AI-based IT project stakeholder influence analysis and communication strategy optimization method

Abstract

This invention relates to the field of multimodal analysis technology and discloses an AI-based method for stakeholder influence analysis and communication strategy optimization in IT projects. The method includes: constructing a unified terminology for an IT domain knowledge graph; achieving multi-source data fusion through anonymization and feature correlation calculation; outputting fused data; realizing comprehensive identification and dynamic role mapping of IT project-specific stakeholders; mining latent stakeholders through trajectory influence values; constructing a dynamic mapping model to track role changes; reusing trajectory influence values ​​and role suitability to complete priority ranking; extracting four types of IT-specific influence features; binding influence and technical risk through correlation strength; calculating a dynamic influence index based on a priority index; constructing and calibrating an IT project-specific communication profile; generating personalized communication strategies; performing quantitative verification of strategy suitability; and outputting communication strategies; collecting dual-channel feedback data according to the influence update cycle and simultaneously adjusting the dynamic influence index and communication strategies.

Description

Technical Field

[0001] This invention relates to the field of multimodal analysis technology, specifically to an AI-based method for stakeholder influence analysis and communication strategy optimization in IT projects. Background Technology

[0002] In existing technologies, stakeholder influence analysis and communication strategy optimization methods are mostly general-purpose and not adapted to the specific scenarios of IT projects, including: technical complexity scenarios, agile iteration scenarios, and large and complex IT project scenarios. They generally adopt conventional data collection, static role definition, simple weighted influence calculation, and fixed communication strategy patterns. When applied to IT projects, they are difficult to match their technical attributes, iterative characteristics, and multi-level cross-organizational collaboration needs. Existing technologies suffer from several drawbacks: poor data collection and processing adaptability, failure to consider IT project characteristics, lack of targeted collection of technical data and cross-organizational stakeholder data, unresolved ambiguities in IT terminology, incomplete semantic analysis, and a lack of scenario-based design for sensitive information processing and multi-source data fusion. Stakeholder identification is incomplete, and role mapping is static: only explicit stakeholders can be identified, failing to uncover IT project-specific implicit stakeholders. Fixed role definitions cannot adapt to dynamic changes in stakeholder roles during project iterations. Influence analysis is detached from the IT context and static: using general influence features, failing to correlate with IT project technical risks, and not adapting to multi-level scenarios in large projects. Static influence calculations cannot respond to real-time changes in agile iteration scenarios. Communication strategies lack specificity, failing to consider differences in IT project stakeholder technical understanding and cross-organizational collaboration characteristics. Static strategy library matching lacks quantitative adaptability verification, making it difficult to overcome cross-organizational and multi-level communication barriers. Furthermore, a lack of dynamic closed-loop optimization mechanisms, a lack of systematic feedback collection and quantitative analysis processes, and an inability to synchronously adjust influence analysis results and communication strategies result in insufficient dynamic optimization capabilities, making it difficult to adapt to the iterative upgrade needs of IT projects. Therefore, there is a need to provide AI-based methods for stakeholder influence analysis and communication strategy optimization in IT projects. Summary of the Invention

[0003] The purpose of this invention is to provide an AI-based method for stakeholder influence analysis and communication strategy optimization in IT projects. To solve the aforementioned problems in the prior art, this invention achieves this through the following technical solution: The first part, the AI-based IT project stakeholder influence analysis and communication strategy optimization method provided by the embodiments of the present invention, specifically includes the following steps: Step 1: Collect data using a scenario-layered and multi-channel linkage model, construct a unified terminology for an IT domain knowledge graph, and achieve multi-source data fusion through anonymization and feature correlation calculation to output fused data; Step 2: Based on the fused data, achieve comprehensive identification and dynamic role mapping of IT project-specific stakeholders, directly identify explicit stakeholders and mine implicit stakeholders through trajectory influence values, build a dynamic mapping model to track role changes, and reuse trajectory influence values ​​and role suitability to complete priority ranking. Step 3: Combining stakeholder identification results with fused data, extract four types of IT-specific influence characteristics, bind influence with technological risk through correlation strength, and calculate a dynamic influence index based on priority index to achieve real-time index updates; Step 4: Based on the dynamic influence index, construct and calibrate a communication profile unique to the IT project, generate personalized communication strategies from four dimensions including channels and frequency, and perform quantitative verification of strategy adaptation in conjunction with the influence index to output the communication strategy. Step 5: Based on the communication strategy, build a closed-loop feedback and dynamic collaborative optimization mechanism. Collect feedback data from both channels according to the influence update cycle, quantify and analyze the feedback effect, and adjust the dynamic influence index and communication strategy in sync. Form a closed loop by verifying the optimization effect.

