Data processing intelligent analysis method and system based on AI algorithm

Abstract

The invention belongs to the technical field of data processing, and discloses a data processing intelligent analysis method and system based on an AI algorithm, and the method comprises the steps: obtaining a to-be-processed original data set which comprises enterprise business system data; performing first classification on the original data set by using a pre-trained first data classification model based on a first AI algorithm to obtain a normal data set and an abnormal data set; performing second classification on the normal data set by using a pre-trained second data classification model based on a second AI algorithm to obtain a plurality of different types of normal data sets; and calculating an information gain index of each type of normal data set, and storing each type of normal data set according to the information gain index. According to the embodiment of the invention, the enterprise data can be classified, analyzed and stored, so that data resources can be better utilized, and better technical support and theoretical basis are provided for realizing data-driven decisions.

Description

Technical Field

[0001] The present invention belongs to the field of data processing technology, and in particular to a data processing intelligent analysis method and system based on AI algorithm. Background Art

[0002] In today's digital age, enterprises are faced with the challenge of generating and processing large amounts of data. With the rapid development of information technology, especially the widespread application of big data and artificial intelligence (AI) technology, enterprises have gained opportunities in data analysis and decision-making. The amount of data generated by enterprise business systems continues to increase, covering customer behavior, sales records, market dynamics, production processes and many other aspects. The complexity and diversity of these data pose major challenges to the accuracy and efficiency of traditional data analysis methods, which is an urgent problem to be solved. Summary of the invention

[0003] The purpose of this invention is to provide a data processing intelligent analysis method and system based on AI algorithm to solve the deficiencies in the prior art, and to be able to classify, analyze and store enterprise data in order to better utilize data resources and provide better technical support and theoretical basis for realizing data-driven decision-making.

[0004] An embodiment of the present application provides a data processing intelligent analysis method based on an AI algorithm, the method comprising: Acquire an original data set to be processed, wherein the original data set includes: enterprise business system data; Using a pre-trained first data classification model based on a first AI algorithm, the original data set is subjected to a first classification to obtain a normal data set and an abnormal data set; Using a pre-trained second data classification model based on a second AI algorithm, the normal data set is subjected to a second classification to obtain a plurality of normal data sets of different types; An information gain index of each type of normal data set is calculated, and each type of normal data set is stored according to the information gain index.

[0005] Optionally, the information gain index is calculated as follows:

[0006] Among them, the The information gain index of the k-th type of normal data set is Original Dataset The entropy of is the normal data set of the kth type The entropy of For normal data set The probability of type j appearing is Normal data set The size of is an adjustment factor greater than 0, and m is the number of types of normal data sets.

[0007] Optionally, storing each type of normal data set according to the information gain index includes: storing the normal data set whose information gain index is greater than a preset threshold in a first storage method; and storing the normal data set whose information gain index is not greater than the preset threshold in a second storage method.

[0008] Optionally, the first storage method is blockchain storage, and the second storage method is cloud platform storage.

[0009] Another embodiment of the present application provides a data processing intelligent analysis system based on an AI algorithm, the system comprising: An acquisition module is used to acquire an original data set to be processed, wherein the original data set includes: enterprise business system data; A first classification module, configured to perform a first classification on the original data set by using a pre-trained first data classification model based on a first AI algorithm to obtain a normal data set and an abnormal data set; A second classification module, used to perform a second classification on the normal data set by using a pre-trained second data classification model based on a second AI algorithm to obtain a plurality of normal data sets of different types; The storage module is used to calculate the information gain index of each type of normal data set, and store each type of normal data set according to the information gain index.

[0010] Yet another embodiment of the present application provides a storage medium, wherein the storage medium stores a computer program, wherein the computer program is configured to execute any of the above methods when running.

[0011] Yet another embodiment of the present application provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute any of the methods described above.

