Cross-department collaboration efficiency optimization method fused with social network analysis

Abstract

The application discloses a cross-department collaboration efficiency optimization method fused with social network analysis, comprising the following steps: multi-modal data acquisition and space-time labeling: collecting communication data, behavior data and physiological signal data in cross-department collaboration, adding time stamp and node identification to each data, and associating the node identification with department attributes and role characteristics in the social network; dynamic identification of cognitive bias: based on the social network node interaction data, adopting natural language processing and sentiment analysis technology to detect logical fallacy of the communication text, and generating bias type label and propagation intensity index in combination with physiological signal wave, the application improves collaboration efficiency, accurately identifies the cognitive bias propagation path in cross-department collaboration through multi-modal data fusion and social network analysis, dynamically intervenes to reduce invalid communication, shortens the task response and dispute resolution time, enhances collaboration resilience, and the space-time enhancement and critical state early warning mechanism of the propagation graph can avoid the risk of network structure imbalance in advance.

Description

Technical Field

[0001] This invention belongs to the field of enterprise management technology, and in particular relates to a method for optimizing cross-departmental collaboration efficiency by integrating social network analysis. Background Technology

[0002] Existing cross-departmental collaboration efficiency optimization technologies are primarily based on Social Network Analysis (SNA) and data-driven methods. By collecting communication data (such as email exchanges and meeting minutes) and behavioral data (such as task response time and resource sharing frequency) from department members, social network models are constructed to analyze indicators such as node centrality and connection strength, identifying key collaboration paths and core coordinators. Simultaneously, information synchronization is achieved by combining process standardization tools (such as BPMN workflow engines) and collaboration platforms (such as WeChat Work and Lark). Communication barriers are reduced by assigning dedicated coordinators and establishing cross-departmental SOPs (Standard Operating Procedures). Some technologies also incorporate machine learning algorithms to cluster historical collaboration data, predict potential collaboration bottlenecks, and provide decision support for resource allocation. These technologies have achieved certain results in improving information transparency and reducing redundant work, and have been widely applied in cross-departmental management in manufacturing, internet, and finance sectors.

[0003] Existing technologies have significant limitations: First, data collection is limited to explicit behavioral data (such as email frequency and task progress), neglecting physiological signals (such as heart rate variability) and implicit cognitive characteristics (such as logical biases). This results in superficial identification of collaboration barriers, failing to pinpoint fundamental cognitive contradictions. Second, social network analysis is mostly static snapshot analysis, lacking dynamic propagation modeling, making it difficult to track the spread of biases in real time, and intervention strategies lag behind changes in collaboration. Third, it ignores the innovative value of anti-consensus connections, simply suppressing differing opinions as conflicts, thus missing optimization opportunities brought about by differentiated cognition. Fourth, it lacks resilience design for extreme scenarios (such as core personnel departures or system crashes), relying on manual contingency plans for network resilience and lacking digital twin simulation and dynamic reorganization capabilities. These shortcomings limit the optimization effects of existing technologies and make them unsuitable for the complex and ever-changing needs of cross-departmental collaboration. Summary of the Invention

[0004] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a method for optimizing cross-departmental collaboration efficiency by integrating social network analysis, which solves the problems of relying on explicit behavioral data, lacking cognitive bias analysis, dynamic intervention and anti-consensus mining, and having weak resilience.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A method for optimizing cross-departmental collaboration efficiency by incorporating social network analysis includes the following steps:

[0007] (1) Multimodal data collection and spatiotemporal annotation: Collect communication data, behavioral data and physiological signal data in cross-departmental collaboration, and attach timestamps and node identifiers to each data point. The node identifiers are associated with departmental attributes and role characteristics in the social network.

[0008] (2) Dynamic identification of cognitive bias: Based on social network node interaction data, natural language processing and sentiment analysis technology are used to detect logical fallacies in communication texts, and physiological signal fluctuations are combined to generate bias type labels and propagation intensity indicators.

[0009] (3) Construction of propagation graph and spatiotemporal enhancement: Based on social network analysis, a dynamic directed graph model is constructed, which maps department members as nodes and deviation propagation paths as edges. The edge weights are calculated by combining the deviation type, propagation frequency and node influence through degree centrality and betweenness centrality. At the same time, the time interval and spatial distribution of collaborative events are transformed into a spatiotemporal weight matrix to dynamically adjust the priority of deviation propagation.