[0004] The second part, the AI-based IT project stakeholder influence analysis and communication strategy optimization system provided in this embodiment of the invention, specifically includes the following modules: Data fusion module: It collects data using a scenario-layered and multi-channel linkage model, constructs a unified terminology for an IT domain knowledge graph, and achieves multi-source data fusion through de-identification and feature correlation calculation, outputting fused data; Mapping and Adaptation Module: Based on fused data, it realizes comprehensive identification and dynamic role mapping of stakeholders unique to IT projects. It directly identifies explicit stakeholders and mines implicit stakeholders through trajectory influence values, builds a dynamic mapping model to track role changes, and reuses trajectory influence values ​​and role adaptability to complete priority ranking. The association update module combines stakeholder identification results with fused data to extract four types of IT-specific influence characteristics. It binds influence and technical risks through association strength and calculates a dynamic influence index based on priority index, enabling real-time index updates. Calibration Decision Module: Based on the dynamic influence index, construct and calibrate a communication profile unique to IT projects, generate personalized communication strategies, combine the influence index to perform quantitative verification of strategy adaptation, and output communication strategies. Collaborative Optimization Module: Based on the output communication strategy, a closed-loop feedback and dynamic collaborative optimization mechanism is constructed. Feedback data from both channels is collected according to the influence update cycle. After quantitatively analyzing the feedback effect, the dynamic influence index and communication strategy are adjusted simultaneously, and a closed loop is formed through the verification of optimization effect.

[0005] The beneficial effects of this invention are: 1. Specifically adapted to the unique scenarios of IT projects characterized by complex technology, agile iteration, and cross-organizational collaboration, a progressive and full-chain innovative technology system was constructed; a complete technology chain was designed, including multi-source data scenario-based preprocessing, dynamic stakeholder identification, risk-related influence analysis, personalized communication strategies, and closed-loop collaborative optimization. In the data processing stage, IT professional terminology was normalized and multi-source feature fusion was achieved. In the stakeholder management stage, the limitations of explicit identification were overcome to achieve implicit stakeholder mining and dynamic role mapping. 2. In the impact analysis phase, a dynamic computing model is constructed that deeply binds IT-specific characteristics and technical risks. In the strategy implementation and optimization phase, a closed-loop mechanism that can feed back to the data source is formed. This deeply integrates innovation in each phase and highly reuses data, effectively solving key defects of existing technologies in IT project scenarios such as poor adaptability, insufficient dynamic response, and weak communication targeting. It forms a unique solution that is different from existing general methods and has substantial technological progress. Moreover, each implementation plan is mutually supportive and indispensable, rather than a simple superposition of technologies, and has significant advantages in scenario adaptability and practical value. Attached Figure Description

[0006] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0007] Figure 1 This is a flowchart of the steps of the AI-based IT project stakeholder influence analysis and communication strategy optimization method provided in Embodiment 1 of the present invention; Figure 2 This is a schematic diagram of the structure of the AI-based IT project stakeholder influence analysis and communication strategy optimization system provided in Embodiment 2 of the present invention. Detailed Implementation

