A data report generation method and system based on artificial intelligence

Abstract

The present invention relates to the technical field of data report generation, specifically to a data report generation method and system based on artificial intelligence, comprising the following steps: obtaining time series data and equipment performance logs from an IDC data center, extracting time tags and performance indicators, sorting the time tags, summarizing the performance indicators by equipment type, and obtaining preliminary analysis data. In the present invention, by structured encoding of data items and creating connection nodes and edge processing, the visualization of the relationship between data is enhanced, the logic of data connection is made clearer, the automated weight allocation and key performance indicator screening process make data analysis more targeted and efficient, and help to quickly identify key information and trends in the data. Further, through dynamic classification and trend extraction, not only the adaptability of the report is improved, but also the real-time and accuracy of the report content is ensured through real-time data content update and data synchronization detection.

Description

Technical Field

[0001] The present invention relates to the technical field of data report generation, and in particular to a data report generation method and system based on artificial intelligence. Background Art

[0002] The field of data report generation technology involves the use of software tools to automate the process of collecting, processing and presenting information from various data sources, combining data science, artificial intelligence and machine learning technologies to discover trends, patterns and insights in the data and automatically convert them into easy-to-understand report formats. Data reporting automation systems usually include data extraction, data cleaning, analysis model construction, and report layout and visual design, which promotes the efficiency of the decision-making process.

[0003] Among them, the data report generation method based on artificial intelligence refers to the use of artificial intelligence technology to automate the process of creating data reports, including the use of machine learning algorithms to parse large amounts of data, identify key information, and generate structured reports based on predefined templates. It has a wide range of uses, such as helping companies quickly extract valuable business insights from big data, improve the speed and accuracy of data analysis, improve strategic decision-making, and provide users with instant and dynamic report updates, enhancing the ability to make data-driven decisions.

[0004] Although existing technologies cover the automation process of data reporting, they often lack the ability to deeply analyze the dynamic relationship between data and update it in real time. Data processing in traditional systems often relies on fixed models and templates, which makes it difficult to adapt to new data structures and business needs when data sources change rapidly. The lack of the ability to flexibly adjust the report structure means that when new trends emerge, these changes cannot be reflected in a timely manner, resulting in decision-making based on outdated information. In addition, traditional technologies are often inefficient when processing and analyzing big data, which may become a bottleneck that restricts the speed and accuracy of decision-making in time-sensitive business scenarios. The lack of real-time data synchronization and update mechanisms means that data reports may not reflect the latest business status, affecting the effectiveness of decision-making and the company's strategic adjustments. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a data report generation method and system based on artificial intelligence.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: a data report generation method based on artificial intelligence, comprising the following steps:

[0007] S1: Obtain time series data and device performance logs from the IDC data center, extract time tags and performance indicators, sort the time tags, summarize the performance indicators by device type, and obtain preliminary analysis data;

[0008] S2: Through the preliminary analysis of the data, the data items are structuredly encoded, connection nodes and edges between devices are created, weights are assigned to each connection, key performance indicators are screened according to the weight assignment results, key time nodes are determined, and importance scores of the nodes are calculated to generate key data graphs;

[0009] S3: classifying the data points in the key data graph, extracting key trends based on the connection strength between the data points, marking the extracted data, classifying and arranging the data according to the marking results, classifying each group of data according to trends and patterns, building a structured trend framework, and obtaining data trend classification records;

[0010] S4: According to the data trend classification records, real-time data content updates are performed, data synchronization detection is performed, report structure and details are automatically adjusted according to input content, key information points are adjusted, and a comprehensive IDC data analysis report is generated.

[0011] As a further solution of the present invention, the step of obtaining the preliminary analysis data is:

[0012] S111: Connect to the IDC data center, use secure network protocols to obtain time series data and equipment performance logs, perform preliminary verification of data integrity and format, exclude damaged or non-compliant records, and generate a verified data set;

[0013] S112: extracting time tags and performance indicators from the verified data set, sorting the time tags, confirming the time sequence correctness of the data, and generating a sorted data set;

[0014] S113: Based on the sorted data set and in combination with the device type, the performance indicators in the data set are summarized using the formula: ;

[0015] Computing device type The average performance index quantified value , ensure consistency in equipment performance assessment and generate preliminary analysis data, including: Represents the device type Middle The performance index of each device is a quantitative value, which reflects the performance of a single device. Represents the device type The number of devices used for averaging calculations.