[0010] (4) Neutralizing node intelligent intervention: The propagation probability of the deviation propagation path is calculated based on the confidence level obtained from social network analysis. When the propagation probability reaches the preset threshold, intervention is triggered. The cognitive flexibility index is calculated based on the cross-departmental coordination records of nodes in the social network. Members that meet the preset standards are used as neutralizing nodes, inserted into the propagation path, and customized intervention strategies are pushed.

[0011] (5) Anti-consensus collaboration activation: Through social network analysis, identify anti-consensus connections in the network where the degree of difference in viewpoints meets the preset value and the frequency of interaction is stable. Based on the node connection strength and knowledge complementarity, extract the intersection and difference features of arguments through structured debate and text comparison algorithms to explore potential innovative value.

[0012] (6) Digital twin simulation and resilience optimization: Construct a virtual twin for cross-departmental collaboration, map the node connection relationship of social networks, simulate extreme scenarios such as core node failure and resource interruption, and form a resilient optimal network structure; and map the neural synchronization index of the physical network in real time. This index is calculated based on the physiological signal synchronization during node interaction to generate highly adaptable team recommendations.

[0013] Preferably, in step (1), the communication data includes meeting recordings, emails and instant messaging content, the behavioral data includes task response time and document sharing frequency, and the physiological signal data includes heart rate variability and skin conductance activity; the collected data is preprocessed, including data cleaning, format conversion and noise removal; the department attribute tags are based on the business department to which they belong, and the role attribute tags include technical expert and coordinator.

[0014] Preferably, the data preprocessing in step (1) also includes outlier identification and correction, using the 3σ principle to remove extreme values ​​from physiological signal data, and using interpolation to complete missing behavioral data; node identifiers are associated one-to-one with node IDs in the social network to ensure data traceability.

[0015] Preferably, in step (2), the logical fallacy includes confirmation bias and anchoring effect; further including:

[0016] 1) Dynamic dictionary expansion: Combining departmental business terminology databases, the deviation detection rules are dynamically updated through comparative learning, and the terminology usage preferences of nodes in the social network are associated when identifying logical fallacies;

[0017] 2) Physiological-cognitive coupling analysis: Establish the mapping relationship between physiological signals and cognitive biases, predict the critical point of bias propagation through a temporal network based on long short-term memory, and trigger an early warning when the anxiety index of a certain node exceeds the threshold for 4 consecutive hours. The anxiety index is quantified based on the fluctuation amplitude of skin conductance.

[0018] Preferably, in step (2) 1), the departmental business terminology library is dynamically updated with the development of departmental business, including professional terms from the R&D department, marketing department, and finance department. R&D department terms include technical parameters and iteration cycles, while marketing department terms include user pain points and conversion rates. The update cycle is once a month, and newly added terms need to be reviewed and confirmed by the cross-departmental expert committee. Terms with a review pass rate of ≥80% are included in the dictionary.

[0019] Preferably, in step (3), the spatiotemporal weight matrix is ​​calculated by weighting a time decay function (based on the historical response delay distribution of nodes) and a spatial distance coefficient (based on the distribution of office areas and the frequency of online interaction), and the time interval is corrected based on the cross-time zone response delay coefficient (calculated according to the working time overlap rate of nodes in different time zones); further including:

[0020] 1) Causal inference and spurious correlation filtering: Causal graphs are used to separate real driving factors from spurious correlations, and to locate the core nodes and connections that truly affect collaboration efficiency in social networks;

[0021] 2) Critical state early warning and network contraction: Monitor the abrupt changes in the betweenness centrality and clustering coefficient of the network. When the betweenness centrality of the core nodes decreases by 40% (based on 3 months of historical data statistics), the network contraction strategy is initiated to retain 20% of the core nodes (which must meet the requirements of information transmission efficiency and connection breadth being in the top 30%) to maintain the transmission of key information.

[0022] Preferably, in step (3) 2), the key information transmission includes cross-departmental task instructions, resource allocation requirements and problem feedback results. The information transmission efficiency is evaluated by the average response time and accuracy. The network contraction strategy has a startup delay of ≤1 hour. The replacement candidate pool of core nodes is sorted by connection strength. The number of candidate nodes is twice that of the reserved nodes to ensure redundancy.