[0008] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0009] Example 1: As Figure 1 As shown in the embodiment of the present invention, the AI-based IT project stakeholder influence analysis and communication strategy optimization method specifically includes the following steps: Step 1: Collect data using a scenario-layered and multi-channel linkage model, construct a unified terminology for an IT domain knowledge graph, and achieve multi-source data fusion through anonymization and feature correlation calculation to output fused data; In this specific implementation, a "scenario-layered + multi-channel linkage" data collection model is adopted, covering all scenarios and all stakeholder types in IT projects. The specific collection scope and channels are as follows: Structured data: Collect IT project documents, organizational structure data, stakeholder basic information, and project progress data in batches through project management platforms, enterprise OA systems, and risk control platforms, with a focus on supplementing data from internal technical teams; Unstructured data: Collect authorized oral communication recordings, informal meeting minutes, collaboration tool messages, technical review comments, and requirement change feedback. Use professional speech-to-text tools to convert the recordings into text, manually remove irrelevant chatter, repetitive statements and other redundant information, and focus on collecting unstructured data related to technical communication. Dynamic scenario data: Collect stakeholder communication behavior data, technical risk correlation data, and iteration scenario data. This data is collected incrementally at fixed intervals using real-time monitoring communication tools and project iteration management tools. Communication behavior data includes: communication frequency, response time, and preferred communication channels. Technical risk correlation data includes: risk level and scope of risk impact for the modules responsible for each stakeholder. Iteration scenario data includes: requirement change records, iteration goal adjustment information, and changes in stakeholder concerns across iterations. IT Terminology Normalization Preprocessing: Construct an IT project domain knowledge graph covering core technical terms such as microservice architecture, database sharding, and cache penetration, associate the business meanings and technical scenarios corresponding to the terms, and use semantic association matching to process the collected unstructured text data. Match the terms in the text data with the standard terms in the knowledge graph to eliminate ambiguous terms and unify terminology expressions. For example, "cache penetration" and "cache invalidation penetration" can be unified into "cache penetration", and "microservice splitting" can be unified into "microservice architecture splitting". Privacy desensitization and multi-source data feature fusion processing: Sensitive information is anonymized and desensitized using feature anonymization mapping and scenario-based feature extraction methods. Core features of various data types are extracted, and multi-source data fusion is achieved through feature correlation calculation. Specific processing steps are as follows: Privacy anonymization: Sensitive stakeholder information is mapped to a unique anonymous identifier, while non-sensitive information such as stakeholder roles, responsibilities, and communication characteristics is retained; Multi-source data feature fusion: Extracting core features from various data types; extracting role, responsible module, and risk level features from structured data; extracting communication sentiment, technical needs, and opinion tendencies from unstructured data; extracting communication frequency, response efficiency, and changes in focus from dynamic data. Multi-source feature fusion is achieved through feature correlation calculation, using the formula: The analysis yielded the feature correlation between the i-th type of data features and the j-th type of data features. The value ranges from [0,1]. The greater the feature correlation, the closer the two types of features are related, and the higher the fusion value. in, This represents the number of stakeholders who are commonly associated with both the i-th and j-th data categories, i.e., the number of stakeholders who exist simultaneously in both data categories. The scenario adaptation coefficient is the matching coefficient between the i-th data feature and the j-th data feature, with a value range of [0.5, 1.0]. It is determined according to the IT project scenario. For example, the scenario adaptation coefficient between technical risk data and technical team communication data is 1.0, and the scenario adaptation coefficient between business requirement data and operation and maintenance team communication data is 0.6. Let be the total number of stakeholders included in the i-th data category; Let be the total number of stakeholders included in the j-th data category; Based on the obtained feature correlation scores, if the feature correlation score is greater than or equal to 0.7, the two types of data features are directly fused; if the feature correlation score is greater than or equal to 0.4 and less than 0.7, the common correlation part of the two types of features is extracted and fused; if the feature correlation score is less than 0.4, the two types of features are retained and the correlation score is labeled. Step 2: Based on the fused data, achieve comprehensive identification and dynamic role mapping of IT project-specific stakeholders, directly identify explicit stakeholders and mine implicit stakeholders through trajectory influence values, build a dynamic mapping model to track role changes, and reuse trajectory influence values ​​and role suitability to complete priority ranking. In a specific embodiment, the correlation influence trajectory analysis method is used to initially screen explicit stakeholders and mine implicit stakeholders. Explicit stakeholders are directly identified through the collected structured data. Based on the fused feature data, the communication trajectory, technical correlation trajectory, and risk correlation trajectory of each stakeholder are extracted, the correlation of the trajectory is analyzed, and implicit stakeholders who are not directly labeled but have potential impact on the project are mined. The communication trajectory includes: the frequency, response time, and content of communication with other stakeholders; the technical correlation trajectory includes: the technical modules the individual is responsible for or participates in, and their participation in technical reviews; the risk correlation trajectory includes: the risk level of the modules the individual is responsible for, and their participation in risk handling. The identification of latent stakeholders is based on the calculation of trajectory influence value, specifically using the following formula: Analysis yielded trajectory influence values ,in, The value represents the trajectory influence of the stakeholders, ranging from [0,2]. The communication trajectory influence coefficient has a value range of [0,1] and is calculated based on the communication frequency and response timeliness between stakeholders and explicit core stakeholders. The influence coefficient of the technology-related trajectory is calculated based on the frequency of stakeholders' participation in technology decision-making and technology review, with a value range of [0,1]. The risk-related trajectory impact coefficient has a value range of [0,1] and is calculated based on the stakeholder's participation in risk handling and risk warning. The trajectory duration coefficient, with a value range of [0.5, 1], is calculated based on the ratio of the duration of the stakeholder's trajectory to the current project cycle. By using the technical correlation trajectory impact coefficient parameter, we can focus on internal technical team stakeholders, and the trajectory impact value is used to prioritize stakeholders. Construct a dynamic mapping model of scenarios, roles, and responsibilities to replace static role definitions. Combine the collected dynamic scenario data and the stakeholder trajectory data mined in this step to define the core set of roles in the IT project, with each role corresponding to clear responsibility characteristics. Real-time monitoring of changes in the responsibilities and characteristics of stakeholders is conducted, and role changes are dynamically identified using role fit calculation. The specific calculation method is as follows: The analysis yields the role fit of the k-th stakeholder at time t. The value range is [0, 1.2]; where is the number of actual duty features of the kth stakeholder at time t, is the duty performance effect coefficient of the kth stakeholder at time t, and is the total number of core duty features corresponding to the initial role of the kth stakeholder; For example, when When the value is greater than or equal to 0.8, the stakeholder's current role is considered unchanged; when... When the value is greater than or equal to 0.4 and less than 0.8, it is determined that the stakeholder's role has undergone partial adjustment, and their duty characteristics are updated; when When the value is less than 0.4, it is determined that there has been a significant change in the stakeholder's role, and the role is rematched and the role information is updated; A scenario-adaptive priority index combined with role-adaptability is used to replace single-dimensional