[0016] As a further solution of the present invention, the step of assigning a weight to each connection is:

[0017] S211: extracting key features through the preliminary analysis of the data, performing hash coding on the features, and generating a feature coding set with a unique identifier;

[0018] S212: constructing a graph structure between devices using the feature coding set, defining a node for each device, and establishing edges according to data association to generate a preliminary connection graph between devices;

[0019] S213: Based on the preliminary connection diagram between the devices, combined with the data interaction frequency and importance allocation weight, using the formula, ;

[0020] Compute Node To Node The weight value , used to measure the importance of connections between nodes, and obtain a weighted connection graph, where Is a node and The frequency of interaction between Is a node The average interaction frequency of is used to normalize the weights, Representation and Node The sum of the absolute differences between the interaction frequencies of all connected nodes and the average value is used to adjust the weight distribution.

[0021] As a further solution of the present invention, the step of obtaining the key data graph is:

[0022] S221: Analyze the weighted connection graph, filter high-weight connections by setting a weight threshold, determine key performance indicators, and generate a key performance indicator list;

[0023] S222: Based on the key performance indicator list, extract the time points with the most significant data changes as key time nodes, and generate a key time node set;

[0024] S223: Score the importance of each node in the key time node set using the formula: ;

[0025] Compute Node Importance Rating , used to evaluate the importance of nodes, build and output key data graphs, where Represents a set of key time nodes, Representative time For Node The influence weight of .

[0026] As a further solution of the present invention, the step of marking the extracted data is:

[0027] S311: Analyze the key data graph and use the formula: ;

[0028] Calculate data points The weighted number of connections , generating data point connectivity metrics results, where It is a data point and An indication value indicating whether there is a connection between them. 0 means no connection, and 1 means connection. It is a data point and The distance between is the attenuation coefficient, which is used to adjust the intensity of the distance effect. Indicates the number of data points;

[0029] S312: Setting a screening threshold based on the data point connection measurement result , by comparison , determine key data points and generate key data point sets;

[0030] S313: Marking the key data point set according to the connection strength and trend distribution of the data points to generate a marked key data point set.

[0031] As a further solution of the present invention, the steps of obtaining the data trend classification record are:

[0032] S321: Analyze and classify the marked key data point set using the formula: ;

[0033] Calculate the total aggregation degree of the classification results , generate preliminary data classification results, where and The data points are and The marking score of It is a data point and The distance between is the attenuation constant, used to adjust for the effect of distance, Indicates the number of data points;

[0034] S322: Analyze the preliminary data classification results, determine the key trends and patterns of each group of data, refine the data classification process through quantitative analysis, and obtain structured data classification results;

[0035] S323: Integrate the structured data classification results, build a structured trend framework, record the classification and characteristics of each data trend, and generate a data trend classification record.

[0036] As a further solution of the present invention, the steps of obtaining the comprehensive IDC data analysis report are:

[0037] S411: Analyze the data trend classification records, identify data points to be updated, and generate a list of data points to be updated;

[0038] S412: Perform synchronization detection on the marked data point list, check data consistency, confirm data integrity, and generate synchronization detection results;

[0039] S413: Based on the synchronous detection results, automatically adjust the report structure and details, update key information points, and use formulas to ;

[0040] Calculating adjusted report quality scores , generate a comprehensive IDC data analysis report, including: Represents key information points The importance weight of Is the input content at the information point The amount of change.

[0041] An artificial intelligence-based data report generation system, comprising:

[0042] The data acquisition and verification module connects to the IDC data center, obtains time series data and equipment performance logs, performs preliminary verification of data integrity and format, extracts time tags and performance indicators and sorts them, summarizes the performance indicators in the data set based on the equipment type, and generates preliminary analysis data;

[0043] The feature extraction and encoding module extracts key features through the preliminary analysis data, performs hash encoding on the features, constructs a graph structure between devices, calculates weight values ​​between nodes, measures the importance of connections between nodes, and obtains a weighted connection graph;

[0044] The high-weight screening module analyzes the weighted connection graph, screens high-weight connections by setting weight thresholds, determines key performance indicators, extracts key time nodes, scores the importance of each node, and constructs and outputs a key data graph;

[0045] The data point measurement and marking module analyzes the key data graph, calculates the weighted connection number of the data points, performs data point connection measurement, sets and compares the screening threshold, determines the key data points, marks the key data point set, and generates a marked key data point set;

[0046] The data classification and trend analysis module analyzes and classifies the key data point set after the marking, calculates the total aggregation degree of the classification results, determines the key trends and patterns of each group of data, refines the data classification processing, records the classification and characteristics of each data trend, and generates data trend classification records;

[0047] The report generation and synchronization detection module analyzes the data trend classification records, identifies the data points that need to be updated, performs synchronization detection, checks data consistency, automatically adjusts the report structure and details, updates key information points, and generates a comprehensive IDC data analysis report.

