A virtual enterprise collaborative decision-making method and system for team heterogeneity identification and management strategy recommendation

Abstract

The application discloses a virtual enterprise collaborative decision-making method and system for team heterogeneity identification and management strategy recommendation, and the method comprises the following steps: collecting team member basic information, dynamic cooperation behavior and periodic psychological measurement data; constructing task and relationship heterogeneity indexes based on the collected data; establishing and training a machine learning model; deploying the model, predicting future high-intensity conflict probability based on the heterogeneity indexes; and generating conflict management strategy suggestions in combination with the heterogeneity indexes and the predicted probability. The application integrates multiple data sources, constructs quantitative task and relationship heterogeneity indexes, and accurately portraits the team state. On this basis, the trained machine learning model is used to dynamically predict the future high-intensity conflict probability. This changes the passive situation of the existing team management system in responding after the event, improves the collaborative decision-making efficiency and overall management efficiency of the virtual team, and realizes the scientization and foresight of team management.

Description

Technical Field

[0001] This invention relates to the field of virtual enterprise team management optimization technology, and in particular to a virtual enterprise collaborative decision-making method and system for team heterogeneity identification and management strategy recommendation. Background Technology

[0002] With the rapid development of information technology, virtual enterprises, as a new form of organization, have demonstrated significant advantages in responding to dynamic market environments and complex innovation tasks due to their ability to quickly integrate high-quality resources across regions and organizations. Virtual teams have become commonplace, especially in fields such as scientific research and collaborative problem-solving. However, members of virtual teams typically possess diverse knowledge backgrounds, professional skills, cultural perspectives, and communication styles; these differences constitute the team's "heterogeneity." Team heterogeneity is a double-edged sword: on the one hand, it can stimulate greater creativity and innovation potential through the clash of ideas; on the other hand, it can also easily trigger cognitive conflicts (stemming from differences in task understanding and solutions) and emotional conflicts (stemming from interpersonal relationships and value differences) within the team.

[0003] Currently, management systems supporting virtual team collaboration (such as various project management software and collaborative office platforms) mainly focus on task assignment, progress tracking, document sharing, and basic communication functions. While these systems improve the efficiency of task management, they have significant shortcomings in dynamic team governance.

[0004] First, there is a lack of in-depth quantification and dynamic monitoring of team heterogeneity: Existing systems typically cannot effectively model and analyze the heterogeneity of team members in the "task dimension" (e.g., differences in knowledge structure and skill expertise) and the "relationship dimension" (e.g., differences in communication patterns and cultural background). Systems largely rely on static member profile information, making it difficult to capture the deeper team state reflected in dynamic behavioral data generated during collaboration. Second, conflict management relies on subjective experience and lacks data-driven decision support: When disagreements or signs of conflict arise within a team, current systems often respond reactively. The selection of conflict management strategies (e.g., confrontation, collaboration, compromise, avoidance, and laissez-faire) largely depends on the personal experience and subjective judgment of team leaders or project managers, lacking objective and personalized strategy recommendations based on real-time team status data and historical conflict pattern analysis. Third, it is difficult to shift from passive response to proactive intervention: Due to a lack of predictive capabilities for potential conflict risks, existing systems cannot provide early warnings and proactive intervention before conflicts erupt. This results in delayed management actions, often requiring higher coordination costs, and may have already damaged team morale and innovation efficiency. Fourth, there is a technology gap between theory and practice: Organizational behavior and management have yielded numerous studies on the impact of team heterogeneity on innovation performance and the crucial moderating role of conflict management strategies. However, these theoretical findings have not yet been systematically engineered into technical solutions capable of running on real-world collaborative platforms and providing intelligent decision support.

[0005] Therefore, there is an urgent need in this field for a solution that can deeply integrate management theory, big data analysis and artificial intelligence technology to achieve accurate identification of the heterogeneity of virtual teams, dynamic prediction of conflict risks, and improve the collaborative decision-making efficiency and innovation performance of virtual enterprises. Summary of the Invention

[0006] To address the aforementioned problems, the present invention aims to provide a collaborative decision-making method and system for virtual enterprises that identifies team heterogeneity and recommends management strategies. This method solves the challenge of dynamic conflict management in virtual teams under heterogeneous backgrounds, achieves multi-dimensional and accurate identification of team heterogeneity characteristics, real-time dynamic early warning of conflict risks, and intelligently pushes suitable management strategies based on individual differences and conflict types. This effectively bridges the technological gap between theoretical research and practical application, and improves the collaborative efficiency and innovation performance of virtual enterprise teams.

[0007] This invention provides a collaborative decision-making method and system for virtual enterprises that identifies team heterogeneity and recommends management strategies.

