Application operation management method and system based on big data

Abstract

The invention relates to the technical field of data processing, in particular to an application operation management method and system based on big data, and the method comprises the steps: obtaining and analyzing registration information corresponding to a plurality of registered applications, and determining initial labels of the plurality of registered applications based on an analysis result; constructing a preset application library based on the plurality of registered applications and the corresponding initial labels; collecting and analyzing user historical behavior data, and determining a user tag according to a data analysis result; performing matching based on the user tag and a preset application library to determine a plurality of initial associated applications; monitoring use parameters of the plurality of initial associated applications in real time, constructing and analyzing a plurality of use parameter curves, and obtaining a plurality of adjustment associated applications based on an analysis result; and collecting actual behavior data of the plurality of adjustment associated applications, and performing analysis based on the actual behavior data and the historical behavior data to update the user tag, thereby updating the adjustment associated applications and obtaining a target associated application. The optimization efficiency of the application software is improved.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a big data-based application operation management method and system. Background Art

[0002] With the rapid development of information technology, big data has become a vital force driving the transformation and upgrading of various industries. In the field of application software, the application of big data not only improves user experience but also greatly enhances the intelligence and personalized service capabilities of software.

[0003] The patent document with Chinese patent application publication number CN102065077A discloses a terminal application software distribution method and system, wherein the method includes: registering a developer as a user of a developer community through an application development terminal; developing application software by the application development terminal; testing the application software by a developer test terminal; and packaging the application software that passes the test and submitting it to the developer community by the application development terminal.

[0004] The operation and management of application software in the existing technology lacks effective data analysis and intelligent decision-making support, resulting in low optimization efficiency of application software. Summary of the Invention

[0005] To this end, the present invention provides an application operation management method and system based on big data, which can solve the problem of low optimization efficiency of application software.

[0006] To achieve the above objectives, the present invention provides an application operation management method based on big data, the method comprising:

[0007] Obtain registration information corresponding to a number of registered applications, analyze the registration information, and construct corresponding initial tags for the number of registered applications based on the analysis results;

[0008] Build a preset application library based on several registered applications and their corresponding initial tags;

[0009] Collect user historical behavior data, analyze the historical behavior data, and determine user tags based on the data analysis results;

[0010] Determine a number of initial associated applications based on matching the user tag and the preset application library;

[0011] monitoring usage parameters of the plurality of initially associated applications in real time at a preset frequency, constructing a plurality of usage parameter curves based on the plurality of usage parameters, analyzing the plurality of usage parameter curves, and screening the plurality of initially associated applications based on the analysis results of the plurality of curves to obtain a plurality of adjusted associated applications;

[0012] Actual behavior data of a number of adjustment-related applications are collected, and analysis is performed based on the actual behavior data and the historical behavior data to update the user tag, thereby updating the adjustment-related applications and obtaining target-related applications.

[0013] Furthermore, the step of constructing corresponding initial tags for a number of registered applications based on the analysis results includes:

[0014] Perform word segmentation processing on any of the registration information to obtain a number of word segments;

[0015] Calculating actual importance levels corresponding to a number of the segmented words, determining the actual importance levels, and determining a number of keywords based on the determination results;

[0016] The initial tag is determined based on a number of the keywords.

[0017] Furthermore, the step of calculating the actual importance corresponding to the plurality of the segmented words includes:

[0018] Counting actual frequencies corresponding to several of the segmented words;

[0019] Determining actual distribution dispersions corresponding to a number of the segmentations;

[0020] The actual importance is calculated based on the actual frequency and the actual distribution dispersion.

[0021] Furthermore, the step of analyzing the historical behavior data includes:

[0022] Statistics on the historical usage frequency and certain historical effective usage duration of any historical application within the historical period;

[0023] Determining historical key applications based on the historical usage frequencies and the historical effective usage durations corresponding to the historical applications;

[0024] The historical key applications are analyzed to determine the user tags.

[0025] Furthermore, the steps of calculating the historical usage frequency and several historical effective usage durations of any historical application within a historical period include:

[0026] determining the historical usage frequency based on the number of start timestamps of any of the historical applications;

[0027] Determine the historical usage duration based on any start timestamp and its corresponding end timestamp corresponding to any historical application;

[0028] Determine the historical interaction duration in any of the historical usage durations by collecting the user's time period behavior data in the historical usage duration corresponding to the historical application;

[0029] The historical effective duration is determined based on any of the historical usage durations and the corresponding historical interaction durations.

[0030] Furthermore, the step of determining the historical interaction duration in any of the historical usage durations by collecting the user's time period behavior data in the historical usage durations corresponding to the historical application includes:

[0031] Collect screen click events, sliding events, and input events during the user's historical application use;

[0032] performing a time series analysis on the plurality of screen click events, the sliding events, and the input events, and determining the user's activity level in each time period based on the sequence analysis results;

[0033] The user historical interaction duration is determined based on a number of the activity levels.

