Team management application OA intelligent processing system and method

Abstract

The invention discloses a team management application OA intelligent processing system and method, and relates to the technical field of enterprise team digital management, and the system comprises a data collection layer, a data storage layer, a data analysis layer, and an application layer. According to the method, team management optimization is realized by sequentially solving the problems of system adaptability of skill dynamic change and task adaptation and employee behavior habit change and task allocation collaboration based on biological rhythm and group psychology; firstly, based on an operation sequence similarity algorithm and the like, the real-time problem of skill dynamic change and task adaptation is solved, skill change is accurately captured, and matching accuracy is improved; based on the result, a deep reinforcement learning model and the like are used for solving the system adaptability problem, so that the system quickly adapts to behavior mode change; and finally, according to the previous two problem achievements, task allocation collaboration is achieved through a multi-dimensional analysis method, employee potential is exerted, multi-aspect deep fusion optimization is achieved, and extra steps and software and hardware investment are not needed in the whole process.

Description

Technical Field

[0001] The present invention relates to the technical field of enterprise team digital management, and specifically to an OA intelligent processing system and method for team management application. Background Art

[0002] In the digital management of enterprise teams, there are many problems that need to be solved urgently.

[0003] First, there is a real-time problem with the dynamic changes in team members' skills and task adaptation. In a rapidly developing business environment, members' skills are constantly changing through training and practice. For example, marketers have mastered data analysis skills, but the OA system is based on static skill assessments and preset task templates. It cannot perceive such changes in real time and still assigns tasks according to the old skill labels, resulting in the inability to give full play to new skills. At the same time, the market environment prompts rapid adjustments in business direction. For example, the transition from offline to online marketing requires new digital marketing skills, but the OA system has difficulty quickly matching new skill requirements with members' existing skills and achieving reasonable task redistribution. Conventional systems lack the ability to perceive skill changes in real time, dynamically analyze business needs, and quickly adjust task assignments. Simply expanding existing technologies cannot solve this problem.

[0004] Secondly, changes in employee behavior habits have resulted in insufficient system adaptability. As companies introduce new work concepts, tools, or business model changes, such as the implementation of agile project management models, employee work rhythms and task collaboration methods have changed significantly. However, OA systems are designed based on previous behavior patterns and business processes, and traditional upgrades are mostly based on clear business demand changes or technology updates. They do not take into account implicit and dynamically changing factors such as employee behavior habits, lack intelligent learning and adaptive capabilities, and are unable to automatically identify changes in behavior patterns and dynamically adjust system functions and process settings. As a result, under agile management, the modules used for traditional project progress tracking cannot accurately reflect the actual progress of the project, affecting managers' decision-making and project advancement efficiency.

[0005] Furthermore, the current system completely ignores biorhythms and group psychological factors when assigning tasks. Each person has a unique biorhythm, and their energy, creativity, and mental agility are different at different times. There are group psychological effects in the team. For example, morale and collaborative atmosphere in the project breakthrough phase affect work efficiency. However, the existing system assigns tasks based only on hard indicators such as employee skills and workload, and arranges highly creative tasks during the low biorhythm period of employees, or assigns tasks in a conventional manner when team morale is low. In addition, the biorhythm cycles of different employees are different, and it is extremely complex to coordinate individual differences with team task requirements and group psychological states. The traditional task allocation method is limited to focusing on the work itself, lacks interdisciplinary thinking, and does not comprehensively use multidisciplinary knowledge to develop new task allocation models and algorithms.

[0006] In summary, it is difficult for existing technologies to effectively solve a series of difficult problems, such as the dynamic changes in team members' skills and the real-time adaptation of tasks, the system adaptability caused by changes in employee behavior habits, and the task allocation and coordination based on biological rhythms and group psychology, without adding any steps and any software and hardware. New technical solutions are urgently needed to optimize the team management OA intelligent processing system.

[0007] In view of this, a team management application OA intelligent processing system and method are provided to overcome the above problems. Summary of the invention

[0008] The purpose of the present invention is to provide an OA intelligent processing system and method for team management application to solve the problems raised in the above background technology.

[0009] In order to solve the above technical problems, the present invention provides a team management application OA intelligent processing system, including:

[0010] Data collection layer: Adopting the native data capture mechanism in the event-driven system, embedding code snippets at the key operation nodes of each business module to capture data, and using data mapping tables to map and integrate the operation data in different business systems in a unified format;

[0011] Data storage layer: We use the MySQL relational database and improve the distributed storage architecture. We use a self-developed distributed storage algorithm to perform shard storage according to business type, time and other dimensions, and dynamically adjust the distribution of storage nodes based on access frequency. At the same time, we use a distributed ledger mechanism based on blockchain technology to perform redundant storage and consistency verification of key data.

[0012] Data analysis layer: Build a multimodal data analysis algorithm model that integrates semantic analysis, behavioral pattern recognition, and time series analysis, and use cross-domain knowledge fusion to integrate knowledge such as psychology and management into the algorithm model;

[0013] Application layer: Design an intelligent interactive interface based on user behavior prediction. By analyzing employees' historical operation behaviors and business needs, predict the employees' next possible operations and display relevant functions and data in advance. At the same time, introduce emotional design concepts to adjust the interface style and prompt information according to the employees' work status and emotional feedback.

[0014] The team management applies OA intelligent processing method, including the following steps:

[0015] Step 1: Basic data collection and integration:

[0016] Adopt a recording method based on operation step decomposition to collect employee operation logs, and use data watermark technology to mark the operation log data;

[0017] The feedback information processing system based on natural language understanding and knowledge graph collects employee feedback information and uses automatic classification and priority sorting algorithms to classify and sort the feedback information;

[0018] Use sentiment analysis combined with topic modeling and social network dynamic modeling technology to collect project discussion area information;

[0019] Step 2:

[0020] Skill change perception: Adopt a skill evaluation algorithm based on the similarity of operation sequences, compare the operation sequences of employees at different times to calculate the similarity, and introduce a dynamic weight evaluation mechanism for operation frequency and time consumption;

[0021] Task adaptation and adjustment: Using a deep task requirement decomposition algorithm based on semantic networks and business rules, the task requirements are converted into semantic networks and each element is refined and weighted, thus building a two-way matching model between employee skills and task requirements;

[0022] Step 3:

[0023] Identification of behavioral pattern changes: Based on deep reinforcement learning, a dynamic behavioral pattern recognition model is constructed, and a multi-dimensional behavioral data fusion analysis method is adopted to integrate employees' operation behavior data, communication behavior data, task collaboration data, etc. for analysis;

[0024] System adaptive adjustment: Dynamically reconstruct related modules based on the dynamic module reconstruction method of microservice architecture, use the interface adaptive adjustment algorithm based on user portraits to generate personalized user portraits and adjust the interface layout and function display according to the behavior patterns and work needs of employees;

[0025] Step 4:

[0026] Biorhythm analysis: Propose a biorhythm analysis algorithm based on time series clustering and cycle prediction, perform clustering analysis and cycle prediction on employees' long-term operation time series data, and introduce correlation analysis between environmental factors and biorhythms;

[0027] Group psychological assessment: Build a group psychological assessment model based on emotional communication and social network structure, adopt a group psychological analysis method based on multi-source data fusion, and integrate employee behavior data, communication data, performance data, etc. with project discussion area information for analysis;

[0028] Task allocation coordination: formulate task allocation strategies based on the task allocation of biological rhythms and group psychology, and establish a dynamic adjustment mechanism to automatically adjust task allocation according to real-time data changes during task execution.

[0029] Furthermore, the data collection layer is based on the event-driven system's native data capture mechanism, and code snippets are implanted in the key operation nodes of each business module.

