Computing network integrated multi-dimensional data sensing system

Abstract

The invention discloses a computing network integrated multi-dimensional data sensing system, which comprises a book login module, a data acquisition module, a data storage module, a non-intrusive observation module and an intelligent behavior prediction module, and is characterized in that the non-intrusive observation module is used for monitoring and analyzing acquired data; and the intelligent behavior prediction module is used for generating action suggestions for the client according to the analysis result obtained by the non-invasive observation module. According to the invention, multi-dimensional analysis is carried out on user demands through the non-invasive observation module, and the analysis result is compared with user data obtained in the data monitoring module, so that the data required to be obtained by the user is analyzed, the user completes attribution analysis of abnormal indexes under the guidance of a product, action suggestions are generated for a client according to the obtained analysis result, and the user experience is improved. Real-time capture, understanding and prediction of computing network resource states are realized, structuring of sensing data and automatic association based on a service view angle are realized, and data support is provided for real-time sensing and dynamic adjustment capability of a computing network.

Description

Technical Field

[0001] The present invention relates to the field of network technology, and in particular to a computing-network integrated multi-dimensional data perception system. Background Art

[0002] With the rapid development of cloud computing, big data, and artificial intelligence, a wave of digital transformation has swept the world. Under the new generation of operating systems, higher requirements are placed on cloud-network integration and cloud-network operations, as well as the quality of network operation data. This has posed a huge challenge to existing network data maintenance methods. Network data collection is an important link in the network system that cannot be ignored and is the top priority of the network system.

[0003] The basic data source of the network data collection method is single. Manual configuration is usually required for the entry / exit of each network element / collection source. There are problems such as many links, high investment, low timeliness and low reliability. It is not convenient for real-time capture, understanding and prediction of the status of computing network resources. It is difficult to achieve structuring of perception data and automatic association based on business perspective, which makes real-time perception and dynamic adjustment of computing networks difficult, reduces the efficiency of network data maintenance, and cannot quickly meet the new generation of cloud network operation requirements under digital transformation. Summary of the Invention

[0004] The purpose of the present invention is to provide a computing-network integrated multi-dimensional data perception system to solve the problems raised in the above background technology.

[0005] To achieve the above-mentioned objectives, the present invention provides the following technical solutions: a computer-network integrated multi-dimensional data perception system, comprising a registration and login module, a data acquisition module, a data storage module, a non-invasive observation module and an intelligent behavior prediction module, wherein the registration and login module, the data acquisition module, the data storage module, the non-invasive observation module and the intelligent behavior prediction module are connected, the registration and login module is used for users to log in to the system and input user data, the data acquisition module is used to connect to an external network and acquire knowledge graphs, data models and user data, the data storage module is used to store the knowledge graphs, data models and user data acquired by the data acquisition module, the non-invasive observation module is used to monitor and analyze the acquired data, and the intelligent behavior prediction module is used to generate action recommendations to customers based on the analysis results obtained by the non-invasive observation module.

[0006] Preferably, the non-invasive observation module includes:

[0007] A data monitoring module, which is used to provide timely and effective feedback on data anomalies;

[0008] The data analysis module takes business scenarios and business goals as the starting point and business decisions as the end point. It decomposes the business scenarios and business goals into several influencing factors and sub-projects, and conducts data-based status analysis around the sub-projects;

[0009] The data warning module detects data anomalies by comparing various data dimensions and issues an alarm for abnormal data.

[0010] Preferably, the data monitoring module includes:

[0011] Basic data indicator monitoring, which is used to monitor the underlying data of the system;

[0012] Registered user number monitoring, which is used to monitor the number of users and online users in the system;

[0013] User basic data monitoring, which is used to detect user information;

[0014] User operation data monitoring, which records and monitors user operation behaviors;

[0015] System response data monitoring, wherein the system response data monitoring is used to monitor the system's feedback on user operation behavior.

[0016] Preferably, the data analysis module includes:

[0017] Single analysis, which is used for trend insights, channel attribution, link tagging, funnel analysis, heat map analysis, clustering analysis, A / B analysis, and retention analysis;

[0018] Combination analysis, which involves performing multi-dimensional combination analysis on a segmentation point;

[0019] User scenario analysis, which is used to analyze the user's time, location, and needs;

[0020] Modeling analysis, which is used to perform churn warning analysis, user activation analysis, and payment decision analysis.