[0034] Furthermore, the step of determining a plurality of initial associated applications based on matching the user tag and the preset application library includes:

[0035] Calculating similarities between the user tag and a plurality of the initial tags to obtain a plurality of actual similarities;

[0036] The actual similarities are compared with the preset similarities, and based on the comparison results, a number of associated tags are determined, thereby determining the initial associated applications.

[0037] Furthermore, the steps of analyzing the plurality of parameter curves include:

[0038] Collecting the real-time response time and real-time CPU usage of any of the initially associated applications in real time at a preset frequency;

[0039] Drawing a time variation curve based on the real-time response times;

[0040] Drawing a usage change curve based on the real-time CPU usage;

[0041] The time variation curve and the usage rate variation curve are analyzed to determine the abnormal time period and the abnormal usage rate period, and the total abnormal period is determined based on the abnormal time period and the abnormal usage rate period.

[0042] Furthermore, the step of screening the initial associated applications based on the curve analysis results includes:

[0043] When the abnormal total period is less than or equal to a preset period, using the initial associated application as the adjusted associated application;

[0044] When the total abnormal period is greater than the preset period, the initial associated application is removed.

[0045] On the other hand, the present invention also provides a system for application operation management method based on big data, the system comprising:

[0046] An application tag determination module is used to obtain registration information corresponding to a number of registered applications, analyze the registration information, and construct corresponding initial tags for the number of registered applications based on the analysis results;

[0047] a construction module, connected to the application tag determination module, for constructing a preset application library based on a number of registered applications and their corresponding initial tags;

[0048] User tag determination module, used to collect user historical behavior data, analyze the historical behavior data, and determine user tags based on the data analysis results;

[0049] A matching module, configured to determine a plurality of initial associated applications based on matching between the user tag and the preset application library;

[0050] an adjustment module, connected to the matching module, configured to monitor usage parameters of the plurality of initially associated applications in real time at a preset frequency, construct a plurality of usage parameter curves based on the plurality of usage parameters, analyze the plurality of usage parameter curves, and screen the plurality of initially associated applications based on the analysis results of the plurality of curves to obtain a plurality of adjusted associated applications;

[0051] An update module is connected to the adjustment module and is used to collect actual behavior data of several adjustment-related applications, analyze the actual behavior data and the historical behavior data to update the user tag, and then update the adjustment-related applications to obtain target-related applications.

[0052] Compared with the prior art, the present invention has the following advantages: by analyzing registration information, the characteristics of each registered application can be accurately understood, laying a solid foundation for subsequent application classification and recommendation. The construction of initial tags helps to quickly identify the basic attributes and potential user groups of applications, providing a basis for subsequent data analysis and application matching. The establishment of a preset application library provides a centralized platform for storing and managing applications with initial tags, facilitating the rapid retrieval and matching of associations between users and applications, thereby improving operational efficiency. The collection and analysis of user historical behavior data helps to gain a deep understanding of user interests and habits, thereby creating accurate user tags, providing a basis for the subsequent determination of target-related applications. By matching user tags with applications in the preset application library, applications related to user interests can be quickly identified, improving system processing efficiency. By monitoring application usage parameters in real time and filtering applications by analyzing usage parameter curves, application usage performance can be analyzed, thereby enabling more precise adjustment and optimization of application recommendation lists to ensure users receive the best application experience. The collection of actual behavior data and analysis combined with historical behavior data to update user tags and application associations gives the recommendation system the ability to self-learn and adapt, dynamically adjust recommendation strategies, and ensure the accuracy and timeliness of recommendations.

[0053] In particular, by performing word segmentation on the registration information and splitting complex long text into multiple independent vocabulary units, it is helpful to have a more detailed understanding of the content of the registration information and provide a basis for subsequent analysis. By calculating the actual importance of each word segment, it is possible to more accurately capture the key points of the registration information and reduce the interference of noise data. After calculating the importance of the word segmentation, the keywords are screened out through the discrimination mechanism, providing accurate data support for the subsequent construction of initial tags. By constructing the initial tags based on keywords, the core features of the registered application are more accurately reflected, providing more accurate data support for subsequent user matching and application recommendations. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 A flowchart of a big data-based application operation management method provided by an embodiment of the present invention;

[0055] Figure 2 A schematic diagram of a process for determining an initial tag of an application in a big data-based application operation management method provided by an embodiment of the present invention;

[0056] Figure 3 A schematic diagram of a process for determining user tags in a big data-based application operation management method provided by an embodiment of the present invention;

[0057] Figure 4 This is a structural block diagram of a big data-based application operation and management system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0058] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below with reference to embodiments. It should be understood that the specific embodiments described herein are merely used to explain the present invention and are not intended to limit the present invention.

[0059] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0060] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside" and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the accompanying drawings. This is only for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it should not be understood as a limitation on the present invention.

