User information sharing method and system based on big data

Abstract

The invention relates to a user information sharing method and system based on big data, and relates to the field of information sharing, and the method comprises the steps: carrying out the feature clustering analysis of false evaluation information, and determining a first abnormal evaluation index; performing feature clustering analysis on the false evaluation information of the plurality of similar platforms based on big data, determining a plurality of associated abnormal evaluation indexes, and performing fusion to determine the abnormal evaluation indexes; and performing data credibility analysis on the evaluation information set according to the abnormal evaluation indexes, setting data meeting a dynamic credibility threshold value as reliable evaluation data, and performing information sharing. The technical problem that the trustworthiness of a user on platform evaluation information is affected due to the fact that false evaluation cannot be effectively filtered because the authenticity of the user evaluation information cannot be effectively identified by an existing method can be solved; according to the method, false good comments and malicious bad comments can be identified more accurately, unreliable evaluation data is filtered out, and real evaluation information is fed back to customers, so that the user experience is optimized, and the customer satisfaction is improved.

Description

Technical Field

[0001] The present invention relates to the field of information sharing, and in particular to a user information sharing method and system based on big data. Background Art

[0002] With the rapid development of e-commerce and online platforms, user evaluation information has become an important reference for consumers to make purchasing decisions. The authenticity of online evaluations directly affects consumers' choice of products or services. Therefore, ensuring the authenticity and credibility of evaluation information has become a key issue in platform operations. However, with the continuous expansion of the platform and the increase in the number of users, the problem of false evaluations has gradually emerged and has had a significant impact on the credibility of the platform.

[0003] False reviews usually include false positive reviews (such as fake orders, buyer shows, etc.) and malicious negative reviews (such as malicious competition, slander, etc.). These reviews are often posted through manual or automated means with the aim of misleading consumers, manipulating product sales or damaging the reputation of competitors. False positive reviews and malicious negative reviews not only mislead consumers' decisions, but also make the platform's evaluation system lose its fairness, affecting the platform's transparency and consumer trust.

[0004] At present, many platforms use traditional rule detection, keyword matching and other technologies to identify the authenticity of evaluation information. Most of these traditional methods rely on manually set rules, lack dynamic adjustment mechanisms, and have many problems. Summary of the invention

[0005] The present invention aims to solve the technical problem that existing methods cannot effectively identify the authenticity of user evaluation information, resulting in failure to effectively filter out false evaluations, which affects users' trust in platform evaluation information. A user information sharing method and system based on big data are provided to solve the problem.

[0006] The technical solution of the present invention to solve the above technical problems is as follows: In a first aspect, the present invention provides a user information sharing method based on big data, comprising: performing feature clustering analysis on false evaluation information of a target platform to determine a first abnormal evaluation index; based on big data, performing feature clustering analysis on false evaluation information of multiple similar platforms of the target platform to determine multiple associated abnormal evaluation indexes; determining an abnormal evaluation index based on a fusion of the first abnormal evaluation index and multiple associated abnormal evaluation indexes; performing data credibility analysis on an evaluation information set of the target platform based on the abnormal evaluation index, setting the evaluation information that meets a dynamic credibility threshold as reliable evaluation data, obtaining a reliable evaluation data set, and sharing user evaluation information of the target platform.

[0007] Preferably, the user information sharing method based on big data also includes: obtaining a historical false evaluation information set of the target platform, wherein the false evaluation information includes false positive evaluation data and malicious negative evaluation data; performing feature extraction and cluster analysis on the historical false positive evaluation data set according to predetermined evaluation feature attributes to obtain a first false positive evaluation index; performing feature extraction and cluster analysis on the historical malicious negative evaluation data set according to predetermined evaluation feature attributes to obtain a first malicious negative evaluation index; and using the first false positive evaluation index and the first malicious negative evaluation index as the first abnormal evaluation index.

[0008] Preferably, the user information sharing method based on big data also includes: obtaining predetermined evaluation feature attributes, wherein the predetermined evaluation feature attributes include at least the number of evaluation words, keywords, modal particles, number of pictures and picture quality; according to the predetermined evaluation feature attributes, performing feature extraction on multiple historical false positive review data in the historical false positive review data set to obtain multiple evaluation feature data sets; and performing feature clustering on the multiple evaluation feature data sets using the K-means clustering algorithm to obtain the first false positive review indicator.