[0030] Furthermore, the shard storage formula of the self-developed distributed storage algorithm in the data storage layer is:

[0031] S i =Hash(B i ,T i )modN

[0032] Among them, S i Indicates the target server node number for data storage, B i Represents the business type code, T i is the timestamp of data generation, N is the total number of server nodes;

[0033] The formula for dynamically adjusting storage nodes is:

[0034]

[0035] Among them, P j represents the probability of data being stored in the jth server node, F j is the access frequency of data on the jth server node.

[0036] Furthermore, the importance evaluation formula of the automatic classification and priority sorting algorithm in the employee feedback information processing system is:

[0037]

[0038] Where I represents the importance score of the feedback information, M is the number of key factors in the knowledge graph related to the feedback information, and w m is the weight of the mth key factor, K m is to extract the content related to the mth key factor from the feedback information, m It is a function that quantitatively evaluates the extracted content according to the key factor type;

[0039] The formula for evaluating the degree of urgency is:

[0040]

[0041] Where E represents the urgency score of the feedback information, N is the number of judgment factors related to the urgency, and v n is the weight of the nth factor, T n is the time information related to urgency extracted from the feedback information, g n It is a function that quantifies the urgency based on time information;

[0042] The comprehensive priority score formula is:

[0043] P=α×I+β×E

[0044] Wherein, P is the comprehensive priority score of the feedback information, α and β are weight coefficients of importance and urgency respectively, and α+β=1.

[0045] Furthermore, the operation sequence similarity of the skill evaluation algorithm based on operation sequence similarity in skill change perception is calculated as:

[0046]

[0047] Among them, Sim(S 1 ,S 2 ) represents two operation sequences S 1 and S 2 The similarity of

[0048] S 1 =[s 11 ,s 12 ,…,s 1L ],S 2 =[s 21 ,s 22 ,…,s 2L ], L is the length of the shorter of the two sequences; Match(s 1i ,s 2i ) is a matching function; |S 1 | and |S 2 | are the operation sequences S 1 and S 2 Length;

[0049] Comprehensive skill evaluation formula considering operation frequency and time consumption:

[0050]

[0051] Among them, Skill(t) represents the skill level of the employee at time, S t and S t-Δt are the operation sequences at time t and t-Δt respectively; w 1 、w 2 、w 3 is the weight coefficient, and w 1 +w 2 +w 3 =1, indicating the importance of operation sequence similarity, operation frequency, and operation time in skill assessment; o i Indicates the i-th operation, Weight(o i ) is the operation o i The weight, Freq(oi ) is the operation o i The frequency, Time(o i ) is the operation o i The average time taken.

[0052] Furthermore, the task requirement semantic network of the task requirement deep decomposition algorithm based on semantic network and business rules in task adaptation adjustment is constructed as follows:

[0053] G=(V,E)

[0054] Among them, G represents the semantic network of task requirements, V is a set of nodes, and each node v i ∈V represents an element of a task, E is a set of edges, and each edge e ij ∈E represents the element v i and v j the relationship between;

[0055] Factor refinement and weight allocation formula:

[0056]

[0057] Among them, W(v i ) is the node v i The weight of node v is k. i The number of associated nodes, Influence(v j ,v i ) represents node v j For node v i Importance(v j ) is the node v j Importance score.

[0058] Furthermore, the time series clustering formula of the biorhythm analysis algorithm based on time series clustering and cycle prediction in biorhythm analysis is:

[0059]

[0060] Among them, C is the number of cluster categories, S i is the i-th cluster, x j is the jth data point in the time series, μ i is the center of the i-th cluster;

[0061] Cycle prediction algorithm formula:

[0062]

[0063] Where y(t) is the predicted employee's work efficiency or ability performance at time, N is the number of periodic components, and A nis the amplitude of the nth periodic component, f n is the frequency of the nth periodic component, is the phase of the nth periodic component, ∈(t) is the error term.

[0064] Furthermore, the intelligent interactive interface based on user behavior prediction in the application layer analyzes employees' historical operation behavior data and uses machine learning algorithms to build a prediction model to predict the employees' next operation.

[0065] Compared with the prior art, the present invention has the following beneficial effects:

[0066] The project focuses on solving three key problems in team management, and optimizes team management by solving the problems of dynamic skill changes and task adaptation, system adaptability of employee behavior habit changes, and task allocation coordination based on biological rhythms and group psychology:

[0067] 1. Solve the problem of dynamic change of skills and real-time adaptation of tasks

[0068] Solution: Skill change perception adopts an algorithm based on operation sequence similarity and a dynamic weight evaluation mechanism for operation frequency and duration; task adaptation adjustment uses a deep decomposition algorithm for task requirements based on semantic networks and business rules and a two-way matching model.

[0069] The traditional method of evaluating skill changes is one-sided, and the decomposition and matching of task requirements are rough. The new method of multi-dimensional evaluation and comprehensive matching is more in line with reality.

[0070] Accurately capture skill changes, save costs, improve the accuracy of task and skill matching, provide data and technical support for subsequent problem solving, and optimize resource and task allocation.

[0071] 2. Solve system adaptability problems based on technical problem results

[0072] Solution: Behavior pattern change identification adopts a deep reinforcement learning model and multi-dimensional behavior data fusion analysis method; the system adaptively adjusts the dynamic module reconstruction based on the microservice architecture and the interface adaptive adjustment algorithm based on the user portrait.

[0073] Traditional behavior pattern recognition and system adjustment methods have poor adaptability and flexibility, while the new method is self-learning, multi-dimensional and highly flexible.

[0074] Accurately identify changes in behavioral patterns, quickly adapt the system, ensure smooth business processes, improve workflow management flexibility and adaptability, and leverage skill problem-solving results without additional investment.

[0075] 3. Realize task allocation collaboration based on the first two problems

[0076] Solution: Biorhythm analysis proposes an algorithm based on time series clustering and cycle prediction and introduces environmental factor correlation analysis; group psychological assessment is constructed based on emotional communication and social network structure models and adopts multi-source data fusion methods; task allocation collaboration is based on biorhythms and group psychology to coordinate tasks and establish a dynamic adjustment mechanism.

[0077] Traditional analysis and evaluation methods are simple and subjective, while new methods conduct multi-dimensional analysis, comprehensively consider individual and team status, and make dynamic adjustments.

[0078] Accurately determine biorhythms and evaluate group psychology to provide a basis for task allocation, unleash employee potential, improve team collaboration and project management efficiency, and achieve deep integration and collaborative optimization in multiple aspects without additional investment.

[0079] Without adding steps or software or hardware, the three key problems in team management are solved one by one to achieve comprehensive optimization of the team from the individual level to the overall level. BRIEF DESCRIPTION OF THE DRAWINGS

[0080] Figure 1 This is a schematic diagram of the OA intelligent processing system and method for team management application of the present invention. DETAILED DESCRIPTION

[0081] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 creative work are within the scope of protection of the present invention.

[0082] See also Figure 1 , the present invention provides a technical solution:

[0083] See also Figure 1 As shown, an embodiment of the team management application OA intelligent processing system and method:

[0084] 1. System architecture construction

[0085] This OA intelligent processing system adopts a unique and improved B / S architecture, which consists of four layers: data collection, storage, analysis and application. Each layer is both independent and collaborative to achieve efficient operation of the system.

[0086] 1. Data Collection Layer

[0087] Data collection method

[0088] Abandon the traditional data collection method that relies on external engines, and adopt a native data capture mechanism in the system based on event-driven. Code snippets are implanted in the key operation nodes of each business module. When employees perform operations, such as recording customer communications and processing orders in the sales business module, the system automatically triggers data capture and records information such as operation time, object and results in detail.

[0089] Based on the operation log structure of the business system, the data mapping table is used to map and integrate the operation data in different business systems in a unified format. For example, the fields of customer follow-up records in the customer relationship management system are mapped with the data structure preset in the OA system to ensure data consistency and availability.