[0061] Furthermore, it should be noted that, in the description of the present invention, unless otherwise expressly specified or limited, the terms "mounted," "connected," and "connected" should be understood in a broad sense. For example, they may refer to fixed connections, detachable connections, or integral connections; they may refer to mechanical connections or electrical connections; they may refer to direct connections or indirect connections through an intermediate medium; and they may refer to internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0062] See also Figure 1 As shown, an embodiment of the present invention provides an application operation management method based on big data, the method comprising:

[0063] Step S100: obtaining registration information corresponding to a number of registered applications, analyzing the registration information, and constructing corresponding initial tags for the number of registered applications based on the analysis results;

[0064] Step S200: Building a preset application library based on a number of registered applications and their corresponding initial tags;

[0065] Step S300: Collect user historical behavior data, analyze the historical behavior data, and determine user tags based on the data analysis results;

[0066] Step S400, determining a number of initial associated applications based on matching the user tag and the preset application library;

[0067] Step S500: monitoring usage parameters of the plurality of initially associated applications in real time at a preset frequency, constructing a plurality of usage parameter curves based on the plurality of usage parameters, analyzing the plurality of usage parameter curves, and screening the plurality of initially associated applications based on the analysis results of the plurality of curves to obtain a plurality of adjusted associated applications;

[0068] Step S600 , collecting actual behavior data of a number of adjustment-related applications, analyzing the actual behavior data and the historical behavior data to update the user tag, and then updating the adjustment-related applications to obtain target-related applications.

[0069] Specifically, the preset frequency in the embodiment of the present invention is once a day.

[0070] Specifically, the usage parameters in the embodiment of the present invention include the real-time CPU usage and real-time response time of the application. In actual applications, the stored usage parameter data is read from the database. According to the type of curve to be constructed (such as the CPU usage curve), the data of the corresponding field and time period are selected. For example, if you want to construct the CPU occupancy curve in the past day, the CPU occupancy data of the application in the past day is read from the database. When drawing the curve, use an open source drawing library such as Matplotlib in Python, JFreeChart in Java, etc. These libraries provide rich drawing functions and can draw smooth curves based on data points. For example, in Matplotlib, you can use the plot() function to use the timestamp as the horizontal coordinate and the parameter value (such as CPU usage) as the vertical coordinate to draw the usage parameter curve.

[0071] Specifically, the embodiments of the present invention accurately understand the characteristics of each registered application through analysis of registration information, laying a solid foundation for subsequent application classification and recommendation. The construction of initial tags helps to quickly identify the basic attributes and potential user groups of applications, providing a basis for subsequent data analysis and application matching. The establishment of a preset application library provides a centralized platform for storing and managing applications with initial tags, facilitating the rapid retrieval and matching of associations between users and applications, thereby improving operational efficiency. The collection and analysis of user historical behavior data helps to gain a deep understanding of user interests and habits, thereby creating accurate user tags and providing a basis for the subsequent determination of target-related applications. By matching user tags with applications in the preset application library, applications related to user interests are quickly identified, improving system processing efficiency. By monitoring application usage parameters in real time and filtering applications by analyzing usage parameter curves, application usage performance can be analyzed, thereby enabling more precise adjustment and optimization of application recommendation lists to ensure that users receive the best application experience. The collection of actual behavior data and analysis combined with historical behavior data to update user tags and application associations enable the recommendation system to have self-learning and adaptive capabilities, dynamically adjust recommendation strategies, and ensure the accuracy and timeliness of recommendations.

[0072] See also Figure 2 As shown, the steps of constructing corresponding initial labels for several registered applications based on the analysis results include:

[0073] Step S110, performing word segmentation processing on any of the registration information to obtain a number of word segments;

[0074] Step S120, calculating actual importance corresponding to a number of the segmented words, determining the actual importance, and determining a number of keywords based on the determination results;

[0075] Step S130: determining the initial tag based on the keywords.

[0076] Specifically, the embodiment of the present invention can perform word segmentation processing on any of the registration information by using a word segmentation library (such as jieba) or an NLP tool (such as spaCy).

[0077] Specifically, the embodiment of the present invention performs word segmentation processing on the registration information by using a word segmentation library (such as jieba) or an NLP tool (such as spaCy), splits the text content in the registration information into vocabulary units with independent meanings, removes meaningless stop words and redundant information, and extracts core vocabulary. It can accurately extract key information in the registration information, avoid the omission of important information due to the length or complexity of the text, and provide high-quality basic data for subsequent keyword extraction and label construction. By calculating the actual importance of the words after word segmentation and judging the importance, we can identify which words are more representative in the registration information and remove words with low correlation with the core functions or features of the application. This can effectively screen out the keywords that best reflect the core functions or features of the registered application, avoid the interference of irrelevant words, and improve the accuracy and pertinence of label construction. The initial labels are determined based on the screened keywords, and the keywords are mapped to the core functions or features of the application to generate initial labels that can accurately describe the application. The initial labels generated for each registered application are more personalized and accurate, which can help users quickly understand the main functions or uses of the application. At the same time, it also provides a high-quality label foundation for subsequent application classification, recommendation and other operations. The entire process from word segmentation to keyword extraction to label construction is completed by automated algorithms, reducing manual intervention, avoiding the subjectivity and inefficiency of manual labeling, and significantly improving processing efficiency. It is especially suitable for large-scale registered application processing scenarios, reducing labor costs and time costs, and improving the scalability of the system.

