Cross-platform enterprise reputation collaborative optimization method and system

Abstract

The invention relates to the technical field of enterprise collaborative management, in particular to a cross-platform enterprise reputation collaborative optimization method and system, and the method comprises the following steps: collecting social media, e-commerce platform and forum reputation data in real time, integrating the data to obtain an initial index, calculating a correction score in combination with time freshness and monthly activity grading distribution weight, and obtaining a final index; identifying a conduction relation tracking path to calculate a risk value, adjusting the weight to improve participation degree dynamic weighting when the risk exceeds warning, and periodically monitoring change to obtain a collaborative optimization record. According to the method, the reputation data of the social media, the e-commerce platform and the forum are integrated in real time, the correction score is calculated in combination with the time freshness parameter and the monthly live user grading distribution weight factor, new data are periodically collected, and changes are monitored in combination with historical records, so that the timeliness and the accuracy of reputation evaluation are improved; the cross-platform risk propagation probability is effectively reduced, the resource configuration efficiency is optimized, the response speed and the overall benefit of enterprise reputation management are enhanced, and more efficient collaborative optimization is realized.

Description

Technical Field

[0001] This invention relates to the field of enterprise collaborative management technology, and in particular to a cross-platform enterprise reputation collaborative optimization method and system. Background Technology

[0002] Enterprise collaborative management technology involves improving the efficiency of collaborative work within and between enterprises through information technology. It aims to optimize resource allocation, improve operational efficiency, and strengthen decision support. Through management systems, process reengineering, and systematic technological means, it achieves efficient attainment of enterprise goals. It not only emphasizes the efficient flow and sharing of information but also stresses the optimization of organizational structure and workflows to enhance overall operational efficiency. Among these methods, enterprise reputation collaborative optimization refers to optimizing the maintenance and improvement of enterprise reputation through collaborative mechanisms across different platforms during the enterprise management process. This is typically achieved through enterprise reputation management systems, evaluation models, data collection and analysis methods, and cross-platform collaborative mechanisms. It primarily relies on the establishment of a reputation scoring system, based on customer feedback, social evaluations, and other data, combined with historical enterprise performance and industry standards for assessment.

[0003] In the current process of corporate reputation optimization, the main reliance is on static reputation scoring systems and historical data evaluation. There is a lack of tracking capabilities for cross-platform reputation transmission mechanisms, and the inability to dynamically adjust weights based on timeliness and platform activity leads to delayed response to reputation events and insensitive risk detection. For example, when negative reviews on social media spread rapidly to e-commerce platforms, timely intervention is not possible, resulting in increased corporate reputation losses, low management efficiency, and impacting the timeliness of resource allocation and the accuracy of decision support. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies by proposing a cross-platform enterprise reputation collaborative optimization method.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a cross-platform enterprise reputation collaborative optimization method, comprising the following steps:

[0006] S1: Obtain corporate reputation scores, user reviews, and activity levels from social media platforms, e-commerce platforms, and forums; collect raw data in real time; and perform preliminary data integration to obtain initial reputation indicator data.

[0007] S2: Based on the reputation score of the initial reputation index data, calculate the standard deviation and mean deviation of the scores of each platform, and combine the time freshness parameter and the platform monthly active user tiered allocation weight factor to calculate the corrected reputation score value and obtain the reputation score correction data.

[0008] S3: Based on the reputation score correction data, identify the reputation transmission relationship from social media platforms to e-commerce platforms, compare the score change trend and platform interaction data to track the propagation path, calculate the reputation propagation risk value, and obtain reputation propagation risk data;

[0009] S4: Based on the reputation propagation risk data, when the reputation propagation risk value exceeds the reputation risk warning value of the digital risk framework, perform the operation of adjusting the recommendation weight coefficient and improving user participation, and perform dynamic weighted calculation to obtain the reputation optimization execution record;

[0010] S5: Periodically collect new reputation data of enterprises on the platform, combine the initial reputation indicator data and the reputation optimization execution record, monitor the dynamic changes in enterprise reputation in real time, and obtain cross-platform collaborative optimization records.

[0011] As a further aspect of the present invention, the initial reputation index data includes social media reputation data, e-commerce platform reputation data, and forum reputation data; the reputation score correction data includes weighted score values, standard deviation adjustment values, and time freshness weighting factors; the reputation propagation risk data includes risk value indicators, propagation path analysis results, and trend change comparison results; the reputation optimization execution record includes weight adjustment parameters, participation enhancement measures, and dynamic weighted calculation results; and the cross-platform collaborative optimization record includes data comparison reports, monitoring logs, and enterprise reputation optimization effect evaluation records.

[0012] As a further aspect of the present invention, the step of obtaining the initial reputation index data specifically comprises:

[0013] S111: Obtain user reviews from social media platforms, e-commerce platforms, and forums; determine the sentiment polarity of the review texts; and conduct differential analysis on the proportion of positive reviews and the proportion of negative reviews to obtain multi-source sentiment bias data.

[0014] S112: For the enterprise reputation scores and user activity collected from various platforms, statistical analysis of user interaction volume and evaluation timeliness data is conducted. Combined with the multi-source sentiment data, the correlation of each indicator is quantified and comprehensively evaluated to establish platform-related reputation heat data.

[0015] S113: Based on the platform-related reputation popularity data of each platform, align and quantify the data from different sources, and perform cross-platform aggregation to generate initial reputation indicator data.

[0016] As a further aspect of the present invention, the reputation score correction data acquisition step specifically comprises:

[0017] S211: Based on the reputation score set in the initial reputation index data, calculate the standard deviation of the reputation scores of all platforms, and calculate the deviation of each platform's score from the mean of all platform scores to obtain the reputation score dispersion measurement result.

[0018] S212: Based on the initial reputation index data, obtain the data collection time and monthly active user data of each platform, calculate the time freshness parameter, and perform hierarchical quantification according to the number of monthly active users to obtain platform dynamic weight data.

[0019] S213: Calculate the reputation scores of each platform, the discrete measurement results of the reputation scores, and the dynamic weight data of the platforms. Combine the negative public opinion impact index and the time conversion constant to obtain the corrected reputation score value and establish the reputation score correction data.

[0020] As a further aspect of the present invention, the step of acquiring reputation dissemination risk data specifically comprises:

[0021] S311: Based on the reputation score correction data, extract the continuously corrected score sequences of social platforms and e-commerce platforms, calculate the Pearson correlation coefficient of the daily score change rate of the two sequences, quantify the synchronicity of the score change trend, and obtain the platform reputation correlation trend record.

