied to a distributed scenario, where multiple devices cooperate with each other to complete it. In such a distributed scenario, one of the multiple devices can only execute one or more steps of the method of one or more embodiments of the present application, and these multiple devices will interact with each other to complete the described method.

[0070] It should be noted that the above specifically describes certain embodiments of the present application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be executed in a different order than in the embodiments and still achieve the desired results. Additionally, the processes depicted in the figures do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain implementations, multitasking and parallel processing are also possible or may be advantageous.

[0071] Based on the same inventive concept, corresponding to the method of any of the above embodiments, the present application also discloses an electronic device; Specifically, Figure 2 The figure shows a schematic hardware structure diagram of an electronic device for a cloud disk space incentive method based on multi-dimensional behavior perception provided in this embodiment. The device may include: a processor 410, a memory 420, an input / output interface 430, a communication interface 440, and a bus 450. Among them, the processor 410, the memory 420, the input / output interface 430, and the communication interface 440 are communicatively connected to each other inside the device through the bus 450.

[0072] The processor 410 can be implemented in a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, etc., and is used to execute relevant programs to implement the technical solutions provided in the embodiments of the present application.

[0073] The memory 420 can be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage devices, dynamic storage devices, etc. The memory 420 can store the operating system and other application programs. When implementing the technical solutions provided in the embodiments of the present application through software or firmware, the relevant program codes are stored in the memory 420 and called and executed by the processor 410.

[0074] The input / output interface 430 is used to connect to the input / output module to implement information input and output. The input / output module can be configured as a component in the device (not shown in the figure) or externally connected to the device to provide corresponding functions. Among them, the input device can include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc., and the output device can include a display, a speaker, a vibrator, an indicator light, etc.

[0075] The communication interface 440 is used to connect to the communication module (not shown in the figure) to implement communication interaction between this device and other devices. Among them, the communication module can implement communication in a wired manner (for example, USB, network cable, etc.) or in a wireless manner (for example, mobile network, WIFI, Bluetooth, etc.).

[0076] The bus 450 includes a path for transmitting information between various components of the device (for example, the processor 410, the memory 420, the input / output interface 430, and the communication interface 440).

[0077] It should be noted that although the above device only shows the processor 410, the memory 420, the input / output interface 430, the communication interface 440, and the bus 450, in the specific implementation process, the device may also include other components necessary for normal operation. In addition, those skilled in the art can understand that the above device may also only include the components necessary for implementing the solution of the embodiments of the present application, and does not necessarily include all the components shown in the figure.

[0078] The electronic device in the above embodiment is used to implement the corresponding cloud disk space incentive method based on multi-dimensional behavior perception in any of the foregoing embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be elaborated here.

[0079] Based on the same inventive concept, corresponding to the method in any of the above embodiments, one or more embodiments of the present application also provide a non-transitory computer-readable storage medium storing computer instructions for causing the computer to execute the cloud disk space incentive method based on multi-dimensional behavior perception as described in any of the foregoing embodiments.

[0080] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette tapes, magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information accessible by a computing device.

[0081] The computer instructions stored in the storage medium of the above embodiment are used to cause the computer to execute the cloud disk space incentive method based on multi-dimensional behavior perception described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be elaborated here.

[0082] Those of ordinary skill in the art should understand that: The discussion of any of the above embodiments is only exemplary and is not intended to imply that the scope of the present application (including the claims) is limited to these examples; Under the idea of the present application, the technical features in the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations in different aspects of one or more embodiments of the present application as described above, and they are not provided in detail for the sake of brevity.

[0083] In addition, for simplicity of explanation and discussion, and in order not to make one or more embodiments of the present application difficult to understand, the well-known power / ground connections to integrated circuit (IC) chips and other components may or may not be shown in the provided drawings. In addition, the device may be shown in block diagram form to avoid making one or more embodiments of the present application difficult to understand, and this also takes into account the fact that the details of the implementation of these block diagram devices are highly dependent on the platform on which one or more embodiments of the present application are to be implemented (i.e., these details should be fully within the understanding of those skilled in the art). In the case where specific details (such as circuits) are set forth to describe the exemplary embodiments of the present application, it will be apparent to those skilled in the art that one or more embodiments of the present application can be implemented without these specific details or with variations of these specific details. Therefore, these descriptions should be considered illustrative rather than restrictive.

[0084] Although the present application has been described in connection with specific embodiments thereof, many alternatives, modifications, and variations of these embodiments will be apparent to those of ordinary skill in the art in light of the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) may be used with the embodiments discussed.