[0048] Compared with the prior art, the advantages and positive effects of the present invention are:

[0049] In the present invention, time series data and equipment performance logs are obtained and processed from the IDC data center to improve the automation and intelligence level of data analysis. By performing structured coding on data items and creating connection nodes and edge processing, the visualization of the relationship between data is enhanced, making the logic of data connection clearer. The automated weight allocation and key performance indicator screening process make data analysis more targeted and efficient, and help to quickly identify key information and trends in the data. Further, through dynamic classification and trend extraction, not only the adaptability of the report is improved, but also the real-time and accuracy of the report content is ensured through real-time data content update and data synchronization detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a schematic diagram of the steps of the present invention;

[0051] Figure 2 A flowchart for obtaining preliminary analysis data of the present invention;

[0052] Figure 3 A flow chart of the steps of assigning weights to each connection according to the present invention;

[0053] Figure 4 It is a flowchart for obtaining the key data diagram of the present invention;

[0054] Figure 5 A flow chart of the steps of marking the extracted data according to the present invention;

[0055] Figure 6 A flowchart for obtaining data trend classification records of the present invention;

[0056] Figure 7 The figure is a flowchart for obtaining the comprehensive IDC data analysis report of the present invention. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0058] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are 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, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0059] See also Figure 1 , a data report generation method based on artificial intelligence, comprising the following steps:

[0060] S1: Obtain time series data and device performance logs from the IDC data center, extract time tags and performance indicators, sort the time tags, summarize the performance indicators by device type, and obtain preliminary analysis data;

[0061] S2: Through preliminary data analysis, data items are structured and encoded, connection nodes and edges between devices are created, weights are assigned to each connection, key performance indicators are screened according to the weight assignment results, key time nodes are determined, and the importance scores of nodes are calculated to generate key data graphs;

[0062] S3: Classify the data points in the key data graph, extract key trends based on the connection strength between the data points, mark the extracted data, classify and sort the data based on the marking results, classify each group of data according to trends and patterns, build a structured trend framework, and obtain data trend classification records;

[0063] S4: Classify records according to data trends, perform real-time data content updates, perform data synchronization detection, automatically adjust report structure and details based on input content, adjust key information points, and generate a comprehensive IDC data analysis report.

[0064] The preliminary analysis data includes time stamp sorting results and performance indicator summary results; the key data graphs include node weight allocation records, key performance indicator screening results and node importance score records; the data trend classification records include data point classification records, key trend extraction results and trend pattern classification results; the comprehensive IDC data analysis report includes report structure automatic adjustment results and key information point adjustment records.

[0065] See also Figure 2 ,The steps for obtaining preliminary analysis data are:

[0066] S111: Connect to the IDC data center, use secure network protocols to obtain time series data and equipment performance logs, perform preliminary verification of data integrity and format, exclude damaged or non-compliant records, and generate a verified data set;

[0067] S112: extracting time tags and performance indicators from the verified data set, sorting the time tags, confirming the time sequence correctness of the data, and generating a sorted data set;

[0068] S113: Based on the sorted data set, combined with the device type, the performance indicators in the data set are summarized using the formula: ;

[0069] Computing device type The average performance index quantified value , ensure consistency in equipment performance assessment and generate preliminary analysis data, including: Represents the device type Middle The performance index of each device is a quantitative value, which reflects the performance of a single device. Represents the device type The number of devices used for averaging calculations.

[0070] For the acquired time series data and device performance logs, the format and integrity of each piece of data is calculated.

[0071] To confirm, and They are the format correctness and completeness index of each piece of data, and the calculation results Used to determine the availability of data.