[0008] The first aspect: A collaborative decision-making method for virtual enterprises that identifies team heterogeneity and recommends management strategies, including:

[0009] S1. Collect basic information data, dynamic collaborative behavior data, and periodic psychological measurement data of team members;

[0010] S2. Construct task heterogeneity indicators and relationship heterogeneity indicators based on the collected data;

[0011] S3. Build and train a machine learning model;

[0012] S4. Deploy the trained machine learning model and predict the probability of high-intensity conflict occurring in the future based on task heterogeneity indicators and relationship heterogeneity indicators.

[0013] S5. Based on the predicted probability of high-intensity conflict, and combined with task heterogeneity indicators and relationship heterogeneity indicators, generate conflict management strategy recommendations.

[0014] In one embodiment of the present invention, the following is stated:

[0015] Basic information data includes: team members' educational background, professional skills, and work experience;

[0016] Dynamic collaborative behavior data includes: semantic analysis of chat logs, frequency of meeting speeches, and contribution to document editing;

[0017] Periodic psychometric data include: MBTI test and team atmosphere perception questionnaire.

[0018] In one embodiment of the present invention:

[0019] Set parameters for skill tag matching degree, knowledge graph distance, and task dependency complexity, and calculate task heterogeneity index based on basic information data, dynamic collaborative behavior data, and periodic psychological measurement data;

[0020] By setting parameters such as language style difference, emotional tendency consistency, and social network density, and calculating relationship heterogeneity index based on basic information data, dynamic collaborative behavior data, and periodic psychological measurement data, a relationship heterogeneity index is obtained.

[0021] In one embodiment of the present invention, the predicted probability results are displayed on a health dashboard.

[0022] In one embodiment of the present invention,

[0023] Based on the predicted probability results, the effectiveness is verified. Based on the verification results, the machine learning model is trained in reverse to continuously optimize the prediction accuracy of the machine learning model.

[0024] In one embodiment of the present invention, the conflict management strategy in S5 includes:

[0025] Countermeasures include encouraging debate and setting challenging goals;

[0026] A laissez-faire strategy, which includes reducing intervention and allowing differences to coexist;

[0027] Harmony strategies include promoting communication and organizing team-building activities;

[0028] Guidance strategies include introducing third-party coordinators.

[0029] In one embodiment of the present invention, the future period of time does not exceed one week.

[0030] The second aspect: A virtual enterprise collaborative decision-making system for team heterogeneity identification and management strategy recommendation, including:

[0031] The data acquisition module is used to collect basic information data, dynamic collaborative behavior data, and periodic psychological measurement data of team members;

[0032] The heterogeneity calculation module is used to construct task heterogeneity indicators and relationship heterogeneity indicators based on the collected data;

[0033] The conflict risk prediction module is equipped with a trained machine learning model to predict the probability of high-intensity conflicts occurring in the future.

[0034] Conflict management strategy recommendation module: Based on the probability of high-intensity conflict, as well as task heterogeneity indicators and relationship heterogeneity indicators, it generates conflict management strategy suggestions;

[0035] The feedback optimization module is used to verify the effectiveness of the predicted probability results and, based on the verification results, to train the machine learning model in reverse.

[0036] The visualization and interaction module is used to display the prediction results on a health dashboard.

[0037] The beneficial effects of this invention are:

[0038] 1. This invention, by fusing multi-source data and constructing quantified indicators of task and relationship heterogeneity, enables precise profiling of team status. Based on this, a trained machine learning model dynamically predicts the probability of high-intensity conflict occurring within a future period. This changes the passive approach of existing team management systems, which can only respond after the fact. Managers can intervene precisely before conflict becomes explicit or escalates, based on early risk warnings provided by the system. This significantly reduces the cost of handling conflicts after they erupt and internal team friction, thereby significantly improving the collaborative decision-making efficiency and overall management efficiency of virtual teams. It achieves scientific and forward-looking team management, transforming from passive response to proactive early warning, and significantly improving the effectiveness of virtual team conflict management.

[0039] 2. This invention establishes a conflict management strategy recommendation module, combining predicted conflict probabilities with specific heterogeneity indicators. Based on a pre-defined algorithm model, it intelligently matches and generates personalized management strategy suggestions. This method engineers and implements management strategy models, making strategy recommendations evidence-based and systematic. It effectively reduces the arbitrariness and risk of misjudgment in managerial decision-making, providing standardized and replicable scientific decision support for virtual enterprises, and improving the pertinence and effectiveness of management measures.