[0022] S312: Based on the reputation correlation trend records of the platform, and by tracking the posting behavior of highly interactive users under the negative topics of the platform on another platform, the number of content forwards and time delays are counted, the weighted interaction index is calculated, and cross-platform reputation transmission strength data is established.

[0023] S313: Based on the cross-platform reputation transmission strength data, combined with the platform reputation correlation trend records, normalized interaction volume of multiple propagation paths, target platform vulnerability and maximum response cycle are introduced to calculate and obtain reputation propagation risk value, and record and integrate to generate reputation propagation risk data.

[0024] As a further aspect of the present invention, the reputation optimization execution record acquisition step specifically comprises:

[0025] S411: Based on the reputation spread risk data, determine whether the reputation spread risk value exceeds the preset digital risk framework reputation risk warning value. If it exceeds the threshold, determine that a risk warning has been triggered and output the risk warning trigger record.

[0026] S412: Based on the risk warning trigger record, perform the recommendation weight coefficient adjustment operation on the downstream e-commerce platform, and perform the user participation improvement operation to integrate and establish a platform intervention strategy combination;

[0027] S413: Execute the aforementioned platform intervention strategy combination, comprehensively analyze the changing trends of reputation score correction data before and after the intervention, the changes in platform activity and user engagement, and at the same time evaluate the strength of the recommendation weight adjustment intervention action, correlate and integrate the information, and archive the evaluation, intervention action and execution time together to generate a reputation optimization execution record.

[0028] As a further aspect of the present invention, the cross-platform collaborative optimization record acquisition step specifically comprises:

[0029] S511: Periodically collect new reputation data of enterprises on social media platforms and e-commerce platforms, obtain the latest reputation score correction data, and compare it with the initial reputation index data to evaluate the direction and magnitude of the reputation score changes on each platform and obtain a record of reputation change differences.

[0030] S512: Based on the recorded differences in reputation changes and in conjunction with the records of reputation optimization implementation, conduct a comprehensive analysis of the overall trend of reputation scores of all monitored platforms, the activity of users jumping between platforms, and the time interval since the last intervention, evaluate the synergistic effect of reputation optimization measures across multiple platforms, and establish a cross-platform reputation synergy index.

[0031] S513: Based on the cross-platform reputation collaboration data, integrate it with the qualitative evaluation conclusions in the reputation change difference record and the reputation optimization execution record to construct a structured document including synergy effect evaluation and multi-dimensional data changes, and obtain cross-platform collaborative optimization records.

[0032] A cross-platform enterprise reputation collaborative optimization system, including:

[0033] The data acquisition module is used to execute S1: acquire corporate reputation scores, user reviews and activity levels from social media platforms, e-commerce platforms and forums, collect raw data in real time, and perform preliminary data integration to obtain initial reputation indicator data;

[0034] The consistency assessment module is used to perform S2: based on the reputation score of the initial reputation index data, the standard deviation and mean deviation of the scores of each platform are calculated, and combined with the time freshness parameter and the platform monthly active user tiered allocation weight factor, the corrected reputation score value is calculated to obtain the corrected reputation score data;

[0035] The transmission path identification module is used to perform S3: based on the reputation score correction data, identify the reputation transmission relationship from the social media platform to the e-commerce platform, compare the score change trend and platform interaction data to track the transmission path, calculate the reputation transmission risk value, and obtain reputation transmission risk data;

[0036] The optimization strategy execution module is used to execute S4: based on the reputation propagation risk data, when the reputation propagation risk value exceeds the reputation risk warning value of the digital risk framework, perform recommendation weight coefficient adjustment and user engagement improvement operations, and perform dynamic weighted calculation to obtain reputation optimization execution records;

[0037] The collaborative optimization module is used to execute S5: periodically collect new reputation data of enterprises on the platform, combine the initial reputation index data and the reputation optimization execution record, monitor the dynamic changes in enterprise reputation in real time, and obtain cross-platform collaborative optimization records.

[0038] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0039] This invention integrates reputation data from social media, e-commerce platforms, and forums in real time, combines time freshness parameters and monthly active user tiered weighting factors to calculate and correct scores, identifies the reputation transmission relationship from social media to e-commerce platforms, tracks the propagation path, calculates risk values, and dynamically adjusts recommendation weight coefficients to increase user engagement and dynamically weights the data when the risk exceeds the warning threshold. Periodically collecting new data and combining it with historical records to monitor changes improves the timeliness and accuracy of reputation assessment, effectively reduces the probability of cross-platform risk propagation, optimizes resource allocation efficiency, enhances the responsiveness and overall effectiveness of enterprise reputation management, and achieves more efficient collaborative optimization. Attached Figure Description

[0040] Figure 1 This is a flowchart of the main steps of the present invention;

[0041] Figure 2 This is a flowchart of the process for obtaining initial reputation index data in this invention;

[0042] Figure 3 This is a flowchart of the reputation score correction data acquisition process for this invention;

[0043] Figure 4 This is a flowchart illustrating the reputation dissemination risk data acquisition process of this invention.

[0044] Figure 5 This is a flowchart of the reputation optimization execution record acquisition process of the present invention;

[0045] Figure 6 This is a flowchart of the cross-platform collaborative optimization record acquisition process of the present invention. Detailed Implementation

[0046] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0047] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0048] Please see Figure 1 A cross-platform enterprise reputation collaborative optimization method includes the following steps:

[0049] S1: Obtain enterprise reputation scores, user reviews, and activity levels from social media platforms, e-commerce platforms, and forums. Collect raw data through real-time API interfaces, perform preliminary data integration, construct a multi-platform reputation data set, and obtain initial reputation indicator data.

[0050] S2: Based on the reputation score of the initial reputation index data, calculate the standard deviation and mean deviation of the scores of each platform, and combine the time freshness parameter (the reciprocal of the difference between the data collection time and the current time (unit: hours⁻¹), the smaller the difference, the larger the parameter value) and the monthly active user classification of the platform (digital platform influence assessment standard (Level I: MAU≥100 million, Level II: 10 million ≤ MAU<100 million, Level III: MAU<10 million)) to allocate weight factors, calculate the corrected reputation score value, and obtain the corrected reputation score data;

[0051] S3: Based on reputation score correction data, identify the reputation transmission relationship from social media platforms to e-commerce platforms, track the transmission path by comparing score change trends and platform interaction data, calculate the reputation transmission risk value (risk indicator defined by international brand assessment standards (0-1 range, 0.35 is the orange warning line)), and obtain reputation transmission risk data;

[0052] S4: Based on reputation propagation risk data, when the reputation propagation risk value exceeds the reputation risk warning value of the digital risk framework (the enterprise reputation risk threshold stipulated by the digital governance framework, 0.35), the recommendation weight coefficient (an adjustable parameter of 0.1-1.0 as defined by the Internet recommendation system standard) is adjusted on the downstream platform, and user engagement is improved simultaneously. The reputation optimization execution record is obtained by dynamically weighting the calculation based on the platform activity and the corrected reputation score.