or weighted summation sorting. The specific calculation method is as follows: The analysis yields the priority index of the kth stakeholder. ,in, Let be the trajectory influence value of the kth stakeholder. For the role fit of the k-th stakeholder at time t, This is the scenario adaptation coefficient, with a value range of [0.5, 1.0], determined based on the current IT project scenario; Based on the priority index, stakeholders are divided into core priority, important priority, and general priority. Step 3: Combining stakeholder identification results with fused data, extract four types of IT-specific influence characteristics, bind influence with technological risk through correlation strength, and calculate a dynamic influence index based on priority index to achieve real-time index updates; In a specific embodiment, the extraction of IT project-specific influence features involves extracting IT project-specific influence features based on the data features fused in step one and the stakeholder identification results in step two. These features replace general features and primarily include the following four categories: Technical solution acceptance characteristics: The degree of acceptance of stakeholders' technical solutions, architecture design, and requirements implementation methods for IT projects. This is identified through extraction of unstructured communication data and semantic analysis to determine three tendencies: acceptance, opposition, and neutrality. Requirements change acceptance characteristics: Stakeholders' acceptance of requirements changes in agile iteration scenarios is determined by extracting dynamic scenario data and counting the number of times stakeholders agree, object, or propose modifications. System integration cooperation characteristics: The degree of cooperation among stakeholders in cross-organizational and cross-team collaborations regarding system integration and module docking is statistically analyzed by extracting communication behavior data and project progress data to determine the number of collaborative tasks completed and the number of collaborative tasks that were delayed. Technical risk control characteristics: the ability of stakeholders to identify and handle technical risks of the modules they are responsible for, and to extract risk-related data to statistically analyze the accuracy of risk warnings and the timeliness of risk handling; Establishing a two-way mapping relationship between influence and technological risk, the specific process is as follows: The collected technical risk data (risk level, affected modules, and scope of risk impact) is used to associate each stakeholder with the corresponding technical risk, extract risk association characteristics, and calculate the risk association strength to achieve a deep binding between influence and technical risk. The specific calculation method is as follows: The analysis yielded the correlation strength between the k-th stakeholder and the corresponding technological risk. ,in, This is the technology risk level coefficient, with a value range of [0.5, 1.0], determined based on the technology risk level of the IT project. This is the coefficient representing the scope of impact of technological risks, with a value range of [0.6, 1.0]. Let be the frequency coefficient of the k-th stakeholder's participation in the handling of technical risks, with a value range of [0,1]. This is the duration coefficient for technological risks, with a value range of [0.5, 1.0]. Risk correlation strength is simultaneously used for dynamic impact calculation and communication strategy generation; Based on the obtained correlation strength, the characteristic correlation-based dynamic influence index calculation formula is adopted: The dynamic influence index of the kth stakeholder was obtained through analysis. The value range is [0, 4.8]. Let k be the priority index of the kth stakeholder. Let be the technical solution approval coefficient for the k-th stakeholder, with a value range of [0,1]. Let be the technical solution approval coefficient for the k-th stakeholder, with a value range of [0,1]. Let be the demand change acceptance coefficient for the k-th stakeholder, with a value range of [0,1]. Let be the system integration cooperation coefficient of the k-th stakeholder, with a value range of [0,1]. Let be the technical risk control coefficient for the k-th stakeholder, with a value ranging from [0,1]. Let be the strength of the association between the k-th stakeholder and the corresponding technological risk. The influence update cycle is dynamically adjusted based on the IT project scenario. This is the stakeholder hierarchy attenuation coefficient, with a value range of [0.5, 1.0], determined based on the stakeholder's hierarchy within the project; We achieve accurate calculation of influence by summing features and correlating multi-dimensional coefficients, realize real-time updates of influence by updating the influence cycle, adapt to multi-level stakeholder scenarios in large and complex projects by using stakeholder hierarchy decay coefficients, and achieve deep binding of influence and technical risks by using risk correlation strength. For large-scale stakeholder networks of hundreds of people in large and complex IT projects, the influence of each stakeholder can be captured through feature extraction and correlation calculation. Without the need for complex network modeling, the influence analysis of large-scale stakeholder networks can be achieved through feature correlation and index calculation of individual stakeholders. Step 4: Based on the dynamic influence index, construct and calibrate a communication profile unique to the IT project, generate personalized communication strategies from four dimensions including channels and frequency, and perform quantitative verification of strategy adaptation in conjunction with the influence index to output the communication strategy. In a specific embodiment, a stakeholder communication profile specific to IT projects is constructed. This profile is tailored to the characteristics of the IT project scenario. The specific implementation process is as follows: Based on the multi-source fusion communication behavior data, the identified stakeholder role information, and the dynamic influence analysis results, the core dimensions of the communication profile are extracted, focusing on four exclusive dimensions that fit the needs of IT projects. Among them, the technology cognition dimension mainly reflects the stakeholder's understanding of IT terminology and the level of technical knowledge reserves; the communication preference dimension focuses on the communication channels that stakeholders prefer to use, the preferred communication frequency, and their needs for the level of detail in the communication content. The scenario adaptation dimension focuses on the changes in the communication priorities and concerns of stakeholders at different stages of IT projects; the cross-organizational collaboration dimension defines the communication permissions, information transmission needs, and collaboration process preferences of stakeholders across organizations. The profile data is processed using a profile similarity calibration method. By comparing the profile features of the target stakeholder with those of similar stakeholders, the similarity and common feature fit between the two are analyzed. If the similarity does not meet a reasonable standard, the profile features of the target stakeholder are calibrated and adjusted in conjunction with the typical profiles of similar stakeholders to ensure that the final communication profile is highly consistent with the actual scenario of the IT project and the real needs of stakeholders. The calibrated profile similarity data can be used for batch optimization of subsequent communication strategies. Generate scenario-based personalized communication strategies. Based on the results of the preliminary dynamic influence analysis and the calibrated communication profile, and combined with the current actual scenario of the IT project, targeted communication strategies are generated. The specific implementation process is divided into four dimensions: In terms of communication channel selection, we combine the communication preferences and dynamic influence levels of stakeholders with the needs of IT project scenarios, prioritize the use of commonly used collaboration tools in IT projects, and adopt different modes such as a combination of primary and backup channels, a single core channel, or batch communication channels according to the differences in the priority of stakeholder influence. In determining the communication frequency, we combine the dynamic influence level of stakeholders with the update cycle of IT project iteration scenarios, match the communication preferences of stakeholders, and set appropriate communication frequencies for stakeholders with different influence priorities. In terms of generating communication content, the content should be tailored to stakeholders’ technical knowledge level, the scenario adaptation requirements of different project stages, and the degree of connection between stakeholders and technical risks. For stakeholders with a high level of technical knowledge, the communication content can include detailed technical details, technical risk analysis, and suggestions for solution optimization. For stakeholders with a low level of technical knowledge, technical terminology should be simplified, and the focus should be on presenting project progress, core results, risk