[0040] 3. The built-in feedback optimization mechanism of this invention constitutes a complete closed-loop system. The system continuously tracks the implementation effect of the recommended strategies, using quantitative indicators to verify the accuracy of predictions and evaluate the effectiveness of the strategies. This real feedback data is used to back-train and fine-tune the parameters of the machine learning model, thereby achieving continuous self-optimization of the model. This enables the system to continuously learn from new team dynamics and intervention cases, adapt to the evolution of different team characteristics and business scenarios, and continuously improve the accuracy of conflict prediction and the applicability of strategy recommendations. Attached Figure Description

[0041] Figure 1 This is a schematic flowchart of the method of the present invention;

[0042] Figure 2 This is a schematic diagram of the system structure of the present invention;

[0043] Figure 3 This is a schematic diagram of the structure of the electronic device of the present invention. Detailed Implementation

[0044] Embodiments of the present invention are described in detail below. Examples of these embodiments are illustrated in the accompanying drawings, wherein the same or similar symbols denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0045] In existing technology-based virtual enterprises, team members often come from different organizations, regions, and professional backgrounds, resulting in significant task and relationship heterogeneity within the team, which can easily lead to cognitive and emotional conflicts. Current collaborative decision-making systems for virtual enterprises lack quantitative modeling and dynamic monitoring of task and relationship heterogeneity, which is detrimental to enterprise management.

[0046] To address the aforementioned issues, this invention discloses a collaborative decision-making method and system for virtual enterprises that identifies team heterogeneity and recommends management strategies. To make the objectives, technical solutions, and advantages of this disclosure clearer, the following detailed description is provided in conjunction with specific embodiments.

[0047] Example 1:

[0048] This embodiment discloses a collaborative decision-making method for virtual enterprises that identifies team heterogeneity and recommends management strategies, such as... Figure 1 As shown, the steps include:

[0049] S1. Collect basic information data, dynamic collaborative behavior data, and periodic psychological measurement data of team members.

[0050] The basic information data of team members includes: team members' educational background, professional skills and work experience.

[0051] Educational background includes academic level (e.g., undergraduate, master's, doctoral), type of university (e.g., comprehensive, science and engineering, liberal arts), and field of study (e.g., computer science, management, mechanical engineering). Professional skills include professional qualification certificates obtained (e.g., PMP project management, software engineer certification), professional tools or technologies mastered (e.g., Python programming language, CAD design software, data analysis tools), and level of professional proficiency in a specific field (e.g., junior, intermediate, senior). Work experience includes the type of industry worked in (e.g., information technology, manufacturing, financial services), years of work experience, job level held (e.g., entry-level employee, department head, project manager), and the type and scale of projects participated in (e.g., large multinational projects, small and medium-sized R&D projects).

[0052] Dynamic collaborative behavior data includes: semantic analysis of chat logs, frequency of meeting speeches, and contribution to document editing.

[0053] The chat log semantic analysis uses natural language processing technology to extract keywords, sentiment (e.g., positive, neutral, negative), and topic distribution (e.g., technical discussions, progress communication, resource coordination) from the dialogue to reflect the communication patterns and collaboration focus among members. The meeting speaking frequency statistics track the number of times team members speak, the percentage of speaking time, and the ratio of active speaking to passive response in various formal and informal meetings, reflecting the members' participation and power distribution in team communication. The document editing contribution is based on the number of times members create, modify, and annotate shared documents (e.g., project plans, requirement documents, technical solutions), the percentage of content modification, and version history tracking, to quantitatively evaluate the individual's specific contribution to knowledge production and information improvement.

[0054] Periodic psychometric data include: MBTI test and team atmosphere perception questionnaire.

[0055] The MBTI test assesses members' preferences across several dimensions, including personality acquisition (introversion / extroversion), information gathering (sensing / intuition), decision-making style (thinking / feeling), and life attitude (judging / perceiving), to determine their personality type. This provides a psychological basis for analyzing differences in cognitive styles, communication styles, and problem-solving approaches among team members. The team atmosphere perception questionnaire, on the other hand, is designed around dimensions such as team trust (e.g., the degree of mutual trust among members and the openness of information sharing), clarity of task goals (e.g., the degree of understanding of the overall team goals and individual responsibilities), effectiveness of conflict handling (e.g., the efficiency and satisfaction in resolving disagreements), and perception of incentive mechanisms (e.g., subjective evaluation of the fairness of team rewards and recognition of personal value). This quantifies members' subjective feelings about the team's collaborative environment, thereby revealing the potential for hidden conflicts or collaborative potential that team heterogeneity may cause at the psychological level.

[0056] S2. Construct task heterogeneity indicators and relationship heterogeneity indicators based on the collected data.