[0090] Traditional data capture engines are highly versatile but lack specificity, making it difficult to accurately meet the real-time data collection needs of enterprises in complex business scenarios. By implanting codes in key operation nodes within the system, data can be directly obtained from the source to ensure the real-time and integrity of the data. The use of data mapping tables solves the problem of data format differences in different business systems and provides a unified standard data foundation for subsequent data analysis.

[0091] This method ensures the timeliness and integrity of the data, and can accurately and real-timely obtain employee behavior information, providing a solid foundation for in-depth analysis of employee skills and behavior patterns. In terms of resource management, accurate and real-time data collection helps companies rationally allocate human and material resources and avoid waste of resources. In terms of personnel management, it provides sufficient data support for subsequent accurate evaluation of employee skills and improves the scientific nature of personnel management.

[0092] (II) Data Storage Layer

[0093] Storage architecture design

[0094] We selected the MySQL relational database and improved its distributed storage architecture. We used a self-developed distributed storage algorithm to store data in shards according to business type, time and other dimensions. For example, nearly one year of employee operation log data is stored on one group of server nodes, and older data is stored on another group of nodes. The storage node distribution is dynamically adjusted according to the access frequency of the data.

[0095] Distributed storage algorithm formula

[0096] Shard storage formula:

[0097] S i =Hash(B i ,T i )modN

[0098] Among them, S i Indicates the target server node number for data storage, Bi Represents the business type code (for example, sales business code is 001, human resources business code is 002, etc.), T i is the timestamp of the data generation, and N is the total number of server nodes. By performing a hash operation on the business type and timestamp and taking the modulus of the total number of nodes, the data can be evenly distributed to different server nodes.

[0099] Dynamically adjust the storage node formula:

[0100]

[0101] Among them, P j represents the probability of data being stored in the jth server node, F j is the access frequency of data on the jth server node. By calculating the proportion of data access frequency of each node to the total access frequency, the storage probability of data in different nodes is determined, so that the distribution of storage nodes can be dynamically adjusted according to the access frequency.

[0102] Traditional distributed storage algorithms are mostly based on simple hash modulo or average distribution by data volume, and do not fully consider the impact of business type and time factors on data access patterns. This algorithm combines business type and timestamp to perform data sharding, which can make data storage more in line with business characteristics and improve the locality of data reading. For example, business data generated recently is often accessed more frequently. In this way, it can be stored in a specific node to improve access efficiency. The formula for dynamically adjusting the distribution of storage nodes based on access frequency breaks through the limitations of traditional static storage allocation and optimizes the storage layout in real time according to the actual access situation of the data. This is an idea that has not appeared in the prior art. The prior art mostly focuses on the balanced storage of data, while this algorithm focuses on the dynamic characteristics of data access to improve the performance of the overall storage system.

[0103] The distributed ledger mechanism based on the concept of blockchain technology is adopted to perform redundant storage and consistency verification of key data. When writing data, the accuracy and integrity of the data are ensured through the consensus mechanism of multiple nodes to prevent data tampering.

[0104] The amount of enterprise data is growing rapidly, and traditional storage methods are prone to storage bottlenecks and reading performance problems when facing massive data. The self-developed distributed storage algorithm combines business characteristics to perform data sharding, which can effectively improve storage efficiency and reading speed. The introduction of blockchain technology enhances data security and reliability, which is crucial when it comes to the storage of key employee information and business data.

[0105] It realizes efficient storage and fast reading of massive data, providing a stable and reliable data source for the data analysis layer. In workflow management, fast data reading can accelerate the flow of business processes and improve work efficiency. In project management, reliable data storage ensures the security of project-related data and provides guarantee for the smooth progress of the project.

[0106] (III) Data Analysis Layer

[0107] Multimodal data analysis algorithm model

[0108] Construct a multimodal data analysis algorithm model that integrates semantic analysis, behavioral pattern recognition, and time series analysis. For employee operation logs, use semantic analysis technology to understand the meaning of the operation content, combine behavioral pattern recognition to explore the potential rules of operation behavior, and use time series analysis to grasp the trend of operation time.

[0109] For example, when analyzing the operational logs of employees writing market research reports, semantic analysis determines the quality and professionalism of the report content, behavioral pattern recognition discovers whether specific materials are frequently referenced during the writing process, and time series analysis determines the changing trends in writing time and efficiency.

[0110] By integrating knowledge from different fields, we can integrate psychology, management and other knowledge into the algorithm model. For example, based on the theory of learning curve in psychology, when evaluating employee skill improvement, we can consider the stage characteristics of skill learning and more accurately judge the degree of skill mastery.

[0111] Traditional analysis methods are single and it is difficult to comprehensively and deeply analyze employees' multi-dimensional data. The integration of multiple analysis technologies can tap into the value of data from different angles and more accurately grasp the changes in employees' skills and behavioral habits. The integration of cross-domain knowledge breaks the limitations of computer science and makes the analysis more in line with the behavioral and psychological characteristics of employees in actual work scenarios.

[0112] Compared with traditional analysis methods, this model has higher accuracy and adaptability, and can more accurately explore the potential patterns of employee skill changes and behavioral habits. In personnel management, accurate skill assessment helps companies make talent training and promotion decisions. In task allocation, it provides a scientific basis for reasonable task allocation and improves the accuracy of task allocation.

[0113] (IV) Application Layer

[0114] Interactive interface design

[0115] Design an intelligent interactive interface based on user behavior prediction. By analyzing employees' historical operation behaviors and business needs, predict the employees' next possible operation, and display relevant functions and data on the interface in advance. For example, when an employee enters the project management module, the system predicts the project progress data that may be viewed or the project report that needs to be submitted based on their past operation habits, and displays the relevant function entrances and data in a prominent position in advance.

[0116] Introducing emotional design concepts, the interface style and prompt information are adjusted according to the work status and emotional feedback of employees. For example, when the system detects that employees are continuously handling complex tasks and their efficiency is reduced, the interface automatically switches to a soothing tone and provides encouraging prompt information to enhance employees' work enthusiasm.

[0117] Traditional interactive interfaces lack intelligent prediction of user behavior, and users need to spend time looking for functions when operating. Design based on user behavior prediction can reduce user operation steps and improve operation efficiency. The emotional design concept takes into account the impact of employee work status on work efficiency. By adjusting the interface style and prompt information, it can improve employees' work experience and enthusiasm.

[0118] Improve the user experience and ensure that employees and managers can use system functions conveniently. In terms of resource management, the efficient operation interface can reduce the time and resources wasted by employees due to cumbersome operations and improve resource utilization efficiency. In workflow management, convenient operations help speed up the execution of business processes and improve overall work efficiency.

[0119] 2. Basic Data Collection and Integration

[0120] (I) Collection of employee operation logs

[0121] Operation log recording mode

[0122] Adopt a recording method based on the decomposition of operation steps. For each business operation, the system decomposes it into multiple atomic operation steps for recording. For example, in the order processing operation, the detailed information of each step such as order information entry, inventory query, logistics information association, etc. is recorded, including operation time, input data, output results, etc.

[0123] Use data watermarking technology to mark operation log data to ensure data integrity and traceability. When data is generated, a watermark containing information such as operation time and employee identification is embedded. In subsequent analysis and use, the authenticity and source of the data can be verified through the watermark.

[0124] Traditional operation log recording methods are too general and cannot accurately reflect the operation details and skill application of employees. Operation step decomposition records can provide more detailed operation information and help to deeply analyze employee skills. Data watermark technology can effectively prevent data from being tampered with and ensure data reliability, which is especially important when data is used as a basis for employee skill evaluation and business decision-making.

[0125] It can directly reflect the work behavior and skill application of employees, and provide rich data support for the perception of skill changes and the recognition of behavior patterns. In personnel management, detailed operation logs help to accurately evaluate the work ability and contribution of employees. In task allocation, it provides strong data support for judging whether employees are suitable for specific tasks.