impact, and countermeasures. For stakeholders across organizations, the communication content should cover collaboration progress, information synchronization lists, and technical details within the scope of authority. In choosing the right time to communicate, we should combine the current stage of the IT project with the dynamic changes in stakeholders' concerns to select the most appropriate time to communicate. For different project stages, we should focus on communicating relevant key content with the corresponding core stakeholders to ensure communication efficiency and effectiveness. Communication strategy adaptability is assessed using a quantitative strategy adaptability verification method to ensure that the generated communication strategy accurately adapts to the IT project scenario and stakeholder needs. The specific implementation process is as follows: By combining previously acquired dynamic influence data, profile similarity data, and data on the correlation strength between stakeholders and technological risks, and incorporating considerations related to tolerance for communication delays, the suitability of communication strategies is evaluated through quantitative calculations. The adaptability assessment will determine whether the communication strategy should be implemented directly, whether details need to be adjusted, or whether it needs to be regenerated. Strategies that meet the adaptability standard will be implemented directly, strategies with average adaptability will only need to adjust details such as communication frequency and content level, and strategies that do not meet the adaptability standard will need to be regenerated. Step 5: Based on the communication strategy for output, build a closed-loop feedback and dynamic collaborative optimization mechanism. Collect feedback data from both channels according to the influence update cycle, quantify and analyze the feedback effect, and adjust the dynamic influence index and communication strategy in sync. Form a closed loop through the optimization effect verification. In a specific embodiment, a closed-loop optimization mechanism of feedback-analysis-adjustment-verification is constructed to achieve synergistic optimization of dynamic influence analysis and communication strategies. The specific implementation method is as follows: It adopts a dual-channel automatic data collection mode that combines active and passive feedback, which can complete the comprehensive collection of feedback data without much human intervention; Among them, proactive feedback data is collected through feedback questionnaires and communication evaluations after the implementation of communication strategies. The questionnaire content is designed in combination with the characteristics of IT project scenarios, focusing on four core dimensions: technical compatibility, reasonableness of content detail, appropriateness of communication frequency, and compatibility of communication channels. The questionnaires are automatically pushed to stakeholders through the project management platform. After stakeholders complete the questionnaires or evaluations, the system automatically collects relevant data. Passive feedback data is automatically collected through real-time monitoring of communication behavior, without requiring active operation by stakeholders. The data focuses on communication response, level of participation, number of feedback opinions, accuracy of understanding of needs, and efficiency of handling technical risks. The accuracy of understanding requirements can be derived by reverse engineering the results of subsequent project execution, and the efficiency of handling technical risks can be derived by reverse engineering the time spent on handling technical risks. The data collection cycle for feedback data should be consistent with the previous dynamic influence update cycle. After collection, the feedback data should be cleaned to remove invalid evaluations, extreme values ​​and other abnormal data, unify the data format, and link and store the cleaned feedback data with the previous multi-source fusion data, dynamic influence data and communication strategy data. The feedback effect is quantitatively analyzed. Based on the cleaned high-quality feedback data, the feedback effect quantitative analysis method is used to comprehensively evaluate the implementation effect of the communication strategy, and at the same time, the rationality of the previous dynamic influence analysis results is evaluated. The evaluation will comprehensively consider proactive feedback, reactive feedback, and the effectiveness of stakeholder participation in technology risk collaboration, while also taking into account the degree of deviation between proactive and reactive feedback. A feedback effectiveness-related index will be obtained through quantitative calculation. This index can reflect both the effectiveness of the communication strategy and the degree of matching between the results of the previous dynamic influence analysis and the actual situation. Based on the quantitative analysis of feedback effectiveness, stakeholder feedback is divided into different levels, with different follow-up handling methods corresponding to different levels. Among them, feedback with excellent effectiveness does not require any follow-up adjustments, feedback with satisfactory effectiveness only requires optimization of details, and feedback with unsatisfactory effectiveness requires simultaneous adjustment of dynamic influence analysis results and communication strategies. Based on the grading results derived from the quantitative analysis of feedback effects, the dynamic influence analysis results and communication strategies are adjusted and optimized simultaneously. Regarding dynamic influence adjustment, the dynamic influence index is adjusted based on the previously obtained dynamic influence index, combined with the quantitative analysis results of feedback effect and communication profile similarity data. After adjustment, the rationality of the index is verified to ensure that the adjusted index is within a reasonable range. The revised dynamic influence index will be updated in the previous influence analysis results. Regarding the adjustment of communication strategies, based on the feedback effect classification results and deviation analysis, we made targeted adjustments to relevant elements such as communication channels, communication frequency, communication content, and communication timing. We reused the previous strategy fit verification method to re-verify the fit of the adjusted communication strategies. The specific adjustment rules are as follows: For cases where the feedback effect is unsatisfactory, the communication channels and communication content should be adjusted first. If the deviation is due to technical understanding and adaptation issues, the technical details in the communication content should be simplified or supplemented. If the deviation is due to communication channel adaptation issues, the core communication channels should be replaced, and the communication strategy should be regenerated and verified after adjustment. If the feedback is satisfactory, only the communication details need to be optimized, without the need to regenerate the communication strategy. At the same time, after each project iteration cycle is completed, based on the accumulated feedback data, the stakeholder communication profile and scenario adaptation parameters are adjusted in sync to ensure that the communication strategy can adapt to the changing needs of cross-iteration scenarios. An optimization effect verification method is used to quantitatively verify the adjusted dynamic influence analysis results and communication strategies to ensure that the optimization effect meets the standards and forms a complete closed loop. During the quantitative verification process, the dynamic influence index after adjustment, the adaptability of the adjusted communication strategy, the feedback effect after adjustment, and other relevant data are comprehensively considered. Combined with the relevant benchmark data before optimization, the relevant index for optimization effect verification is obtained through quantitative calculation, and the effectiveness of the optimization effect is determined based on the index. The specific verification rules are as follows: if the optimization effect verification index reaches a reasonable standard, the optimization is deemed effective, the adjustment results are retained, and the next round of feedback data collection cycle begins; if the optimization effect verification index is within the critical range, the optimization is deemed effective, the relevant parameters are fine-tuned, and the verification is performed again; if the optimization effect verification index does not reach a reasonable standard, the optimization is deemed invalid, and the process is traced back to the feedback effect quantitative analysis stage to re-identify the cause of the deviation and make targeted adjustments. For each complete process of feedback collection, effect analysis, collaborative adjustment and optimization verification, a detailed optimization log is recorded, which includes information such as adjustment parameters, reasons for adjustment, and verification results. Log data will be synchronously added to the multi-source fusion data from the previous stage, and used for data reuse and model optimization in subsequent steps to achieve continuous accumulation and optimization of data throughout the entire process.

