Multi-dimensional data management and analysis method and system of cloud platform data center

Abstract

The invention belongs to the technical field of data processing, and particularly relates to a multi-dimensional data management and analysis method and system for a cloud platform data center, and the system comprises a data collection and preprocessing module which is connected with a data source of the cloud platform data center, collects data, and sends the data to the cloud platform data center; cleaning and converting the collected data; the multi-dimensional data storage module is used for receiving the data from the data acquisition and preprocessing module and constructing a multi-dimensional storage model for data storage; the multi-dimensional data analysis module is used for carrying out clustering processing and classification prediction on the data based on the data in the multi-dimensional data storage module, and carrying out modeling prediction on the time sequence data and the unstructured data; and the data visualization module is used for displaying a prediction result of the multi-dimensional data analysis module. Delay is reduced, resource waste is avoided, data consistency is guaranteed, and the problems of redundancy and slow retrieval of traditional storage are solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of data processing, and particularly relates to a multi-dimensional data management and analysis method and system for a cloud platform data center. Background Art

[0002] With the rapid development of cloud computing, the amount of data stored in the cloud platform data center has increased explosively. Traditional data management and analysis means are difficult to meet the current requirements of the cloud platform data center for efficient management and in-depth analysis of massive data. On the one hand, in terms of data management, it is extremely difficult to integrate data from different sources and formats, and it is difficult to ensure the consistency and integrity of the data, and the data storage redundancy is high, resulting in waste of storage resources. On the other hand, in the field of data analysis, existing analysis methods usually only process data from a single dimension or a limited number of dimensions, and cannot fully explore the complex relationships and potential values hidden behind the data, making it difficult to help the cloud platform data center provide more accurate and intelligent services for users. Therefore, a multi-dimensional data management and analysis method and system for a cloud platform data center have been improved and designed. Summary of the Invention

[0003] In view of the above deficiencies existing in the prior art, the present invention provides a multi-dimensional data management and analysis method and system for a cloud platform data center to solve the problems in the above background art.

[0004] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0005] A multi-dimensional data management and analysis system for a cloud platform data center, including a data collection and preprocessing module, which establishes a connection with the data source of the cloud platform data center to collect data, and cleans and transforms the collected data;

[0006] A multi-dimensional data storage module, which receives the data from the data collection and preprocessing module and constructs a multi-dimensional storage model for data storage;

[0007] A multi-dimensional data analysis module, which performs clustering processing and classification prediction on the data based on the data in the multi-dimensional data storage module, and models and predicts time series data and unstructured data;

[0008] A data visualization module, which is used to display the results predicted by the multi-dimensional data analysis module.

[0009] Further, the data acquisition and preprocessing module includes a data acquisition interface unit, a data cleaning unit, and a data conversion unit. The input end of the data acquisition interface unit is connected to the data source of the cloud platform. The output end of the data acquisition interface unit is connected to the data cleaning unit. The data cleaning unit transports the cleaned data to the data conversion unit. After receiving the data, the data conversion unit uses encoding conversion technology to unify the structured data encoding format for different types of data, converts semi-structured data into a relational data structure according to the schema mapping rules, and extracts unstructured data information and converts it into a structured feature vector using natural language processing technology.

[0010] Further, the data acquisition interface unit uses a data communication protocol to connect to relational databases, NoSQL databases, log file systems, and message queues within the cloud platform to collect structured, semi-structured, and unstructured data.

[0011] Further, the data cleaning unit removes noise data through a rule-based filtering method, uses a hash algorithm to detect and remove duplicate data, fills numerical missing values by calculating the mean and median of similar data, and fills text missing values by inferring based on the semantic context of the data.

[0012] Further, the multi-dimensional data storage module includes a multi-dimensional data model construction unit, a distributed storage unit, and a cache management unit. The multi-dimensional data model construction unit builds a data storage model with multiple dimensions of business, time, and user. The distributed storage unit uses the distributed file system HDFS to store large-scale unstructured and semi-structured data, uses the columnar database Apache Hive to store structured data, and realizes fast data retrieval through a metadata management mechanism. The cache management unit uses Redis cache technology, sets the LRU algorithm to manage the cache space, ensures the effective use of the cache space, and regularly updates the cache data to ensure data consistency.