[0012] Compared with the prior art, the present invention provides a data processing intelligent analysis method based on an AI algorithm, which obtains an original data set to be processed, wherein the original data set includes: enterprise business system data; using a pre-trained first data classification model based on a first AI algorithm, the original data set is first classified to obtain a normal data set and an abnormal data set; using a pre-trained second data classification model based on a second AI algorithm, the normal data set is second classified to obtain multiple normal data sets of different types; the information gain index of each type of normal data set is calculated, and each type of normal data set is stored according to the information gain index, so that enterprise data can be classified, analyzed and stored, in order to better utilize data resources and provide better technical support and theoretical basis for data-driven decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 A hardware structure block diagram of a computer terminal for a data processing intelligent analysis method based on an AI algorithm provided in an embodiment of the present invention; Figure 2 A flowchart of a data processing intelligent analysis method based on an AI algorithm provided in an embodiment of the present invention; Figure 3 A schematic diagram of the structure of a data processing intelligent analysis system based on an AI algorithm provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0014] The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, but should not be construed as limiting the present invention.

[0015] The embodiment of the present invention first provides a data processing intelligent analysis method based on AI algorithm, which can be applied to electronic devices such as computer terminals, specifically ordinary computers, etc.

[0016] The following describes it in detail by taking running on a computer terminal as an example. Figure 1 The hardware structure block diagram of a computer terminal for a data processing intelligent analysis method based on an AI algorithm provided in an embodiment of the present invention. Figure 1 As shown, the computer device includes a processor, a memory and a network interface connected via a system bus, wherein the memory may include a non-volatile storage medium and an internal memory. The non-volatile storage medium may store an operating system and a computer program. The computer program includes program instructions, which, when executed, may enable the processor to execute any data processing intelligent analysis method based on an AI algorithm.

[0017] The processor is used to provide computing and control capabilities and support the operation of the entire computer equipment.

[0018] The internal memory provides an environment for the operation of computer programs in non-volatile storage media. When the computer program is executed by the processor, the processor can execute any data processing intelligent analysis method based on AI algorithm.

[0019] The network interface is used for network communication, such as sending assigned tasks, etc. Those skilled in the art will understand that Figure 1 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0020] It should be understood that the processor may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0021] See also Figure 2 , an embodiment of the present invention provides a data processing intelligent analysis method based on AI algorithm, which may include the following steps: S201, obtaining an original data set to be processed, wherein the original data set includes: enterprise business system data; In the present invention, the first step involves obtaining the raw data set to be processed, especially the data generated by the enterprise business system. These raw data sets usually come from multiple business systems, including but not limited to customer relationship management (CRM) system, enterprise resource planning (ERP) system, financial management system, supply chain management system, etc. These systems continuously record the operational data of the enterprise, covering a wide range of content, such as customer information, sales records, financial statements, inventory management, and market analysis data. Therefore, the types and structures of the raw data sets are highly diverse and complex.

[0022] Specifically, first, enterprises need to identify all relevant business systems and their data sources to ensure that all important data that may affect business decisions are covered. This includes internal systems, external market data and third-party data sources. Use appropriate APIs (application programming interfaces), data export tools or data crawler technology to extract the required data from the identified systems. During the data extraction process, it is necessary to ensure the integrity and consistency of the data to avoid the impact of missing or inconsistent data on subsequent analysis. The extracted raw data needs to be cleaned and formatted to remove duplicate values, fill in missing values ​​and unify the data format. This step ensures that the acquired data set is high-quality and valuable for subsequent analysis. Data extracted from different systems often have differences in format and structure. Data integration involves unified processing of data from multiple sources, integrating them into a complete data set for analysis, and ensuring data availability. While obtaining the raw data, relevant metadata can be recorded, such as the source of the data, the time of data extraction, the frequency of updates, and other information, to facilitate subsequent data tracking and management.

[0023] The step of obtaining enterprise business system data is of fundamental importance to the entire data processing intelligent analysis method. This step not only involves the extensive collection of data, but also directly affects the accuracy and effectiveness of subsequent data classification, analysis and decision-making. First, the original data set reflects the real situation of the enterprise's operations and can provide the enterprise with a comprehensive perspective to identify potential risks and opportunities. Secondly, good data construction can lay a solid foundation for subsequent AI algorithm model training and ensure the learning effect and accuracy of the classification model. In addition, in-depth mining and acquisition of high-quality original data can help enterprises understand market trends and customer needs, thereby optimizing the decision-making process and enhancing market competitiveness.

