Media resource optimization management system based on intelligent server

Abstract

The invention discloses a media resource optimization management system based on an intelligent server, which comprises a resource acquisition module, a resource analysis module, a resource storage module, a resource transmission module, a resource management module and an intelligent decision module. And intelligent management of media resources is realized. The intelligent server can quickly process a large amount of media data, powerful support is provided for operation of various optimization algorithms, the system can operate efficiently and accurately, and the high requirement of modern media services for resource management is met.

Description

Technical Field

[0001] The present invention relates to the technical field of resource optimization systems, and in particular to a media resource optimization management system based on an intelligent server. Background Art

[0002] Media resources refer to a collection of digital content that can be stored, transmitted and processed in a digital environment for various purposes such as information dissemination, entertainment, and education. It covers a variety of forms, such as text, images, audio, video, animation, etc. These resources are widely used in many fields such as Internet media platforms, digital libraries, online educational institutions, video conferencing systems, etc., and are an important part of the modern information society. With the rapid development of information technology, the number of media resources has exploded, and the demand for their efficient management and optimal utilization has become increasingly urgent.

[0003] There are defects in the current media resource optimization. First, there is a serious waste of resource storage: when existing intelligent servers store media resources, due to the lack of accurate analysis of resource popularity and usage frequency, a large amount of cold data that is not often used occupies valuable storage space. For example, in the server of a video website, some videos that are very old and have very low viewing volume still occupy the same storage location as popular videos, resulting in inefficient storage resource utilization. When faced with high concurrent access to popular resources, due to unreasonable storage distribution, slow reading or even freezing is prone to occur, which seriously affects the user experience; second, low resource transmission efficiency: in the process of media resource transmission, the existing system fails to fully consider the dynamic changes of the network environment. When the network is congested, the transmission strategy cannot be adjusted adaptively, resulting in increased resource transmission delay and poor user experience. At the same time, for different types of media resources (such as pictures, audio, and video), no differentiated transmission optimization solutions are adopted, and the advantages of network bandwidth cannot be fully utilized. For example, in a mobile network environment, the transmission of large-size high-definition videos is still carried out at a fixed bit rate, without dynamic adjustment based on network signal strength and available bandwidth, and video loading is often slow or interrupted; Third, the quality of resources is uneven: in the acquisition and processing of media resources, the existing technology lacks an effective quality monitoring and optimization mechanism. Some low-quality media resources are stored in the server without screening and processing, affecting the user's viewing and usage experience. In addition, when transcoding and compressing media resources, it is easy to lose picture quality and sound quality, further reducing the quality of resources. For example, after multiple compressions, some pictures have obvious distortion, which affects the visual effect of the picture.

[0004] In addition, there are problems with the management mode being cumbersome and complex. The traditional media resource management mode relies on manual operations. From the input, classification, update to deletion of resources, a large amount of repetitive work needs to be done by management personnel. This not only consumes a large amount of human and time costs, but also is prone to human errors. For example, when classifying and managing a large number of picture resources, it is difficult to ensure the accuracy of manual annotation, resulting in difficulties in subsequent retrieval and use. At the same time, there is also a lack of real-time monitoring and early warning mechanisms. The existing management mode cannot monitor the running status of media resources in real time and cannot detect problems such as resource failures, insufficient storage, and abnormal access in a timely manner. When problems occur, they often have an impact on the business, lacking the ability of early warning and automatic processing. For example, when the storage capacity of the server is about to be exhausted, management personnel cannot be informed in time, resulting in new media resources being unable to be stored normally and affecting the normal development of the business. Summary of the Invention

[0005] In view of the problems disclosed in the background art, the present invention provides an optimized media resource management system based on an intelligent server.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] An optimized media resource management system based on an intelligent server, characterized in that it includes a resource collection module, a resource analysis module, a resource storage module, a resource transmission module, a resource management module, and an intelligent decision-making module;

[0008] The resource collection module: is used to collect various media resources and transmit the collected resources to the resource analysis module;

[0009] The resource analysis module: is used to deeply analyze the collected resources. On the one hand, the analysis results are fed back to the resource collection module to guide the subsequent collection direction of the resource collection module. On the other hand, they are transmitted to the intelligent decision-making module;

[0010] The intelligent decision-making module: combines the resource analysis results and the overall system strategy to issue instructions to the resource storage module, the resource transmission module, and the resource management module;

[0011] The resource storage module: stores and allocates resources according to the instructions;

[0012] The resource transmission module: is used for efficient resource transmission;

[0013] The resource management module: is used for overall coordinated management of resources.