[0053] S5: Periodically collect new reputation data of enterprises on various platforms, combine it with the initial reputation indicator data and reputation optimization execution records, compare the data, monitor the dynamic changes in enterprise reputation in real time according to the security monitoring plan, and obtain cross-platform collaborative optimization records.

[0054] Initial reputation metrics data include social media reputation data, e-commerce platform reputation data, and forum reputation data. Reputation score correction data includes weighted score values, standard deviation adjustment values, and time freshness weighting factors. Reputation dissemination risk data includes risk value indicators, dissemination path analysis results, and trend change comparison results. Reputation optimization execution records include weight adjustment parameters, participation enhancement measures, and dynamic weighted calculation results. Cross-platform collaborative optimization records include data comparison reports, monitoring logs, and enterprise reputation optimization effect evaluation records.

[0055] Please see Figure 2 Step S1 is as follows:

[0056] S111: Obtain user reviews from social media platforms, e-commerce platforms, and forums; determine the sentiment polarity of the review texts; and conduct differential analysis on the proportion of positive reviews and the proportion of negative reviews to obtain multi-source sentiment bias data.

[0057] The system acquires user reviews from social media platforms, e-commerce platforms, and forums. For example, for a newly released phone model, "New Pro1," user reviews were collected from social media platform A ("Excellent battery life, fast charging"), e-commerce platform B ("Smooth system, but average camera performance"), and a tech forum C ("New Pro1 gaming performance test: noticeable overheating in high frame rate mode"). These unstructured review texts are first processed using a pre-defined sentiment dictionary containing positive sentiment words (such as "excellent," "smooth," "excellent") and negative sentiment words (such as "average," "noticeable overheating," "lag"). Each review is then judged for sentiment polarity. For instance, the review "Excellent battery life, fast charging" is judged as having a strong sentiment polarity. A comment containing 2 positive sentiment words and 0 negative sentiment words is considered a positive review. The comment "The system is smooth, but the photo quality is average" is judged to contain 1 positive sentiment word and 1 negative sentiment word. By setting a rule that "when the number of positive sentiment words is greater than the number of negative sentiment words, the overall review is positive," this comment is considered neutral to slightly positive, and for simplicity, it is counted as a positive review. The post "Significant heat generation in high frame rate mode" is judged as a negative review. After completing the polarity judgment of all 15,000 reviews collected from platforms A, B, and C, the number of positive reviews and negative reviews on each platform are counted. If platform A collected 5,000 reviews, of which 4,000 are positive and 1,000 are negative, then its positive review ratio is... The proportion of negative reviews is Subsequently, a difference analysis was conducted on these two ratios, that is, the difference between the proportion of positive reviews and the proportion of negative reviews was calculated as a measure of the platform's sentiment. For platform A, the difference was [value missing]. Similarly, if e-commerce platform B has a 75% positive review rate and a 25% negative review rate, then the difference is... If the positive review rate for the C Technology Forum is 40% and the negative review rate is 60%, then the difference is... In this way, the evaluation data from the three platforms are integrated and quantified to obtain multi-source sentiment bias data.

[0058] S112: For enterprise reputation scores and user activity collected from various platforms, statistical analysis of user interaction volume and evaluation timeliness data is conducted. Combined with multi-source sentiment data, correlation quantification and comprehensive evaluation of various indicators are carried out to establish platform-related reputation heat data.

[0059] For example, the product rating of "New Pro1" on e-commerce platform B is 4.8 out of 5, and the daily active users on social media platform A related to this topic are 5 million. Simultaneously, user interaction data, such as the total number of likes, comments, and shares of relevant content, as well as the timeliness of evaluations, are collected. Based on the multi-source sentiment data obtained from the aforementioned steps, the correlation of various indicators is quantified and comprehensively evaluated. This process uses specific formulas to ensure the objectivity of the results. Before starting the calculation, some raw data needs to be standardized. For example, corporate reputation ratings are usually on a 5-point or 10-point scale. To eliminate differences in units, they need to be uniformly converted to values ​​between 0 and 1. For example, the 4.8 points (out of 5) on platform B are standardized to... Evaluation of timeliness The calculation is based on the time elapsed since the evaluation was published, with a maximum timeframe of 365 days. The calculation method is as follows: If a review was posted 10 days ago, its timeliness is [not specified]. Concentration of negative reviews This is used to measure the severity of negative feedback. Its setting references the coefficient of variation of the number of negative reviews within a specific time window (e.g., the past 7 days). Through backtesting analysis of historical data, it sets a threshold when the coefficient of variation is within a certain range. When the interval is, The value is 0.2, when... When the interval is, The value is 0.5. When it is greater than 1.5, The value is set to 1.0. This setting is based on the following experiment: 50 different products were selected, and data were observed within 90 days after their respective new product launches. It was found that when a public relations crisis or serious quality problem occurred, negative reviews surged in a short period of time, and the corresponding coefficient of variation all exceeded 1.5. Under normal fluctuation conditions, the value is less than 0.5. Therefore, this is used as the dividing standard.

[0060] Table 1. Sample evaluation data of "New Pro1" on B e-commerce platform.

[0061]

[0062] Table 1 shows three evaluation samples and related data collected from e-commerce platform B. The multi-source sentiment index directly uses the calculation result of platform B from the previous step, which is 0.50. It is assumed that the user activity level of e-commerce platform B... 12 million, user interaction volume The total number of likes and comments in the past month is 500,000, and the coefficient of variation for the number of negative reviews in the past 7 days is 0.8. Therefore, the concentration of negative reviews is high. The value is set to 0.5. Based on the above data and formula, platform-related reputation popularity data is established.