[0126] (II) Collection of employee feedback information

[0127] Feedback information processing mechanism

[0128] Develop a feedback information processing system based on natural language understanding and knowledge graph. Employees input feedback content through free text, and the system first uses natural language understanding technology to extract key information, such as descriptions of skill improvement and key points of work experience. Then, this information is associated with the knowledge graph in the system, for example, new skills are associated with related business areas, project types, etc., to determine their position and application scenarios in the enterprise business system.

[0129] Automatic classification and prioritization algorithms are used to classify and prioritize feedback information according to the importance and urgency of the feedback content. For example, urgent feedback involving business process optimization is marked as high priority, while general suggestions are marked as low priority, so that the system and managers can prioritize important feedback.

[0130] Feedback information automatic classification and priority sorting algorithm formula

[0131] Importance assessment formula:

[0132]

[0133] Where I represents the importance score of the feedback information, M is the number of key factors in the knowledge graph related to the feedback information, and w m is the weight of the mth key factor (for example, factors related to business process optimization have higher weights than general recommendation factors), K m is to extract the content related to the mth key factor from the feedback information, m is a function that quantitatively evaluates the extracted content according to the key factor type. For example, for business process optimization feedback, f m This may be quantified based on the number of process steps mentioned in the feedback and the potential scope of impact.

[0134] Urgency assessment formula:

[0135]

[0136] Where E represents the urgency score of the feedback information, N is the number of judgment factors related to the urgency, and v n is the weight of the nth factor (e.g., factors involving business interruption risk have high weights), T n is the time information related to the urgency extracted from the feedback information (such as the time point when the business is expected to be affected), n It is a function that quantifies the degree of urgency based on time information. For example, if the feedback mentions that a certain business will have serious problems within 24 hours, g n This time information is converted into a higher urgency score.

[0137] Comprehensive priority score formula:

[0138] P=α×I+β×E

[0139] Among them, P is the comprehensive priority score of the feedback information, α and β are the weight coefficients of importance and urgency respectively, and α+β = 1. According to the management strategies of different enterprises, the values ​​of and can be adjusted. For example, enterprises that pay more attention to the handling of urgent issues can appropriately increase the values.

[0140] Traditional feedback information classification and sorting mostly rely on manual experience or simple keyword matching, which cannot accurately and comprehensively evaluate the importance and urgency of feedback content. This algorithm quantitatively evaluates feedback by constructing a multi-factor comprehensive evaluation model, combining key factors in the knowledge graph and key information such as time in the feedback information. For example, in the importance assessment, considering the different degrees of impact of different business factors in the knowledge graph on the enterprise, a comprehensive calculation is performed through weights and quantification functions, which is not adopted by traditional methods. In the urgency assessment, the time information is quantified and included in the calculation to more accurately reflect the urgent nature of the feedback. This multi-dimensional, multi-factor quantitative evaluation method breaks the limitations of traditional simple classification and sorting, and provides an innovative method for the efficient processing of feedback information.

[0141] Traditional employee feedback processing methods are inefficient and difficult to quickly and accurately extract useful information and make effective use of. The combination of natural language understanding and knowledge graph technology can better understand employee feedback content and integrate it into the enterprise business system. Automatic classification and priority sorting algorithms can improve the efficiency of feedback processing and ensure that important feedback is processed in a timely manner.

[0142] Make full use of employees' subjective information to provide a more comprehensive perspective for system decision-making. In personnel management, timely handling of employee feedback helps improve employee satisfaction and loyalty. In workflow management, employee feedback on business processes can help companies optimize processes in a timely manner and improve work efficiency.

[0143] (III) Information collection in project discussion area

[0144] Information mining technology

[0145] The project discussion area information is mined by combining sentiment analysis with topic modeling. Sentiment analysis is used to determine the sentiment tendency of speech content, such as positive, negative or neutral. Topic modeling is used to extract discussion topics, such as project technical problem discussions, team collaboration issues, etc.

[0146] Using social network dynamic modeling technology, we can build a social network relationship diagram of team members in the discussion area in real time. By analyzing the speech interactions, likes and comments between members, we can determine the influence of members in the team and the information dissemination path. For example, members who frequently receive likes and replies are regarded as opinion leaders in the team, and their speeches may have a greater impact on team decisions and atmosphere.

[0147] Traditional analysis of project discussion forum information is relatively simple and cannot deeply explore the team collaboration status and group psychology information contained therein. The combination of sentiment analysis and topic modeling can fully understand the discussion forum information from two dimensions: content and sentiment. Social network dynamic modeling technology can intuitively display the interactive relationship between team members and provide a powerful tool for analyzing group psychology.

[0148] Provide a strong basis for task allocation based on group psychology. In project management, understanding the collaboration status and group psychology of team members can help managers to reasonably allocate tasks and improve team collaboration efficiency. In personnel management, discovering potential leaders in the team can provide reference for team formation and management.

[0149] 3. Solve the real-time problem of dynamic changes in team members’ skills and task adaptation

[0150] 1. Perception of skill changes

[0151] Operation log analysis method

[0152] A skill evaluation algorithm based on the similarity of operation sequences is proposed. The operation sequences of employees at different times are compared and their similarity is calculated. For example, for data analysis skills, the operation sequence of employees using data analysis tools at different time periods is compared. If the similarity of the operation sequence increases, it means that their mastery of data analysis skills is more proficient and stable.

[0153] Skill evaluation algorithm formula based on operation sequence similarity

[0154] Operation sequence similarity calculation:

[0155]

[0156] Among them, Sim(S 1 ,S 2 ) represents two operation sequences S 1 and S 2 The similarity of

[0157] S 1 =[s 11 ,s 12 ,…,s 1L ],S 2 =[s 21 ,s 22 ,…,s 2L ], L is the length of the shorter of the two sequences. Match(s 1i ,s 2i ) is a matching function, if the operation s 1i and 2i If they are the same or belong to the same semantic category (for example, in data analysis, "opening a data file" and "importing data" belong to the same semantic category), it returns 1, otherwise it returns 0. 1 | and |S 2 | are the operation sequences S 1 and S 2 Length.

[0158] Comprehensive skill evaluation formula considering operation frequency and time consumption:

[0159]

[0160] Among them, Skill(t) represents the skill level of the employee at time, S t and S t-Δt are the operation sequences at time t and t-Δt (Δt is a time interval, such as a week or a month). 1 、w 2 、w 3 is the weight coefficient, and w 1 +w 2 +w 3 =1, indicating the importance of operation sequence similarity, operation frequency, and operation time in skill assessment. i Indicates the i-th operation, Weight(o i ) is the operation o i The weight of the operation (determined by the complexity and importance of the operation, such as complex data mining operations with high weights and simple data cleaning operations with low weights), Freq(oi ) is the operation o i The frequency, Time(o i ) is the operation o i The average time taken.

[0161] Traditional skill assessment algorithms often only focus on the results of operations or the number of simple operations, and are unable to assess the degree of skill mastery from the perspective of the consistency and logic of operations. The operation sequence similarity calculation of this algorithm breaks through the limitation of traditional skill assessment only from the results, and measures the stability and proficiency of skills from the perspective of the operation process. For example, in data analysis, even if the final analysis results are the same, the optimization of the sequence of operation steps may reflect the employee's in-depth understanding of data analysis skills. The formula that comprehensively considers the frequency and time consumption of operations innovatively incorporates operation characteristics into skill assessment. Traditional methods do not distinguish between the complexity and importance of different operations. This algorithm sets weights to make the assessment more in line with the actual skill improvement process. Different types of operations contribute differently to skill assessment. This method can more comprehensively and accurately assess the skill level of employees.

[0162] Introduce a dynamic weight evaluation mechanism for operation frequency and time consumption. When evaluating skills, the weights of operation frequency and time consumption are dynamically adjusted according to the complexity and importance of the operation. For example, for complex data mining operations, the weights of operation accuracy and time consumption are relatively high; while for simple data cleaning operations, the weight of operation frequency is relatively high.