[0024] S202, using a pre-trained first data classification model based on a first AI algorithm, performing a first classification on the original data set to obtain a normal data set and an abnormal data set; At this stage, the model helps companies identify data quality and potential risks by identifying and distinguishing normal data from abnormal data. Normal data usually refers to data that meets preset standards or expected behavior, while abnormal data may indicate errors, fraud or critical business risks.

[0025] Specifically, a suitable first AI algorithm (such as decision tree, random forest or support vector machine, etc.) can be selected according to business needs, and the first data classification model can be trained. During the training process, the labeled data set is used for learning so that the first data classification model can effectively identify the characteristics of normal and abnormal data. Before the first data classification model is applied, the original data set needs to be preprocessed, including data cleaning, feature selection and format conversion, to ensure that the data format meets the model input requirements. Input the prepared original data set into the trained first data classification model. The first data classification model will analyze and judge the input data according to the learned rules and output the classification results. After the classification is completed, the classification results of the first data classification model can be evaluated to check whether the division of normal data sets and abnormal data sets conforms to business reality. When it is found that the classification effect is not ideal, the first data classification model can be adjusted and retrained according to the evaluation results to improve the classification accuracy.

[0026] By classifying data into normal and abnormal, enterprises can quickly identify potential risk points, such as system errors, data input errors, or abnormal transactions. This instant feedback mechanism can help enterprises take timely measures to reduce losses or avoid risks; the identification of abnormal data allows enterprises to focus on data quality issues and ensure that reliable data is used in subsequent data analysis and decision-making processes. This provides a basis for efficient data governance; by focusing resources on the investigation and processing of abnormal data, enterprises can more effectively allocate personnel and technical resources and improve overall data management efficiency; normal data sets will provide a reliable basis for subsequent data analysis, which can reflect the true face of enterprise operations and support decision makers in formulating more targeted strategies.

[0027] S203, using a pre-trained second data classification model based on a second AI algorithm, performing a second classification on the normal data set to obtain a plurality of normal data sets of different types; In the steps of the present invention, a pre-trained second data classification model based on a second AI algorithm is used to further classify the normal data set to identify multiple different types of normal data sets. Based on the normal data set obtained in the first classification step, the goal of this process is to further subdivide these data in order to better understand their internal structure and characteristics. For example, assume that the initial normal data set includes customer transaction records, website access logs, inventory information and other data. After applying the second data classification model, these data can be subdivided into multiple types, such as: high-value customer transaction records, low-value customer transaction records, normal browsing behavior, abnormal access behavior, inventory high-turnover goods and low-turnover goods, etc. Through such segmentation, enterprises can more effectively perform personalized analysis and processing for different types of data.

[0028] Specifically, before classifying a normal data set, it is necessary to first select and extract representative features. These features can include numerical (such as transaction amount, access time) and categorical (such as customer type, product category) features. Feature engineering can be performed to ensure data quality, eliminate outliers, and select dimensions that best express data characteristics through data preprocessing techniques (such as standardization, normalization, and missing value processing).

[0029] According to business needs and data characteristics, select a suitable second AI algorithm, such as support vector machine (SVM), random forest, deep learning model (such as neural network), etc. Configure model hyperparameters, such as learning rate, tree depth, etc., to optimize model performance. You can use the cross-validation method to tune parameters to find the best hyperparameter combination.

[0030] The normal data set is divided into a training set and a test set. Usually, the training set accounts for 70%-80% and the test set accounts for 20%-30%. In this way, the generalization ability of the model can be effectively evaluated. To ensure the consistency of data distribution, stratified sampling can be used when dividing to ensure a balanced distribution of each category.

[0031] The training set is input into the pre-configured second data classification model for training. The model constructs the classification decision boundary or cluster center by learning the characteristics of the training data. During the training process, methods such as early stopping can be used to prevent the model from overfitting and ensure its excellent performance on unknown data.

[0032] The test set is used to evaluate the model performance. The main evaluation indicators include classification accuracy, precision, recall, and F1 score. These indicators can be used to judge the classification effect of the model. The confusion matrix can be visualized to intuitively display the classification effect of different types of normal data, find misclassified samples, and further analyze the reasons.