[0072] set up This is a data set obtained from the IDC data center, where each piece of data The format and integrity of each piece of data must be verified. and integrity Perform a numerical evaluation, where It is 0 or 1, 0 means the format is incorrect, 1 means the format is correct; The same is 0 or 1, 0 means the data is incomplete, 1 means the data is complete. By accumulating these two indices, we get , this result is used to determine the availability of each piece of data; if The data is both well-formatted and complete. This indicates a problem with the data; if If the format and integrity checks of show is available. If The format check is 1 and the integrity check is 0. show There are problems and it is not available. The results show that the preliminary screening of the data has been completed, and only the data with a score of 2 are considered valid. The valid data will enter the next processing stage to ensure the quality and accuracy of data processing.

[0073] In the sorting process of time stamps, according to the formula, ;

[0074] The logical order of the time stamps is determined by multiplying the time stamp index by the weighted sum of the time values; where, represents the weighted sum of the sorted time stamps, Indicates The value of the time stamp, is the index of the time stamp in the array, is the total number of time stamps.

[0075] Get timestamp array (Simplified to years), where the time marker 2023 is in the first position, 2021 is in the second position, and 2022 is in the third position. The formula is used to calculate the index multiplied by the time value for each time marker:

[0076] for (The first position, ),calculate ;

[0077] for (The second position, ),calculate ;

[0078] for (The third position, ),calculate ;

[0079] Add value to get:

[0080] ;

[0081] The results show that by calculating the weighted sum of time tags, the logical order of the time series can be effectively confirmed, supporting the accuracy of subsequent data processing. The results show that the logical order of time tags has been correctly set, providing a strict temporal basis for subsequent data aggregation and analysis.

[0082] In the performance index summary, the performance index of each type of equipment needs to be averaged. The average performance index of the equipment type is calculated according to the formula , calculate the average performance value of each device type.

[0083] Get the device type of a certain model Contains performance data for three devices , , , then the number of devices ;

[0084] Calculate according to the formula: ;

[0085] Indicates the device type The average performance indicator is quantified as 100. This process ensures a fair evaluation of the performance indicators, reflecting the type of equipment The results show that through precise calculation, the device type has been obtained The average performance indicators provide data support for further performance comparison and decision-making.

[0086] See also Figure 3 , the steps to assign weights to each connection are:

[0087] S211: extract key features through preliminary data analysis, perform hash coding on the features, and generate a feature coding set with a unique identifier;

[0088] S212: constructing a graph structure between devices using the feature coding set, defining a node for each device, and establishing edges based on data association to generate a preliminary connection graph between devices;

[0089] S213: Based on the preliminary connection diagram between devices, combined with the frequency and importance of data interaction, weights are assigned using the formula, ;

[0090] Compute Node To Node The weight value , used to measure the importance of connections between nodes, and obtain a weighted connection graph, where Is a node and The frequency of interaction between Is a node The average interaction frequency of is used to normalize the weights, Representation and Node The sum of the absolute differences between the interaction frequencies of all connected nodes and the average value is used to adjust the weight distribution.

[0091] The hash encoding process of data features is based on the formula , calculate the unique code of each feature. Representative features The hash code value of is the hash coefficient, is the modulus used to determine the size of the hash table to ensure uniform distribution of the codes.

[0092] Determine the characteristics , converted into digital form , select and , then the hash code is calculated as: ;

[0093] The results showed that the characteristics The position in the hash table is 95, which ensures fast retrieval and uniqueness of the data feature.

[0094] The creation of the equipment room diagram structure is based on the formula , construct nodes and edges. In the formula, Represents a graph structure, including a set of nodes and edge sets , Is a node and nodes The conditional expression for whether there is a connection between them is determined based on the data association.

[0095] With nodes , where the connection condition between nodes is determined by the data similarity. and If the data similarity exceeds the threshold value of 0.8, then , otherwise 0. OK and The similarity is 0.9, then ,express and There are edges between them. The results show that and Directly connected, reflecting the direct data association between devices.

[0096] For each connection edge between devices, according to the formula Calculate weights.

[0097] Get Node With three nodes , , Interaction frequency , , , then the average interaction frequency .

[0098] Then the sum of the differences is calculated as: ;

[0099] Weight Calculated as: ;

[0100] The result represents the node To Node The weight value is 2.24, which can measure the importance of the connection between nodes and obtain a weighted connection graph.

[0101] See also Figure 4 , the steps to obtain the key data graph are:

[0102] S221: Analyze the weighted connection graph, filter high-weight connections by setting a weight threshold, determine key performance indicators, and generate a key performance indicator list;

[0103] S222: Based on the key performance indicator list, extract the time points with the most significant data changes as key time nodes, and generate a key time node set;

[0104] S223: Score the importance of each node in the key time node set using the formula: ;

[0105] Compute Node Importance Rating , used to evaluate the importance of nodes, build and output key data graphs, where Represents a set of key time nodes, Representative time For Node The influence weight of .