[0013] Further, the multi-dimensional data analysis module includes an association rule mining unit, a clustering analysis unit, a classification prediction unit, and a deep learning prediction unit. The association rule mining unit uses the Apriori or FP-Growth algorithm to explore potential association relationships in multi-dimensional data. The clustering analysis unit uses the K-Means or DBSCAN algorithm to cluster data with similar features based on the characteristics of the data in the dimensions of business, time, and user. The classification prediction unit uses decision trees, support vector machines, and naive Bayes algorithms to classify and predict data. The deep learning prediction unit uses neural network technology to model and predict time series data and unstructured data.

[0014] Further, in the data visualization module, users can perform visualization configuration through a custom configuration unit.

[0015] A multi-dimensional data management and analysis method for a cloud platform data center, comprising the following steps;

[0016] Step 1: Data collection and preprocessing, collecting data with different structures from the data sources of the cloud platform, and cleaning and transforming the collected data;

[0017] Step 2: Multi-dimensional data storage, constructing a multi-dimensional storage model based on the data after collection and preprocessing, and storing different types of data using a distributed storage method;

[0018] Step 3: Multi-dimensional data analysis, on the basis of the completion of data storage, exploring the potential relationships between data through association rule mining, grouping according to data characteristics by clustering analysis, judging the category to which the data belongs by classification prediction, and predicting and mining complex data rules by deep learning;

[0019] Step 4: Data visualization, presenting the results of multi-dimensional data analysis to users in an intuitive form, and users can customize the display method.

[0020] Further, in the said Step 1: Data collection and preprocessing, it includes using interfaces that support HTTP, FTP, and JDBC protocols to collect structured, semi-structured, and unstructured data from relational databases, NoSQL databases, log file systems, and message queue data sources of the cloud platform. After the data collection is completed, noise and duplicate parts in the data are removed by means of rule filtering and hash algorithms, numerical missing values are filled by calculating the mean or median, text missing values are filled according to semantics, the encoding of structured data is uniformly converted to UTF-8, semi-structured data is converted according to pattern mapping rules, and unstructured text data is processed using natural language processing techniques.

[0021] Further, the said Step 2: Multi-dimensional data storage includes, after the data collection and preprocessing are completed, constructing a multi-dimensional storage model based on business, time, and users. Based on the constructed model, HDFS is used to store unstructured and semi-structured data, Hive is used to store structured data, hot data is judged according to the constructed model, Redis cache is adopted, and the cache space is managed by the LRU algorithm and updated regularly.

[0022] Compared with the prior art, the present invention has the following beneficial effects:

[0023] 1. The present invention achieves efficient data management: In the data collection phase, data collection interfaces using various protocols such as HTTP, FTP, and JDBC are utilized to widely adapt to data sources such as relational databases, NoSQL databases, log file systems, and message queues within the cloud platform, comprehensively collecting structured, semi-structured, and unstructured data. After data cleaning, rule-based filtering and hash algorithms are used to remove noise and duplicate data. For numerical and text missing values, the mean / median is calculated and context semantic filling is adopted respectively to ensure data quality. Through data transformation, structured data is encoded uniformly as UTF-8, semi-structured data is transformed according to the schema mapping rules, and unstructured text data is transformed using natural language processing techniques, laying a good foundation for subsequent storage. A multi-dimensional storage model including business, time, user, etc. is constructed. The distributed file system HDFS is used to store large-scale unstructured and semi-structured data, and the columnar database Apache Hive is combined to store structured data, significantly improving storage efficiency. The metadata management mechanism can quickly locate and obtain data, enabling fast retrieval. For frequently accessed hot data, the Redis caching technology is adopted, achieving sub-millisecond access speed based on in-memory storage. The LRU algorithm is used to manage the cache space and update data regularly, reducing latency, avoiding resource waste, and ensuring data consistency, thus solving the problems of traditional storage redundancy and slow retrieval.