[0063] S113: Based on the platform-related reputation popularity data of each platform, align and quantify the data from different sources, and perform cross-platform aggregation to generate initial reputation indicator data;

[0064] Based on the platform-related reputation heat data of each platform, for example, through the quantitative evaluation and comprehensive analysis in the aforementioned steps, the platform-related reputation heat of three source channels—Social Media Platform A, E-commerce Platform B, and Tech Forum C—has been obtained, assuming their values ​​are 0.75, 0.88, and 0.62 respectively. At this point, it is necessary to align and quantify the data from these three sources. Considering the differences in user composition and speaking habits across different platforms, the influence of their reputation heat is not equal. For example, the user group of Tech Forum C is more professional, and its evaluations have a more direct impact on core consumers, while Social Media Platform A, although having a large volume of voices, contains more diverse information. Therefore, it is necessary to assign differentiated influence weights to the platform-related reputation heat of different platforms. The weight setting is based on the platform's professionalism. The study considered factors such as the match between the platform's user base and the product, as well as the correlation between platform reputation and actual sales in historical data. Using a combination of expert scoring and the analytic hierarchy process, the weights for C (Tech Forum), B (E-commerce Platform), and A (Social Platform) were set at 0.4, 0.4, and 0.2, respectively. The experimental process for setting these weights involved inviting 10 industry analysts to independently score the influence of each platform (from 1 to 10). The highest and lowest scores were removed, and the average was taken. This average was then combined with regression analysis of the platform voice volume and sales data of 20 similar products in the past three months after launch to fine-tune the weights. The final weight allocation scheme was then determined, and cross-platform aggregation was performed to quantify the company's overall reputation across multiple platforms. Initial reputation indicator data was generated through integrated processing.

[0065] Please see Figure 3 Step S2 is as follows:

[0066] S211: Based on the reputation score set in the initial reputation index data, calculate the standard deviation of the reputation scores of all platforms, and calculate the deviation of each platform's score from the mean of all platform scores to obtain the discrete measure of reputation scores.

[0067] Based on the reputation score set in the initial reputation indicator data, for example, if the reputation scores of three platforms—social media platform A, e-commerce platform B, and technology forum C—are 7.8, 8.5, and 6.5 respectively, the first step is to statistically process this score set and calculate the mean of the reputation scores for all platforms. Then, the variance of the scores for that group is calculated. The calculation process is as follows: Then, take the square root of the variance to obtain the standard deviation, which is [value missing]. This standard deviation reflects the overall fluctuation of reputation scores across different platforms. Next, we calculate the deviation of each platform's score from the calculated mean score of 7.6. For platform A, the mean deviation is... For platform B, its mean deviation is... For the C platform, its mean deviation is... This set of deviation values ​​visually shows the position of each platform's rating relative to the average level. The calculated standard deviation of 0.829 is integrated with the set of mean deviation values ​​[0.2, 0.9, -1.1] for each platform to obtain the discrete measure of reputation rating.

[0068] S212: Based on the initial reputation indicator data, obtain the data collection time and monthly active user data of each platform, calculate the time freshness parameter, and perform hierarchical quantification according to the number of monthly active users to obtain the platform dynamic weight data.

[0069] The system retrieves data collection times and monthly active user data from the initial reputation metrics for each platform and calculates a time freshness parameter. This parameter is defined as the reciprocal of the difference between the data collection time and the current time, in hours. For example, if the current time is 3 PM on August 7, 2024, and platform A's data was collected 24 hours ago, then its time freshness parameter is: The data from platform B was collected 72 hours ago, and its parameter value is... The data from platform C was collected 120 hours ago, and its parameter values ​​are... Next, the influence of a platform is quantified and graded based on its monthly active users (MAU). This quantification is based on the Digital Platform Influence Assessment Standard, which was developed after long-term tracking and data regression analysis of the user scale and information dissemination effectiveness of more than 200 mainstream digital platforms in the market. Platforms with MAU greater than or equal to 100 million are classified as Level I, for example, Platform A has 500 million MAU and belongs to Level I. Platforms with 10 million to 100 million MAU are classified as Level II, for example, Platform B has 80 million MAU and belongs to Level II. Platforms with less than 10 million MAU are classified as Level III, for example, Platform C has 5 million MAU and belongs to Level III. This classification result will serve as a reference for subsequent judgment of platform weight. The set of time freshness parameter values ​​[0.0417, 0.0139, 0.0083] obtained above are used as the core quantitative indicator to obtain the platform's dynamic weight data.

[0070] S213: Statistically analyze the reputation scores, reputation score dispersion metrics, and platform dynamic weight data for each platform. Combine this with the negative public opinion impact index and time conversion constant, using the following formula:

[0071] ;

[0072] The corrected reputation score is obtained through calculation, and the corrected reputation score data is established. This represents the revised reputation score. Represents the platform's original reputation score. Represents the average reputation score across all platforms. The standard deviation of reputation scores across all platforms. The platform's dynamic weight value represents the platform's overall weight. Represents the time conversion constant. The negative public opinion impact index represents the platform;

[0073] Based on the original reputation scores of each platform, and the reputation score dispersion metric results and platform dynamic weight data obtained in the aforementioned steps, a negative public opinion impact index and a time conversion constant are introduced. The original reputation scores are then corrected using a formula to obtain the corrected reputation score value. The logic of this formula is as follows: first, the original platform score is calculated... With average score The gap The gap term determines the direction of the basic adjustment; secondly, a dynamic adjustment coefficient is constructed. The standard deviation This represents the overall dispersion of the data. The greater the dispersion, the stronger the need for adjustment. The platform dynamic weight value in the denominator... This represents the timeliness of the data; the higher the timeliness, the better. The larger the value, the smaller the dynamic adjustment coefficient, indicating a reduced degree of modification to fresh data and a smaller time conversion constant. The introduction of is to make To make it dimensionless so as to correlate with the standard deviation The addition operation is performed, and the absolute value sign ensures that the adjustment coefficient is positive. Finally, the negative public opinion impact index after the square root is subtracted. The square root operation here is used to smooth the extreme impact of the impact index, making its punishment effect more stable. By comprehensively considering the relative position of the score, the overall dispersion of the data, the timeliness of individual data, and the negative public opinion risk of the platform, the original score is dynamically corrected.

[0074] Assign values ​​to the parameters in the formula, including the time conversion constant. Set to 24 hours, this value represents a standard statistical daily period, used to standardize the time freshness parameter in hourly units, and is the negative public opinion impact index. The data is obtained by monitoring the growth rate of negative comments and the frequency of highly negative sentiment words within a specific period, such as the past 7 days, through a monitoring platform. These two indicators are normalized and then weighted and summed. The value range is limited to 0 to 1 after backtesting data from 100 historical public opinion events, where 0 represents no negative impact and 1 represents the most serious public opinion crisis.

[0075] Table 2. Platform Reputation Score Correction Parameters

[0076]

[0077] Table 2 lists the parameter values ​​required for rating adjustments on the three platforms. A detailed calculation is performed using platform B as an example, substituting its parameters: Original Reputation Score Average reputation score across all platforms Standard deviation of reputation scores across all platforms The platform's dynamic weight value Time conversion constant And the platform's negative public opinion impact index ,

[0078] The calculation process for the revised reputation score is as follows:

[0079] ;

[0080] The results show that the original score of platform B, 8.5, decreased to 7.4712 after correction. This is because the platform's score was higher than the average level and was pulled towards the mean. At the same time, the negative public opinion impact index also had a negative effect on it. The corrected reputation scores of platforms A, B, and C, obtained through the same calculation, were summarized to establish reputation score correction data.