[0023] Preferably, in step (4), the propagation probability threshold is ≥70% (based on statistics of 500+ historical collaboration cases), and the cognitive flexibility index is ≥0.8 (calculated based on node-based cross-departmental task completion rate and knowledge sharing frequency); further including:

[0024] 1) Real-time verification of intervention effect: The difference-in-differences method was used to compare the changes in collaboration efficiency between the intervention group and the control group. When the causal effect value was less than 5%, the intervention was withdrawn and a new plan was generated.

[0025] 2) Memory Curve Activation: Combining the Ebbinghaus forgetting curve, push weak activation tasks (such as 10-minute quick synchronization meetings) to connections in social networks with knowledge transfer intervals exceeding 7 days to improve knowledge retention rate.

[0026] Preferably, in step (5), the difference in opinions in the anti-consensus connection is ≥60% and the interaction frequency is stable (≥3 times per week); the text comparison algorithm uses a bidirectional LSTM model to extract argument feature vectors, and calculates the intersection and difference of arguments through cosine similarity; further including:

[0027] 1) Quantification of the value of controversy: Calculate the innovation potential index of anti-consensus connection (difference in viewpoints × 0.3 + frequency of interaction × 0.2 + reuse rate of historical solutions × 0.5). When the index exceeds 0.6, multi-agent collaborative simulation is initiated.

[0028] 2) Intelligent Arbitration and Rights Allocation: Based on the Nash bargaining model, the residual value of collaboration is automatically allocated. Cross-departmental rights are automatically transferred through smart contracts, and the allocation weight is related to the contribution of nodes in the social network.

[0029] Preferably, in step (6), resilience is assessed based on the network information transmission efficiency retention rate (≥80%) after node failure; the synchronization coefficient of the neural synchronization index is ≥0.6; further including:

[0030] 1) Crisis response plan generation: Pre-store network reorganization strategies for more than 100 crisis scenarios, predict the optimal response path through a time-series graph neural network, and design the reorganization strategy based on the node redundancy and connection resilience of the social network.

[0031] 2) Neural synchronization analysis: Combining EEG phase synchronization with memory curves, a personnel fit score and knowledge activation plan are generated to optimize the node pairing in the social network.

[0032] The technical effects and advantages of this invention's cross-departmental collaboration efficiency optimization method integrating social network analysis are as follows:

[0033] 1. This invention improves collaboration efficiency by accurately identifying the propagation path of cognitive biases in cross-departmental collaboration through multimodal data fusion and social network analysis, dynamically intervening to reduce ineffective communication, and shortening task response and dispute resolution time.

[0034] 2. This invention enhances collaborative resilience and provides a spatiotemporal enhancement and critical state early warning mechanism for the propagation graph, which can avoid the risk of network structure imbalance in advance and maintain the efficiency of key information transmission through core node optimization in extreme scenarios.

[0035] 3. This invention taps into innovative value, and the anti-consensus collaboration activation mechanism transforms departmental cognitive differences into innovative driving forces. Through structured analysis, it extracts complementary arguments, thereby improving the innovation and implementation rate of the solution.

[0036] 4. This invention optimizes resource allocation, combines intelligent intervention at nodes with digital twin simulation, and achieves precise personnel matching and resource scheduling, reducing cross-departmental resource waste.

[0037] 5. This invention ensures the sustainability of collaboration, filters out spurious factors through causal inference, focuses on the core elements that truly drive collaboration efficiency, and forms an iterative optimization loop to adapt to the dynamic development needs of enterprises. Attached Figure Description

[0038] Figure 1 This is a flowchart illustrating the cross-departmental collaboration efficiency optimization method that integrates social network analysis proposed in this invention. Detailed Implementation

[0039] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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 are within the scope of protection of the present invention.

[0040] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0041] The cross-departmental collaboration efficiency optimization method of this invention is achieved through six core steps: multimodal data acquisition and spatiotemporal annotation, dynamic identification of cognitive biases, propagation graph construction and spatiotemporal enhancement, intelligent intervention of neutralizing nodes, activation of anti-consensus collaboration, and digital twin simulation and resilience optimization. Each step is integrated through Social Network Analysis (SNA), forming a closed loop of data-driven, dynamic intervention, and simulation optimization. The specific implementation process is as follows: Figure 1 As shown.