[0014] For the above-mentioned optimized media resource management system based on an intelligent server, its media resource optimization management method is as follows:

[0015] S1. Resource Collection and Preliminary Processing: The resource collection module collects media resources from various sources, and uses format conversion software tools to uniformly convert video files in different formats into MP4 format. Subsequently, image recognition algorithms are used to detect video parameters to ensure that the collected resources meet the system requirements. The collected resources are transmitted to the resource analysis module in the form of a data stream;

[0016] S2. Resource Analysis and Deep Processing: The resource analysis module uses artificial intelligence algorithms to perform image recognition on the content of pictures and videos to determine their themes and categories, and by statistically analyzing the user's access records and applying time series analysis algorithms, predicts the popularity of the resources.

[0017] S3. Resource Storage Optimization: According to the results of the resource analysis module, the resource storage module optimizes the storage of media resources; for popular resources, uses the high-speed interface of the storage device to store them in high-speed storage devices to improve access speed; for cold data, migrates it to low-cost large-capacity storage devices through data migration tools; at the same time, adopts distributed storage technology to disperse the resources and store them on multiple server nodes.

[0018] S4. Resource Transmission Optimization: In the resource transmission module, uses network monitoring hardware devices and software programs to obtain network parameters in real time. For video resources, adopts adaptive bitrate transmission technology, and dynamically adjusts the bitrate of the video according to the network conditions through the transmission rate algorithm.

[0019] S5. Intelligent Resource Management: The resource management module is responsible for the full life cycle management of media resources; when resources are entered, collaborates with the resource collection module to ensure that resources are accurately entered into the system; at the same time, by establishing resource indexes and tags, and using the retrieval positioning weight algorithm to achieve fast retrieval and accurate positioning of resources.

[0020] S6. Intelligent Decision-Driven Optimization: The intelligent decision module comprehensively analyzes various data in the processes of resource collection, analysis, storage, transmission, and management; uses deep reinforcement learning algorithms to optimize system decisions. When optimizing system decisions, it is achieved by constructing a deep neural network architecture containing an Actor network and a Critic network. The Actor network outputs continuous actions, and the Critic network evaluates the value of these actions. Its learning process is carried out by maximizing the long-term cumulative reward.

[0021] For the above media resource optimization management system based on an intelligent server, in step S2, the formula of the time series analysis algorithm is as follows:

[0022]

[0023] In algorithm formula ①, y tRepresents the resource heat value at time t, and θ j are the autoregressive coefficient and the moving average coefficient respectively, p and q are the orders of the model, and ∈ t is a white noise sequence.

[0024] In the above media resource optimization management system based on an intelligent server, in step S3, when the resources are stored dispersedly, the allocation of the storage locations is optimized through the following algorithm;

[0025]

[0026] In algorithm ②, S represents the optimal storage location, C i represents the cost of storing in the i-th storage device, F represents the heat of the resource, and T i represents the access speed of the i-th storage device.

[0027] In the above media resource optimization management system based on an intelligent server, in step S4, the transmission rate algorithm is:

[0028]

[0029] In algorithm ③, R represents the transmission rate, B represents the available network bandwidth at present, S represents the resource size, and T represents the transmission time expected by the user.

[0030] In the above media resource optimization management system based on an intelligent server, in step S5, the retrieval and positioning weight algorithm is:

[0031] TF-IDF(w, d) = TF(w, d) × IDF(w) ④;

[0032] In ④, w is the word, d is the document, TF(w, d) is the frequency of occurrence of the word w in the document d, n w is the total number of documents, and w is the number of documents containing the word.

[0033] In the above media resource optimization management system based on an intelligent server, in step S6, the reward algorithm for maximizing the long-term cumulative reward is as follows:

[0034]

[0035] In ⑤, let the state be s t and the action executed be a t and the reward obtained be r t , and the discount factor be γ, then the long-term cumulative reward J can be obtained.