[0163] Traditional analysis methods only analyze operation logs from a single dimension, making it difficult to comprehensively and accurately evaluate employee skill changes. The operation sequence similarity algorithm evaluates skill mastery from the perspective of operation coherence and logic, which is more in line with the actual process of skill improvement. The dynamic weight evaluation mechanism takes into account the characteristics of different operations, making the evaluation results more scientific and accurate.

[0164] It can accurately capture the dynamic changes of employee skills without relying on additional skill tests, saving time and labor costs while ensuring the authenticity and real-time nature of the data. In resource management, accurate skill assessment helps companies rationally allocate human resources and assign the right talents to the right positions. In task allocation, it provides an accurate basis for timely adjustment of task allocation and improves the rationality of task allocation.

[0165] 2. Task Adaptation and Adjustment

[0166] Task requirement decomposition and matching algorithm

[0167] Adopt a deep task requirement decomposition algorithm based on semantic networks and business rules. Transform task requirements into semantic networks, where nodes represent the elements of the task, such as data processing, result presentation, etc., and edges represent the relationship between elements. According to business rules, refine and weight each element. For example, for the data analysis task of an online marketing project, subdivide the data processing element into sub-elements such as data cleaning and data mining, and assign weights to each sub-element according to the project goals.

[0168] Algorithm formula for deep decomposition of task requirements based on semantic network and business rules

[0169] Task requirements: semantic network construction:

[0170] G=(V,E)

[0171] Among them, G represents the semantic network of task requirements, V is a set of nodes, and each node v i ∈V represents an element of a task (such as data processing, result presentation, etc.), E is a set of edges, and each edge e ij ∈E represents the element v i and v j relationships between data and results (e.g., there may be an “influence” relationship between data processing and result presentation).

[0172] Factor refinement and weight allocation formula:

[0173]

[0174] Among them, W(v i ) is the node v i (i.e., the weight of the task element), k is the weight of the node v i The number of associated nodes, Influence(v j ,v i ) represents node v j For node v i The degree of influence of data cleaning on data mining (for example, in data analysis tasks, the degree of influence of data cleaning on data mining can be determined by business rules), Importance (v j ) is the node v j The importance score of the data (determined according to business rules and project goals. For example, for a project with the goal of accurate analysis, the importance score of nodes related to data accuracy is high).

[0175] Traditional task requirement decomposition methods are simple and crude, mostly based on manual experience judgment or simple task classification, which cannot accurately reflect the complex requirements of tasks and the relationship between various elements. This algorithm converts task requirements into a structured model by constructing a semantic network, clearly showing the relationship between various elements. In the weight distribution formula, the degree of influence between elements and the importance of the elements themselves are comprehensively considered, which is not available in traditional methods. For example, in a project, different task elements have different effects on the realization of the final goal. This quantitative method can more scientifically determine the weight of each element. This deep decomposition method based on semantic networks and business rules provides a more accurate description of task requirements for the precise matching of subsequent tasks and employee skills, breaking through the limitations of traditional task requirement analysis.

[0176] Build a two-way matching model between employee skills and task requirements. Consider not only the requirements of the task requirements for employee skills, but also the adaptability of employee skills to the task and the room for improvement. For example, for an employee who has a certain data analysis foundation but needs to further improve his skills, give priority to assigning data analysis tasks that are challenging but can help him improve his skills.

[0177] Traditional task requirement decomposition and matching methods are simple and crude, and cannot accurately reflect the complex requirements of tasks and the skill characteristics of employees. The decomposition algorithm that combines semantic networks and business rules can analyze task requirements more comprehensively and accurately. The two-way matching model breaks the limitations of traditional one-way matching, comprehensively considers both tasks and employees, and makes task allocation more scientific and reasonable.

[0178] It improves the accuracy of matching tasks with employee skills, realizes scientific and rational task allocation, and fully stimulates the work potential of employees. In personnel management, reasonable task allocation helps employees' career development and skill improvement. In project management, accurate task allocation can improve the execution efficiency and quality of projects.

[0179] 4. Solve the system adaptability problem caused by changes in employee behavior habits

[0180] 1. Identification of changes in behavior patterns

[0181] Behavioral pattern recognition technology

[0182] A dynamic behavioral pattern recognition model based on deep reinforcement learning. The model automatically adjusts model parameters to adapt to changes in behavioral patterns by continuously interacting with employee operational behavior data. For example, when an enterprise implements an agile project management model, employees' task switching and collaborative behaviors change. The model can automatically identify these changes and update its understanding of normal behavioral patterns.

[0183] Adopt a multi-dimensional behavior data fusion analysis method. Fusion analysis is performed on employees’ operation behavior data, communication behavior data, task collaboration data, etc. to determine whether the behavior pattern has changed from multiple perspectives. For example, based on the frequency and content of employees’ speeches in the project discussion area and their operation behavior in the task allocation module, a comprehensive judgment is made as to whether their work behavior pattern meets the requirements of agile project management.

[0184] Traditional behavior pattern recognition methods have poor adaptability and are difficult to quickly and accurately identify dynamic changes in employee behavior patterns. Deep reinforcement learning models have self-learning and adaptive capabilities, and can keep up with changes in employee behavior patterns in a timely manner. Multi-dimensional behavior data fusion analysis methods make comprehensive judgments from multiple aspects, avoiding the one-sidedness of single data dimension analysis and improving recognition accuracy.

[0185] It can accurately identify changes in employee behavior patterns and provide accurate judgment basis for system adaptive adjustment. In workflow management, timely identification of changes in behavior patterns can help companies adjust business processes in a timely manner and improve work efficiency. In personnel management, understanding changes in employee behavior patterns can help managers better manage and guide employees.

[0186] (II) System Adaptive Adjustment

[0187] System Tuning Strategy

[0188] Dynamic module reconstruction method based on microservice architecture. When the system recognizes changes in employee behavior patterns, the microservice architecture is used to dynamically reconstruct related modules. For example, for the progress tracking module, according to the needs of agile project management, the original linear progress display method is reconstructed into an iterative progress display, with task iteration as the basic unit, and each iteration cycle is divided into planning, execution, review and other stages, and the progress status is updated in real time.

[0189] Use an adaptive interface adjustment algorithm based on user portraits. Generate personalized user portraits based on employee behavior patterns and work needs. For example, for employees who frequently perform data analysis, highlight data analysis-related functions and data entry points on the system interface. When employee behavior patterns change, automatically update user portraits, and adjust interface layout and function display accordingly.

[0190] Traditional system adjustment methods are inflexible and difficult to quickly adapt to changes in employee behavior patterns. The dynamic module reconstruction method of the microservice architecture can flexibly adjust specific modules and quickly adapt to changes in business needs. The interface adaptive adjustment algorithm based on user portraits improves the system's ease of use and adaptability based on user personalized needs.

[0191] The system can quickly adapt to changes in employee behavior habits, improving the system's adaptability and user experience. In terms of resource management, the system's rapid adaptation reduces resource waste and inefficiency caused by changes in employee behavior. In workflow management, it ensures the smooth progress of business processes and improves work efficiency.

[0192] 5. The problem of achieving collaborative task allocation based on biological rhythms and group psychology

[0193] (I) Biorhythm analysis

[0194] Biorhythm Analysis Methods

[0195] A biorhythm analysis algorithm based on time series clustering and cycle prediction is proposed. Cluster analysis is performed on the long-term operation time series data of employees, and the operation time is divided into different time periods, such as peak work period and trough period. Then, the cycle prediction algorithm is used to predict the work efficiency and ability performance of employees in different time periods. For example, through analysis, it is found that an employee has a higher work efficiency between 10-12 am and 3-5 pm, and this pattern has a certain periodicity.