[0081] The revised reputation score is not an isolated original score, but a comprehensive, risk-adjusted quantitative indicator of reputation obtained through multi-dimensional dynamic adjustments. First, by bringing the original score of a single platform closer to the average score across all platforms, it smooths and calibrates extreme high or low scores caused by differences in user preferences or evaluation systems on specific platforms, thus avoiding bias from a single data source. Second, the degree of this adjustment is dynamic, adjusting inversely based on the freshness of the data itself, assigning higher trust to newer data and reducing its averaging effect. Finally, a risk penalty related to the intensity of recent negative public opinion is directly deducted, quantifying the immediate impact of sudden negative events on reputation. Therefore, the revised reputation score comprehensively reflects the entity's reputation performance on a specific platform, its relative position in overall online reputation, the timeliness of information, and current public opinion risk, thus providing a more stable and valuable reputation assessment result than the original score.

[0082] Please see Figure 4 Step S3 is as follows:

[0083] S311: Based on reputation score correction data, extract the continuously corrected score sequences of social platforms and e-commerce platforms, calculate the Pearson correlation coefficient of the daily score change rate of the two sequences, quantify the synchronicity of score change trends, and obtain the platform reputation correlation trend record.

[0084] To identify the reputation transmission relationship from social media platform A to e-commerce platform B, based on reputation score correction data, we first extract the daily reputation score correction data for both platforms over a continuous observation period, such as July 1st to July 7th, forming two score sequences. The score sequence for platform A is as follows: The rating sequence of platform B is as follows Next, the rate of change of daily ratings for each platform is calculated. For platform A, the rate of change sequence since July 2nd is calculated. Calculated as ,get Similarly, calculate the rate of change sequence of platform B. Subsequently, the two rate of change sequences of length 6 were analyzed. and The Pearson correlation coefficient is calculated. This calculation process includes calculating the means of the two series separately. The mean rate of change for platform A is -0.031, and the mean rate of change for platform B is -0.021. Then, the standard deviations of each series and the covariance of the two series are calculated. By dividing the covariance by the product of the two standard deviations, a quantitative indicator of the synchronicity of the two rating trends is obtained. Assuming the calculated Pearson correlation coefficient is 0.75, this value indicates that the downward or upward trend of the reputation score of platform A is positively correlated with the trend of the score of platform B. This coefficient value is used as the final quantitative result to obtain the platform reputation correlation trend record.

[0085] S312: Based on the platform reputation correlation trend record, and track the posting behavior of highly interactive users under the negative topic of the platform on another platform, count the number of content forwards and time delay, calculate the weighted interaction index, and establish cross-platform reputation transmission strength data;

[0086] The quantitative result of the platform reputation correlation trend record was 0.75. For a specific negative topic about the quality of a certain brand's products on social media platform A, which had already been identified, the system tracked users whose interaction behavior on this topic exceeded a certain threshold. This threshold was set at more than 10 comments or reposts within 24 hours. Three highly interactive users, User 1, User 2, and User 3, were selected. Then, during a set observation period, such as 72 hours, the system tracked the public posting behavior of these three users on e-commerce platform B, especially their reviews or discussions about the same brand's products. The system monitored and statistically analyzed the direct reposting of their posts. The interaction intensity is measured by a quantitative or equivalent interaction metric, along with the time delay between posting relevant content on platform A and posting related content on platform B. For example, tracking user 1's posting of a negative review on platform B with 150 reposts and an 8-hour delay, user 2's 80 reposts with a 12-hour delay, and user 3's 50 reposts with a 16-hour delay. This data is then integrated to calculate a weighted interaction metric. This weighting is not based on pre-defined weights, but rather by dividing the total number of reposts by the number of users to obtain the average interaction intensity, which is then combined with the average time delay for a comprehensive evaluation. The total number of reposts is calculated as follows: The average time delay is [time] times. By combining these two core metrics, we can establish cross-platform reputation transmission strength data within a given hour.

[0087] S313: Based on cross-platform reputation transmission strength data, combined with platform reputation correlation trend records, normalized interaction volume from multiple propagation paths, target platform vulnerability, and maximum response cycle are introduced, using the following formula:

[0088] ;

[0089] The reputation dissemination risk value is calculated, and the data is recorded and integrated to generate reputation dissemination risk data. Represents the risk value of reputation spread. Represents the normalized interaction quantity for the k-th propagation path. Represents the total number of identified propagation paths. This represents the trend of platform reputation association from social media platform A to e-commerce platform B. The vulnerability index represents the target platform B. This represents the average propagation time delay from platform A to platform B. This represents the maximum response period used for normalization;

[0090] Based on cross-platform reputation transmission strength data and combined with platform reputation correlation trend records, normalized interaction volume from multiple propagation paths from platform A to platform B, target platform vulnerability, and maximum response cycle are introduced. A formula is used to calculate reputation transmission risk, first by using a summation sign... Summarize the interaction volume of all identified propagation paths, then calculate the square root. Processing is performed to smooth out the impact of extreme interaction volumes, followed by correlation with platform relationships. and the inherent vulnerabilities of the target platform Multiply the sums, and finally subtract a factor determined by the average propagation delay. With maximum response period The time decay term obtained by division is used to quantify the propagation speed. The advantage of the formula is that it does not use fixed weighting coefficients, but dynamically combines four dimensions: the breadth of propagation, the strength of association, the vulnerability of the target, and the timeliness of propagation.

[0091] Table 3. Parameters for Calculating Reputation Transmission Risk

[0092]

[0093] As shown in Table 3, substitute the values ​​of each parameter into the formula for calculation.

[0094] ;

[0095] The result indicates that the calculated reputation spread risk value is 0.8438, which exceeds the 0.35 orange warning line defined in the international brand assessment standard, falling into the high-risk range. This calculation result is one data point in the reputation spread risk data. This calculation will be performed on all identified platform transmission relationships, and finally integrated to generate complete reputation spread risk data.