[0036] The technical effects and advantages of the present invention:

[0037] The present invention discloses a media resource optimization management system based on an intelligent server, including a resource collection module, a resource analysis module, a resource storage module, a resource transmission module, a resource management module, and an intelligent decision-making module. This system makes full use of the powerful computing power and data analysis ability of the intelligent server to achieve the intelligent management of media resources. The intelligent server can quickly process a large amount of media data, providing strong support for the operation of various optimization algorithms, enabling the system to operate efficiently and accurately, and meeting the high requirements of modern media services for resource management;

[0038] Specifically, it has the following advantages:

[0039] 1. Storage optimization: Through the accurate analysis of resource popularity and usage frequency, the reasonable allocation of storage resources is realized, greatly improving the storage utilization rate and reducing the storage cost. Popular resources can be accessed quickly, and cold data is also properly stored, avoiding the waste of storage resources.

[0040] 2. Transmission optimization: Dynamically adjust the transmission strategy according to the network environment, improving the resource transmission efficiency, reducing the transmission delay, and enhancing the user experience. Different types of media resources can be targeted for transmission optimization, making full use of the network bandwidth and reducing problems such as video stuttering and slow loading.

[0041] 3. Quality optimization: Adopt advanced quality monitoring and optimization mechanisms to ensure the high quality of media resources. During the resource collection, processing, and transmission processes, the resource quality is improved through algorithms, reducing picture quality and audio quality losses, and providing a better viewing and usage experience for users.

[0042] 4. Process simplification: The automated management process greatly reduces manual operations, improves management efficiency, and reduces the probability of human errors. Managers can invest more time and energy in more valuable work, such as resource planning and operation.

[0043] 5. Real-time monitoring and early warning: Real-time monitor the operation status of media resources, be able to detect problems in a timely manner and give early warnings and handle them, ensuring the continuity and stability of the business. Prevent problems such as resource failures and insufficient storage in advance, avoiding the impact on the business, and improving the reliability of the system.

[0044] In summary, the media resource optimization management system based on an intelligent server of the present invention has significant advantages in media resource optimization and management, can effectively improve the performance of the intelligent server and the user experience, and has broad application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a schematic diagram of the present invention. Detailed implementation mode

[0046] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0047] This embodiment discloses a media resource optimization management system based on an intelligent server, which is characterized in that it includes a resource collection module, a resource analysis module, a resource storage module, a resource transmission module, a resource management module, and an intelligent decision-making module;

[0048] The resource collection module: is used to collect various media resources and transmit the collected resources to the resource analysis module;

[0049] The resource analysis module: is used to deeply analyze the collected resources. On the one hand, the analysis results are fed back to the resource collection module to guide the subsequent collection direction of the resource collection module. On the other hand, they are transmitted to the intelligent decision-making module;

[0050] The intelligent decision-making module: combines the resource analysis results and the overall system strategy to issue instructions to the resource storage module, the resource transmission module, and the resource management module;

[0051] The resource storage module: stores and allocates resources according to the instructions;

[0052] The resource transmission module: is used for efficient resource transmission;

[0053] The resource management module: is used for overall planning and management of resources.

[0054] For the above media resource optimization management system based on an intelligent server, its media resource optimization management method is as follows:

[0055] S1. Resource collection and preliminary processing: The resource collection module collects media resources from various sources (such as cameras, sensors, web crawlers, etc.), and uses format conversion software tools to uniformly convert video files in different formats into the MP4 format. Subsequently, the resolution, frame rate and other parameters of the video are detected through an image recognition algorithm to ensure that the collected resources meet the system requirements. The collected resources are transmitted to the resource analysis module in the form of a data stream;

[0056] S2. Resource analysis and in-depth processing: The resource analysis module uses artificial intelligence algorithms to perform image recognition on the content of pictures and videos to determine their themes and categories, and by statistically analyzing the user's access records and applying time series analysis algorithms, predicts the popularity of the resources.

[0057] S3. Resource storage optimization: According to the results of the resource analysis module, the resource storage module optimizes the storage of media resources. For popular resources, it uses the high-speed interface of the storage device to store them in high-speed storage devices to improve access speed. For cold data, it migrates them to low-cost large-capacity storage devices through data migration tools. At the same time, it adopts distributed storage technology to disperse the resources and store them on multiple server nodes.

[0058] S4. Resource transmission optimization: In the resource transmission module, it uses network monitoring hardware devices and software programs to obtain parameters such as network bandwidth and latency in real time. For video resources, it adopts adaptive bitrate transmission technology and dynamically adjusts the bitrate of the video according to the network conditions through a transmission rate algorithm to ensure smooth video playback.