[0078] Specifically, the embodiment of the present invention determines several keywords based on the discrimination results, including:

[0079] Comparing the importance levels with a preset importance threshold to obtain a comparison result;

[0080] When the importance is greater than or equal to the importance threshold, the corresponding segmented word is used as a keyword;

[0081] When the importance is less than the importance threshold, the corresponding word is not used as a keyword.

[0082] Specifically, the preset importance threshold in the embodiment of the present invention is 80%.

[0083] Specifically, the embodiment of the present invention performs word segmentation on the registration information, splitting complex long text into multiple independent vocabulary units, which helps to understand the content of the registration information more finely and provides a basis for subsequent analysis. By calculating the actual importance of each word segmentation, it can more accurately capture the key points of the registration information and reduce the interference of noise data. After calculating the importance of the word segmentation, the discrimination mechanism is used to screen out keywords, providing accurate data support for the subsequent construction of initial tags. Through the process of constructing initial tags based on keywords, the core features of the registered application are more accurately reflected, providing more accurate data support for subsequent user matching and application recommendations.

[0084] Specifically, the step of calculating the actual importance corresponding to the plurality of segmentations includes:

[0085] Counting actual frequencies corresponding to several of the segmented words;

[0086] Determining actual distribution dispersions corresponding to a number of the segmentations;

[0087] The actual importance is calculated based on the actual frequency and the actual distribution dispersion.

[0088] Specifically, the embodiment of the present invention determines the actual distribution dispersion corresponding to the plurality of segmentations, including:

[0089] Dividing any of the registration information into a front section, a middle section, and a back section;

[0090] Counting the frequencies of any of the segmented words in the first segment, the middle segment, and the last segment as the first frequency, the second frequency, and the third frequency;

[0091] The differences among the first frequency, the second frequency, and the third frequency are calculated, the average of the differences is calculated, the average is normalized, and the processed value is used as the actual distribution dispersion.

[0092] Specifically, the embodiment of the present invention divides any registration information into the front section, the middle section, and the back section, including:

[0093] Identify a number of sentences corresponding to any of the registration information;

[0094] The number of sentences is counted, and the registration information is divided according to a ratio of 1:1:1 to obtain a front section, a middle section, and a back section.

[0095] Specifically, in the embodiment of the present invention, the step of calculating the actual importance based on the actual frequency and the actual distribution dispersion includes:

[0096] Calculating a frequency score based on the actual frequency and the preset frequency;

[0097] Calculating a dispersion score based on the actual distribution dispersion and the preset distribution dispersion;

[0098] The actual importance is determined based on the frequency score and the dispersion score.

[0099] Specifically, the actual importance level in the embodiment of the present invention = α × frequency score + β × dispersion score, where α and β are weight coefficients, and α + β = 1;

[0100] The frequency score = (actual frequency - preset frequency) / preset frequency. When the actual frequency is less than the preset frequency, the frequency score = 0;

[0101] The dispersion score = (preset distribution dispersion - actual distribution dispersion) / preset distribution dispersion. When the actual distribution dispersion is greater than the preset distribution dispersion, the dispersion score = 0.

[0102] The preset frequency is 2 / 3 of the maximum frequency among the occurrence frequencies of several word segmentation statistics;

[0103] The preset distribution discreteness is 1 / 3.

[0104] Specifically, the embodiment of the present invention compares the actual frequency of the words with the preset frequency, quantifies the frequency of occurrence of each word, and thus distinguishes between high-frequency words and low-frequency words. It can quickly screen out words that frequently appear in the registration information. These high-frequency words are more likely to be key words that describe the core functions or features of the application, thereby providing an important basis for subsequent keyword extraction. By evaluating the uniformity of the distribution of words in the registration information, it can distinguish between those that appear in a concentrated manner (possibly redundant information) and those that are evenly distributed (possibly key information). It can effectively identify those words that are more evenly distributed in the text. These words usually better reflect the overall theme and core content of the registration information, avoiding misjudging their importance due to the words that appear in a local concentrated manner. Combining the frequency score and the dispersion score, the frequency of occurrence and distribution uniformity of the words are comprehensively considered, thereby more comprehensively evaluating the importance of the words, and more accurately determining the actual importance of each word, avoiding the deviation caused by relying solely on a single indicator (such as frequency). This comprehensive evaluation method can more accurately screen out keywords and provide high-quality input for subsequent label construction. Through the comprehensive calculation of frequency score and dispersion score, words with high actual importance are screened out as keywords, ensuring that the extracted keywords have both a high frequency of occurrence and are evenly distributed in the text, and can fully reflect the core content of the registration information. It improves the accuracy and reliability of keyword extraction, reduces the interference of misjudgment and redundant information, and enables the constructed initial labels to more accurately describe the functions and features of the registered application. Keywords are extracted based on the accurately determined actual importance, and the initial labels are constructed accordingly, ensuring that the initial labels can accurately reflect the core functions and features of the registered application, avoiding user misunderstandings or management confusion due to inaccurate labels, optimizing the quality of the initial labels, and improving the user experience, so that users can find applications that meet their needs more quickly. At the same time, it also provides a more efficient application classification and management method for the application management platform, improves the overall management efficiency, and evaluates the importance of vocabulary by comprehensively considering the two dimensions of frequency and discreteness, so that the system can adapt to different types of registration information. Whether it is long text or short text, key information can be screened out through the comprehensive evaluation of frequency and discreteness, which enhances the adaptability and robustness of the system, and can handle diverse registration information, avoiding the failure or inaccuracy of keyword extraction due to differences in text structure or content.