[0057] The purpose of the task heterogeneity metric is to quantify the degree of difference among team members in the knowledge, skills, and task relevance required to complete a task. Its construction includes:

[0058] A skills tag library is built based on basic information data, mapping key information from educational background (professional field), professional skills (certificates, tools, technologies) and work experience (industry, project type) to a unified skills tag library (e.g., Python, project management, machine learning).

[0059] A multi-dimensional skill vector is generated for each team member. Each dimension of the vector represents a skill tag, and its weight is determined by a combination of factors such as proficiency level, years of work experience, and project size.

[0060] Skill vectors are validated and enhanced through dynamic collaborative behavior data. By analyzing document editing contributions, members' editing behavior across different document types (technical solutions, design drafts, test reports) is used to validate and dynamically adjust the weights of their skill vectors. Semantic analysis of chat logs extracts technical terms and project challenges from discussions, serving as supplementary sources for skill tags.

[0061] Cognitive style characteristics are extracted from periodic psychometric data. The MBTI test converts the assessment results into a cognitive preference vector to measure differences in members' problem-solving styles. The team atmosphere perception questionnaire extracts members' ratings on task goal clarity, innovation support, etc., to indirectly assess the team's consistency in task cognition.

[0062] Skill tag matching is based on the skill vectors of team members. It calculates the similarity of skill vectors between all team members and takes the average value. The higher the value, the more similar the skills are and the lower the task heterogeneity.

[0063] The knowledge graph distance is based on members' skill tags, educational background, and project experience keywords. These tags are mapped to nodes in a pre-built domain knowledge graph, and the average shortest path distance between all pairs of corresponding nodes is calculated. The larger the distance value, the greater the difference in knowledge background and the higher the task heterogeneity.

[0064] Tasks depend on complexity. By modeling the task set as a directed graph, graph theory metrics such as graph density and critical path length are calculated. The higher the complexity, the tighter the coupling between tasks and the greater the difficulty of collaboration.

[0065] The task heterogeneity index is generated using a weighted method and represented as follows:

[0066] in, For skill tag matching degree, For knowledge graph distance, For task-dependent complexity, , , For weights.

[0067] The purpose of the relationship heterogeneity index is to quantify the differences among team members in communication patterns, social network structures, and psychological feelings. Its construction process is as follows:

[0068] Language style differences measure the differences in communication habits and expression among team members. Sources include dynamic collaborative behavior data (semantic analysis of chat logs) and periodic psychometric data (MBTI).

[0069] From the chat logs, NLP techniques were used to extract language feature vectors for each member, including lexical and grammatical features. MBTI was used as a supplement: MBTI classes were converted into numerical features and incorporated into the vectors.

[0070] To calculate the average style difference of the team, calculate the cosine distance or Euclidean distance between the language feature vectors of all members in the team. The higher the distance value, the greater the language style difference between members and the higher the heterogeneity of the relationship.

[0071] Sentiment consistency measures the synchronicity or divergence of emotional atmosphere among team members during collaboration. It can be obtained through dynamic collaborative behavior data (such as sentiment analysis of chat logs and meeting speeches).

[0072] For each member, sentiment analysis is performed on their chat and speech content over a period of time to obtain the daily average sentiment score, forming an individual sentiment sequence. The correlation coefficients between each pair of members' sentiment time series are calculated. The average of all correlation coefficients is then calculated; a lower average indicates greater asynchrony in members' emotional responses and higher heterogeneity in relationships.

[0073] Social network density measures the overall tightness of social interactions among team members. Data is derived from dynamic collaborative behavior data (chat frequency, meeting interactions, etc.). An undirected weighted graph is constructed by assigning weights to edges based on the communication frequency among members (e.g., number of chats per week). The network density is calculated; a lower social network density value indicates sparser direct communication connections between members, with interactions concentrated among a small number of people, and higher relationship heterogeneity.

[0074] The heterogeneity indicators of relationships are integrated into a single comprehensive indicator, expressed by the following formula:

[0075]

[0076] in, As an indicator of relational heterogeneity, For differences in language style, for

[0077] Consistency of emotional tendencies and For social network density parameters, , , For weights.

[0078] Through the above process, the system can transform multi-source, heterogeneous raw data into two core, quantifiable heterogeneity indicators, providing a solid data foundation for subsequent conflict risk prediction and intelligent strategy recommendation.

[0079] S3. Build and train a machine learning model.

[0080] A dataset is constructed using task heterogeneity metrics and relationship heterogeneity metrics. The machine learning model can be the XGBoost model, which is then trained using the dataset.

[0081] During model training, the constructed dataset is preprocessed, including data cleaning to remove outliers and missing values, and Z-score normalization is used to normalize the feature data to ensure that task heterogeneity indicators and relational heterogeneity indicators of different magnitudes can play an equally important role in model training.