[0042] Example 1

[0043] This embodiment provides a method for optimizing cross-departmental collaboration efficiency by incorporating social network analysis, applicable to optimizing cross-departmental production collaboration in manufacturing. Specific implementation details include:

[0044] Objective: To resolve the issues of low component compatibility (historical compatibility rate of 68%) and response delay (average 48 hours) caused by discrepancies in the transmission of technical parameters between the R&D and production departments of an automobile manufacturing company.

[0045] Implementation steps:

[0046] Step (1): Multimodal data acquisition and spatiotemporal annotation

[0047] Content collected:

[0048] Communication data: Meeting recordings between the R&D and Production departments (3 cross-departmental meetings per week), email exchanges (an average of 20 emails per day), and WeChat chat logs;

[0049] Behavioral data: Production department's response time to R&D parameters (recorded continuously for 30 days), drawing sharing log (5 times per week);

[0050] Physiological signal data: Smart bracelets were worn by 10 core members to collect heart rate variability (60-100ms is the normal range) and skin conductance (resting value ≤5μS);

[0051] Preprocessing: Use Python scripts to clean up fuzzy recordings (remove 3 files with excessive noise) and deduplicate emails (delete 8 duplicate emails). Use linear interpolation to fill in the 3 missing response times in the production department.

[0052] Node identifiers: Label roles such as "R&D-Structural Engineer" and "Production-Assembly Team Leader" and associate them with department attributes (such as "R&D Department" and "Production Department").

[0053] Step (2): Dynamic identification of cognitive biases:

[0054] 1) Dynamic dictionary expansion:

[0055] Establish a “R&D-Production Terminology Comparison Table” to map the “tolerance ±0.1mm” in R&D to the “assembly accuracy grade C” in production. Update the table weekly through comparative learning (add 5 new sets of terminology matching in the first week).

[0056] 2) Physiological-cognitive coupling analysis:

[0057] Using LSTM time series network analysis, it was found that when the R&D engineer's heart rate variability dropped to 45ms (below the threshold) and skin conductance activity increased by 30%, it corresponded to the confirmation bias of "rejecting parameter adjustment" (a total of 12 such events were identified); the production department repeatedly cited "historical assembly experience" in 8 meetings and judged it to be the anchoring effect.

[0058] Step (3): Construction of propagation graph and spatiotemporal enhancement dynamic directed graph model: with 20 members as nodes and deviation propagation path as edges (e.g., the “parameter solidification” deviation edge weight of R&D team leader → production team leader is 0.7);

[0059] 1) Causal inference: After filtering out the spurious correlation of "high document sharing frequency → high adaptation rate" (correlation coefficient 0.3), it was found that "existence of dedicated coordinators" was the real driving factor (contribution 72%).

[0060] 2) Spatiotemporal weight matrix: Time decay coefficient for cross-time zone collaboration (files sent by R&D night shift) is 0.6, and spatial distance coefficient between workshop and office building is 0.8.

[0061] Step (4): Intelligent intervention at the neutralization node:

[0062] Production department process engineers with a cognitive flexibility index ≥ 0.8 (historical coordination success rate 85%, cross-domain knowledge matching degree 0.82) were selected as neutralization nodes;

[0063] 1) Intervention strategy: Organize "parameter adaptation workshops", push out 5 cross-departmental tolerance adjustment cases, and monitor in real time that the probability of deviation propagation has decreased from 75% to 30%;

[0064] 2) Memory Curve Activation: For connections with a knowledge transfer interval of more than 7 days (such as the transfer of "material strength parameters" between R&D and production), a notification will be sent 10 minutes after the push (triggering twice in the first week).

[0065] Step (5): Activate anti-consensus collaboration:

[0066] Identify anti-consensus connections (R&D advocates "technology first," while production advocates "efficiency first") with 65% difference in viewpoints but 5 interactions per week. Extract the intersection through structured debate ("ensure accuracy while compressing assembly time") and generate 3 compromise solutions.

[0067] Step (6): Digital twin simulation and resilience optimization:

[0068] A virtual twin was constructed to simulate the scenario of "core R&D members leaving the company". The optimal reorganization scheme (using 2 reserve coordinators) was predicted when the network betweenness centrality decreased by 35%. The results showed that the adaptation rate stability was improved by 20% after the intervention.