[0024] 2. In-depth data analysis: Based on the data collected, preprocessed, and stored in the early stage, association rule mining algorithms such as Apriori and FP-Growth are used to mine potential association relationships among multi-dimensional data. Through clustering analysis algorithms such as K-Means and DBSCAN, data is clustered according to its characteristics in different dimensions. Classification algorithms such as decision trees, support vector machines, and naive Bayes are used for classification prediction. Combining neural network technologies such as RNN, LSTM, and CNN in deep learning, RNN and its variant LSTM are used to model time series data, and CNN is used to extract and analyze features of unstructured data such as images and videos. These technical means cooperate with each other to break through the limitations of a single dimension and comprehensively mine the potential value of data, providing an accurate basis for decision-making.

[0025] 3. Optimize the user experience: The data visualization module provides a rich library of visualization components such as bar charts, line charts, pie charts, heat maps, scatter plots, and radar charts based on the results of multi-dimensional data analysis, supporting users to customize configurations such as data display dimensions, metrics, and chart styles. Through the dynamic update function, data changes are presented in real time, visually presenting the data, reducing the user's understanding difficulty, and improving the interaction efficiency between the cloud platform and users, thereby enhancing the service quality. Description of the Drawings

[0026] Figure 1Flowchart of a multi-dimensional data management and analysis method for a cloud platform data center according to the present invention;

[0027] Figure 2 Constitution diagram of a multi-dimensional data management and analysis system for a cloud platform data center according to the present invention; Specific implementation manners

[0028] In order to enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0029] Among them, the accompanying drawings are only for illustrative purposes, showing only schematic diagrams, rather than physical diagrams, and should not be construed as a limitation of this patent; in order to better illustrate the embodiments of the present invention, some components in the accompanying drawings will be omitted, enlarged or reduced, and do not represent the dimensions of actual products; for those skilled in the art, it is understandable that some well-known structures and their descriptions in the accompanying drawings may be omitted.

[0030] In the accompanying drawings of the embodiments of the present invention, the same or similar reference numerals correspond to the same or similar components; in the description of the present invention, it should be understood that if terms such as "upper", "lower", "left", "right", "inner", "outer", etc. are used to indicate the orientation or positional relationship, it is based on the orientation or positional relationship shown in the accompanying drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, so the terms describing the positional relationship in the accompanying drawings are only for illustrative purposes and should not be construed as a limitation of this patent. For those of ordinary skill in the art, the specific meanings of the above terms can be understood according to specific circumstances.

[0031] In the description of the present invention, unless otherwise clearly defined and limited, if terms such as "connection" are used to indicate the connection relationship between components, this term should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or integrated; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and can be the communication inside two components or the interaction relationship between two components. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0032] Embodiment 1:

[0033] As Figure 1-2 shown, the present invention. Specifically, a multi-dimensional data management and analysis system for a cloud platform data center includes a data collection and preprocessing module that establishes a connection with the data source of the cloud platform data center, collects data, and cleans and transforms the collected data;

[0034] The multi-dimensional data storage module receives data from the data acquisition and preprocessing module and constructs a multi-dimensional storage model for data storage;

[0035] The multi-dimensional data analysis module performs clustering processing and classification prediction on the data based on the data in the multi-dimensional data storage module, and models and predicts time series data and unstructured data;

[0036] The data visualization module is used to display the results predicted by the multi-dimensional data analysis module.

[0037] Further, the data acquisition and preprocessing module includes a data acquisition interface unit, a data cleaning unit, and a data conversion unit. The input end of the data acquisition interface unit is connected to the data source of the cloud platform. The output end of the data acquisition interface unit is connected to the data cleaning unit. The data cleaning unit transports the cleaned data to the data conversion unit. After receiving the data, the data conversion unit uses encoding conversion technology to unify the structured data encoding format for different types of data, converts semi-structured data into a relational data structure according to the pattern mapping rules, and extracts unstructured data information and converts it into a structured feature vector using natural language processing technology.