[0033] The normal data set is classified using the trained second data classification model. The model assigns each sample in the normal data set to the corresponding type based on the learned features and decision boundaries. For example, the model may classify normal data into multiple types, such as "high-value customer transaction records", "low-value customer transaction records", "frequent purchases", "occasional visitors", etc.

[0034] Post-processing of classification results may include screening classification results, removing noise and inconsistent data types, and ensuring the accuracy and usability of classification results. Integrate and store multiple normal data sets of different types after classification to facilitate subsequent data analysis and application. The classification results can also be further analyzed to evaluate the characteristic distribution, relative proportion and potential value of various normal data sets. Display classification results through data visualization tools (such as charts and dashboards) to provide intuitive information for decision makers. Based on the classification results, enterprises can carry out targeted market behavior analysis, customer segmentation, personalized marketing strategies, etc.

[0035] By segmenting normal data into multiple different types, enterprises can analyze the characteristics and behaviors of each type of data more deeply, thereby revealing the potential relationships and patterns in the data. This helps enterprises identify important information such as market trends and customer preferences; different types of normal data sets can provide more detailed business insights and help decision makers formulate more accurate strategies and operational plans. For example, analyzing the transaction records of high-value customers can provide data support for marketing activities, thereby formulating more effective customer relationship management strategies; by identifying different types of data sets, enterprises can adjust resource allocation and work priorities in a targeted manner. For example, if a certain type of data (such as abnormal access behavior) shows high risk, the enterprise can focus on this part of the data to reduce potential business risks; after segmenting normal data, enterprises can formulate targeted processing strategies, such as dynamic pricing and promotion strategies for high-turnover products, thereby improving overall operational efficiency and profit margins.

[0036] S204, calculating the information gain index of each type of normal data set, and storing each type of normal data set according to the information gain index.

[0037] After the second classification of the normal data set, multiple normal data sets of different types are obtained. Next, the information gain index (IG) of each type of normal data set is calculated to evaluate the distinguishing ability and potential value of each type of data. The information gain index is a metric that reflects the effectiveness of a certain type of data in providing information. A higher information gain index means that the data set has more information in classification or decision-making.

[0038] In this process, the entropy of the entire normal data set is first calculated to obtain its uncertainty. This entropy measure reflects the degree of confusion of information in the data. Then, the entropy of each type of normal data set is calculated. By comparing these two entropy values, the value of a certain type of data set in information transmission can be obtained. This value measurement will help companies identify which types of data have a higher contribution to decision-making and help focus limited storage and analysis resources on more representative types.

[0039] According to the calculation results of the information gain index, enterprises can use the first storage method (such as blockchain storage) for data sets with information gain index higher than the preset threshold. This storage method can ensure the security and immutability of the data, so that important data can be properly protected. For data sets with information gain index lower than the threshold, the second storage method (such as cloud platform storage) is used. This method is low-cost and convenient for subsequent quick access, which is suitable for situations where the importance of data is not high.

[0040] Specifically, a calculation formula for the information gain index can be:

[0041] The formula is designed to comprehensively evaluate the effectiveness of a type of data set in information transmission and decision support. By introducing the concept of entropy and probability distribution, the formula captures the diversity and information concentration of the data set, ensuring that the information gain index can effectively reflect the unique contribution of the data set.

[0042] Among them, the is the information gain index of the k-th type of normal data set, which measures the effectiveness of this type of data in information transmission. The higher the value, the more information this type of data can provide in classification or decision making. For the original data set The entropy of reflects the uncertainty and complexity of the original data set. A high entropy value indicates that the data is diverse and contains a large amount of information, while a low entropy value indicates that the data is relatively concentrated. For example, , P(j) is the probability of normal data belonging to the jth type appearing in the original data set. Said is the normal data set of the kth type The entropy of reflects the purity of this type of data. If the entropy value of N_k is low, it means that this type of data is relatively homogeneous and has less information, and vice versa. For example, Said For normal data set The probability of type k appearing in .

[0043] Suppose there is an original data set D containing 100 samples. After the second classification, the normal data set N contains 80 samples, of which 50 samples belong to category A, 20 samples belong to category B, and 10 samples belong to category C. Then, P(A) = 50 / 100 = 0.5, P(B) = 20 / 100 = 0.2, P(C) = 10 / 100 = 0.1, substitute into the H(D) formula to get H(D); P(A|N) = 50 / 80 = 0.625, P(B|N) = 20 / 80 = 0.25, P(C|N) = 10 / 80 = 0.125, substitute into the H(N_k) formula to get the entropy of each type of normal data set.