[0069] Implementation results:

[0070] The component compatibility rate has increased from 68% to 92% (an improvement of 35%).

[0071] The production department's response time has been reduced from 48 hours to 24 hours (a 50% reduction);

[0072] Deviation propagation events decreased from 12 times per week to 4 times (a reduction of 67%).

[0073] Example 2

[0074] This embodiment provides a method for optimizing cross-departmental collaboration efficiency by integrating social network analysis, which is used for optimizing cross-departmental product iteration collaboration in internet companies. Specific implementation details include:

[0075] Objective: To resolve the problem of long iteration cycles (average 30 days) and low solution approval rate (60%) caused by differences in understanding of requirements between the product department and the R&D department of an internet company.

[0076] Implementation steps:

[0077] Steps (1)-(4): Basic Process

[0078] Refer to steps (1)-(4) of Example 1 to complete data collection (collecting DingTalk chat records and requirement document modification logs), deviation identification (discovering the product department's "user experience is absolutely the priority" confirmation error), graph construction (nodes are 15 core members), and neutral node screening (product director, cognitive flexibility index 0.85).

[0079] Step (5): Activate anti-consensus collaboration:

[0080] 1) Quantification of the value of the controversy:

[0081] Calculate the innovation potential index of anti-consensus connection: 70% of opinion difference × 0.3 + 12 times of weekly interaction frequency × 0.2 + 50% of historical solution reuse rate × 0.5 = 0.62 (exceeding the 0.6 threshold), and start multi-agent collaborative simulation;

[0082] 2) Intelligent Arbitration and Rights Allocation:

[0083] The text comparison algorithm (based on word vector cosine similarity) is used to extract the intersection of the product's "button path simplification" and the R&D's "interface compatibility" ("reducing user operation error rate"), and the remaining value of collaboration is allocated through the Nash bargaining model (55% for the product department and 45% for the R&D department).

[0084] Step (6): Digital Twin Simulation:

[0085] Simulate a "requirement change" scenario to verify the robustness of the anti-consensus solution (fluctuation amplitude ≤10%) and output the optimal collaboration path (product requirements → R&D interface design → joint review).

[0086] Implementation results:

[0087] The iteration cycle was shortened from 30 days to 22 days (a reduction of 27%).

[0088] The approval rate of the proposal increased from 60% to 84% (an increase of 40%).

[0089] The anti-consensus connection has been transformed into three core functional optimizations (such as the "one-click return" function).

[0090] Example 3

[0091] This embodiment provides a method for optimizing cross-departmental collaboration efficiency by integrating social network analysis, which is used to optimize cross-departmental risk control collaboration in financial institutions. Specific implementation details include:

[0092] Objective: To resolve the problem of low loan approval efficiency (50 loans per day) and high dispute rate (35%) caused by discrepancies in risk assessment standards between the risk control department and the credit department of a certain bank.

[0093] Implementation steps:

[0094] Steps (1)-(5): Basic Process

[0095] Complete data collection (collecting approval documents, meeting minutes, and credit system operation logs), deviation identification (anchoring effect of the risk control department's "over-reliance on historical bad debt rates"), graph construction (nodes are 20 risk control / credit specialists), intervention (neutralizing node is the deputy head of risk control), and anti-consensus activation (identifying connections with 62% difference in viewpoints).

[0096] Step (6): Digital twin simulation and resilience optimization:

[0097] 1) Crisis response plan generation:

[0098] Pre-store 100+ scenarios such as "risk control supervisor resignation", and use time sequence graph neural network to predict the optimal response path (enable senior loan officer to coordinate temporarily);

[0099] 2) Neural synchronicity analysis:

[0100] Using EEG equipment to collect brainwaves, it was found that the phase synchronization coefficient between risk control specialist C and credit specialist D was 0.65 (≥0.6 indicates high compatibility), and the approval efficiency improved after pairing. "Risk assessment consensus cases" were pushed to members with a synchronization coefficient of 0.4, and the synchronization coefficient rose to 0.6 after 3 weeks.

[0101] Implementation results:

[0102] The average daily number of approvals increased from 50 to 80 (a 60% increase);

[0103] The dispute rate decreased from 35% to 15.75% (a 55% decrease);

[0104] The approval time for highly adaptable team combinations has been reduced by 40%.