[0096] The Reputation Spread Risk Value is a comprehensive and dynamic assessment result that quantifies the spillover effect of reputation across different digital platforms. Its construction integrates several key dimensions: First, it measures the breadth and intensity of cross-platform dissemination, i.e., assessing the average interaction volume of related topics, to determine the scale and impact of reputation information dissemination; second, it considers the inherent connections between platforms and the vulnerability of the target platform. This includes not only the synchronicity of reputation score changes between two platforms calculated based on historical data, but also an assessment of whether the structure of the information receiving platform itself is prone to amplifying external influences; finally, the risk value also uses dissemination timeliness as a core moderating factor. The faster a reputation event spreads across platforms, the higher the corresponding risk, and vice versa. Therefore, the Reputation Spread Risk Value ultimately presents a multi-dimensional risk assessment combining dissemination effectiveness, platform coupling, target carrying capacity, and timeliness, providing a quantitative basis for the cross-platform transmission of reputation with a forward-looking early warning function.

[0097] Please see Figure 5 Step S4 is as follows:

[0098] S411: Based on the reputation spread risk data, determine whether the reputation spread risk value exceeds the preset digital risk framework reputation risk warning value. If it has exceeded the threshold, determine that the risk warning has been triggered and output the risk warning trigger record.

[0099] Based on reputation risk data, to initiate the reputation risk response process, the urgency of the risk must first be assessed. This process retrieves the reputation risk value recorded in the reputation risk data regarding the transmission of reputation risk from social media platform A to e-commerce platform B, which is 0.8438. Subsequently, this value is compared with the preset digital risk framework reputation risk warning value, which is set at 0.35. This value is not arbitrarily set but is based on a retrospective analysis of the full-cycle data of 50 major corporate reputation crisis events. Statistical analysis revealed that within the 72-hour window before the full outbreak of the crisis, more than 80% of the events... The risk value for reputation transmission exceeded 0.35, so this value was used as a key warning trigger point. The setting of this value has undergone multiple rounds of stress testing and historical data verification to ensure its stability and effectiveness in practical applications. The specific judgment operation is as follows: the obtained risk value of 0.8438 is compared with the warning value of 0.35. Since 0.8438 is greater than 0.35, the comparison result is true. The system then determines that the current situation has met the triggering conditions for risk warning, records the judgment result in the form of a Boolean value, such as "True", and generates the corresponding risk warning trigger record.

[0100] S412: Based on the risk warning trigger record, perform the recommendation weight coefficient adjustment operation on the downstream e-commerce platform, and perform the user engagement improvement operation to integrate and establish a combination of platform intervention strategies;

[0101] The system invoked the risk warning trigger record. Because the result was "True," the system automatically executed a series of intervention operations on e-commerce platform B, downstream of the reputation transmission path. First, it adjusted the recommendation weight coefficients. Within this platform's recommendation system, the default recommendation weight coefficients for positive, neutral, and negative emotional content are 1.0, 0.7, and 0.5, respectively. Now, for specific products or brands related to this negative reputation event, the recommendation weight coefficient for their negative emotional content is lowered from 0.5 to 0.2. This adjustment aims to reduce the exposure of negative information within the platform. Simultaneously, it also adjusted user engagement... The enhancement operation is not a generalized execution, but rather consists of two specific actions. The first action is to target approximately 5,000 users who have posted 4-star or 5-star reviews on the platform within the past 90 days and are related to the same brand or product category, by pushing a no-threshold coupon worth 20 yuan. The second action is to collaborate with the brand to launch a themed discussion activity called "Show Off Your Quality Life Moments" in the platform's official community, and to offer brand-related gifts to the first 100 participants. The above-mentioned adjustment of the recommendation weight coefficient and the two user participation enhancement actions are programmatically combined and recorded as shown in Table 4 below.

[0102] Table 4. List of Platform Intervention Strategies

[0103]

[0104] As shown in Table 4, this table details the specific content of each intervention action. These actions together constitute a complete response plan. This plan should be archived to establish a platform intervention strategy combination.

[0105] S413: Implement a combination of platform intervention strategies, conduct a comprehensive analysis of the changing trends of reputation score correction data before and after the intervention, changes in platform activity and user engagement, evaluate the strength of the recommendation weight adjustment intervention, correlate and integrate the information, and archive the evaluation, intervention actions and execution time together to generate a reputation optimization execution record.

[0106] To evaluate the effectiveness of the platform intervention strategy combination, a comprehensive analysis of changes in multiple core indicators was conducted after a complete monitoring period of 72 hours following the intervention. First, reputation score correction data was collected before and after the intervention. The score was 7.4712 immediately before the intervention and 7.6200 72 hours later, showing an increase of 0.1488. Second, platform activity data was collected. The average daily active users were 8.5 million before the intervention and 8.55 million after, a change of 50,000, indicating that the overall platform traffic did not fluctuate drastically due to the intervention. Third, the change in user engagement was evaluated. User engagement was quantified by statistically analyzing the number of participants in topical discussion activities and the coupon redemption rate of the target user group. The number of participants was 8,000, and the coupon redemption rate was 45%, representing an increase of approximately 15 percentage points in participation compared to similar activities before the intervention. Meanwhile, the intensity of the intervention, adjusting the recommendation weight, was assessed at a coefficient of -0.3, classifying it as a moderate-intensity intervention. Finally, all the above analytical information, including the growth trend of reputation score correction data, the stability of platform activity, changes in user participation, and the assessment of the intensity of the intervention, was correlated and integrated to form a qualitative evaluation conclusion, such as "This intervention effectively improved the reputation score while stimulating positive user participation without negatively impacting the platform's basic activity." This evaluation conclusion, along with the specific details of the platform intervention strategy and the execution timestamp, was archived to generate a reputation optimization execution record.

[0107] Please see Figure 6 The S5 steps are as follows:

[0108] S511: Periodically collect new reputation data of enterprises on social media platforms and e-commerce platforms, obtain the latest reputation score correction data, and compare it with the initial reputation index data to evaluate the direction and magnitude of reputation score changes on each platform and obtain a record of reputation change differences.

[0109] Periodically collect new reputation data for businesses on social media platform A and e-commerce platform B. This collection operation is set to run once every 24 hours. By calling the publicly available data interfaces of the two platforms, the latest user-generated content related to the businesses is obtained in batches, including but not limited to user comment text, product ratings, and post sentiment. After obtaining the above data, it needs to be processed to generate the latest reputation score correction data. For example, in 1000 new mentions collected on social media platform A on August 5, 2025, natural language processing identifies 150 positive, 600 neutral, and 250 negative content, and assigns sentiment scores of 1, 0, and -1 respectively. Then, the sentiment mean for this period is (1501 + 6000 + 250⁻¹) / 1000 = -0.1. This sentiment mean is then normalized using a function, for example... This is converted into a reputation score of 1 to 10. This is the latest corrected reputation score data for platform A. Then, it is combined with the initial reputation score data recorded 24 hours prior—platform A's initial score was 5.20, and e-commerce platform B's initial score was 7.80—and compared with the newly acquired scores of 5.05 for platform A and 7.75 for platform B. This comparison is a direct difference operation, and the change for platform A is [value missing]. The change value of platform B By assessing that the reputation scores of both platforms changed in a negative direction, and that the decline in platform A was greater than that of platform B, this quantitative result was combined with qualitative judgment to obtain a record of the differences in reputation changes.