[0038] Further, the data acquisition interface unit uses a data communication protocol to connect to relational databases, NoSQL databases, log file systems, and message queues in the cloud platform to collect structured, semi-structured, and unstructured data.

[0039] Further, the data cleaning unit removes noise data through a rule-based filtering method, detects and removes duplicate data using a hash algorithm, fills numerical missing values by calculating the mean and median of similar data, and fills text missing values by inferring based on the semantic context of the data.

[0040] Further, the multi-dimensional data storage module includes a multi-dimensional data model construction unit, a distributed storage unit, and a cache management unit. The multi-dimensional data model construction unit builds a data storage model with multiple dimensions of business, time, and users. The distributed storage unit uses the distributed file system HDFS to store large-scale unstructured and semi-structured data, uses the columnar database Apache Hive to store structured data, and realizes fast data retrieval through a metadata management mechanism. The cache management unit uses Redis cache technology, sets the LRU algorithm to manage the cache space to ensure the effective use of the cache space, and updates the cache data regularly to ensure data consistency.

[0041] Further, the multi-dimensional data analysis module includes an association rule mining unit, a clustering analysis unit, a classification prediction unit, and a deep learning prediction unit. Among them, the association rule mining unit uses the Apriori or FP-Growth algorithm to explore potential association relationships in multi-dimensional data; the clustering analysis unit uses the K-Means or DBSCAN algorithm to cluster data with similar characteristics according to the characteristics of data in the business, time, and user dimensions; the classification prediction unit uses decision trees, support vector machines, and naive Bayes algorithms to classify and predict data; the deep learning prediction unit uses neural network technology to model and predict time series data and unstructured data.

[0042] Further, in the data visualization module, users can perform visualization configuration through the custom configuration unit.

[0043] This embodiment also provides a multi-dimensional data management and analysis method for a cloud platform data center, including the following steps;

[0044] Step 1: Data collection and preprocessing, collect data with different structures from the data sources of the cloud platform, and clean and transform the collected data;

[0045] Step 2: Multi-dimensional data storage, build a multi-dimensional storage model based on the data after collection and preprocessing, and use the distributed storage method to store different types of data;

[0046] Step 3: Multi-dimensional data analysis, on the basis of completing data storage, explore potential relationships between data through association rule mining, group data according to data characteristics through clustering analysis, judge the category to which the data belongs through classification prediction, and mine complex data rules through deep learning prediction;

[0047] Step 4: Data visualization, present the results of multi-dimensional data analysis to users in an intuitive form, and users can customize the display method.

[0048] Further, in Step 1: Data collection and preprocessing, it includes using interfaces that support HTTP, FTP, and JDBC protocols to collect structured, semi-structured, and unstructured data from relational databases, NoSQL databases, log file systems, and message queue data sources of the cloud platform. After completing data collection, use rule filtering and hash algorithms to remove noise and duplicate parts in the data, fill numerical missing values by calculating the mean or median, fill text missing values according to semantics, uniformly convert the encoding of structured data to UTF-8, convert semi-structured data according to schema mapping rules, and process unstructured text data using natural language processing technology.

[0049] Further, Step 2: Multidimensional data storage includes, after the data collection and preprocessing are completed, constructing a multidimensional storage model based on business, time, and users. Based on the constructed model, use HDFS to store unstructured and semi-structured data, use Hive to store structured data, judge hot data according to the constructed model, adopt Redis cache, manage the cache space through the LRU algorithm, and update it regularly.

[0050] The data collection interface unit uses protocols such as HTTP, FTP, and JDBC to connect to the relational database (storing structured data such as orders and user basic information), NoSQL database (storing semi-structured data such as product details), log file system (recording unstructured data such as user operation logs), and message queue (processing real-time message data) within the e-commerce cloud platform. For example, collect the operation logs of users browsing product pages from the log file system, and collect the order information and basic attribute data of users from the relational database.