[0105] Example 4

[0106] This embodiment provides a method for optimizing cross-departmental collaboration efficiency by integrating social network analysis, applicable to optimizing cross-departmental marketing collaboration in fast-moving consumer goods (FMCG) companies. Specific implementation details include:

[0107] Purpose of implementation: To resolve the high stockout rate (15%) caused by a dispute between the marketing department and supply chain department of a fast-moving consumer goods company regarding "promotional inventory".

[0108] Implementation steps:

[0109] Step (2): Dynamic identification of cognitive biases:

[0110] 1) Dynamic dictionary expansion:

[0111] Establish a "Marketing-Supply Chain Terminology Database" and define "sold out" as "inventory turnover days < 3 days", which will be updated monthly (a new "pre-sale conversion" will be added with a "safety stock coefficient of 0.8").

[0112] 2) Physiological-cognitive coupling analysis:

[0113] Identify the "optimistic expectation bias" in the marketing department (triggered when skin conductance increases by 30%) and the "conservative inventory bias" in the supply chain department (triggered when heart rate variability drops to 50ms).

[0114] Implementation results:

[0115] The out-of-stock rate decreased from 15% to 5% (a 67% decrease);

[0116] The frequency of cross-departmental email disputes decreased from 8 times per week to 3 times (a reduction of 62.5%).

[0117] Example 5

[0118] This embodiment provides a method for optimizing cross-departmental collaboration efficiency by incorporating social network analysis, applicable to optimizing cross-departmental diagnostic and treatment collaboration in medical institutions. Specific implementation details include:

[0119] Objective: To address the problem of long consultation time (60 minutes) and low accuracy (78%) caused by information transmission delays between the emergency department and laboratory department of a certain hospital.

[0120] Implementation steps:

[0121] Step (3): Propagation map construction and spatiotemporal enhancement:

[0122] 1) Critical state early warning and network contraction:

[0123] Monitoring revealed a 40% decrease in the betweenness centrality of the connection between emergency department doctors and laboratory technicians (due to delays in laboratory reports), triggering network contraction and retaining 20% ​​of the core nodes (head nurses in the emergency department and team leaders in the laboratory).

[0124] Key information transmission: By transmitting task instructions through the "emergency inspection green channel", the information transmission efficiency (average response time) has been reduced from 25 minutes to 10 minutes.

[0125] Implementation results:

[0126] The average consultation time was reduced from 60 minutes to 35 minutes (a reduction of 42%).

[0127] The accuracy of information delivery increased from 78% to 96% (an improvement of 23%).

[0128] Comparative Example 1

[0129] This comparison provides a traditional collaborative optimization method.

[0130] Implementation objective: An electronics company is attempting to optimize R&D and procurement collaboration using traditional methods (goal: shorten the material selection cycle).

[0131] Implementation steps:

[0132] A cross-departmental meeting is held once a week, and only meeting minutes are recorded.

[0133] Shared material documents (without deviation analysis);

[0134] Information is relayed by department managers (without graph construction or node intervention).

[0135] Implementation results:

[0136] The dispute over material selection has lasted for 15 days (without any optimization).

[0137] Information delay rate: 30% (no improvement);

[0138] The collaboration efficiency improvement is only 12% (far lower than the 35% in Embodiment 1 of the present invention).

[0139] Compared with Examples 1-5 and Comparative Example 1, the method of the present invention has significant advantages in optimizing cross-departmental collaboration efficiency.

[0140] Differences in technical approaches: Examples 1-5 all adopt the six-step process of claim 1, integrating multimodal data (physiological signals + behavioral data + communication data), cognitive bias propagation maps, dynamic intervention on social networks, and other technologies. For example, Example 1 identifies biases through physiological-cognitive coupling analysis, and Example 3 uses digital twins to simulate extreme scenarios. Comparative Example 1 relies only on traditional meetings and document sharing, without involving social network analysis and bias identification.

[0141] Comparison of effects: The improvement in collaboration efficiency in Examples 1-5 (27%-35%) far exceeds that of Comparative Example 1 (12%). Specifically:

[0142] Dispute resolution: The iteration cycle of Example 2 was shortened by 27%, while the dispute in Comparative Example 1 lasted for more than 15 days because it failed to explore the value of anti-consensus.