[0010] Example 2: As Figure 2 As shown in the embodiment of the present invention, the AI-based IT project stakeholder influence analysis and communication strategy optimization system specifically includes the following modules: Data fusion module: It collects data using a scenario-layered and multi-channel linkage model, constructs a unified terminology for an IT domain knowledge graph, and achieves multi-source data fusion through de-identification and feature correlation calculation, outputting fused data; Mapping and Adaptation Module: Based on fused data, it realizes comprehensive identification and dynamic role mapping of stakeholders unique to IT projects. It directly identifies explicit stakeholders and mines implicit stakeholders through trajectory influence values, builds a dynamic mapping model to track role changes, and reuses trajectory influence values ​​and role adaptability to complete priority ranking. The association update module combines stakeholder identification results with fused data to extract four types of IT-specific influence characteristics. It binds influence and technical risks through association strength and calculates a dynamic influence index based on priority index, enabling real-time index updates. Calibration Decision Module: Based on the dynamic influence index, construct and calibrate a communication profile unique to IT projects, generate personalized communication strategies, combine the influence index to perform quantitative verification of strategy adaptation, and output communication strategies. Collaborative Optimization Module: Based on the output communication strategy, a closed-loop feedback and dynamic collaborative optimization mechanism is constructed. Feedback data from both channels is collected according to the influence update cycle. After quantitatively analyzing the feedback effect, the dynamic influence index and communication strategy are adjusted simultaneously, and a closed loop is formed through the verification of optimization effect.