[0021] Preferably, the data early warning module:

[0022] Determination of magnitude indicators and conversion indicators, which are used to determine the data indicators of the links and the conversion status of the links;

[0023] Determine the normal fluctuation range of the indicator, which sets a normal floating range for each indicator based on historical data;

[0024] Trigger condition determination, wherein the trigger condition determination is used to determine the trigger condition for data early warning;

[0025] Determine the warning period and frequency, which is used to determine the period and frequency of warnings;

[0026] An early warning method is used to alert data anomalies in different ways.

[0027] Preferably, the intelligent behavior prediction module includes:

[0028] A capability base, which serves as the infrastructure of the intelligent behavior prediction module;

[0029] Channel management, which is used to configure question and answer and permissions for user requests;

[0030] Dialogue management, which uses a dialogue engine to adapt and select large models;

[0031] Conversational applications, wherein the conversational applications are used to perform conversational data query, data analysis, and data retrieval.

[0032] Preferably, the intelligent behavior prediction module includes:

[0033] A chart recommendation engine, which generates visual charts from data based on query results;

[0034] A semantic parser is used to parse the user input language through rule model parsing and large language model parsing.

[0035] Preferably, the capability base has the basic capabilities of data connection, modeling calculation, rendering construction, and permission management. The dialogue management is used to support the dialogue engine's intent classification, intent understanding, idea decomposition, and context management capabilities. The dialogue management uses the generalization capabilities of large models to complete the dialogue interaction between the product and the user.

[0036] Preferably, the chart recommendation engine is connected to the data storage module and acquires the knowledge graph, data model and user behavior information, and the chart recommendation engine is connected to the data analysis module and outputs the analysis results.

[0037] A method for operating a computing-network integrated multi-dimensional data perception system includes the following steps:

[0038] Step S1: The user enters the system through the registration and login module, and the data acquisition module obtains the user data and stores it in the data storage module;

[0039] Step S2: The user inputs the requirements, the non-intrusive observation module analyzes the user requirements, the data monitoring module monitors the user's basic data and operation data, and provides feedback on data anomalies;

[0040] Step S3: Perform multi-dimensional analysis of user needs through the data analysis module, compare it with the user data obtained from the data monitoring module, and parse out the data that the user needs to obtain based on the user's historical data;

[0041] Step S4: The intelligent behavior prediction module performs intent classification, intent understanding, thought analysis, and context management on the user input language. Combined with the analysis of user needs in step S3, the generalization capability of the large model is used to complete the dialogue interaction between the product and the user.

[0042] Step S5: Based on the abnormal data detected in step S2, the correct data problem is recommended to the user by breaking down and analyzing the business problem raised by the user;

[0043] Step S6: Based on the recommended questions selected by the user, detect anomalies in the data, generate attribution questions, and recommend the attribution questions to the user;

[0044] Step S7: Complete the attribution analysis of abnormal data indicators based on the attribution problem selected by the user, and generate an analysis report and action recommendations.

[0045] Technical effects and advantages of the present invention:

[0046] The present invention utilizes a setting method that cooperates with a non-intrusive observation module and an intelligent behavior prediction module. The non-intrusive observation module performs multi-dimensional analysis of user needs and compares them with the user data obtained in the data monitoring module. Based on the user's historical data, the data that the user needs to obtain is parsed, and the intelligent behavior prediction module is used to perform intent classification, intent understanding, thought decomposition and context management on the user's input language. Combined with the analysis of user needs, the generalization ability of the large model is used to complete the dialogue interaction between the product and the user. The user can further complete the attribution analysis of abnormal indicators under the guidance of the product, and generate action suggestions for the customer based on the obtained analysis results, realizing real-time capture, understanding and prediction of the computing network resource status, realizing the structuring of perception data and automatic association based on the business perspective, and providing data support for the real-time perception and dynamic adjustment capabilities of the computing network. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 Schematic diagram of the overall system of the present invention.

[0048] Figure 2 This is a schematic diagram of the non-invasive observation module of the present invention.

[0049] Figure 3 Schematic diagram of the data monitoring module of the present invention.

[0050] Figure 4 Schematic diagram of the data analysis module of the present invention.

[0051] Figure 5 This is a schematic diagram of the data early warning module of the present invention.

[0052] Figure 6 Schematic diagram of the intelligent behavior prediction module of the present invention.

[0053] Figure 7 This is a flow chart of the intelligent behavior prediction module of the present invention.