[0105] See also Figure 3 As shown, the steps of analyzing the historical behavior data include:

[0106] Step S310, counting the historical usage frequency and several historical effective usage durations of any historical application within a historical period;

[0107] Step S320, determining a historical key application based on the historical usage frequencies and the historical effective usage durations corresponding to the historical applications;

[0108] Step S330: Analyze the historical key applications to determine the user tag.

[0109] Specifically, the embodiments of the present invention can comprehensively and objectively reflect the user's usage of various applications within a certain period by counting the usage frequency and effective usage time of historical applications, providing an accurate data basis for the subsequent determination of user tags. Based on the analysis of usage frequency and effective usage time, it can accurately identify the historical key applications that users use most frequently, which is an important basis for constructing user tags.

[0110] Specifically, in an embodiment of the present invention, determining historical key applications based on the historical usage frequencies and the historical effective usage durations corresponding to the historical applications includes:

[0111] When the historical usage frequency is greater than the preset frequency and the historical effective usage duration is greater than the preset usage duration, the corresponding historical application is regarded as a historical key application.

[0112] Specifically, the preset frequency in the embodiment of the present invention is 1 / 3 of the total usage frequencies corresponding to the plurality of historical applications;

[0113] The preset usage time is 1 / 3 of the historical period;

[0114] The historical period is 1 day.

[0115] Specifically, the embodiments of the present invention filter out applications with high usage frequency through preset frequency, and filter out applications with long usage time through preset usage time, and can quickly filter out applications that are used frequently and for a long time during user use. These applications are more likely to meet the core needs of users or key applications in high-frequency usage scenarios. By comprehensively considering historical usage frequency and historical effective usage time, misjudgment caused by relying on a single indicator (such as only looking at frequency or only looking at time) is avoided, and it is ensured that the filtered applications meet both the user's usage frequency requirements and the usage time requirements. It can more accurately identify the user's real needs and key applications in high-frequency usage scenarios, and avoid misjudging the importance of certain applications due to accidental or short-term use. Historical applications that meet the conditions are marked as historical key applications, which helps the system better understand the user's core needs, optimize the application recommendation algorithm, and prioritize the allocation of resources to key applications that users really need, thereby improving user experience and system resource utilization efficiency. Historical key applications are filtered out through preset frequency and preset usage time, and those with low usage frequency or short usage time are removed. Applications, avoid these applications from ineffectively occupying system resources, reduce the interference of ineffective applications, improve the system's operating efficiency, ensure that the system can more efficiently process and optimize the performance of key applications, dynamically screen historical key applications based on historical usage frequency and historical effective usage time, and dynamically adjust the identification of key applications according to changes in user usage behavior to adapt to changes in user needs, enhance the system's adaptability and dynamic adjustment capabilities, and be able to respond to changes in user needs in a timely manner, continuously optimize application recommendations and resource allocation strategies, and improve user experience by accurately identifying and optimizing key applications, ensuring that users can quickly find and efficiently use the applications they really need during use, improve user satisfaction and dependence on the system, enhance the stickiness between users and the system, and promote long-term user use and retention. Through the automated screening mechanism with preset frequency and preset usage time, the workload of manual screening and analysis is reduced, the subjectivity and inconsistency of manual operations are avoided, and the system's automated processing efficiency is improved. It is especially suitable for processing scenarios of large-scale application data, reducing labor costs and time costs.

[0116] Specifically, the steps for counting the historical usage frequency and several historical effective usage durations of any historical application within a historical period include:

[0117] determining the historical usage frequency based on the number of start timestamps of any of the historical applications;

[0118] Determine the historical usage duration based on any start timestamp and its corresponding end timestamp corresponding to any historical application;

[0119] Determine the historical interaction duration in any of the historical usage durations by collecting the user's time period behavior data in the historical usage duration corresponding to the historical application;

[0120] The historical effective duration is determined based on any of the historical usage durations and the corresponding historical interaction durations.

[0121] Specifically, the embodiment of the present invention determines the historical usage frequency by counting the number of start timestamps of historical applications, and calculates the historical usage time based on the start timestamp and end timestamp. At the same time, it collects the user's time period behavior data within the historical usage time to judge the interaction time, and determines the historical effective time accordingly. It plays a role in accurately quantifying the application usage, and realizes the accurate capture and evaluation of the user's real usage behavior, thereby providing reliable data support for the subsequent screening of key applications and optimization of resource allocation, and improving the system's analysis accuracy of user behavior and the intelligence level of application management.