[0196] Biological rhythm analysis algorithm formula based on time series clustering and cycle prediction

[0197] Time series clustering formula:

[0198]

[0199] Where C is the number of clustering categories (for example, dividing working hours into peak hours, trough hours, etc.), S i is the i-th cluster, x j is the jth data point in the time series (i.e., the employee’s operation time point), μ i is the center of the ith cluster. The operation time series is divided into different time periods by minimizing the sum of the squares of the distances from each data point to the center of the cluster to which it belongs.

[0200] Cycle prediction algorithm formula:

[0201]

[0202] Where y(t) is the predicted employee's work efficiency or ability performance at time, N is the number of periodic components, and A n is the amplitude of the nth periodic component, f n is the frequency of the nth periodic component, is the phase of the nth periodic component, ∈(t) is the error term. The periodic changes of employee work efficiency and ability performance are predicted by fitting multiple sinusoidal functions.

[0203] Traditional biorhythm analysis methods are simple, mostly based on experience or simple time statistics, and do not make full use of data mining and prediction technology to mine patterns from time series data. The time series clustering formula of this algorithm draws on the clustering algorithm in data mining, but is innovatively applied to the analysis of employee operation time, dividing the time series into meaningful time periods, which does not appear in traditional biorhythm analysis. The period prediction algorithm predicts employee work efficiency and performance by constructing a sinusoidal function model and taking into account the possible periodic changes in biorhythms. Traditional methods do not take into account this periodic quantitative prediction. This algorithm introduces a mathematical model to more accurately analyze employee biorhythms and provide a more scientific basis for task allocation. At the same time, when considering the relationship between environmental factors and biorhythms, the model can be further expanded and environmental factors can be included in the formula as adjustment parameters to better meet actual work scenarios.

[0204] Introduce the correlation analysis between environmental factors and biorhythms. Consider the impact of factors such as work environment and task type on employees' biorhythms. For example, in a noisy working environment, employees' biorhythms may be disturbed and their work efficiency may decrease. By analyzing these correlations, the biorhythm status of employees in different environments can be more accurately evaluated.

[0205] Traditional biorhythm analysis methods are simple and do not fully consider the actual working conditions and environmental factors of employees. Time series clustering and cycle prediction algorithms can mine the laws of employees' biorhythms from data. The correlation analysis between environmental factors and biorhythms makes the analysis results more in line with actual working scenarios and improves the accuracy and practicality of biorhythm analysis.

[0206] It can accurately determine the peak and trough periods of employees' biorhythms, providing a scientific basis for task allocation. In task allocation, assigning tasks according to employees' biorhythms can give full play to employees' advantages and improve work efficiency and quality. In personnel management, understanding employees' biorhythms can help to reasonably arrange working hours and rest time, and improve employees' job satisfaction and physical and mental health.

[0207] 2. Group Psychological Assessment

[0208] Group Psychological Assessment Model

[0209] Construct a group psychological assessment model based on emotional communication and social network structure. Analyze the communication path and scope of emotional information in the project discussion area, and judge the group psychological state in combination with the social network structure. For example, when positive emotions spread rapidly and widely in the social network, it means that the team morale is high; conversely, when negative emotions spread, the team morale may be low.

[0210] Adopt the group psychology analysis method of multi-source data fusion. Integrate and analyze the employee's behavior data, communication data, performance data, etc. with the project discussion area information to comprehensively evaluate the group psychology from multiple perspectives. For example, combine the employee's task completion and speaking attitude in the discussion area to judge the overall work atmosphere and psychological state of the team.

[0211] Traditional group psychology assessment mainly relies on subjective judgment or a single data source, which lacks accuracy and comprehensiveness. The assessment model that combines emotional communication with social network structure reveals the dynamic changes of group psychology from the perspective of information communication. The analysis method of multi-source data fusion integrates multi-faceted data and comprehensively considers multiple factors to more accurately grasp the group psychological state.

[0212] Comprehensively evaluate the group psychological state from multiple dimensions to make the evaluation results more accurate and reliable. In team management, accurately understanding the group psychology helps managers take timely measures to adjust the team atmosphere and improve team cohesion and collaboration efficiency. In the process of project advancement, a good group psychological state can stimulate the enthusiasm and creativity of team members and increase the probability of project success.

[0213] 3. Task Allocation and Collaboration

[0214] Task allocation strategy

[0215] Task allocation based on biorhythms and group psychology: After the system determines the peak and trough periods of employees' biorhythms and evaluates the group psychological state of the team, it formulates a unique task allocation strategy. For example, when team morale is low, tasks that require high creativity and enthusiasm will be assigned to members who have a positive influence in the team during the peak period of employees' biorhythms, so that their outstanding performance can drive the team atmosphere. At the same time, other members are assigned some relatively easy tasks with clear goals and quick results, which will enhance their sense of accomplishment and gradually improve team morale.

[0216] Dynamic adjustment mechanism: During the execution of tasks, the system continuously monitors the employee's biorhythm-related data, task completion progress, and team member interactions. By establishing a dynamic adjustment model, task allocation is automatically adjusted according to real-time data changes. For example, if it is found that the employee originally assigned to the creative task has a decreased work efficiency due to biorhythm changes or other reasons, and another employee is at the peak of his biorhythm and has a certain interest and ability in the task, the system will automatically transfer part of the task to the latter to ensure efficient progress of the task.

[0217] Traditional task allocation ignores biological rhythms and group psychological factors, making it difficult to fully tap the potential of employees and improve the overall effectiveness of the team. The task allocation strategy of this system fully considers the individual differences of employees and the overall status of the team. By rationally allocating tasks, it can achieve the best use of talents and improve team collaboration. The dynamic adjustment mechanism is based on real-time data and can respond to various changes in a timely manner to ensure that task allocation is always in the optimal state, which is difficult to achieve with traditional task allocation systems.

[0218] The task allocation based on biological rhythm and group psychology can give full play to the potential of employees and improve the efficiency of teamwork and work quality. In terms of resource management, reasonable task allocation avoids the waste of human resources and improves the efficiency of resource utilization. In project management, it significantly improves the success rate and execution efficiency of projects, creating greater value for enterprises.

[0219] 6. Problem Solving Order and Relevance Analysis

[0220] Prioritize solving the problems of dynamic skill changes and task adaptation

[0221] Reason: Employee skills are the core basis for task allocation. In a rapidly changing business environment, real-time understanding of the dynamic changes in employee skills is the basis for achieving reasonable task allocation. By solving this problem, the system has established a complete employee skill data collection, analysis and evaluation system. For example, the operation log analysis method and employee feedback association mechanism developed in the process of solving this problem provide key data support and technical foundation for solving other problems in the future. This not only ensures the accuracy of current task allocation, but also lays a solid foundation for analyzing the relationship between employee behavior habits and skill improvement and the impact of biorhythms on skill performance.

[0222] It has achieved a preliminary and reasonable allocation of tasks, improved the efficiency and quality of employees in task execution, and avoided resource waste and work delays caused by the mismatch between skills and tasks. In terms of personnel management, it has promoted the effective use and improvement of employees' skills and is conducive to their career development planning.

[0223] Solve the problem of changing behavior habits based on the results of solving skill problems

[0224] Implementation method: The operation log analysis methods, data processing technologies, and in-depth understanding of employee behavior data accumulated when solving the problem of dynamic skill changes and task adaptation can be directly applied to the identification of changes in employee behavior habits and system adaptive adjustments. For example, the monitoring and analysis methods of operation behavior data are used for both the perception of skill changes and the identification of changes in behavior patterns. Through in-depth mining of operation behavior data, effective management of employee skills and behavior habits can be achieved at the same time. This process does not require additional steps or software and hardware investment, and makes full use of existing data processing processes and technologies, reducing system improvement costs and improving the overall operating efficiency of the system.