[0054] Figure 8 Schematic diagram of the system flow of the present invention. DETAILED DESCRIPTION

[0055] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0056] The present invention provides Figure 1-6 The multi-dimensional data perception system of the computing network integration shown in the figure includes a registration module, a data acquisition module, a data storage module, a non-invasive observation module and an intelligent behavior prediction module. The registration module, the data acquisition module, the data storage module, the non-invasive observation module and the intelligent behavior prediction module are connected. The registration module is used for users to log in to the system and input user data. Users enter the system through the registration module and input user requirements. The data acquisition module is used to connect to the external network and obtain knowledge graphs, data models and user data. The data acquisition module is based on the big data system and can obtain data through manual settings and automatic settings, which is convenient for data and The data storage module is used to store the knowledge graph, data model and user data obtained by the data acquisition module. The non-intrusive observation module is used to monitor and analyze the obtained data. The intelligent behavior prediction module is used to generate action suggestions to customers based on the analysis results obtained by the non-intrusive observation module. User needs are analyzed through the non-intrusive observation module and the intelligent behavior prediction module, and business data is monitored and analyzed. Action suggestions are generated to customers based on the obtained analysis results, realizing real-time capture, understanding and prediction of the computing network resource status, realizing the structuring of perception data and automatic association based on business perspective, and providing data support for the real-time perception and dynamic adjustment capabilities of the computing network.

[0057] The non-intrusive observation module includes a data monitoring module, a data analysis module, and a data early warning module. The data monitoring module is used to provide timely and effective feedback on data anomalies. It monitors the data to observe whether there are any anomalies and then analyzes the data. The data analysis module takes business scenarios and business goals as the starting point for thinking and business decisions as the end point. It decomposes the business scenarios and business goals into several influencing factors and sub-projects, conducts data status analysis around the sub-projects, and finds ways to improve the status quo. The data early warning module discovers data anomalies through comparison of various data dimensions and issues alarms for abnormal data.

[0058] The data monitoring module includes basic data indicator monitoring, registered user number monitoring, user basic data monitoring, user operation data monitoring and system response data monitoring. Basic data indicator monitoring is used to monitor the underlying data of the system, registered user number monitoring is used to monitor the number of users and online users in the system, user basic data monitoring is used to detect user information, user operation data monitoring records and monitors user operation behaviors, and system response data monitoring is used to monitor the system's feedback on user operation behaviors.

[0059] The data analysis module includes single item analysis, combination analysis, user scenario analysis and modeling analysis, single potential insight, channel attribution, link tagging, funnel analysis, heat map analysis, cluster analysis, A / B analysis and retention analysis. The combination analysis conducts multi-dimensional combination analysis on a segmentation point. The user scenario analysis is used to analyze the user's time, place and needs. The modeling analysis is used for churn warning analysis, user activation analysis and payment decision analysis. The data analysis module can automatically graph the desired data by brushing out from the data background, and each section is made into a chart for our quick review. For example, each triggering behavior of the user is added with a tracking point, and the data we want is queried by the time dimension. Data monitoring and data analysis are very important for operations. Good data monitoring can reduce product bugs, affect user experience, and reduce the occurrence of major accidents.

[0060] The data early warning module includes the determination of magnitude indicators and conversion indicators, the determination of normal fluctuation range of indicators, the determination of trigger conditions, the determination of warning cycle and frequency, and the early warning method. The magnitude indicators and conversion indicators are used to determine the data indicators of the links and the conversion conditions of the links. The magnitude indicators are the data indicators of each link. The significance of the magnitude indicators is that they can be processed into the data we want. The conversion indicators are the conversion of each link. By observing the conversion indicators, we can quickly locate the link where the problem occurs. The normal fluctuation range of indicators is determined by setting a normal floating range for each indicator based on historical data. Each indicator must set a normal floating range based on historical data. It can be determined from the following four data dimensions. Identify the normal fluctuation range, year-on-year data, that is, compare with the same day and same period last week, month-on-month data, that is, compare with the average value of the same period of the previous three days, the conversion of each link, that is, compare with the conversion of each link in the previous N days, each hourly increase, that is, compare with the increase of each hour in the previous N days, and the trigger conditions are used to determine the trigger conditions for data early warning. Usually, an early warning will be issued if it is lower than the normal floating range. The early warning cycle and frequency are used to determine the cycle and frequency of the early warning. The early warning method uses different methods to alert data anomalies, and uses magnitude indicators and conversion indicators to monitor the data indicators and link conversion conditions of the links to quickly locate the data and links with problems.