[0059] S5. Intelligent resource management: The resource management module is responsible for the full life cycle management of media resources. When resources are entered, it collaborates with the resource collection module to ensure accurate entry of resources into the system. At the same time, by establishing resource indexes and tags and using a retrieval and positioning weight algorithm, it realizes the rapid retrieval and accurate positioning of resources.

[0060] S6. Intelligent decision-driven optimization: The intelligent decision module comprehensively analyzes various data in the processes of resource collection, analysis, storage, transmission, and management. It uses deep reinforcement learning algorithms to optimize system decisions. When optimizing system decisions, it is achieved by constructing a deep neural network architecture that includes an Actor network and a Critic network. The Actor network outputs continuous actions (such as adjusting resource storage strategies, optimizing transmission paths, etc.), and the Critic network evaluates the value of these actions. Its learning process is carried out by maximizing the long-term cumulative reward.

[0061] Among them, in step S2, the formula of the time series analysis algorithm is as follows:

[0062]

[0063] In algorithm formula ①, y t represents the resource popularity value at time t, and θ j are the autoregressive coefficient and the moving average coefficient respectively, p and q are the orders of the model, and ∈ t is a white noise sequence;

[0064] Algorithm formula ① performs a weighted sum on historical popularity values (y t-j ), combines the current and historical white noise (∈ t-j)To predict the resource popularity value at time t, the autoregressive part captures the long-term trend of resource popularity, and the moving average part considers the recent random fluctuations, so as to realize the prediction of resource popularity.

[0065] Among them, in step S3, when the resources are stored dispersedly, the allocation of storage locations is optimized through the following algorithm;

[0066]

[0067] In algorithm ②, S represents the optimal storage location, C i represents the cost of storing in the i-th storage device, F represents the popularity of the resource, and T i represents the access speed of the i-th storage device;

[0068] Through algorithm ②, for each storage device i, calculate the storage cost C i and the sum of the resource popularity F divided by the access speed T i At the same time, considering the two parameters F and T i comprehensively, take the i that makes the value the smallest. At this time, the corresponding storage device is the optimal storage location S.

[0069] This formula calculates the optimal storage location by comprehensively considering the storage cost, resource popularity, and access speed of the storage device. When the resource popularity is high, the impact of the access speed on the overall result is more significant, thus guiding the system to preferentially select a storage device with a faster access speed.

[0070] Among them, in step S4, the transmission rate algorithm is:

[0071]

[0072] In algorithm ③, R represents the transmission rate, B represents the available network bandwidth at present, S represents the resource size, and T represents the transmission time expected by the user;

[0073] In the formula, by calculating the rate required to transmit the resource within the user's expected time and then comparing it with the current available network bandwidth B, take the smaller value of the two as the actual transmission rate R; at this time, if the network bandwidth is sufficient to meet the requirement of transmitting the resource within the user's expected time, then use as the transmission rate; if the network bandwidth is insufficient and the transmission cannot be completed within the expected time, then use the current available network bandwidth B as the transmission rate.

[0074] Among them, in step S5, the retrieval and positioning weight algorithm is:

[0075] TF-IDF(w,d) = TF(w,d) × IDF(w) ④;

[0076] In ④, w is a word, d is a document, and TF(w, d) is the frequency of occurrence of word w in document d. n w is the total number of documents, and w is the number of documents containing the word;

[0077] The term frequency TF(w, d) reflects the frequency of occurrence of word w in document d. The more times it appears, the larger this value is, indicating that the word is more important within the document; In, n w the smaller it is, that is, the fewer documents a certain word appears in, the larger IDF(w) is, indicating that the word has stronger distinctiveness. The product of the two, TF-IDF(w, d), comprehensively considers the frequency of the word within the document and its distinctiveness within the entire document set. The higher the TF-IDF(w, d) value, the more important word w is to document d.

[0078] Among them, in step S6, the reward algorithm for maximizing the long-term cumulative reward is as follows:

[0079]

[0080] In ⑤, let the state be s t perform action a t and the obtained reward be r t , and the discount factor be γ, then the long-term cumulative reward J can be obtained; the long-term cumulative reward J can be used as an important indicator to evaluate the pros and cons of the media resource optimization strategy, helping the system select the optimal resource management strategy to maximize the long-term cumulative reward.