[0122] Specifically, in an embodiment of the present invention, determining the historical effective duration based on any of the historical usage durations and the corresponding historical interaction durations includes:

[0123] Calculating the actual difference between the historical usage duration and the historical interaction duration;

[0124] Comparing the actual difference with the preset difference, and when the actual difference is greater than or equal to the preset difference, taking the historical interaction duration as the effective usage duration;

[0125] When the actual difference is less than the preset difference, the historical usage duration is used as the effective usage duration.

[0126] Specifically, the preset difference in the embodiment of the present invention is 1 / 10 of the historical usage time.

[0127] Specifically, the embodiment of the present invention accurately distinguishes between active user interaction and passive occupancy time by calculating the actual difference between historical usage time and historical interaction time, and comparing it with the preset difference. When the actual difference is greater than or equal to the preset difference, it means that the user has less interaction in the application and may be in a passive or inactive state. At this time, using the historical interaction time as the effective usage time can avoid the inflated effective time caused by a long period of no interaction; and when the actual difference is less than the preset difference, it means that the user interacts more frequently in the application. At this time, using the historical usage time as the effective usage time can more comprehensively reflect the user's actual usage. This method realizes the accurate evaluation of the user's actual usage behavior, provides more reliable data support for subsequent analysis and decision-making based on effective usage time, and improves the accuracy of the system's user behavior analysis and the rationality of resource allocation.

[0128] Specifically, the step of determining the historical interaction duration in any of the historical usage durations by collecting the user's time period behavior data in the historical usage durations corresponding to the historical application includes:

[0129] Collect screen click events, sliding events, and input events during the user's historical application use;

[0130] performing a time series analysis on the plurality of screen click events, the sliding events, and the input events, and determining the user's activity level in each time period based on the sequence analysis results;

[0131] The user historical interaction duration is determined based on a number of the activity levels.

[0132] Specifically, in an embodiment of the present invention, collecting the user's behavior data during the use of historical applications can be achieved through an embedded SDK (Software Development Kit), which can record the user's behavior events in real time and send them to the server for analysis.

[0133] Specifically, the embodiment of the present invention collects screen click events, sliding events, and input events during the user's historical application use, performs time series analysis on these events, and then determines the user's activity level in each time period, and based on this, determines the user's historical interaction duration. The role of finely capturing user behavior can accurately distinguish between the user's active interaction period and inactive period in the application. Through this precise behavioral analysis, a deep quantification of the user's actual usage behavior is achieved, providing more accurate data support for the subsequent analysis of the user's actual use of the application, thereby improving the system's ability to understand user behavior and the accuracy of evaluating the application's usage effect.

[0134] Specifically, the step of performing time series analysis on the screen click events, the sliding events, and the input events in the embodiment of the present invention includes:

[0135] sorting the screen click events, the sliding events, and the input events according to timestamps;

[0136] Divide the historical usage duration into several time periods;

[0137] Counting the number of the screen click events, the sliding events, and the input events in several time periods respectively, and determining the activity levels of the several time periods based on the statistical results;

[0138] When the activity level is greater than a preset activity level threshold, this time period is marked as a valid time period, and several valid time periods are summed to obtain the historical interaction duration.

[0139] Specifically, in an embodiment of the present invention, the activity level of several time periods is determined based on the result of quantity statistics, and the total number of the screen click events, the sliding events, and the input events may be used as the activity level.

[0140] Specifically, the activity level threshold in this embodiment of the present invention is 3.

[0141] Specifically, the embodiment of the present invention performs time series analysis on screen click events, sliding events, and input events, sorts them by timestamp, divides them into time periods, counts the number of events in each time period to determine the level of activity, and marks time periods with activity levels above a threshold as valid time periods. Finally, the historical interaction duration is obtained by summing the total. This structured processing of user behavior data and quantification of user activity levels allows for precise identification of users' effective usage periods within an application, thereby enabling accurate assessment of users' real-world interaction behaviors. This provides more reliable data support for subsequent analysis of user behavior patterns and application usage effects, improving the system's accuracy in analyzing user behavior and the intelligence level of application management.

[0142] Specifically, the step of determining a plurality of initial associated applications based on matching the user tag and the preset application library includes:

[0143] Calculating similarities between the user tag and a plurality of the initial tags to obtain a plurality of actual similarities;

[0144] The actual similarities are compared with the preset similarities, and based on the comparison results, a number of associated tags are determined, thereby determining the initial associated applications.

[0145] Specifically, in the embodiment of the present invention, the similarity between the user tag and the plurality of initial tags may be calculated by using a cosine similarity algorithm.

[0146] Specifically, the preset similarity in the embodiment of the present invention is 4 / 5.

[0147] Specifically, the embodiments of the present invention quantify the degree of match between user needs and application features by calculating the similarity between user tags and initial tags and comparing the actual similarity with a preset similarity. This allows for accurate screening of initially associated applications that are highly relevant to the user tags, avoiding inaccurate results caused by subjective judgment or simple matching. This approach achieves efficient and accurate matching of user needs and applications, providing users with application recommendations that better meet their personalized needs while enhancing the intelligence level of the application recommendation system and user experience.