[0106] Filter the edges in the weighted connection graph according to the formula To decide the side Whether to keep. In the formula, Representative side Whether it is selected as a key edge, It is the edge The weight of It is the preset weight threshold used to filter out key performance indicators from the graph.

[0107] Set threshold , if the node and nodes The weight between , then according to the formula , we can conclude The results show that the edge is a key edge, retained in the final screening graph, representing and The connection has a significant impact on system performance.

[0108] Among them, the key time nodes are extracted from the key performance indicators according to the formula , select the moment with the largest change. In the formula, Represents a key time node, is the performance indicator in time The critical moment of system performance change is determined by this formula.

[0109] Collection time , , The changes in performance indicators are , , According to the formula ,Sure The results showed that the time point is a critical time node, because at this moment, the change in performance indicators is the largest, indicating that the system performance is The most critical change has occurred.

[0110] Among them, the importance score of each key time node is calculated according to the formula Perform calculations.

[0111] node In time , , The weights are , , , and these times are all critical times, then: ;

[0112] The results show that the node The average importance score of is 0.333, which represents its average influence in the system.

[0113] See also Figure 5 , the steps to mark the extracted data are:

[0114] S311: Analyze key data graphs and use formulas to ;

[0115] Calculate data points The weighted number of connections , generating data point connectivity metrics results, where It is a data point and An indication value indicating whether there is a connection between them. 0 means no connection, and 1 means connection. It is a data point and The distance between is the attenuation coefficient, which is used to adjust the intensity of the distance effect. Indicates the number of data points;

[0116] S312: Set the screening threshold based on the data point connection measurement results , by comparison , determine key data points and generate key data point sets;

[0117] S313: Marking the key data point set according to the connection strength and trend distribution of the data points to generate a marked key data point set.

[0118] In the process of calculating the weighted connection number of each data point, according to the formula , calculate each data point connection strength.

[0119] There are three data points , , ,in and and Connection, the connection status indicator value and ,distance and , attenuation coefficient , substitute into the formula to calculate: ;

[0120] The calculation results reflect the data points The weighted connection strength to its connection points, taking distance into account, provides a more accurate measure of connection strength.

[0121] Filter key data points based on connection strength, according to the formula Screening is performed to ensure that the selected data points are key nodes, that is, those points with particularly high connection strength.

[0122] Setting Thresholds For the previously calculated data points , ,but The threshold is not exceeded. This threshold screening process is the basis for key trend extraction, ensuring that only data points with high connection strength are analyzed, thus more accurately reflecting the main structures and trends in the data graph.

[0123] Mark the selected key data point set and use the formula , calculate the data points Total marks , and mark the data points, where It is a data point and Whether there is a connection indication value between them, It is a data point and The connection coefficient between .

[0124] Combined with the above results, the connection status and , connection coefficient and ,but: ;

[0125] The calculation results show that the data points The total score of the tag is 1, which can be used for further classification and analysis of subsequent data points, ensuring that each data point is labeled according to its actual importance in the network.

[0126] See also Figure 6 , the steps to obtain data trend classification records are:

[0127] S321: Based on the marked key data point set, analyze and classify it using the formula, ;

[0128] Calculate the total aggregation degree of the classification results , generate preliminary data classification results, where and The data points are and The marking score of It is a data point and The distance between is the attenuation constant, used to adjust for the effect of distance, Indicates the number of data points;

[0129] S322: Analyze the preliminary data classification results, determine the key trends and patterns of each group of data, refine the data classification process through quantitative analysis, and obtain structured data classification results;

[0130] S323: Integrate the structured data classification results, build a structured trend framework, record the classification and characteristics of each data trend, and generate data trend classification records.

[0131] Among them, the labeled data point set is analyzed and classified using the formula , to quantify the degree of clustering of the classification.

[0132] Three data points , , , marking score , , ,distance , , the decay constant , substitute into the formula to calculate: ;

[0133] The value reflects the degree of aggregation among data points, and a low degree of aggregation indicates that data points with similar labels are farther apart in space, which helps to build a structured trend framework.

[0134] Among them, the data classification processing is refined according to the formula , each data point is further processed according to its category, where Representative data points The classification score of It is a data point In category The voting value in Yes Category Standard or trend score.