[0081] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent server-based media resource optimization management system, characterized in that It includes a resource collection module, a resource analysis module, a resource storage module, a resource transmission module, a resource management module, and an intelligent decision-making module; The resource collection module: is used to collect various media resources and transmit the collected resources to the resource analysis module; The resource analysis module: is used to deeply analyze the collected resources. On the one hand, the analysis results are fed back to the resource collection module to guide the subsequent collection direction of the resource collection module. On the other hand, they are transmitted to the intelligent decision-making module; The intelligent decision-making module: combines the resource analysis results and the overall system strategy to issue instructions to the resource storage module, the resource transmission module, and the resource management module; The resource storage module: stores and allocates resources according to the instructions; The resource transmission module: is used for efficient resource transmission; The resource management module: is used to comprehensively plan and manage resources in all aspects.

2. The media resource optimization management system based on an intelligent server according to claim 1, characterized in that, Its method for optimizing the management of media resources is as follows: S1. Resource collection and preliminary processing: The resource collection module collects media resources from various sources, and uses a format conversion software tool to uniformly convert video files in different formats into the MP4 format. Subsequently, the video parameters are detected through an image recognition algorithm to ensure that the collected resources meet the system requirements. The collected resources are transmitted to the resource analysis module in the form of a data stream; S2. Resource analysis and in-depth processing: The resource analysis module uses artificial intelligence algorithms to perform image recognition on the content of pictures and videos to determine their themes and categories, and by statistically analyzing the user's access records and applying time series analysis algorithms, predicts the popularity of the resources. S3. Resource storage optimization: According to the results of the resource analysis module, the resource storage module optimizes the storage of media resources; for popular resources, they are stored in a high-speed storage device using the high-speed interface of the storage device to improve the access speed; For cold data, it is migrated to a low-cost large-capacity storage device through a data migration tool; at the same time, distributed storage technology is adopted to disperse the resources and store them on multiple server nodes. S4. Resource transmission optimization: In the resource transmission module, network monitoring hardware devices and software programs are used to obtain network parameters in real time. For video resources, adaptive bitrate transmission technology is adopted, and the bitrate of the video is dynamically adjusted according to the network conditions through a transmission rate algorithm. S5. Intelligent resource management: The resource management module is responsible for the full life cycle management of media resources; when resources are entered, it cooperates with the resource collection module to ensure that the resources are accurately entered into the system; at the same time, by establishing resource indexes and tags, and using a retrieval and positioning weight algorithm, fast retrieval and accurate positioning of resources are realized. S6. Intelligent decision-making-driven optimization: The intelligent decision-making module comprehensively analyzes various data in the processes of resource collection, analysis, storage, transmission, and management; uses a deep reinforcement learning algorithm to optimize system decisions. When optimizing system decisions, it is achieved by constructing a deep neural network architecture that includes an Actor network and a Critic network. The Actor network outputs continuous actions, and the Critic network evaluates the value of these actions. Its learning process is carried out by maximizing the long-term cumulative reward.

3. The media resource optimization management system based on an intelligent server according to claim 2, wherein In step S2, the formula of the time series analysis algorithm is as follows: In algorithm formula ①, y t represents the resource heat value at time t, and θ j are the autoregressive coefficient and the moving average coefficient respectively, p and q are the orders of the model, and ∈ t is a white noise sequence.

4. The media resource optimization management system based on an intelligent server according to claim 2, wherein In step S3, when resources are stored dispersedly, the allocation of storage locations is optimized by the following algorithm; In Algorithm ②, S represents the optimal storage location, and C i represents the cost of storing in the i-th storage device, F represents the popularity of the resource, and T i represents the access speed of the i-th storage device.

5. The media resource optimization management system based on an intelligent server according to claim 2, wherein In step S4, the transmission rate algorithm is: In algorithm ③, R represents the transmission rate, B represents the available bandwidth of the current network, S represents the resource size, and T represents the transmission time expected by the user.

6. The media resource optimization management system based on an intelligent server according to claim 2, wherein In step S5, the retrieval and positioning weight algorithm is: TF-IDF(w,d) = TF(w,d) × IDF(w) ④; In ④, w is a word, d is a document, and TF(w, d) is the frequency of occurrence of word w in document d. n w is the total number of documents, and w is the number of documents containing the word.

7. The media resource optimization management system based on an intelligent server according to claim 2, wherein In step S6, the reward algorithm for maximizing the long-term cumulative reward is as follows: ⑤, let the state be s t Execute action a t The obtained reward is r t , with the discount factor being γ, the long-term cumulative reward J can be obtained.

Images