[0082] The preprocessed dataset was divided into a training set and a test set in a 7:3 ratio. The training set was used to learn the model parameters, and the test set was used to evaluate the model's generalization ability.

[0083] In terms of parameter tuning for the XGBoost model, a grid search combined with 5-fold cross-validation can be used to optimize key parameters such as learning rate, maximum tree depth, number of leaf nodes, and regularization coefficient, in order to minimize the mean squared error of the model on the validation set.

[0084] During training, the model is monitored for overfitting or underfitting by monitoring the changes in the loss function of the training and validation sets. If overfitting occurs, the regularization coefficient is increased or the complexity of the tree is reduced. If underfitting occurs, the depth of the tree is increased or the learning rate is increased.

[0085] After multiple rounds of parameter adjustments and model iterations, the XGBoost model with high prediction accuracy and stability was finally obtained. This model can predict the conflict risks that virtual enterprise teams may face in the process of collaborative decision-making based on the input task heterogeneity index and relationship heterogeneity index.

[0086] S4. Deploy the trained machine learning model and predict the probability of high-intensity conflict occurring in the future based on task heterogeneity indicators and relationship heterogeneity indicators.

[0087] The latest task heterogeneity and relationship heterogeneity indicators of the collected virtual enterprise teams are standardized to ensure consistency with the data format and dimensions used during model training.

[0088] Subsequently, the standardized indicator data is input into the pre-trained XGBoost model. The model calculates and outputs the probability of a high-intensity conflict occurring within the team in the future (e.g., within the next week) by non-linearly combining and mapping the input features.

[0089] The probability value ranges from 0 to 1, with values ​​closer to 1 indicating a higher risk of high-intensity conflict and values ​​closer to 0 indicating a lower risk. The model also outputs a ranking of the feature importance of each input indicator to the conflict probability prediction result, enabling subsequent analysis of which heterogeneous factors have a more significant impact on conflict risk and providing data support for developing targeted management strategies.

[0090] S5. Based on the predicted probability of high-intensity conflict, and combined with task heterogeneity indicators and relationship heterogeneity indicators, generate conflict management strategy recommendations.

[0091] Conflict management strategies include: confrontational strategies, laissez-faire strategies, reconciliation strategies, and guidance strategies.

[0092] Countermeasures include encouraging debate and setting challenging goals.

[0093] By fostering an open and inclusive discussion atmosphere, team members are guided to engage in rational debates on core issues such as task objectives and implementation plans. This encourages the collision and integration of different viewpoints, transforming task heterogeneity into a source of innovation. Setting shared goals with a certain degree of difficulty and challenge can stimulate team members' sense of collective honor and willingness to collaborate, prompting them to proactively communicate and coordinate actions to achieve the goal, thereby reducing conflicts arising from differences in task perception.

[0094] A laissez-faire strategy includes reducing intervention and allowing differences to coexist.

[0095] When team heterogeneity manifests as differences in members' work styles, communication preferences, or minor task details, and these differences do not significantly negatively impact the achievement of core objectives, or may even spark potential innovation due to diverse perspectives, a laissez-faire approach can be adopted. Reduced intervention means that managers do not need to intervene and mediate every disagreement within the team, but rather give members a certain degree of autonomy, allowing them to naturally adapt to each other through interaction and explore collaborative methods that accommodate their differences.

[0096] Allowing for disagreements to coexist emphasizes respecting the value of individual differences and recognizing that not all disagreements require a unified solution; in fact, the existence of some disagreements can maintain the team's vitality and diversity of thought. The key to this strategy lies in accurately assessing the nature and extent of disagreements, ensuring they do not escalate into destructive conflicts that hinder teamwork, and trusting in team members' self-regulation abilities.

[0097] Harmony strategies include promoting communication and organizing team-building activities.

[0098] Facilitating communication aims to build an open and transparent dialogue platform, encouraging team members to proactively express their views, needs, and concerns, while actively listening to the ideas of others. Managers can guide members to focus on the problem itself, rather than personal emotions, advocating for empathy and understanding the logic and value behind others' perspectives. Furthermore, collaborative office tools and instant messaging software can break down time and space limitations, facilitating information sharing and exchange among members at any time, ensuring the timeliness and accuracy of information transmission, and avoiding new disagreements caused by information asymmetry. Organizing team-building activities enhances emotional connections and rapport among members through relaxed and enjoyable group interactions. By simultaneously promoting communication and organizing team-building activities, this reconciliation strategy can effectively improve interpersonal relationships within the team, reduce friction caused by heterogeneity, enhance team cohesion and centripetal force, and make team members more willing to work together towards common goals.