[0110] S512: Based on the records of reputation change differences and combined with the records of reputation optimization implementation, conduct a comprehensive analysis of the overall trend of reputation score changes of all monitored platforms, the activity of related users jumping between platforms, and the time interval since the last intervention, evaluate the synergistic effect of reputation optimization measures across multiple platforms, and establish a cross-platform reputation synergy index.

[0111] Based on the reputation change difference records for each platform and the archived reputation optimization execution records, a comprehensive analysis and evaluation of the current cross-platform collaborative status of the enterprise's reputation is conducted according to the preset security monitoring plan. This process does not rely on a single numerical calculation, but rather interprets the correlation of data from multiple dimensions. First, the overall trend of reputation score changes across all monitored platforms is analyzed. Based on the reputation change difference data, Platform A's score decreased by 0.15, and Platform B's score decreased by 0.05, indicating that the overall reputation is in a clear downward trend. Second, the activity of user jumps between platforms is evaluated. By monitoring the user behavior logs in the website backend, it was found that in the past 24-hour observation period, 1,500 users jumped from Social Platform A to the enterprise-related pages on E-commerce Platform B, while 1,500 users jumped from E-commerce Platform B back to Social Platform A. With only 250 users on the platform, this asymmetry in traffic indicates that the impact is mainly transmitted unidirectionally from social platform A to e-commerce platform B. Platform B has failed to form an effective positive reputation feedback loop. Furthermore, by assessing the time interval since the last intervention and its current effectiveness, and reviewing the reputation optimization execution records, it can be seen that the last targeted coupon distribution and trending topic discussion on platform B was 72 hours ago. Combined with the fact that the reputation scores of both platforms are still declining, it can be determined that the positive impact of the last intervention has diminished and cannot completely offset the negative impact originating from platform A. Finally, by integrating the judgment of the overall decline in reputation, the significant asymmetry in user flow, and the assessment conclusion of the weakening effectiveness of historical intervention measures, the system conducts a comprehensive qualitative assessment of the current collaborative status and establishes cross-platform reputation collaborative data, which is rated as "negative imbalance".

[0112] S513: Based on cross-platform reputation collaboration data, integrate it with qualitative assessment conclusions in reputation change difference records and reputation optimization execution records to construct a structured document that includes synergy effect assessment and multi-dimensional data changes, and obtain cross-platform collaborative optimization records;

[0113] Based on cross-platform reputation collaboration data rated as "negative imbalance," this rating indicates that current cross-platform reputation management is in a negative collaborative state, with the collaborative health status below the preset "neutral equilibrium" benchmark. Subsequently, this rating is integrated with the reputation change difference records obtained in previous steps and the qualitative assessment conclusions in the reputation optimization execution records. This integration process is a structured information compilation operation, specifically creating a new data record entry containing the following fields: Collaboration Index Rating, which is "negative imbalance"; Change Details for Each Platform, which shows that the score of Platform A decreased by 0.15 and the score of Platform B decreased by 0.05; Historical Intervention Assessment, which shows that the intervention measures implemented on Platform B 72 hours ago failed to effectively curb the overall reputation decline trend. The information from these three parts is combined into a comprehensive description, as shown in Table 5 below.

[0114] Table 5 Collaborative Optimization Data Record Table

[0115]

[0116] As shown in Table 5, this table clearly presents the combination of qualitative rating and quantitative analysis. The structured data entries are saved to the database to obtain cross-platform collaborative optimization records.

[0117] A cross-platform enterprise reputation collaborative optimization system, including:

[0118] The data acquisition module is used to execute S1: acquire corporate reputation scores, user reviews and activity levels from social media platforms, e-commerce platforms and forums, collect raw data in real time, and perform preliminary data integration to obtain initial reputation indicator data;

[0119] The consistency assessment module is used to execute S2: Based on the reputation score of the initial reputation indicator data, the standard deviation and mean deviation of the scores of each platform are calculated. Combined with the time freshness parameter and the weighting factor of the monthly active users of the platform, the corrected reputation score value is calculated to obtain the corrected reputation score data.

[0120] The transmission path identification module is used to perform S3: based on reputation score correction data, identify the reputation transmission relationship from social media platforms to e-commerce platforms, compare the score change trend and platform interaction data to track the transmission path, calculate the reputation transmission risk value, and obtain reputation transmission risk data;

[0121] The optimization strategy execution module is used to execute S4: based on reputation propagation risk data, when the reputation propagation risk value exceeds the reputation risk warning value of the digital risk framework, it performs recommendation weight coefficient adjustment and user engagement improvement operations, and performs dynamic weighted calculation to obtain reputation optimization execution records;

[0122] The collaborative optimization module is used to execute S5: periodically collect new reputation data of enterprises on the platform, combine it with the initial reputation indicator data and reputation optimization execution records, monitor the dynamic changes in enterprise reputation in real time, and obtain cross-platform collaborative optimization records.

[0123] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A cross-platform enterprise reputation collaborative optimization method, characterized in that, Includes the following steps: S1: Obtain corporate reputation scores, user reviews, and activity levels from social media platforms, e-commerce platforms, and forums; collect raw data in real time; and perform preliminary data integration to obtain initial reputation indicator data. S2: Based on the reputation score of the initial reputation index data, calculate the standard deviation and mean deviation of the scores of each platform, and combine the time freshness parameter and the platform monthly active user tiered allocation weight factor to calculate the corrected reputation score value and obtain the reputation score correction data. S3: Based on the reputation score correction data, identify the reputation transmission relationship from social media platforms to e-commerce platforms, compare the score change trend and platform interaction data to track the propagation path, calculate the reputation propagation risk value, and obtain reputation propagation risk data; S4: Based on the reputation propagation risk data, when the reputation propagation risk value exceeds the reputation risk warning value of the digital risk framework, perform the operation of adjusting the recommendation weight coefficient and improving user participation, and perform dynamic weighted calculation to obtain the reputation optimization execution record; S5: Periodically collect new reputation data of enterprises on the platform, combine the initial reputation indicator data and the reputation optimization execution record, monitor the dynamic changes in enterprise reputation in real time, and obtain cross-platform collaborative optimization records.