[0225] The system can quickly adapt to changes in employee behavior habits, ensuring the smooth progress of business processes and avoiding reduced work efficiency and management chaos caused by the system not adapting to changes in employee behavior. In workflow management, the flexibility and adaptability of business processes are improved, and the ability of enterprises to cope with internal changes is improved.

[0226] Based on the solution of the first two problems, the synergy between biological rhythm and group psychology is achieved

[0227] Collaborative implementation process: The solution to the first two problems enables the system to have rich employee operation time information, task completion data, and team member interaction information. These data provide sufficient data sources for biorhythm analysis and group psychology assessment. Through the cross-domain integration and analysis of these data, innovative task allocation collaboration based on biorhythms and group psychology is achieved. For example, the biorhythm is determined using employee operation time information, and the group psychology is evaluated in combination with task completion status and team member interaction information, and then a scientific and reasonable task allocation strategy is formulated. This process further improves the scientificity and rationality of task allocation without adding any additional steps and software and hardware investment, fully taps the potential of employees, and improves team collaboration efficiency and work quality.

[0228] It has achieved comprehensive optimization from the individual to the team level, greatly improving the overall operational efficiency and competitiveness of the enterprise. It has achieved deep integration and collaborative optimization in multiple aspects such as resources, workflow, personnel and project management, bringing significant economic and management benefits to the enterprise.

[0229] Summarize:

[0230] First solve the real-time problem of dynamic changes in team members' skills and task adaptation

[0231] Solution:

[0232] Perception of skill changes: A skill evaluation algorithm based on the similarity of operation sequences is used to compare the operation sequences of employees at different times to calculate similarities, such as analyzing the sequence of operation steps of employees using data analysis tools to determine their mastery of data analysis skills. At the same time, a dynamic weight evaluation mechanism for operation frequency and time consumption is introduced to dynamically adjust the weights of operation frequency and time consumption based on the complexity and importance of the operation. For example, complex data mining operations have high accuracy and time consumption weights, while simple data cleaning operations have high operation frequency weights.

[0233] Task adaptation and adjustment: Use a deep task demand decomposition algorithm based on semantic networks and business rules to convert task requirements into semantic networks, refine each element and assign weights. For example, for online marketing project data analysis tasks, subdivide data processing elements and assign weights. Build a two-way matching model between employee skills and task requirements, taking into account the task's requirements for employee skills and the adaptability and improvement space of employee skills to tasks, and assign appropriate tasks to employees.

[0234] The traditional method only analyzes the operation log from a single dimension, which makes it difficult to comprehensively and accurately evaluate the changes in employee skills. In addition, the task requirements are simply broken down and matched roughly, and cannot reflect the complex requirements of tasks and the characteristics of employee skills. The new method evaluates skills from multiple dimensions such as operation consistency, logic, and operation characteristics, and comprehensively matches from both the task and employee perspectives, which is more in line with actual needs.

[0235] Accurately capture the dynamic changes of employee skills without the need for additional skill tests, saving costs and ensuring that data is real and real. Improve the accuracy of matching tasks with employee skills, achieve scientific and reasonable task allocation, fully stimulate employee work potential, and provide key data support and technical foundation for subsequent problem solving, such as providing operational behavior data for behavioral habit analysis, and providing employee skills and task allocation data for biorhythm and group psychology analysis. In resource management, rationally allocate human resources; in task allocation, improve rationality.

[0236] Based on the results of solving the skill problem, the solution to the system adaptability problem caused by the change of employee behavior habits is as follows:

[0237] Identification of behavioral pattern changes: Develop a dynamic behavioral pattern identification model based on deep reinforcement learning, so that it can continuously interact and learn with employee operation behavior data, and automatically adjust parameters to adapt to changes in behavioral patterns. For example, after the company implements the agile project management model, the model can identify changes in employee task switching and collaborative behavior. Use a multi-dimensional behavioral data fusion analysis method to integrate operational behavior, communication behavior, task collaboration and other data to comprehensively judge whether the behavioral pattern has changed, such as combining the project discussion area speech and task allocation module operation to judge the employee's work behavior pattern.

[0238] System adaptive adjustment: Based on the dynamic module reconstruction method of microservice architecture, when the employee behavior pattern changes are identified, the relevant modules are dynamically reconstructed, such as the progress tracking module is reconstructed according to the requirements of agile project management. The user portrait-based interface adaptive adjustment algorithm is used to generate personalized user portraits according to employee behavior patterns and work requirements. When the behavior pattern changes, the portrait is automatically updated and the interface layout and function display are adjusted.

[0239] Traditional behavior pattern recognition methods have poor adaptability, are difficult to quickly and accurately identify dynamic changes, and have poor flexibility in system adjustment. The new method uses deep reinforcement learning self-learning and adaptive capabilities, multi-dimensional data fusion to avoid one-sidedness, microservice architecture to achieve flexible adjustment, and user portraits to meet personalized needs.

[0240] Accurately identify changes in employee behavior patterns and provide accurate judgment basis for system adaptive adjustment. The system quickly adapts to changes in employee behavior habits, improves adaptability and user experience, ensures smooth business processes, and avoids reduced efficiency and management chaos caused by system inadaptability. In workflow management, improve business process flexibility and adaptability, and enhance the company's ability to respond to internal changes. This process uses the operation log analysis methods, data processing technology, and understanding of employee behavior data accumulated when solving skill problems, without the need for additional steps or software and hardware investment.

[0241] Based on the solutions to the first two problems, a collaborative solution to task allocation based on biological rhythms and group psychology is achieved:

[0242] Biorhythm analysis: A biorhythm analysis algorithm based on time series clustering and cycle prediction is proposed. The clustering analysis of the long-term operation time series data of employees is carried out to divide the operation time periods, and the cycle prediction algorithm is used to predict work efficiency and ability performance, such as determining the time period when employees are most efficient. The correlation analysis between environmental factors and biorhythms is introduced, and the impact of work environment, task type, etc. on biorhythms is considered to more accurately evaluate the biorhythm status of employees.

[0243] Group psychology assessment: Build a group psychology assessment model based on emotional communication and social network structure, analyze the path and scope of emotional information dissemination in the project discussion area, and judge the group psychological state in combination with the social network structure, such as judging team morale through positive or negative emotional communication. Use a group psychology analysis method that integrates multi-source data, integrates employee behavior, communication, performance and other data with project discussion area information, and comprehensively evaluates group psychology.

[0244] Task allocation coordination: Based on the biological rhythm and group psychology, the system determines the employee's biological rhythm period and the team's group psychological state, and then formulates a task allocation strategy. For example, when team morale is low, highly creative tasks are assigned to members with positive influence during the peak period of the employee's biological rhythm, and other members are assigned easy tasks to boost morale. Dynamic adjustment mechanism: During task execution, the system continuously monitors information such as employee biological rhythms, task progress, and team interaction, and automatically adjusts task allocation according to real-time data changes through dynamic adjustment models.

[0245] Traditional biorhythm analysis methods are simple and do not take into account actual work conditions and environmental factors. Group psychological assessment relies on subjective judgment or a single data source, lacking accuracy and comprehensiveness. Traditional task allocation ignores biorhythm and group psychological factors. The new method analyzes biorhythm and group psychology from multiple dimensions, comprehensively considers individual differences and the overall status of the team to allocate tasks, and dynamically adjusts based on real-time data to achieve collaborative optimization.

[0246] Accurately determine the peak and trough periods of employees' biorhythms and evaluate the psychological state of the group, providing a scientific basis for task allocation. Based on the coordination of task allocation based on biorhythms and group psychology, fully tap the potential of employees and improve team collaboration efficiency and work quality. In resource management, avoid waste of human resources and improve resource utilization efficiency; in project management, improve project success rate and execution efficiency. Achieve comprehensive optimization from individual to team level, and achieve deep integration and collaborative optimization in multiple aspects without adding additional steps and software and hardware investment.