[0051] After the data is collected, the data cleaning unit starts to work. In the log data, through a rule-based filtering method, eliminate noise data such as invalid page access records caused by network fluctuations of users; use the hash algorithm to detect and remove duplicate user login logs, etc. For numerical missing values, such as the missing product discount values in some orders, fill them by calculating the average value of similar products in similar orders; for text missing values, such as incomplete filling of some user address information, speculate and fill them according to the common address format and context semantics of the user's location.

[0052] The data conversion unit processes different types of data. For structured data, unify the encoding formats of fields such as user names in the database to UTF-8 encoding. For semi-structured product JSON data, according to the predefined schema mapping rules, convert information such as the name, price, and description of the product into a relational data structure for subsequent storage and analysis. For unstructured user review text data, use lexical analysis in natural language processing technology to identify keywords, and then determine the relationships between words through syntactic analysis, extract key information and convert it into a structured feature vector. For example, convert the description of product quality in user reviews into structured label vectors such as "quality good review", "quality average", "quality bad review", etc.

[0053] Multidimensional data storage stage

[0054] The multi-dimensional data model construction unit constructs a storage model mainly based on dimensions such as business, time, and users. In the business dimension, the e-commerce business is subdivided into sections such as product management, order processing, and user marketing; the time dimension is accurate to the second level to accurately record the time of user actions such as placing orders and browsing products; the user dimension covers the basic attributes of users (age, gender, region, etc.), behavioral characteristics (number of times of browsing products, purchase frequency, etc.), and consumption preferences (favorite product categories, price ranges, etc.).

[0055] The distributed storage unit stores large-scale unstructured and semi-structured data based on the constructed model, such as storing a large amount of unstructured data such as user evaluation text data and product pictures in Hadoop Distributed File System (HDFS). It uses the columnar database Apache Hive to store structured data, such as user order data and user basic information data. Through the metadata management mechanism, indexes are established for the data stored in HDFS and Hive to achieve fast retrieval. For example, when querying all orders of a certain user within a specific time period, the relevant data stored in Hive can be quickly located through the metadata.

[0056] The cache management unit adopts the Redis cache technology for hot data with frequent access. For example, for the detailed data of popular products, it is cached in Redis, and sub-millisecond data access speed is achieved based on in-memory storage. The cache space is managed by the LRU (Least Recently Used) algorithm. When the cache space is insufficient, the least recently used cache data is eliminated to ensure the effective utilization of the cache space. At the same time, the latest popular product data is obtained from the distributed storage unit regularly to update the Redis cache to ensure data consistency.

[0057] Multi-dimensional data analysis stage

[0058] The association rule mining unit uses algorithms such as the Apriori algorithm and the FP-Growth algorithm to mine potential association relationships among multi-dimensional data. By setting the minimum support and minimum confidence, association rules such as users who have purchased baby milk powder are likely to purchase baby diapers are discovered, providing a basis for product recommendations.

[0059] The clustering analysis unit adopts algorithms such as K-Means and DBSCAN to cluster according to the characteristics of data in different dimensions. For example, according to characteristics such as the consumption amount, purchase frequency, and favorite product categories of users, users are clustered into different groups such as high-consumption active user groups and low-consumption low-frequency user groups to implement personalized marketing strategies.

[0060] The classification prediction unit uses algorithms such as decision trees, support vector machines, and naive Bayes to perform classification prediction based on the user's historical data. Based on data such as the user's historical purchase records, browsing behavior, and credit rating, it predicts whether the user will make a purchase within the next week, providing a reference for the precise placement of marketing activities.

[0061] The deep learning prediction unit uses recurrent neural networks (RNNs) and their variant, long short-term memory networks (LSTMs), to model time series data and predict the future trend of commodity sales volume in the data center of the e-commerce cloud platform, so as to make inventory allocation in advance. It uses convolutional neural networks (CNNs) to extract features and analyze unstructured data such as product pictures uploaded by users, and predicts the preference trends of users for product pictures of different styles, helping merchants optimize product display pictures.