2. The cross-platform enterprise reputation collaborative optimization method according to claim 1, characterized in that, The initial reputation indicator data includes social media reputation data, e-commerce platform reputation data, and forum reputation data. The reputation score correction data includes weighted score values, standard deviation adjustment values, and time freshness weight factors. The reputation dissemination risk data includes risk value indicators, dissemination path analysis results, and trend change comparison results. The reputation optimization execution record includes weight adjustment parameters, participation improvement measures, and dynamic weighted calculation results. The cross-platform collaborative optimization record includes data comparison reports, monitoring logs, and enterprise reputation optimization effect evaluation records.

3. The cross-platform enterprise reputation collaborative optimization method according to claim 1, characterized in that, The specific steps for obtaining the initial reputation indicator data are as follows: S111: Obtain user reviews from social media platforms, e-commerce platforms, and forums; determine the sentiment polarity of the review texts; and conduct a differential analysis of the proportion of positive reviews and the proportion of negative reviews to obtain multi-source sentiment bias data. S112: For the enterprise reputation scores and user activity collected from various platforms, statistical analysis of user interaction volume and evaluation timeliness data is conducted. Combined with the aforementioned multi-source sentiment data, correlation quantification and comprehensive evaluation of various indicators are performed to establish platform-related reputation heat data. S113: Based on the platform-related reputation popularity data of each platform, align and quantify the data from different sources, and perform cross-platform aggregation to generate initial reputation indicator data.

4. The cross-platform enterprise reputation collaborative optimization method according to claim 1, characterized in that, The specific steps for obtaining the reputation score correction data are as follows: S211: Based on the reputation score set in the initial reputation index data, calculate the standard deviation of the reputation scores of all platforms, and calculate the deviation of each platform's score from the mean of all platform scores to obtain the reputation score dispersion measurement result. S212: Based on the initial reputation index data, obtain the data collection time and monthly active user data of each platform, calculate the time freshness parameter, and perform hierarchical quantification according to the number of monthly active users to obtain platform dynamic weight data. S213: Calculate the reputation scores of each platform, the discrete measurement results of the reputation scores, and the dynamic weight data of the platforms. Combine the negative public opinion impact index and the time conversion constant to obtain the corrected reputation score value and establish the reputation score correction data.

5. The cross-platform enterprise reputation collaborative optimization method according to claim 1, characterized in that, The specific steps for obtaining reputation dissemination risk data are as follows: S311: Based on the reputation score correction data, extract the continuously corrected score sequences of social platforms and e-commerce platforms, calculate the Pearson correlation coefficient of the daily score change rate of the two sequences, quantify the synchronicity of the score change trend, and obtain the platform reputation correlation trend record. S312: Based on the reputation correlation trend records of the platform, and by tracking the posting behavior of highly interactive users under the negative topics of the platform on another platform, the number of content forwards and time delays are counted, the weighted interaction index is calculated, and cross-platform reputation transmission strength data is established. S313: Based on the cross-platform reputation transmission strength data, combined with the platform reputation correlation trend records, normalized interaction volume of multiple propagation paths, target platform vulnerability and maximum response cycle are introduced to calculate and obtain reputation propagation risk value, and record and integrate to generate reputation propagation risk data.

6. The cross-platform enterprise reputation collaborative optimization method according to claim 1, characterized in that, The specific steps for obtaining reputation optimization execution records are as follows: S411: Based on the reputation spread risk data, determine whether the reputation spread risk value exceeds the preset digital risk framework reputation risk warning value. If it exceeds the threshold, determine that a risk warning has been triggered and output the risk warning trigger record. S412: Based on the risk warning trigger record, perform the recommendation weight coefficient adjustment operation on the downstream e-commerce platform, and perform the user participation improvement operation to integrate and establish a platform intervention strategy combination; S413: Execute the aforementioned platform intervention strategy combination, comprehensively analyze the changing trends of reputation score correction data before and after the intervention, the changes in platform activity and user engagement, and at the same time evaluate the strength of the recommendation weight adjustment intervention action, correlate and integrate the information, and archive the evaluation, intervention action and execution time together to generate a reputation optimization execution record.

7. The cross-platform enterprise reputation collaborative optimization method according to claim 1, characterized in that, The specific steps for obtaining cross-platform collaborative optimization records are as follows: S511: Periodically collect new reputation data of enterprises on social media platforms and e-commerce platforms, obtain the latest reputation score correction data, and compare it with the initial reputation index data to evaluate the direction and magnitude of the reputation score changes on each platform and obtain a record of reputation change differences. S512: Based on the recorded differences in reputation changes and in conjunction with the records of reputation optimization implementation, conduct a comprehensive analysis of the overall trend of reputation scores of all monitored platforms, the activity of users jumping between platforms, and the time interval since the last intervention, evaluate the synergistic effect of reputation optimization measures across multiple platforms, and establish a cross-platform reputation synergy index. S513: Based on the cross-platform reputation collaboration data, integrate it with the qualitative evaluation conclusions in the reputation change difference record and the reputation optimization execution record to construct a structured document including synergy effect evaluation and multi-dimensional data changes, and obtain cross-platform collaborative optimization records.

8. A cross-platform enterprise reputation collaborative optimization system, characterized in that: The system is used to implement the cross-platform enterprise reputation collaborative optimization method according to any one of claims 1-7, including: The data acquisition module is used to execute S1: acquire corporate reputation scores, user reviews and activity levels from social media platforms, e-commerce platforms and forums, collect raw data in real time, and perform preliminary data integration to obtain initial reputation indicator data; The consistency assessment module is used to perform S2: based on the reputation score of the initial reputation index data, the standard deviation and mean deviation of the scores of each platform are calculated, and combined with the time freshness parameter and the platform monthly active user tiered allocation weight factor, the corrected reputation score value is calculated to obtain the corrected reputation score data; The transmission path identification module is used to perform S3: based on the reputation score correction data, identify the reputation transmission relationship from the social media platform to the e-commerce platform, compare the score change trend and platform interaction data to track the transmission path, calculate the reputation transmission risk value, and obtain reputation transmission risk data; The optimization strategy execution module is used to execute S4: based on the reputation propagation risk data, when the reputation propagation risk value exceeds the reputation risk warning value of the digital risk framework, perform recommendation weight coefficient adjustment and user engagement improvement operations, and perform dynamic weighted calculation to obtain reputation optimization execution records; The collaborative optimization module is used to execute S5: periodically collect new reputation data of enterprises on the platform, combine the initial reputation index data and the reputation optimization execution record, monitor the dynamic changes in enterprise reputation in real time, and obtain cross-platform collaborative optimization records.

Images