[0247] Through the above method, we first solve the real-time problem of dynamic changes in team members' skills and task adaptation, then solve the system adaptability problem caused by changes in employees' behavioral habits, and finally achieve task allocation coordination based on biological rhythms and group psychology, which satisfies the three problems of solving the real-time problem of dynamic changes in team members' skills and task adaptation, the system adaptability problem caused by changes in employees' behavioral habits, and the task allocation coordination problem based on biological rhythms and group psychology without adding any steps or any software or hardware.

Claims

1. The team management application OA intelligent processing system is characterized by: include: Data collection layer: Adopting the native data capture mechanism in the event-driven system, embedding code snippets at the key operation nodes of each business module to capture data, and using data mapping tables to map and integrate the operation data in different business systems in a unified format; Data storage layer: We use the MySQL relational database and improve the distributed storage architecture. We use a self-developed distributed storage algorithm to perform shard storage according to business type, time and other dimensions, and dynamically adjust the distribution of storage nodes based on access frequency. At the same time, we use a distributed ledger mechanism based on blockchain technology to perform redundant storage and consistency verification of key data. Data analysis layer: Build a multimodal data analysis algorithm model that integrates semantic analysis, behavioral pattern recognition, and time series analysis, and use cross-domain knowledge fusion to integrate knowledge such as psychology and management into the algorithm model; Application layer: Design an intelligent interactive interface based on user behavior prediction. By analyzing employees' historical operation behaviors and business needs, predict the employees' next operation and display relevant functions and data in advance. At the same time, introduce emotional design concepts to adjust the interface style and prompt information according to the employees' work status and emotional feedback.

2. The team management application OA intelligent processing method is characterized by: The following steps are involved: Step 1: Basic data collection and integration: Adopt a recording method based on operation step decomposition to collect employee operation logs, and use data watermark technology to mark the operation log data; The feedback information processing system based on natural language understanding and knowledge graph collects employee feedback information and uses automatic classification and priority sorting algorithms to classify and sort the feedback information; Use sentiment analysis combined with topic modeling and social network dynamic modeling technology to collect project discussion area information; Step 2: Skill change perception: Adopt a skill evaluation algorithm based on the similarity of operation sequences, compare the operation sequences of employees at different times to calculate the similarity, and introduce a dynamic weight evaluation mechanism for operation frequency and time consumption; Task adaptation and adjustment: Using a deep task requirement decomposition algorithm based on semantic networks and business rules, the task requirements are converted into semantic networks and each element is refined and weighted, thus building a two-way matching model between employee skills and task requirements; Step 3: Identification of behavioral pattern changes: Based on deep reinforcement learning, a dynamic behavioral pattern recognition model is constructed, and a multi-dimensional behavioral data fusion analysis method is adopted to integrate employees' operation behavior data, communication behavior data, task collaboration data, etc. for analysis; System adaptive adjustment: Dynamically reconstruct related modules based on the dynamic module reconstruction method of microservice architecture, use the interface adaptive adjustment algorithm based on user portraits to generate personalized user portraits and adjust the interface layout and function display according to the behavior patterns and work needs of employees; Step 4: Biorhythm analysis: Propose a biorhythm analysis algorithm based on time series clustering and cycle prediction, perform clustering analysis and cycle prediction on employees' long-term operation time series data, and introduce correlation analysis between environmental factors and biorhythms; Group psychological assessment: Build a group psychological assessment model based on emotional communication and social network structure, adopt a group psychological analysis method based on multi-source data fusion, and integrate employee behavior data, communication data, performance data, etc. with project discussion area information for analysis; Task allocation coordination: formulate task allocation strategies based on the task allocation of biological rhythms and group psychology, and establish a dynamic adjustment mechanism to automatically adjust task allocation according to real-time data changes during task execution.

3. The team management application OA intelligent processing system according to claim 1, characterized in that: The data collection layer is based on the event-driven system's native data capture mechanism, and code snippets are implanted in the key operation nodes of each business module.

4. The team management application OA intelligent processing method according to claim 1, characterized in that: The shard storage formula of the self-developed distributed storage algorithm in the data storage layer is: S i =Hash(B i ,T i )modN Among them, S i Indicates the target server node number for data storage, B i Represents the business type code, T i is the timestamp of data generation, N is the total number of server nodes; The formula for dynamically adjusting storage nodes is: Among them, P j represents the probability of data being stored in the jth server node, F j is the access frequency of data on the jth server node.

5. The team management application OA intelligent processing method according to claim 2, characterized in that: The importance evaluation formula of the automatic classification and prioritization algorithm in the employee feedback information processing system is: Where I represents the importance score of the feedback information, M is the number of key factors in the knowledge graph related to the feedback information, and w m is the weight of the mth key factor, K m is to extract the content related to the mth key factor from the feedback information, m It is a function that quantitatively evaluates the extracted content according to the key factor type; The formula for evaluating the degree of urgency is: Where E represents the urgency score of the feedback information, N is the number of judgment factors related to the urgency, and v n is the weight of the nth factor, T n is the time information related to urgency extracted from the feedback information, g n It is a function that quantifies the urgency based on time information; The comprehensive priority score formula is: P=α×I+β×E Wherein, P is the comprehensive priority score of the feedback information, α and β are weight coefficients of importance and urgency respectively, and α+β=1.

6. The team management application OA intelligent processing method according to claim 2, characterized in that: The operation sequence similarity of the skill evaluation algorithm based on operation sequence similarity in skill change perception is calculated as: Among them, Sim(S1,S2) represents the similarity of two operation sequences S1 and S2, S1 = [s 11 ,s 12 ,…,s 1L ],S2=[s 21 ,s 22 ,…,s 2L ], L is the length of the shorter of the two sequences; Match(s 1i ,s 2i ) is a matching function; |S1| and |S2| are the lengths of the operation sequences S1 and S2 respectively; Comprehensive skill evaluation formula considering operation frequency and time consumption: Among them, Skill(t) represents the skill level of the employee at time, S t and S t-Δt are the operation sequences at time t and t-Δt respectively; w1, w2, w3 are weight coefficients, and w1+w2+w3=1, which respectively represent the importance of operation sequence similarity, operation frequency, and operation time in skill evaluation; o i Indicates the i-th operation, Weight(o i ) is the operation o i The weight, Freq(o i ) is the operation o i The frequency, Time(o i ) is the operation o i The average time taken.

7. The team management application OA intelligent processing method according to claim 2, characterized in that: The task requirement semantic network of the task requirement deep decomposition algorithm based on semantic network and business rules in task adaptation adjustment is constructed as follows: G=(V,E) Among them, G represents the semantic network of task requirements, V is a set of nodes, and each node v i ∈V represents an element of a task, E is a set of edges, and each edge e ij ∈E represents the element v i and v j the relationship between; Factor refinement and weight allocation formula: Among them, W(v i ) is the node v i The weight of node v is k. i The number of associated nodes, Influence(v j ,v i ) represents node v j For node v i Importance(v j ) is the node v j Importance score.

8. The team management application OA intelligent processing method according to claim 2, characterized in that: The time series clustering formula of the biorhythm analysis algorithm based on time series clustering and cycle prediction in biorhythm analysis is: Among them, C is the number of cluster categories, S i is the i-th cluster, x j is the jth data point in the time series, μ i is the center of the i-th cluster; Cycle prediction algorithm formula: Where y(t) is the predicted employee's work efficiency or ability performance at time, N is the number of periodic components, and A n is the amplitude of the nth periodic component, f n is the frequency of the nth periodic component, is the phase of the nth periodic component, ∈(t) is the error term.

9. The team management application OA intelligent processing system according to claim 1, characterized in that: The intelligent interactive interface based on user behavior prediction in the application layer analyzes employees' historical operation behavior data and uses machine learning algorithms to build a prediction model to predict the next operation of employees.

Images