[0062] Data visualization stage

[0063] The visualization component library unit provides rich visualization chart components such as bar charts, line charts, pie charts, heat maps, scatter plots, and radar charts. For example, a bar chart is used to show the sales volume comparison of different commodity categories, and a line chart is used to present the change trend of commodity sales volume over time.

[0064] The custom configuration unit supports users to customize data display dimensions, metrics, chart styles, etc. through operations such as dragging. Users can, according to their own needs, choose to display only the purchase data of users in specific regions, or set the chart style to their favorite style.

[0065] The dynamic update unit updates the data visualization results in real time to show the data changes. For example, when new order data is generated, the statistical chart of the order volume is updated in real time, and the data update frequency can be set according to user needs, enabling users to keep abreast of the latest developments in e-commerce business.

[0066] Introduction to the working principle

[0067] In the multi-dimensional data management and analysis system of the present invention, each module collaborates closely to form an organic whole. The data collection and preprocessing module is responsible for collecting data from various data sources of the cloud platform, and cleaning and transforming it to provide high-quality and uniformly formatted data for subsequent processing. The multi-dimensional data storage module reasonably stores data according to the constructed multi-dimensional model, and improves the data access speed through a caching mechanism. The multi-dimensional data analysis module uses a variety of algorithms to deeply mine and analyze the stored data from different perspectives to discover the potential value in the data. The data visualization module presents the analysis results to users in an intuitive and easy-to-understand manner. Users perform interactive operations through the visualization interface, and their feedback can prompt the previous modules to optimize and adjust, so as to achieve the efficient management and multi-dimensional in-depth analysis of the massive data in the cloud platform data center, and continuously improve the service quality and operation efficiency of the cloud platform.

[0068] The above are only embodiments of the present invention. The circuits, electronic components, and modules involved are all prior arts, which can be fully implemented by those skilled in the art without further elaboration. The content protected by this application does not involve improvements to software and methods. Common knowledge such as specific structures and characteristics in the solution is not described in detail here. Those of ordinary skill in the art know all the common technical knowledge in the technical field to which the invention belongs before the application date or priority date, can know all the prior arts in this field, and have the ability to apply conventional experimental means before this date. Those of ordinary skill in the art can, under the inspiration given by this application and combined with their own abilities, improve and implement this solution. Some typical well-known structures or well-known methods should not become obstacles for those of ordinary skill in the art to implement this application. It should be noted that for those skilled in the art, without departing from the structure of the present invention, several deformations and improvements can still be made, and these should also be regarded as the protection scope of the present invention, and these will not affect the implementation effect of the present invention and the practicality of the patent.

Claims

1. A multi-dimensional data management and analysis system for a cloud platform data center, characterized in that: It includes a data acquisition and preprocessing module that establishes a connection with the data source in the cloud platform data center to acquire data, and cleans and transforms the acquired data; A multi-dimensional data storage module that receives data from the data acquisition and preprocessing module and constructs a multi-dimensional storage model for data storage; A multi-dimensional data analysis module that performs clustering processing and classification prediction on the data based on the data in the multi-dimensional data storage module, and models and predicts time series data and unstructured data; A data visualization module for displaying the results predicted by the multi-dimensional data analysis module.

2. The multi-dimensional data management and analysis system for a cloud platform data center according to claim 1, characterized in that: The data acquisition and preprocessing module includes a data acquisition interface unit, a data cleaning unit, and a data conversion unit. The input end of the data acquisition interface unit is connected to the data source of the cloud platform, the output end of the data acquisition interface unit is connected to the data cleaning unit, the data cleaning unit transports the cleaned data to the data conversion unit. After receiving the data, the data conversion unit uses encoding conversion technology to unify the structured data encoding format for different types of data, converts semi-structured data into a relational data structure according to the pattern mapping rules, and uses natural language processing technology to extract unstructured data information and convert it into a structured feature vector.