[0135] Data Points There are three possible categories , , In the case of , , , standard scores for each category , , .but The calculation is: ;

[0136] The results show that the data points The classification scores according to their possible categories are , which can be used for precise classification and trend analysis.

[0137] Among them, construct a structured trend framework according to the formula , integrate the classification results of all data points to build a framework, where, represents the score of the constructed structured trend framework, is the total number of data points, Each data point The classification score of .

[0138] have data points, each of which As the above calculation results , then the framework The score is calculated as:

[0139] ;

[0140] The results show the average score of the overall data trend framework, reflecting the overall trend quality of all data points by classification results.

[0141] See also Figure 7 , the steps to obtain the comprehensive IDC data analysis report are:

[0142] S411: Analyze data trend classification records, identify data points that need to be updated, and generate a list of data points to be updated;

[0143] S412: Perform synchronization detection on the marked data point list, check data consistency, confirm data integrity, and generate synchronization detection results;

[0144] S413: Based on the synchronous detection results, automatically adjust the report structure and details, update key information points, use formulas, ;

[0145] Calculating adjusted report quality scores , generate a comprehensive IDC data analysis report, including: Represents key information points The importance weight of Is the input content at the information point The amount of change.

[0146] Among them, identify the data points that need to be updated, and score them according to the data change amount and importance according to the formula , calculate the update priority of each data point and generate a list of data points to be updated. Representative data points The update priority of is the amount of data change, is the importance score of the data point.

[0147] The changes in the three data points collected are , , , and their importance scores are , , , then the update priority is calculated as:

[0148] ;

[0149] The value represents the priority of each data point update, and the data update operation is performed according to this priority.

[0150] Among them, the marked data point list is synchronously detected, using the data consistency score and data integrity coefficient according to the formula , calculate the comprehensive score of synchronization detection and generate the synchronization detection result. represents the comprehensive score of synchronous detection, is the data consistency score, is the data integrity factor.

[0151] Setting Data Consistency Score and data integrity factor , then the comprehensive score is calculated as: ;

[0152] The score reflects the synchronization efficiency and accuracy of the data, ensuring the reliability and integrity of the data.

[0153] Among them, for adjusting the report structure and details, the adjusted report quality score is calculated based on the change in input content and the importance of key information points, and a comprehensive IDC data analysis report is generated.

[0154] There are three key information points, and their weights are , , The corresponding change is , , . Then the report quality score is calculated as: ;

[0155] The results show that the report quality score, calculated by weighted changes in key information points, is 16.5, reflecting the effective adjustment of the report structure and details.

[0156] An artificial intelligence-based data report generation system, comprising:

[0157] The data acquisition and verification module connects to the IDC data center, obtains time series data and equipment performance logs, performs preliminary verification of data integrity and format, extracts time tags and performance indicators and sorts them, summarizes the performance indicators in the data set based on the equipment type, and generates preliminary analysis data;

[0158] The feature extraction and encoding module extracts key features through preliminary data analysis, performs hash encoding on the features, builds a graph structure between devices, calculates the weight values ​​between nodes, measures the importance of connections between nodes, and obtains a weighted connection graph;

[0159] The high-weight screening module analyzes the weighted connection graph, screens high-weight connections by setting weight thresholds, determines key performance indicators, extracts key time nodes, scores the importance of each node, and constructs and outputs key data graphs;

[0160] The data point measurement and marking module analyzes the key data graph, calculates the weighted connection number of the data points, measures the data point connection, sets the screening threshold and compares it, determines the key data points, marks the key data point set, and generates the marked key data point set;

[0161] The data classification and trend analysis module analyzes and classifies the marked key data point set, calculates the total aggregation degree of the classification results, determines the key trends and patterns of each group of data, refines the data classification process, records the classification and characteristics of each data trend, and generates data trend classification records;

[0162] The report generation and synchronization detection module analyzes data trend classification records, identifies data points that need to be updated, performs synchronization detection, checks data consistency, automatically adjusts report structure and details, updates key information points, and generates comprehensive IDC data analysis reports.