[0148] Specifically, the embodiment of the present invention determines several associated tags based on the comparison result, including:

[0149] When the actual similarity is greater than or equal to the preset similarity, the corresponding initial label is used as the associated label;

[0150] When the actual similarity is less than the preset similarity, it is not used as the associated label.

[0151] Specifically, the embodiment of the present invention screens tags that are highly relevant to user needs by comparing the actual similarity with the preset similarity. When the actual similarity is greater than or equal to the preset similarity, the corresponding initial tag is determined as an associated tag, thereby ensuring that the filtered tags can accurately reflect user needs; and when the actual similarity is less than the preset similarity, it is not used as an associated tag to avoid introducing irrelevant or weakly related tags that interfere with subsequent application recommendations. This method realizes the precise screening and filtering of tags, improves the accuracy of matching tags with user needs, and thus improves the accuracy and reliability of application recommendations, enhancing the user experience and the intelligence level of the system.

[0152] Specifically, the steps for analyzing several parameter curves include:

[0153] Collecting the real-time response time and real-time CPU usage of any of the initially associated applications in real time at a preset frequency;

[0154] Drawing a time variation curve based on the real-time response times;

[0155] Drawing a usage change curve based on the real-time CPU usage;

[0156] The time variation curve and the usage rate variation curve are analyzed to determine the abnormal time period and the abnormal usage rate period, and the total abnormal period is determined based on the abnormal time period and the abnormal usage rate period.

[0157] Specifically, the steps of determining the abnormal time period and the abnormal usage period in the embodiment of the present invention include:

[0158] Marking the curve segment corresponding to the real-time response time greater than or equal to the preset response time in the time variation curve as the time abnormality period;

[0159] The curve segment corresponding to the utilization rate change curve where the real-time CPU utilization rate is greater than or equal to the preset CPU utilization rate is marked as the utilization rate abnormal period.

[0160] Specifically, the preset response time in the embodiment of the present invention is 200 milliseconds, and the response time of the web application should be less than 200-500 milliseconds to ensure a good user experience;

[0161] The preset CPU usage can be set according to the total CPU capacity of the system and the importance of the application. For example, if the system has 8 CPU cores and it is desired that an application occupy resources of no more than 2 cores, the preset CPU usage can be 25%.

[0162] Specifically, the embodiment of the present invention is

[0163] Specifically, the steps of screening a number of initial related applications based on a number of curve analysis results include:

[0164] When the abnormal total period is less than or equal to a preset period, using the initial associated application as the adjusted associated application;

[0165] When the total abnormal period is greater than the preset period, the initial associated application is removed.

[0166] Specifically, the preset time period in the embodiment of the present invention is 2 / 3 of the total time period corresponding to the time variation curve.

[0167] Specifically, the present invention dynamically monitors application performance by collecting the real-time response time and real-time CPU usage of initially associated applications and plotting time and usage change curves, respectively. By marking curve segments exceeding preset thresholds as abnormal periods, it can accurately locate performance issues in applications related to response time and CPU usage. Ultimately, the total abnormal period is determined based on the abnormal time and usage periods, enabling rapid identification and accurate assessment of application performance anomalies. This provides strong support for timely optimization and adjustment of application performance, improving system stability and reliability.

[0168] See also Figure 4 As shown, an embodiment of the present invention further provides a system for an application operation management method based on big data, the system comprising:

[0169] An application tag determination module 10 is configured to obtain registration information corresponding to a plurality of registered applications, analyze the registration information, and construct corresponding initial tags for the plurality of registered applications based on the analysis results;

[0170] A construction module 20, connected to the application tag determination module 10, is used to construct a preset application library based on a number of registered applications and their corresponding initial tags;

[0171] A user tag determination module 30 is used to collect user historical behavior data, analyze the historical behavior data, and determine user tags based on the data analysis results;

[0172] A matching module 40 is configured to determine a plurality of initial associated applications based on the user tag and the preset application library;

[0173] an adjustment module 50 connected to the matching module 40 and configured to monitor usage parameters of the plurality of initially associated applications in real time at a preset frequency, construct a plurality of usage parameter curves based on the plurality of usage parameters, analyze the plurality of usage parameter curves, and screen the plurality of initially associated applications based on the analysis results of the plurality of curves to obtain a plurality of adjusted associated applications;

[0174] The updating module 60 is connected to the adjusting module 50 and is used to collect actual behavior data of several adjustment-related applications, analyze the actual behavior data and the historical behavior data to update the user tag, and then update the adjustment-related applications to obtain target-related applications.

[0175] Specifically, the system of the application operation management method based on big data provided by the embodiment of the present invention can execute the above-mentioned application operation management method based on big data to achieve the same technical effect, which will not be repeated here.

[0176] Thus far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art may make equivalent changes or substitutions to the relevant technical features, and the technical solutions after such changes or substitutions will fall within the scope of protection of the present invention.