[0044] Said is the probability of type j appearing in a normal data set, reflecting the importance and representativeness of this type in the overall data. By analyzing these probabilities, the distribution of each type in the overall data can be identified. For normal data set The size of Num ensures that the information gain can be standardized when it is calculated. The introduction of Num makes the information gain proportional to the sample size, avoiding the interference of the sample size on the results. is a regulatory factor greater than 0, used to control the flexibility and robustness of the model. The value of can better balance the different components of information gain and ensure that the information gain index finally calculated has good adaptability. The m is the number of types of normal data sets.

[0045] Specifically, the first storage method can be used to store normal data sets whose information gain index is greater than a preset threshold; in this step, the information gain index calculated for each type of normal data set is first evaluated. If the information gain index of a certain type of data set exceeds the preset threshold, it means that the data set has significant value in providing key information, supporting decision-making processes, and enhancing data insights. Based on this judgment, these high-value data sets will be stored in the first storage method, which usually involves a more secure storage solution with a protection mechanism, such as blockchain storage, to ensure the integrity and traceability of the data.

[0046] It is of great strategic significance to use data sets with high information gain index for priority storage. This storage method can protect the core data assets of enterprises and help enterprises achieve long-term data preservation and tamper-proof. At the same time, through advanced storage technologies such as blockchain, the transparency and trust of data sharing can be improved, which is particularly important for industries that require compliance and audit tracking. This approach plays a key role in promoting data-driven decision-making and enhancing data asset management.

[0047] For the normal data set whose information gain index is not greater than the preset threshold, the second storage method is adopted for storage.

[0048] In this step, the information gain index is used as an indicator to determine which normal data sets have lower value. When the information gain index of a certain type of normal data set does not exceed the preset threshold, it indicates that it has a smaller role in information transmission or decision support, so it can be stored in a second storage method. Usually, this storage method may be based on a cloud platform, providing more flexible storage options and effective cost management.

[0049] By storing data sets with low information gain index in more cost-effective cloud platforms, enterprises can reduce storage costs while ensuring data accessibility. This strategy ensures that high-value data is managed separately from relatively low-value data, optimizing resource allocation.

[0050] Exemplarily, the first storage method may be blockchain storage, and the second storage method may be cloud platform storage.

[0051] In this method, two different storage methods are used to ensure the effective management and utilization of normal data sets with different information gain indexes. First, for normal data sets with information gain index greater than the preset threshold, these data are stored using blockchain technology due to their significant importance and value. Blockchain, with its decentralized, transparent and tamper-proof characteristics, can effectively protect these key data and ensure the integrity and security of the data. Secondly, for normal data sets with information gain index not greater than the preset threshold, cloud platform storage is selected considering that the value of these data for decision support is relatively low. Cloud storage provides higher flexibility and scalability, and can meet the needs of enterprises for large-scale data storage at a lower cost.

[0052] The implementation of this storage strategy not only improves the efficiency of data storage, but also optimizes the use of resources. By storing high-value data in the blockchain, enterprises can ensure transparency and trust in the sharing and use of data, which is especially important in industries with high compliance requirements. This practice helps maintain the reputation of the enterprise and reduce the risk of data leakage and tampering. At the same time, using cloud platforms to store low-value data can effectively reduce the storage costs of enterprises and ensure that enterprises can maximize the use of data resources without sacrificing business flexibility.

[0053] In order to effectively implement this storage solution, a data storage management system can be established first, which combines data classification with storage strategy. In the data processing process, the data set is first classified, and then the information gain index is calculated. After the data classification is completed, the system automatically evaluates the information gain index of each data set and makes storage decisions based on the results.

[0054] In terms of technical implementation, blockchain platforms (such as Ethereum or Hyperledger) can be used to store high-value data to ensure data security and transparency. At the same time, cloud service providers (such as AWS, Google Cloud or Azure) can be used to store low-value data, providing flexible storage options and efficient access capabilities.