[0085] One or more embodiments of the present application are intended to cover all such alternatives, modifications, and variations that fall within the broad scope of the appended claims. Accordingly, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of one or more embodiments of the present application shall be included within the protection scope of the present application.

Claims

1. A cloud disk space incentive method based on multi-dimensional behavior perception, characterized in that The method includes: The data collection layer monitors user behavior data in real time, where the user behavior data includes storage requirements, behavior activity, storage expansion requests, and lifecycle status; The state quantification layer quantifies the user behavior based on the user behavior data to obtain multiple state quantification formulas corresponding to the user behavior data; The intelligent decision-making engine generates an incentive strategy based on multiple state quantification formulas; The feedback closed-loop layer dynamically updates the policy parameters based on immediate reward feedback to achieve continuous optimization of the system.

2. The cloud disk space incentive method based on multi-dimensional behavior perception according to claim 1, wherein The state quantification formula includes a storage requirement index calculation formula, and the storage requirement index calculation formula is as follows: ， Among them, , sigmod represents the normalization function, S used represents the storage usage, S total represents the total storage of the package, Trend(H30) represents the linear regression trend slope based on 30-day historical data, represents the proportion of video files in the total files, and α, β, and γ all represent weight coefficients.

3. The cloud disk space incentive method based on multi-dimensional behavior perception according to claim 1, wherein, The state quantification formula further includes a behavior activity calculation formula, and the behavior activity calculation formula is as follows: ， Among them, w type represents the weight of the operation type, represents the file size, i represents the i th file, i has a value range of 1 to n , n is a positive integer, and both ρ and θ represent weight coefficients, represents the session density, = number of consecutive operations / total number of operations.

4. The method for cloud disk space incentive based on multi-dimensional behavior perception according to claim 1, characterized in that The state quantification formula further includes a storage expansion request intensity calculation formula, and the storage expansion request intensity calculation formula is as follows: ， Wherein, ， Among them, S used represents the storage usage; S total represents the total storage of the package, represents the explicit request intensity. If is greater than 95%, the explicit request intensity is 1, represents the latent demand score, represents the average number of files uploaded per day.

5. The method for cloud disk space incentive based on multi-dimensional behavior perception according to claim 1, wherein The state quantification formula further includes a lifecycle status calculation formula, and the lifecycle status calculation formula is as follows: ， Among them, Treg represents the number of days since the user registered, H represents the number of operations of the user's core functions in the past 7 days, P represents the user's payment status, P = 0 indicates unpaid, P = 1 indicates paid, S represents the user's storage utilization rate, ,S used represents the storage usage amount, S total represents the total storage of the package, The values of 0, 1, 2, and 3 represent new users, growth stage, maturity stage, and inactive users respectively.

6. The method for cloud disk space incentive based on multi-dimensional behavior perception according to claim 1, wherein The incentive strategy includes a coupon type decision function, and the coupon type decision function is as follows: ， Among them, represents a set of different coupon types, represents the coupon type feature vector parameter, , represents a four-dimensional real number space, represents the historical average reward, and ε is 1e -5 , representing a parameter used to prevent the divisor from being zero, represents the total number of decision-making times, n c is the number of selections for various coupons, , represents the reward calculation model, c represents a specific coupon, represents the optimal c obtained by maximizing the reward.

7. The method for cloud disk space incentive based on multi-dimensional behavior perception according to claim 6, characterized in that, The incentive strategy further includes a reward calculation model, and the reward calculation model is as follows: ， Among them, represents the user's behavior of using coupons, ∈{-1, 1}, = 1 indicates receipt, = -1 indicates non-receipt, represents the change rate of storage space usage, , represents the one-dimensional real number space, and both μ1 and μ2 represent weight coefficients, satisfying μ1 + μ2 = 1.

8. The method for cloud disk space incentive based on multi-dimensional behavior perception according to claim 7, characterized in that, The feedback closed-loop layer dynamically updates the policy parameters based on immediate reward feedback, including updating the parameters in real time according to the following formula , and : ， Wherein, ， Among them, represents the actual reward, represents the learning rate, represents the parameter of the new coupon type feature vector, represents the new historical average reward, represents the number of selections of new various coupons, represents the derivative symbol.

9. An electronic device, the electronic device comprising: A memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor implements the method according to any one of claims 1 to 8 when executing the computer program.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the method according to any one of claims 1 to 8.

Images