[0061] The intelligent behavior prediction module includes a capability base, channel management, dialogue management, conversational applications, a chart recommendation engine, and a semantic parser. The capability base serves as the module's foundational structure. Channel management configures questions and answers and permissions for user requests. Dialogue management adapts and selects large models through the dialogue engine. Conversational applications perform conversational data query, data analysis, and data retrieval. The chart recommendation engine generates visual charts based on query results. The semantic parser parses user input using rule-based models and large language models. Business users only need to ask the right business questions. The intelligent behavior prediction module combines data and business knowledge to break down analytical ideas and recommend the right data questions. After users click to select the corresponding data question, the system automatically detects data anomalies based on the generated data results and recommends corresponding attribution questions. Users can then, under the guidance of the product, complete attribution analysis of abnormal indicators and generate analysis reports and action recommendations. In this way, users can complete the entire analytical closed loop for a specific business problem, from descriptive analysis to diagnostic analysis and then to prescriptive analysis, step by step.

[0062] The capability base has the basic capabilities of data connection, modeling calculation, rendering construction, and permission management. Dialogue management is used to support the dialogue engine's intent classification, intent understanding, idea decomposition and context management capabilities. Dialogue management uses the generalization capabilities of large models to complete dialogue interactions between products and users. The chart recommendation engine is connected to the data storage module and obtains knowledge graphs, data models and user behavior information. The chart recommendation engine is connected to the data analysis module and outputs the analysis results.

[0063] The present invention provides Figure 7-8 The method for operating a computing-network integrated multi-dimensional data perception system includes the following steps:

[0064] Step S1: The user enters the system through the registration and login module, and the data acquisition module obtains the user data and stores it in the data storage module;

[0065] Step S2: The user inputs the requirements, the non-intrusive observation module analyzes the user requirements, the data monitoring module monitors the user's basic data and operation data, and provides feedback on data anomalies;

[0066] Step S3: Perform multi-dimensional analysis of user needs through the data analysis module, compare it with the user data obtained from the data monitoring module, and parse out the data that the user needs to obtain based on the user's historical data;

[0067] Step S4: The intelligent behavior prediction module performs intent classification, intent understanding, thought analysis, and context management on the user input language. Combined with the analysis of user needs in step S3, the generalization capability of the large model is used to complete the dialogue interaction between the product and the user.

[0068] Step S5: Based on the abnormal data detected in step S2, the correct data problem is recommended to the user by breaking down and analyzing the business problem raised by the user;

[0069] Step S6: Based on the recommended questions selected by the user, detect anomalies in the data, generate attribution questions, and recommend the attribution questions to the user;

[0070] Step S7: Complete the attribution analysis of abnormal data indicators based on the attribution problem selected by the user, and generate an analysis report and action recommendations.

[0071] The working principle of the present invention is as follows: a non-invasive observation module is used to perform multi-dimensional analysis of user needs, and compare them with the user data obtained in the data monitoring module. Based on the user's historical data, the data that the user needs to obtain is parsed, and the intelligent behavior prediction module is used to perform intent classification, intent understanding, idea decomposition and context management capabilities on the user's input language. Combined with the analysis of user needs, the generalization capability of the large model is used to complete the dialogue interaction between the product and the user, decompose the analysis ideas and recommend the correct data problems. After the user clicks to select the corresponding data problem, the system can also automatically perform data anomaly detection based on the generated data results and recommend corresponding attribution problems. The user can further complete the attribution analysis of the abnormal indicators under the guidance of the product, and generate action suggestions for the customer based on the obtained analysis results, so as to realize the real-time capture, understanding and prediction of the computing network resource status, realize the structuring of perception data and automatic association based on the business perspective, and provide data support for the real-time perception and dynamic adjustment capabilities of the computing network.

[0072] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A computing-network integrated multi-dimensional data perception system, characterized by: It includes a registration and login module, a data acquisition module, a data storage module, a non-invasive observation module and an intelligent behavior prediction module. The registration and login module, the data acquisition module, the data storage module, the non-invasive observation module and the intelligent behavior prediction module are connected. The registration and login module is used for users to log in to the system and input user data. The data acquisition module is used to connect to an external network and acquire knowledge graphs, data models and user data. The data storage module is used to store the knowledge graphs, data models and user data acquired by the data acquisition module. The non-invasive observation module is used to monitor and analyze the acquired data. The intelligent behavior prediction module generates action suggestions to customers based on the analysis results obtained by the non-invasive observation module.