[0055] To avoid data silos, blockchain storage can also be integrated with cloud storage, and data interaction and access between the two can be achieved through API interfaces, thereby creating a more coherent and unified data management ecosystem. This integration can ensure that high-value data can be used efficiently, while low-value data can also be stored and accessed in a more cost-effective environment. Ultimately, enterprises can achieve more intelligent decision support and effectively improve overall operational efficiency.

[0056] The core significance of this step is that through the calculation of the information gain index, enterprises can efficiently screen out valuable data types, provide a scientific basis for subsequent data storage and analysis, and ensure the rational use of resources. In addition, it can also promote the implementation of data-driven decision-making, so that enterprises can flexibly respond to market changes, seize business opportunities, and improve overall operational efficiency during real-time data analysis and storage.

[0057] It can be seen that the original data set to be processed is obtained, wherein the original data set includes: enterprise business system data; using a pre-trained first data classification model based on a first AI algorithm, the original data set is first classified to obtain a normal data set and an abnormal data set; using a pre-trained second data classification model based on a second AI algorithm, the normal data set is second classified to obtain multiple different types of normal data sets; the information gain index of each type of normal data set is calculated, and each type of normal data set is stored according to the information gain index, so that the enterprise data can be classified, analyzed and stored, in order to better utilize data resources and provide better technical support and theoretical basis for data-driven decision-making.

[0058] Another embodiment of the present invention provides a data processing intelligent analysis system based on AI algorithm, see Figure 3 , the system may include: The acquisition module 301 is used to acquire the original data set to be processed, wherein the original data set includes: enterprise business system data; A first classification module 302 is used to perform a first classification on the original data set by using a pre-trained first data classification model based on a first AI algorithm to obtain a normal data set and an abnormal data set; A second classification module 303 is used to perform a second classification on the normal data set by using a pre-trained second data classification model based on a second AI algorithm to obtain a plurality of normal data sets of different types; The storage module 304 is used to calculate the information gain index of each type of normal data set, and store each type of normal data set according to the information gain index.

[0059] It can be seen that the original data set to be processed is obtained, wherein the original data set includes: enterprise business system data; using a pre-trained first data classification model based on a first AI algorithm, the original data set is first classified to obtain a normal data set and an abnormal data set; using a pre-trained second data classification model based on a second AI algorithm, the normal data set is second classified to obtain multiple different types of normal data sets; the information gain index of each type of normal data set is calculated, and each type of normal data set is stored according to the information gain index, so that the enterprise data can be classified, analyzed and stored, in order to better utilize data resources and provide better technical support and theoretical basis for data-driven decision-making.

[0060] An embodiment of the present invention further provides a storage medium, in which a computer program is stored, wherein the computer program is configured to execute the steps of any of the above method embodiments when running.

[0061] Specifically, in this embodiment, the above storage medium may be configured to store a computer program for performing the following steps: S201, obtaining an original data set to be processed, wherein the original data set includes: enterprise business system data; S202, using a pre-trained first data classification model based on a first AI algorithm, performing a first classification on the original data set to obtain a normal data set and an abnormal data set; S203, performing a second classification on the normal data set using a pre-trained second data classification model based on a second AI algorithm to obtain a plurality of normal data sets of different types; S204, calculating the information gain index of each type of normal data set, and storing each type of normal data set according to the information gain index.

[0062] It can be seen that the original data set to be processed is obtained, wherein the original data set includes: enterprise business system data; using a pre-trained first data classification model based on a first AI algorithm, the original data set is first classified to obtain a normal data set and an abnormal data set; using a pre-trained second data classification model based on a second AI algorithm, the normal data set is second classified to obtain multiple different types of normal data sets; the information gain index of each type of normal data set is calculated, and each type of normal data set is stored according to the information gain index, so that the enterprise data can be classified, analyzed and stored, in order to better utilize data resources and provide better technical support and theoretical basis for data-driven decision-making.

[0063] An embodiment of the present invention further provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0064] Specifically, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0065] Specifically, in this embodiment, the processor may be configured to perform the following steps through a computer program: S201, obtaining an original data set to be processed, wherein the original data set includes: enterprise business system data; S202, using a pre-trained first data classification model based on a first AI algorithm, performing a first classification on the original data set to obtain a normal data set and an abnormal data set; S203, using a pre-trained second data classification model based on a second AI algorithm, performing a second classification on the normal data set to obtain a plurality of normal data sets of different types; S204, calculating the information gain index of each type of normal data set, and storing each type of normal data set according to the information gain index.