[0177] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that the present invention is susceptible to various modifications and variations. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A big data-based application operation management method, characterized in that: include: Obtain registration information corresponding to a number of registered applications, analyze the registration information, and construct corresponding initial tags for the number of registered applications based on the analysis results; Build a preset application library based on several registered applications and their corresponding initial tags; Collect user historical behavior data, analyze the historical behavior data, and determine user tags based on the data analysis results; Determine a number of initial associated applications based on matching the user tag and the preset application library; monitoring usage parameters of the plurality of initially associated applications in real time at a preset frequency, constructing a plurality of usage parameter curves based on the plurality of usage parameters, analyzing the plurality of usage parameter curves, and screening the plurality of initially associated applications based on the analysis results of the plurality of curves to obtain a plurality of adjusted associated applications; Actual behavior data of a number of adjustment-related applications are collected, and analysis is performed based on the actual behavior data and the historical behavior data to update the user tag, thereby updating the adjustment-related applications and obtaining target-related applications.

2. The application operation management method based on big data according to claim 1, characterized in that: The step of constructing corresponding initial tags for a number of registered applications based on the analysis results includes: Perform word segmentation processing on any of the registration information to obtain a number of word segments; Calculating actual importance levels corresponding to a number of the segmented words, determining the actual importance levels, and determining a number of keywords based on the determination results; The initial tag is determined based on a number of the keywords.

3. The application operation management method based on big data according to claim 2, characterized in that: The step of calculating the actual importance corresponding to the plurality of segmentations includes: Counting actual frequencies corresponding to several of the segmented words; Determining actual distribution dispersions corresponding to a number of the segmentations; The actual importance is calculated based on the actual frequency and the actual distribution dispersion.

4. The application operation management method based on big data according to claim 3 is characterized in that: The step of analyzing the historical behavior data includes: Statistics on the historical usage frequency and certain historical effective usage duration of any historical application within the historical period; Determining historical key applications based on the historical usage frequencies and the historical effective usage durations corresponding to the historical applications; The historical key applications are analyzed to determine the user tags.

5. The application operation management method based on big data according to claim 4, characterized in that: The steps for calculating the historical usage frequency and several historical effective usage durations of any historical application within a historical period include: determining the historical usage frequency based on the number of start timestamps of any of the historical applications; Determine the historical usage duration based on any start timestamp and its corresponding end timestamp corresponding to any historical application; Determine the historical interaction duration in any of the historical usage durations by collecting the user's time period behavior data in the historical usage duration corresponding to the historical application; The historical effective duration is determined based on any of the historical usage durations and the corresponding historical interaction durations.

6. The application operation management method based on big data according to claim 5, characterized in that: The step of determining the historical interaction duration in any of the historical usage durations by collecting the user's time period behavior data in the historical usage durations corresponding to the historical application includes: Collect screen click events, sliding events, and input events during the user's historical application use; performing a time series analysis on the plurality of screen click events, the sliding events, and the input events, and determining the user's activity level in each time period based on the sequence analysis results; The user historical interaction duration is determined based on a number of the activity levels.

7. The application operation management method based on big data according to claim 6, characterized in that: The step of determining a plurality of initial associated applications based on matching the user tag and the preset application library includes: Calculating similarities between the user tag and a plurality of the initial tags to obtain a plurality of actual similarities; The actual similarities are compared with the preset similarities, and based on the comparison results, a number of associated tags are determined, thereby determining the initial associated applications.

8. The application operation management method based on big data according to claim 7, characterized in that: The steps for analyzing several parametric curves include: Collecting the real-time response time and real-time CPU usage of any of the initially associated applications in real time at a preset frequency; Drawing a time variation curve based on the real-time response times; Drawing a usage change curve based on the real-time CPU usage; The time variation curve and the usage rate variation curve are analyzed to determine the abnormal time period and the abnormal usage rate period, and the total abnormal period is determined based on the abnormal time period and the abnormal usage rate period.

9. The application operation management method based on big data according to claim 8, characterized in that: The steps of screening a number of initial related applications based on a number of curve analysis results include: When the abnormal total period is less than or equal to a preset period, using the initial associated application as the adjusted associated application; When the total abnormal period is greater than the preset period, the initial associated application is removed.

10. A system applied to the big data-based application operation management method according to any one of claims 1 to 9, characterized in that: include: An application tag determination module is used to obtain registration information corresponding to a number of registered applications, analyze the registration information, and construct corresponding initial tags for the number of registered applications based on the analysis results; a construction module, connected to the application tag determination module, for constructing a preset application library based on a number of registered applications and their corresponding initial tags; User tag determination module, used to collect user historical behavior data, analyze the historical behavior data, and determine user tags based on the data analysis results; A matching module, configured to determine a plurality of initial associated applications based on matching between the user tag and the preset application library; an adjustment module, connected to the matching module, configured to monitor usage parameters of the plurality of initially associated applications in real time at a preset frequency, construct a plurality of usage parameter curves based on the plurality of usage parameters, analyze the plurality of usage parameter curves, and screen the plurality of initially associated applications based on the analysis results of the plurality of curves to obtain a plurality of adjusted associated applications; An update module is connected to the adjustment module and is used to collect actual behavior data of several adjustment-related applications, analyze the actual behavior data and the historical behavior data to update the user tag, and then update the adjustment-related applications to obtain target-related applications.

Images