2. The computing-network integrated multi-dimensional data perception system according to claim 1, characterized in that: The non-intrusive observation module includes: A data monitoring module, which is used to provide timely and effective feedback on data anomalies; The data analysis module takes business scenarios and business goals as the starting point and business decisions as the end point. It decomposes the business scenarios and business goals into several influencing factors and sub-projects, and conducts data-based status analysis around the sub-projects; The data warning module detects data anomalies by comparing various data dimensions and issues an alarm for abnormal data.

3. The computing-network integrated multi-dimensional data perception system according to claim 2, characterized in that: The data monitoring module includes: Basic data indicator monitoring, which is used to monitor the underlying data of the system; Registered user number monitoring, which is used to monitor the number of users and online users in the system; User basic data monitoring, which is used to detect user information; User operation data monitoring, which records and monitors user operation behaviors; System response data monitoring, wherein the system response data monitoring is used to monitor the system's feedback on user operation behavior.

4. The computing-network integrated multi-dimensional data perception system according to claim 2, characterized in that: The data analysis module includes: Single analysis, which is used for trend insights, channel attribution, link tagging, funnel analysis, heat map analysis, clustering analysis, A / B analysis, and retention analysis; Combination analysis, which involves performing multi-dimensional combination analysis on a segmentation point; User scenario analysis, which is used to analyze the user's time, location, and needs; Modeling analysis, which is used to perform churn warning analysis, user activation analysis, and payment decision analysis.

5. The computing-network integrated multi-dimensional data perception system according to claim 2, characterized in that: The data early warning module includes: Determination of magnitude indicators and conversion indicators, which are used to determine the data indicators of the links and the conversion status of the links; Determine the normal fluctuation range of the indicator, which sets a normal floating range for each indicator based on historical data; Trigger condition determination, wherein the trigger condition determination is used to determine the trigger condition for data early warning; Determine the warning period and frequency, which is used to determine the period and frequency of warnings; An early warning method is used to alert data anomalies in different ways.

6. The computing-network integrated multi-dimensional data perception system according to claim 1, characterized in that: The intelligent behavior prediction module includes: A capability base, which serves as the infrastructure of the intelligent behavior prediction module; Channel management, which is used to configure question and answer and permissions for user requests; Dialogue management, which uses a dialogue engine to adapt and select large models; Conversational applications, wherein the conversational applications are used to perform conversational data query, data analysis, and data retrieval.

7. The computing-network integrated multi-dimensional data perception system according to claim 1, characterized in that: The intelligent behavior prediction module includes: A chart recommendation engine, which generates visual charts from data based on query results; A semantic parser is used to parse the user input language through rule model parsing and large language model parsing.

8. The computing-network integrated multi-dimensional data perception system according to claim 6, characterized in that: The capability base has the basic capabilities of data connection, modeling calculation, rendering construction, and permission management. The dialogue management is used to support the dialogue engine's intent classification, intent understanding, idea decomposition and context management capabilities. The dialogue management uses the generalization capabilities of large models to complete the dialogue interaction between the product and the user.

9. The computing-network integrated multi-dimensional data perception system according to claim 7, characterized in that: The chart recommendation engine is connected to the data storage module and acquires the knowledge graph, data model and user behavior information. The chart recommendation engine is connected to the data analysis module and outputs the analysis results.

10. A method for operating a computing-network integrated multi-dimensional data perception system, characterized in that: The following steps are involved: Step S1: The user enters the system through the registration and login module, and the data acquisition module obtains the user data and stores it in the data storage module; Step S2: The user inputs the requirements, the non-intrusive observation module analyzes the user requirements, the data monitoring module monitors the user's basic data and operation data, and provides feedback on data anomalies; Step S3: Perform multi-dimensional analysis of user needs through the data analysis module, compare it with the user data obtained from the data monitoring module, and parse out the data that the user needs to obtain based on the user's historical data; Step S4: The intelligent behavior prediction module performs intent classification, intent understanding, thought analysis, and context management on the user input language. Combined with the analysis of user needs in step S3, the generalization capability of the large model is used to complete the dialogue interaction between the product and the user. Step S5: Based on the abnormal data detected in step S2, the correct data problem is recommended to the user by breaking down and analyzing the business problem raised by the user; Step S6: Based on the recommended questions selected by the user, detect anomalies in the data, generate attribution questions, and recommend the attribution questions to the user; Step S7: Complete the attribution analysis of abnormal data indicators based on the attribution problem selected by the user, and generate an analysis report and action recommendations.

Images