[0099] Guidance strategies, including the introduction of third-party coordinators,

[0100] The third-party coordinator must possess neutrality, professional conflict management skills, and extensive teamwork experience, enabling them to objectively examine issues arising from heterogeneity within the team. When irreconcilable conflicts or communication barriers arise within the team, the coordinator can intervene to organize specialized communication meetings, guiding team members to express their views and demands in an orderly manner, avoiding emotional arguments. By utilizing professional communication skills and conflict resolution models, the coordinator helps members understand the logic and stances behind differing opinions, promoting empathy. Furthermore, the coordinator can provide personalized communication and collaboration suggestions tailored to the characteristics of team members, helping them improve their ability to interact with colleagues from diverse backgrounds, thereby fundamentally enhancing the team's overall ability to cope with challenges posed by heterogeneity.

[0101] Furthermore, the effectiveness is verified based on the predicted probability results. Based on the verification results, the machine learning model is trained in reverse to continuously optimize the prediction accuracy of the machine learning model.

[0102] The effectiveness verification involves comparing and analyzing actual data, such as the degree of improvement in team heterogeneity issues and changes in team performance indicators after the recommended management strategies are applied in practice, with the improvement effect expected by the machine learning model based on predicted probabilities during the strategy recommendation stage.

[0103] By setting clear evaluation indicators, such as the percentage reduction in team conflict rate, the improvement in member collaboration efficiency, and the improvement in project task completion quality, the actual effect of strategy application is quantitatively evaluated. If there is a significant deviation between the actual effect and the model prediction results, such as the actual improvement effect of some strategies being far lower than expected, or negative effects not predicted by the model occurring, these deviation data, along with relevant strategy application scenarios, team characteristics, and other detailed information, are added to the training set as new training samples for backpropagation training in the machine learning model. During backpropagation training, the model will readjust the weight parameters of each feature variable, correct potential errors in the prediction algorithm, and thus continuously improve the model's ability to capture the mapping relationship between the heterogeneous characteristics of different virtual enterprise teams and the effects of management strategies. This ensures that the model can continuously evolve with the accumulation of real-world application data, providing more accurate and reliable strategy recommendation support for collaborative decision-making in virtual enterprises.

[0104] Furthermore, the predicted probability results are displayed on a health dashboard.

[0105] The key information, such as the heterogeneity characteristics of virtual enterprise teams, the matching probability of management strategies, and the expected effect evaluation data, is presented in a visual format.

[0106] The dashboard can be designed in layers. The upper layer displays the team's overall health score and key risk warning indicators in real time. The middle layer uses a radar chart to visually show the distribution of heterogeneous characteristics in different dimensions. The lower layer uses bar charts to compare the recommendation probability and potential implementation effect of different management strategies, including specific quantitative data such as the expected value of conflict mitigation and the percentage increase in collaboration efficiency.

[0107] Meanwhile, by clicking on any data module on the dashboard, the detailed calculation logic, historical trend, and case reference data of similar teams for that indicator can be displayed in a linked manner. This helps decision-makers quickly locate the core issues of team heterogeneity and understand the underlying basis of strategy recommendations, thereby making more targeted management decisions in complex virtual collaborative environments.

[0108] Example 2:

[0109] In conjunction with the method of Embodiment 1, this embodiment discloses a virtual enterprise collaborative decision-making system, which is applied to the above method.

[0110] like Figure 2 As shown, the system includes: a data acquisition module, a heterogeneity calculation module, a conflict risk prediction module, a conflict management strategy recommendation module, a feedback optimization module, and a visualization interaction module.

[0111] The data acquisition module is used to collect basic information data, dynamic collaborative behavior data, and periodic psychological measurement data of team members.

[0112] The basic information data includes static attributes such as members' age, education, professional background, skill level, job role, and department; dynamic collaborative behavior data includes real-time data such as members' communication frequency, information interaction duration, task allocation and completion, document collaborative editing records, meeting participation, and decision voting results in the virtual collaboration platform; and periodic psychological measurement data is collected regularly through standardized scales, including psychological state indicators such as members' job satisfaction, team belonging, role conflict perception, stress level, and willingness to innovate.

[0113] The data acquisition module supports interface integration with existing enterprise OA systems, project management tools, instant messaging platforms, and HRM systems to achieve automated collection and real-time synchronization of multi-source data, ensuring the comprehensiveness and timeliness of the data and providing a reliable data foundation for subsequent heterogeneity analysis and strategy recommendations.

[0114] The heterogeneity calculation module is used to construct task heterogeneity indicators and relationship heterogeneity indicators based on the collected data.