[0163] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A data report generation method based on artificial intelligence, characterized in that: The following steps are involved: Obtain time series data and device performance logs from the IDC data center, extract time stamps and performance indicators, sort the time stamps, summarize the performance indicators by device type, and obtain preliminary analysis data; By means of the preliminary analysis data, the data items are structuredly encoded, connection nodes and edges between devices are created, weights are assigned to each connection, key performance indicators are screened according to the weight assignment results, key time nodes are determined, and importance scores of the nodes are calculated to generate key data graphs; Classify the data points in the key data graph, extract key trends based on the connection strength between the data points, mark the extracted data, classify and sort the data based on the marking results, classify each group of data based on trends and patterns, build a structured trend framework, and obtain data trend classification records; According to the data trend classification records, real-time data content updates are performed, data synchronization detection is performed, report structure and details are automatically adjusted according to input content, key information points are adjusted, and a comprehensive IDC data analysis report is generated; The steps of assigning weights to each connection are: By means of the preliminary analysis of the data, key features are extracted, the features are hash-coded, and a feature coding set with a unique identifier is generated; Using the feature coding set to construct a graph structure between devices, defining a node for each device, and establishing edges based on data association to generate a preliminary connection graph between devices; Based on the preliminary connection diagram between the devices, combined with the data interaction frequency and importance allocation weights, the formula is adopted: ； Compute Node To Node The weight value , used to measure the importance of connections between nodes, and obtain a weighted connection graph, where Is a node and The frequency of interaction between Is a node The average interaction frequency of is used to normalize the weights, Representation and Node The sum of the absolute differences between the interaction frequencies of all connected nodes and the average value is used to adjust the weight distribution; The steps for obtaining the key data graph are: Analyze the weighted connection graph, filter high-weight connections by setting weight thresholds, determine key performance indicators, and generate a key performance indicator list; Based on the key performance indicator list, extract the time points with the most significant data changes as key time nodes, and generate a key time node set; The importance of each node in the key time node set is scored using the formula: ； Compute Node Importance Rating , used to evaluate the importance of nodes, build and output key data graphs, where Represents a set of key time nodes, Representative time For Node The weight of influence; The step of marking the extracted data is: Analyze the key data graph and use the formula: ； Calculate data points The weighted number of connections , generating data point connectivity metrics results, where It is a data point and An indication value indicating whether there is a connection between them. 0 means no connection, and 1 means connection. It is a data point and The distance between is the attenuation coefficient, which is used to adjust the intensity of the distance effect. Indicates the number of data points; Based on the data point connection measurement results, set the screening threshold , by comparison , determine key data points and generate key data point sets; Marking the key data point set according to the connection strength and trend distribution of the data points to generate a marked key data point set; The steps for obtaining the data trend classification record are as follows: Based on the key data point set after the labeling, analysis and classification are performed, using the formula: ； Calculate the total aggregation degree of the classification results , generate preliminary data classification results, where and The data points are and The marking score of It is a data point and The distance between is the attenuation constant, used to adjust for the effect of distance, Indicates the number of data points; Analyze the preliminary data classification results, determine the key trends and patterns of each group of data, refine the data classification process through quantitative analysis, and obtain structured data classification results; Integrate the structured data classification results, build a structured trend framework, record the classification and characteristics of each data trend, and generate a data trend classification record.

2. The method for generating data reports based on artificial intelligence according to claim 1, characterized in that: The steps for obtaining the preliminary analysis data are: Connect to the IDC data center, use secure network protocols to obtain time series data and equipment performance logs, perform preliminary verification of data integrity and format, exclude damaged or non-compliant records, and generate verified data sets; Extracting time stamps and performance indicators from the verified data set, sorting the time stamps, confirming the time sequence correctness of the data, and generating a sorted data set; Based on the sorted data set, combined with the device type, the performance indicators in the data set are summarized using the formula: ； Computing device type The average performance index quantified value , ensure consistency in equipment performance assessment and generate preliminary analysis data, including: Represents the device type Middle The performance index of each device is a quantitative value, which reflects the performance of a single device. Represents the device type The number of devices used for averaging calculations.

3. The method for generating data reports based on artificial intelligence according to claim 1, characterized in that: The steps for obtaining the comprehensive IDC data analysis report are as follows: Analyze the data trend classification records, identify data points that need to be updated, and generate a list of data points to be updated; Perform synchronization detection on the marked data point list, check data consistency, confirm data integrity, and generate synchronization detection results; Based on the synchronous detection results, the report structure and details are automatically adjusted, key information points are updated, and formulas are used. ； Calculating adjusted report quality scores , generate a comprehensive IDC data analysis report, including: Represents key information points The importance weight of Is the input content at the information point The amount of change.

Images