[0143] Information transmission: Example 5 reduced the information latency rate from 30% to 10% through a network contraction strategy, while Comparative Example 1 had a latency rate of 30% due to path rigidity;

[0144] Innovation Value: Example 2 activates anti-consensus connection to generate 3 innovative solutions, while Comparative Example 1 misses potential optimization opportunities.

[0145] In summary, this invention, through the deep coupling of social network analysis and dynamic intervention, addresses the neglect of implicit collaboration barriers by traditional methods, and significantly improves the resilience and efficiency of cross-departmental collaboration.

[0146] The above embodiments can be implemented in whole or in part by software, hardware, firmware or other arbitrary combinations. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product.

[0147] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0148] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0149] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of protection of the claims.

[0150] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for optimizing cross-departmental collaboration efficiency by integrating social network analysis, characterized in that, Includes the following steps: (1) Multimodal data collection and spatiotemporal annotation: Collect communication data, behavioral data and physiological signal data in cross-departmental collaboration, and attach timestamps and node identifiers to each data point. The node identifiers are associated with departmental attributes and role characteristics in the social network. (2) Dynamic identification of cognitive bias: Based on social network node interaction data, natural language processing and sentiment analysis technology are used to detect logical fallacies in communication texts, and physiological signal fluctuations are combined to generate bias type labels and propagation intensity indicators. (3) Construction of propagation graph and spatiotemporal enhancement: Based on social network analysis, a dynamic directed graph model is constructed, which maps department members as nodes and deviation propagation paths as edges. The edge weights are calculated by combining the deviation type, propagation frequency and node influence through degree centrality and betweenness centrality. At the same time, the time interval and spatial distribution of collaborative events are transformed into a spatiotemporal weight matrix to dynamically adjust the priority of deviation propagation. (4) Neutralizing node intelligent intervention: The propagation probability of the deviation propagation path is calculated based on the confidence level obtained from social network analysis. When the propagation probability reaches the preset threshold, intervention is triggered. The cognitive flexibility index is calculated based on the cross-departmental coordination records of nodes in the social network. Members that meet the preset standards are used as neutralizing nodes, inserted into the propagation path, and customized intervention strategies are pushed. (5) Anti-consensus collaboration activation: Through social network analysis, identify anti-consensus connections in the network where the degree of difference in viewpoints meets the preset value and the frequency of interaction is stable. Based on the node connection strength and knowledge complementarity, extract the intersection and difference features of arguments through structured debate and text comparison algorithms to explore potential innovative value. (6) Digital twin simulation and resilience optimization: Construct a virtual twin for cross-departmental collaboration, map the node connection relationship of social networks, simulate extreme scenarios such as core node failure and resource interruption, and form a resilient optimal network structure; and map the neural synchronization index of the physical network in real time. This index is calculated based on the physiological signal synchronization during node interaction to generate highly adaptable team recommendations.

2. The method for optimizing cross-departmental collaboration efficiency by incorporating social network analysis as described in claim 1, characterized in that, In step (1), the communication data includes meeting recordings, emails and instant messaging content, the behavioral data includes task response time and document sharing frequency, and the physiological signal data includes heart rate variability and skin conductance activity; the collected data is preprocessed, including data cleaning, format conversion and noise removal. Department attribute tags are based on the business department to which they belong, and role attribute tags include technical expert and coordinator.

3. The method for optimizing cross-departmental collaboration efficiency by incorporating social network analysis as described in claim 2, characterized in that, The data preprocessing in step (1) also includes outlier identification and correction, using the 3σ principle to remove extreme values ​​from physiological signal data, and using interpolation to complete missing behavioral data; node identifiers are associated one-to-one with node IDs in the social network to ensure data traceability.

4. The method for optimizing cross-departmental collaboration efficiency by incorporating social network analysis as described in claim 1, characterized in that, In step (2), logical fallacies include confirmation bias and anchoring effect; further including: 1) Dynamic dictionary expansion: Combining departmental business terminology databases, the deviation detection rules are dynamically updated through comparative learning, and the terminology usage preferences of nodes in the social network are associated when identifying logical fallacies; 2) Physiological-cognitive coupling analysis: Establish the mapping relationship between physiological signals and cognitive biases, predict the critical point of bias propagation through a temporal network based on long short-term memory, and trigger an early warning when the anxiety index of a certain node exceeds the threshold for 4 consecutive hours. The anxiety index is quantified based on the fluctuation amplitude of skin conductance.