[0115] The basic information data, including education level, professional background, and skill level, is quantified. For example, different education levels, professional fields, and skill levels are assigned standardized weight values. The heterogeneity of knowledge and skills is characterized by calculating the variance or entropy values ​​of team members in these dimensions. At the same time, the task allocation and completion status in the dynamic collaborative behavior data are combined to analyze the participation and proficiency of members in different types of tasks (such as technology research and development, marketing, project management, etc.), construct a dispersion index of task ability distribution, and comprehensively form a quantitative evaluation result of task heterogeneity index.

[0116] The construction of relational heterogeneity indicators focuses on communication frequency, information interaction duration, and meeting participation in dynamic collaborative behavior data. By calculating the differences in interaction density among members, the diversity of information transmission paths, and the proportion of interaction frequency among members at different levels, the heterogeneity characteristics of the interpersonal relationship network within the team are reflected. At the same time, team belonging and role conflict perception indicators in periodic psychometric data are also included in the consideration of relational heterogeneity. By analyzing the degree of difference in these psychological indicators among different members, the evaluation dimensions of relational heterogeneity are further enriched, thereby comprehensively and accurately depicting the heterogeneity of the team at the task and relational levels.

[0117] The conflict risk prediction module is equipped with a trained machine learning model to predict the probability of high-intensity conflicts occurring in the future.

[0118] Specifically, the XGBoost model can be used. After multiple rounds of parameter adjustment and model iteration, an XGBoost model with high prediction accuracy and stability can be obtained. This model can predict the probability of high-intensity conflict occurring in the collaborative decision-making process of a virtual enterprise team in the future, based on the input task heterogeneity indicators and relationship heterogeneity indicators. When the predicted risk probability reaches the preset warning threshold, the system will automatically trigger the subsequent management strategy recommendation process.

[0119] The conflict management strategy recommendation module generates conflict management strategy suggestions based on the probability of high-intensity conflicts, as well as task heterogeneity indicators and relationship heterogeneity indicators.

[0120] The system comprehensively analyzes the input task heterogeneity and relationship heterogeneity indicators to identify the key drivers of conflict risk. It has pre-built multiple conflict management strategy template libraries, covering multiple dimensions such as task restructuring, role redefinition, communication mechanism optimization, team building activities, and conflict mediation skills training. Based on the probability level of conflict risk and the type and intensity of key drivers, the module intelligently matches and combines templates from the strategy library to generate targeted strategy recommendations.

[0121] These recommendations are not simply a list of strategies, but include specific implementation steps, expected results, responsible parties, and resource inputs, so that virtual enterprise teams can quickly take effective intervention measures based on the recommended strategies, reduce the possibility of high-intensity conflicts, and ensure the smooth progress of collaborative decision-making.

[0122] The feedback optimization module is used to verify the effectiveness of the predicted probability results and, based on the verification results, to train the machine learning model in reverse.

[0123] This module continuously collects data on actual conflict occurrences and the effects of strategy implementation during the virtual enterprise collaborative decision-making process. It compares this real-world data with previously predicted high-intensity conflict probabilities to evaluate the accuracy of the model's predictions and the effectiveness of its strategy recommendations. Specifically, it uses pre-defined evaluation metrics, such as prediction accuracy, the reduction in conflict incidence, and the improvement in team collaboration efficiency after strategy implementation, to quantitatively assess model performance and strategy effectiveness. If a significant deviation is found between the prediction results and the actual situation, or if the strategy implementation fails to achieve the expected results, the feedback optimization module automatically triggers the model's reverse training process. It uses this deviation data and effect evaluation results as new training samples to readjust the key parameters and weights in the machine learning model, continuously optimizing the model's conflict risk prediction algorithm and strategy matching logic.

[0124] Through this continuous closed-loop feedback and model iteration, the system can continuously improve the accuracy of conflict prediction and the applicability of management strategies as the virtual enterprise team changes dynamically and the actual business scenarios evolve, ensuring that the entire collaborative decision-making system always maintains a good operating status and decision support capability.

[0125] The visualization and interaction module is used to display the prediction results on a health dashboard, intuitively presenting key information such as the real-time status of various indicators of heterogeneity of virtual enterprise teams, high-intensity conflict risk warning levels, simulation of the expected effects of different management strategies, and trend analysis of historical conflict events.

[0126] The dashboard can combine color coding and dynamic charts. For example, green, yellow and red can be used to indicate low, medium and high risk statuses, respectively. A line graph can be used to show the change in conflict rate over time, and a bar chart can be used to compare the percentage increase in collaboration efficiency after the implementation of different strategies. This helps decision-makers quickly grasp the overall health status of the team and potential risk points.