[0011] The above provides a detailed description of one embodiment of the present invention, but the content described is only a preferred embodiment of the present invention and should not be considered as limiting the scope of the present invention. The above formulas are all dimensionless numerical calculations, and the formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world situation. The preset parameters in the formulas are set by those skilled in the art based on actual conditions and historical experience, and can be adjusted according to actual conditions. The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. All equivalent changes and improvements made in accordance with the scope of the present invention should still fall within the patent coverage of the present invention.

Claims

1. An AI-based method for stakeholder influence analysis and communication strategy optimization in IT projects, characterized by: Includes the following steps: Data is collected using a scenario-layered and multi-channel linkage model. A unified terminology for an IT domain knowledge graph is constructed. Multi-source data fusion is achieved through de-identification and feature correlation calculation, and fused data is output. Based on fused data, we can achieve comprehensive identification and dynamic role mapping of stakeholders unique to IT projects. We can directly identify explicit stakeholders and mine implicit stakeholders through trajectory influence values. We can build a dynamic mapping model to track role changes and reuse trajectory influence values ​​and role suitability to complete priority ranking. By combining stakeholder identification results with fused data, four types of IT-specific influence characteristics are extracted. Influence is linked to technological risks through the strength of association, and a dynamic influence index is calculated by combining a priority index to achieve real-time index updates. Based on the dynamic influence index, a unique communication profile for IT projects is constructed and calibrated to generate personalized communication strategies. The influence index is then used to quantify and validate the strategy adaptation, and the communication strategies are output. Based on the communication strategy for output, a closed-loop feedback and dynamic collaborative optimization mechanism is constructed. Feedback data from both channels is collected according to the influence update cycle. After quantitatively analyzing the feedback effect, the dynamic influence index and communication strategy are adjusted simultaneously, and a closed loop is formed through the verification of optimization effect.

2. The AI-based IT project stakeholder influence analysis and communication strategy optimization method according to claim 1, characterized in that, The method for constructing a unified terminology for an IT domain knowledge graph is as follows: It adopts a scenario-layered and multi-channel linkage data collection mode to cover all scenarios and all stakeholder types in IT projects, collecting structured data, unstructured data and dynamic scenario data respectively; Structured data is collected in batches through project management platforms, enterprise OA systems, and risk control platforms, with a focus on supplementing relevant data from internal technical teams; Unstructured data is collected after authorization from stakeholders. Recordings are converted into text and redundant information is removed using professional speech-to-text tools. The focus is on collecting technical communication data. Dynamic scene data is collected incrementally at fixed intervals through real-time monitoring and communication tools and project iteration management tools. Simultaneously, an IT project domain knowledge graph covering core technical terms is constructed, linking the business meanings and technical scenarios corresponding to the terms. Semantic association matching is used to process unstructured text data, eliminating ambiguous terms and unifying their expressions.

3. The AI-based IT project stakeholder influence analysis and communication strategy optimization method according to claim 1, characterized in that, The method for achieving multi-source data fusion is as follows: A feature anonymization mapping and contextual feature extraction method is used to anonymize and desensitize sensitive information of stakeholders while retaining non-sensitive information. Extract the core features of various types of data. For structured data, extract features such as roles, responsible modules, and risk levels. For unstructured data, extract features such as communication sentiment, technical demands, and opinion tendencies. For dynamic data, extract features such as communication frequency, response efficiency, and changes in focus. Multi-source feature fusion is achieved by calculating feature correlation. Depending on the magnitude of feature correlation, methods such as direct fusion, fusion by extracting common related parts, or fusion by retaining features and labeling correlation are adopted to complete the fusion of multi-source data and output fused data.

4. The AI-based IT project stakeholder influence analysis and communication strategy optimization method according to claim 1, characterized in that, The method for identifying latent stakeholders is as follows: By directly identifying explicit stakeholders through the collected structured data, and based on the output fused data, extracting the communication trajectory, technology association trajectory, and risk association trajectory of each stakeholder; The communication trajectory includes the frequency of communication with other stakeholders, response time, and content of communication; the technology-related trajectory includes the technical modules that the person is responsible for or participates in and the participation in technical reviews; and the risk-related trajectory includes the risk level of the modules the person is responsible for and the degree of participation in risk handling. The correlation of three types of trajectories was analyzed using the correlation impact trajectory analysis method. Implicit stakeholders were identified by calculating the trajectory impact value, with a focus on internal technical team stakeholders. Implicit stakeholders who were not directly labeled but had potential impact on the project were also identified. The trajectory impact value was used in subsequent stakeholder priority ranking.