5. The method for optimizing cross-departmental collaboration efficiency by incorporating social network analysis as described in claim 4, characterized in that, In step (2) 1), the departmental business terminology library is dynamically updated with the development of departmental business. It includes professional terms from the R&D department, marketing department, and finance department. R&D department terms include technical parameters and iteration cycles, while marketing department terms include user pain points and conversion rates. The update cycle is once a month. New terms need to be reviewed and confirmed by the cross-departmental expert committee. Terms with a review pass rate of ≥80% are included in the dictionary.

6. The method for optimizing cross-departmental collaboration efficiency by incorporating social network analysis as described in claim 1, characterized in that, In step (3), the spatiotemporal weight matrix is ​​calculated by weighting the time decay function with the spatial distance coefficient, and the time interval is corrected based on the cross-time zone response delay coefficient; further including: 1) Causal inference and spurious correlation filtering: Causal graphs are used to separate real driving factors from spurious correlations, and to locate the core nodes and connections that truly affect collaboration efficiency in social networks; 2) Critical state early warning and network contraction: Monitor abrupt changes in the network betweenness centrality and clustering coefficient. When the betweenness centrality of core nodes decreases by 40%, initiate a network contraction strategy to retain 20% of the core nodes to maintain the transmission of critical information.

7. The method for optimizing cross-departmental collaboration efficiency by incorporating social network analysis as described in claim 6, characterized in that, In step (3) 2), the key information transmission includes cross-departmental task instructions, resource allocation requirements and problem feedback results. The information transmission efficiency is evaluated by the average response time and accuracy. The network contraction strategy has a startup delay of ≤1 hour. The replacement candidate pool of core nodes is sorted by connection strength. The number of candidate nodes is twice that of the reserved nodes to ensure redundancy.

8. The method for optimizing cross-departmental collaboration efficiency by incorporating social network analysis as described in claim 1, characterized in that, In step (4), the propagation probability threshold is ≥70%, and the cognitive flexibility index is ≥0.8; further including: 1) Real-time verification of intervention effect: The difference-in-differences method was used to compare the changes in collaboration efficiency between the intervention group and the control group. When the causal effect value was less than 5%, the intervention was withdrawn and a new plan was generated. 2) Memory Curve Activation: Combining the Ebbinghaus forgetting curve, push weak activation tasks to connections in social networks with knowledge transfer intervals exceeding 7 days to improve knowledge retention rate.

9. The method for optimizing cross-departmental collaboration efficiency by incorporating social network analysis as described in claim 1, characterized in that, In step (5), the difference in opinions in the anti-consensus connection is ≥60% and the interaction frequency is stable, ≥3 times per week; the text comparison algorithm uses a bidirectional LSTM model to extract argument feature vectors, and calculates the intersection and difference of arguments through cosine similarity; further including: 1) Quantification of the value of controversy: Calculate the innovation potential index of anti-consensus connection. Potential index = difference of opinion × 0.3 + interaction frequency × 0.2 + reuse rate of historical solutions × 0.

5. When the index exceeds 0.6, multi-agent collaborative simulation is initiated. 2) Intelligent Arbitration and Rights Allocation: Based on the Nash bargaining model, the residual value of collaboration is automatically allocated. Cross-departmental rights are automatically transferred through smart contracts, and the allocation weight is related to the contribution of nodes in the social network.

10. The method for optimizing cross-departmental collaboration efficiency by incorporating social network analysis as described in claim 1, characterized in that, In step (6), resilience is assessed based on the network information transmission efficiency retention rate after node failure; The synchronization coefficient of the neural synchronization index is ≥0.6; further including: 1) Crisis response plan generation: Pre-store network reorganization strategies for more than 100 crisis scenarios, predict the optimal response path through a time-series graph neural network, and design the reorganization strategy based on the node redundancy and connection resilience of the social network. 2) Neural synchronization analysis: Combining EEG phase synchronization with memory curves, a personnel fit score and knowledge activation plan are generated to optimize the node pairing in the social network.

Images