3. The multi-dimensional data management and analysis system for a cloud platform data center according to claim 2, characterized in that: The data acquisition interface unit uses a data communication protocol to connect to relational databases, NoSQL databases, log file systems, and message queues in the cloud platform to acquire structured, semi-structured, and unstructured data.

4. The multi-dimensional data management and analysis system of a cloud platform data center according to claim 3, characterized in that: The data cleaning unit removes noise data through a rule-based filtering method, uses a hash algorithm to detect and remove duplicate data, fills numerical missing values by calculating the mean and median of similar data, and fills text missing values by inferring based on the data context semantics.

5. The multi-dimensional data management and analysis system for a cloud platform data center according to claim 1, characterized in that: The multi-dimensional data storage module includes a multi-dimensional data model construction unit, a distributed storage unit, and a cache management unit. The multi-dimensional data model construction unit builds a data storage model with multiple dimensions of business, time, and users; the distributed storage unit uses the distributed file system HDFS to store large-scale unstructured and semi-structured data, uses the columnar database Apache Hive to store structured data, and realizes fast data retrieval through a metadata management mechanism; the cache management unit adopts Redis cache technology, sets the LRU algorithm to manage the cache space to ensure the effective use of the cache space, and regularly updates the cache data to ensure data consistency.

6. The multi-dimensional data management and analysis system of a cloud platform data center according to claim 1, characterized in that: The multi-dimensional data analysis module includes an association rule mining unit, a clustering analysis unit, a classification prediction unit, and a deep learning prediction unit. The association rule mining unit uses the Apriori or FP-Growth algorithm to explore potential association relationships in multi-dimensional data; the clustering analysis unit uses the K-Means or DBSCAN algorithm to cluster data with similar features based on the characteristics of the data in the dimensions of business, time, and users; the classification prediction unit uses decision trees, support vector machines, and naive Bayes algorithms to perform classification prediction on the data; the deep learning prediction unit uses neural network technology to model and predict time series data and unstructured data.

7. The multi-dimensional data management and analysis system for a cloud platform data center according to claim 1, wherein: In the data visualization module, users can customize the configuration unit for visualization configuration.

8. A multi-dimensional data management and analysis method for a cloud platform data center, characterized in that: It includes the following steps: Step 1: Data collection and preprocessing. Collect data with different structures from the data sources of the cloud platform, and clean and transform the collected data. Step 2: Multi-dimensional data storage. Build a multi-dimensional storage model based on the data after collection and preprocessing, and use the distributed storage method to store different types of data. Step 3: Multi-dimensional data analysis. On the basis of the completion of data storage, explore the potential relationships between data through association rule mining, group data according to data characteristics by clustering analysis, judge the category to which the data belongs by classification prediction, and predict and mine complex data rules by deep learning. Step 4: Data visualization. Present the results of multi-dimensional data analysis to users in an intuitive form, and users can customize the display method.

9. The multi-dimensional data management and analysis method for a cloud platform data center according to claim 8, characterized in that: The said Step 1: Data collection and preprocessing includes using interfaces that support HTTP, FTP, and JDBC protocols to collect structured, semi-structured, and unstructured data from relational databases, NoSQL databases, log file systems, and message queue data sources of the cloud platform. After the data collection is completed, use rule filtering and hash algorithms to remove the noise and duplicate parts in the data, fill the numeric missing values by calculating the mean or median, fill the text missing values according to semantics, uniformly convert the encoding of structured data to UTF-8, convert semi-structured data according to the schema mapping rules, and process unstructured text data using natural language processing technology.

10. A multi-dimensional data management and analysis system for a cloud platform data center according to claim 9, characterized in that: The said Step 2: Multi-dimensional data storage includes, after the data collection and preprocessing are completed, building a multi-dimensional storage model with business, time, and users. Based on the built model, use HDFS to store unstructured and semi-structured data, use Hive to store structured data, judge hot data according to the built model, adopt Redis cache, and manage the cache space through the LRU algorithm and update it regularly.

Images