5. The AI-based IT project stakeholder influence analysis and communication strategy optimization method according to claim 1, characterized in that, The method for constructing a dynamic mapping model to track role changes is as follows: Construct a dynamic mapping model of scenarios, roles, and responsibilities to replace the traditional static role definition. Combine the collected dynamic scenario data and the stakeholder trajectory data mined in this step to define the core role set of the IT project and set clear responsibility characteristics for each core role. Real-time monitoring of changes in the responsibilities and characteristics of stakeholders is conducted. A dynamic calculation method for role fit is used to identify role changes. By calculating the fit between the actual responsibilities and characteristics of stakeholders and the initial core responsibilities and characteristics, it is determined whether the stakeholder's role has not changed, has been partially adjusted, or has undergone significant changes. Corresponding actions are taken to perform operations such as no need to update, updating responsibilities and characteristics, or rematching roles and updating information, thereby achieving dynamic role tracking.

6. The AI-based IT project stakeholder influence analysis and communication strategy optimization method according to claim 1, characterized in that, The method for completing the priority sorting is as follows: The scenario-adaptive priority index calculation method is adopted to replace the single-dimensional or weighted summation sorting method. The stakeholder trajectory influence value and role adaptability data obtained in this step are reused, and the scenario adaptability coefficient determined according to the current IT project scenario is combined to calculate the priority index of each stakeholder. Based on the priority index, stakeholders are divided into three levels: core priority, important priority, and general priority, thus completing the stakeholder priority ranking. The ranking results simultaneously support the influence analysis in step three and the communication strategy generation in step four.

7. The AI-based IT project stakeholder influence analysis and communication strategy optimization method according to claim 1, characterized in that, The method for linking influence and technological risk through the strength of association is as follows: Based on the output fusion data features and stakeholder identification results, four types of IT project-specific influence features are extracted: technical solution acceptance features, requirement change acceptance features, system integration cooperation features, and technical risk control features, to replace traditional general influence features. By combining the collected technical risk data, each stakeholder is associated with the corresponding technical risk, risk association characteristics are extracted, and a risk association strength calculation method is adopted. Combining the technical risk level, the scope of impact, and the stakeholder's participation in risk management, the association strength between the stakeholder and the corresponding technical risk is calculated, thus achieving a deep binding between influence and technical risk.

8. The AI-based IT project stakeholder influence analysis and communication strategy optimization method according to claim 1, characterized in that, The method for achieving real-time index updates is as follows: The dynamic influence index calculation method based on feature correlation is adopted. The dynamic influence index of each stakeholder is calculated by using the stakeholder priority index, the four types of IT-specific influence feature coefficients extracted in this step, and the risk correlation strength data, combined with the influence update cycle dynamically adjusted according to the IT project scenario and the hierarchical decay coefficient determined according to the stakeholder project level. The system achieves real-time updates of the dynamic influence index through an influence update cycle, adapts to multi-level stakeholder scenarios in large and complex projects through a hierarchical decay coefficient, and realizes influence analysis of large-scale stakeholder networks through the correlation of individual stakeholder characteristics and index calculation.

9. The AI-based IT project stakeholder influence analysis and communication strategy optimization method according to claim 1, characterized in that, The method for generating personalized communication strategies is as follows: Based on integrated communication behavior data, stakeholder role information, and dynamic influence index, four core dimensions are extracted: technology cognition, communication preference, scenario adaptation, and cross-organizational collaboration, to construct a unique stakeholder communication profile for IT projects. A portrait similarity calibration method is used to compare the portrait features of the target related person with those of similar related persons, and to calibrate and adjust the portraits that do not reach a reasonable similarity. By combining the dynamic influence index with the calibrated communication profile, personalized and scenario-based communication strategies are generated for stakeholders with different influence priorities, technical awareness levels, and cross-organizational attributes from four dimensions: communication channels, communication frequency, communication content, and communication timing.

10. The AI-based IT project stakeholder influence analysis and communication strategy optimization method according to claim 1, characterized in that, The method for forming a closed loop through optimization effect verification is as follows: A dual-channel automatic data collection mode combining proactive and passive feedback is adopted. Feedback data is collected according to the impact update cycle. Proactive feedback data is collected through questionnaires and evaluations, while passive feedback data is automatically collected by monitoring communication behavior. After collection, abnormal data is cleaned and stored in association. The feedback effect quantitative analysis method is used to evaluate the effectiveness of the communication strategy and the rationality of the dynamic influence index, classify the feedback effect level, and adjust the dynamic influence index and communication strategy in sync. The optimization effect verification method is used to quantitatively verify the adjustment results, determine whether the optimization is effective, partially effective or ineffective and handle it accordingly. After each round of process is completed, the optimization log is recorded and fed back to the previous data fusion to form a complete closed-loop optimization mechanism.

Images