[0066] It can be seen that the original data set to be processed is obtained, wherein the original data set includes: enterprise business system data; using a pre-trained first data classification model based on a first AI algorithm, the original data set is first classified to obtain a normal data set and an abnormal data set; using a pre-trained second data classification model based on a second AI algorithm, the normal data set is second classified to obtain multiple different types of normal data sets; the information gain index of each type of normal data set is calculated, and each type of normal data set is stored according to the information gain index, so that the enterprise data can be classified, analyzed and stored, in order to better utilize data resources and provide better technical support and theoretical basis for data-driven decision-making.

[0067] The above describes in detail the structure, features and effects of the present invention based on the embodiments shown in the drawings. The above is only a preferred embodiment of the present invention, but the present invention is not limited to the scope of implementation shown in the drawings. Any changes made according to the concept of the present invention, or modifications to equivalent embodiments with equivalent changes, which still do not exceed the spirit covered by the description and drawings, should be within the protection scope of the present invention.

Claims

1. A data processing intelligent analysis method based on AI algorithm, It is characterized in that The method comprises: Acquire an original data set to be processed, wherein the original data set includes: enterprise business system data; Using a pre-trained first data classification model based on a first AI algorithm, the original data set is subjected to a first classification to obtain a normal data set and an abnormal data set; Using a pre-trained second data classification model based on a second AI algorithm, the normal data set is subjected to a second classification to obtain a plurality of normal data sets of different types; An information gain index of each type of normal data set is calculated, and each type of normal data set is stored according to the information gain index.

2. The method according to claim 1, It is characterized in that The calculation formula of the information gain index is: Among them, the is the information gain index of the k-th type of normal data set, For the original data set The entropy of is the normal data set of the kth type The entropy of For normal data set The probability of type j appearing in For normal data set The size of is an adjustment factor greater than 0, and m is the number of types of normal data sets.

3. The method according to claim 2, It is characterized in that The storing of each type of normal data set according to the information gain index includes: For the normal data set whose information gain index is greater than the preset threshold, a first storage method is adopted for storage; For the normal data set whose information gain index is not greater than the preset threshold, the second storage method is adopted for storage.

4. The method according to claim 3, It is characterized in that The first storage method is blockchain storage, and the second storage method is cloud platform storage.

5. A data processing intelligent analysis system based on AI algorithm, It is characterized in that The system comprises: An acquisition module is used to acquire an original data set to be processed, wherein the original data set includes: enterprise business system data; A first classification module, configured to perform a first classification on the original data set by using a pre-trained first data classification model based on a first AI algorithm to obtain a normal data set and an abnormal data set; A second classification module, configured to perform a second classification on the normal data set by using a pre-trained second data classification model based on a second AI algorithm to obtain a plurality of normal data sets of different types; The storage module is used to calculate the information gain index of each type of normal data set, and store each type of normal data set according to the information gain index.

6. The system according to claim 5, It is characterized in that The calculation formula of the information gain index is: Among them, the is the information gain index of the k-th type of normal data set, Original Dataset The entropy of is the normal data set of the kth type The entropy of For normal data set The probability of type j appearing in For normal data set The size of is an adjustment factor greater than 0, and m is the number of types of normal data sets.

7. The system according to claim 6, It is characterized in that The storage module is specifically used for: For the normal data set whose information gain index is greater than the preset threshold, a first storage method is adopted for storage; For the normal data set whose information gain index is not greater than the preset threshold, the second storage method is adopted for storage.

8. The system according to claim 7, It is characterized in that The first storage method is blockchain storage, and the second storage method is cloud platform storage.

9. A storage medium, It is characterized in that The storage medium stores a computer program, wherein the computer program is configured to execute the method according to any one of claims 1 to 4 when executed.

10. An electronic device comprising a memory and a processor, It is characterized in that A computer program is stored in the memory, and the processor is configured to run the computer program to perform the method according to any one of claims 1 to 4.

Images