[0127] The present invention also provides an electronic device, Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention, such as... Figure 3As shown, the electronic device may include a processor, a communications interface, memory, and a communication bus, wherein the processor, communications interface, and memory communicate with each other via the communication bus. The processor can invoke logical instructions from the memory, for example, to execute the following method:

[0128] S1. Collect basic information data, dynamic collaborative behavior data, and periodic psychological measurement data of team members;

[0129] S2. Construct task heterogeneity indicators and relationship heterogeneity indicators based on the collected data;

[0130] S3. Build and train a machine learning model;

[0131] S4. Deploy the trained machine learning model and predict the probability of high-intensity conflict occurring in the future based on task heterogeneity indicators and relationship heterogeneity indicators.

[0132] S5. Based on the predicted probability of high-intensity conflict, and combined with task heterogeneity indicators and relationship heterogeneity indicators, generate conflict management strategy recommendations.

[0133] Furthermore, the logical instructions in the aforementioned memory can be implemented as software functional units and sold or used as independent products, and can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0134] This invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, is implemented to perform the methods provided in the above embodiments, including, for example:

[0135] S1. Collect basic information data, dynamic collaborative behavior data, and periodic psychological measurement data of team members;

[0136] S2. Construct task heterogeneity indicators and relationship heterogeneity indicators based on the collected data;

[0137] S3. Build and train a machine learning model;

[0138] S4. Deploy the trained machine learning model and predict the probability of high-intensity conflict occurring in the future based on task heterogeneity indicators and relationship heterogeneity indicators.

[0139] S5. Based on the predicted probability of high-intensity conflict, and combined with task heterogeneity indicators and relationship heterogeneity indicators, generate conflict management strategy recommendations.

[0140] The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0141] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0142] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A collaborative decision-making method for virtual enterprises based on team heterogeneity identification and management strategy recommendation, characterized in that, include: S1. Collect basic information data, dynamic collaborative behavior data, and periodic psychological measurement data of team members; S2. Construct task heterogeneity indicators and relationship heterogeneity indicators based on the collected data; S3. Build and train a machine learning model; S4. Deploy the trained machine learning model and predict the probability of high-intensity conflict occurring in the future based on task heterogeneity indicators and relationship heterogeneity indicators. S5. Based on the predicted probability of high-intensity conflict, and combined with task heterogeneity indicators and relationship heterogeneity indicators, generate conflict management strategy recommendations.

2. The method according to claim 1, characterized in that, The following is stated: Basic information data includes: team members' educational background, professional skills, and work experience; Dynamic collaborative behavior data includes: semantic analysis of chat logs, frequency of meeting speeches, and contribution to document editing; Periodic psychometric data include: MBTI test and team atmosphere perception questionnaire.

3. The method according to claim 1, characterized in that: Set parameters for skill tag matching degree, knowledge graph distance, and task dependency complexity, and calculate task heterogeneity index based on basic information data, dynamic collaborative behavior data, and periodic psychological measurement data; By setting parameters such as language style difference, emotional tendency consistency, and social network density, and calculating relationship heterogeneity index based on basic information data, dynamic collaborative behavior data, and periodic psychological measurement data, a relationship heterogeneity index is obtained.

4. The method according to claim 1, characterized in that, Display the predicted probability results on a health dashboard.

5. The method according to claim 1, characterized in that, Based on the predicted probability results, the effectiveness is verified. Based on the verification results, the machine learning model is trained in reverse to continuously optimize the prediction accuracy of the machine learning model.

6. The method according to claim 1, characterized in that, The conflict management strategies in S5 include: Countermeasures include encouraging debate and setting challenging goals; A laissez-faire strategy, which includes reducing intervention and allowing differences to coexist; Harmony strategies include promoting communication and organizing team-building activities; Guidance strategies include introducing third-party coordinators.

7. The method according to claim 1, characterized in that, The "future period" is no more than one week.

8. A virtual enterprise collaborative decision-making system, characterized in that, The system, applicable to the method of any one of claims 1 to 7, comprises: The data acquisition module is used to collect basic information data, dynamic collaborative behavior data, and periodic psychological measurement data of team members; The heterogeneity calculation module is used to construct task heterogeneity indicators and relationship heterogeneity indicators based on the collected data; The conflict risk prediction module is equipped with a trained machine learning model to predict the probability of high-intensity conflicts occurring in the future. Conflict management strategy recommendation module: Based on the probability of high-intensity conflict, as well as task heterogeneity indicators and relationship heterogeneity indicators, it generates conflict management strategy suggestions; The feedback optimization module is used to verify the effectiveness of the predicted probability results and, based on the verification results, to train the machine learning model in reverse. The visualization and interaction module is used to display the prediction results on a health dashboard.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the method as described in any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 7.

Images