[0009] Preferably, the user information sharing method based on big data also includes: taking the platform attribute characteristics of the target platform as constraints, performing associated retrieval based on big data, determining multiple similar platforms, and obtaining multiple similar historical false evaluation information sets of the multiple similar platforms, wherein the similar historical false evaluation information includes similar false positive review data and similar malicious negative review data; according to predetermined evaluation feature attributes, performing feature extraction and cluster analysis on multiple similar historical false positive review data sets and similar historical malicious negative review data sets, respectively, to obtain multiple associated abnormal evaluation indicators.

[0010] Preferably, the user information sharing method based on big data also includes: based on platform attribute characteristics, respectively analyzing the correlation between the target platform and multiple similar platforms to determine the correlations of multiple platforms; respectively calculating the ratio of the correlation of each platform to the sum of the correlations of multiple platforms, setting it as the indicator fusion weight, and obtaining multiple indicator fusion weights; weightedly fusing the multiple associated abnormality evaluation indicators according to the multiple indicator fusion weights to determine a first associated abnormality evaluation indicator; performing secondary weighted fusion on the first abnormality evaluation indicator and the first associated abnormality evaluation indicator to obtain the abnormality evaluation indicator, wherein the fusion weight of the first abnormality evaluation indicator is 0.7, and the fusion weight of the first associated abnormality evaluation indicator is 0.3.

[0011] Preferably, the user information sharing method based on big data also includes: obtaining the abnormal evaluation index, wherein the abnormal evaluation index includes multiple abnormal evaluation attribute data; randomly selecting first evaluation information from the evaluation information set, and performing similarity analysis on the first evaluation information according to the multiple abnormal evaluation attribute data to obtain a first similarity set, and calculating the mean to obtain a first similarity coefficient; subtracting the first similarity coefficient from 1 to obtain a first data credibility coefficient, and sequentially analyzing to obtain multiple data credibility coefficients of multiple evaluation information in the evaluation information set.

[0012] Preferably, the user information sharing method based on big data also includes: based on big data, obtaining a sequence of the number of false evaluation information and a sequence of the proportion of false evaluation information on a target platform within a preset historical time zone, as well as a sequence of the number of multiple similar false evaluation information and a sequence of the proportion of multiple similar false evaluation information on multiple similar platforms within a preset time zone; pre-training an evaluation trend analysis plug-in, and inputting the sequence of the number of false evaluation information, the sequence of the proportion of false evaluation information, the correlation of multiple platforms, the sequence of the number of multiple similar false evaluation information and the sequence of the proportion of multiple similar false evaluation information into the evaluation trend analysis plug-in, and outputting an evaluation trend coefficient; optimizing and adjusting a predetermined credibility threshold according to the evaluation trend coefficient to obtain a dynamic credibility threshold; judging the multiple data credibility coefficients according to the dynamic credibility threshold, setting the evaluation information corresponding to the data credibility coefficient greater than the dynamic credibility threshold as reliable evaluation data, and obtaining a reliable evaluation data set.

[0013] Preferably, the user information sharing method based on big data also includes: based on big data, collecting a sample training data set, wherein the sample training data includes a sample false evaluation information quantity sequence, a sample false evaluation information ratio sequence, a sample platform association set, a sample similar false evaluation information quantity sequence set, a sample similar false evaluation information ratio sequence set and a sample evaluation trend coefficient, and the sample evaluation trend coefficient can be obtained by manual labeling; using the sample training data set, supervised training and testing of the BP neural network until the loss function converges, to obtain a trained evaluation trend analysis plug-in.

[0014] In a second aspect, the present invention provides a user information sharing system based on big data, comprising: a first abnormal evaluation index determination module, used to perform feature clustering analysis on false evaluation information of a target platform to determine a first abnormal evaluation index; an associated abnormal evaluation index determination module, used to perform feature clustering analysis on false evaluation information of multiple similar platforms of the target platform based on big data to determine multiple associated abnormal evaluation indicators; an abnormal evaluation index fusion module, used to determine the abnormal evaluation index based on the fusion of the first abnormal evaluation index and multiple associated abnormal evaluation indicators; a data credibility analysis module, used to perform data credibility analysis on the evaluation information set of the target platform based on the abnormal evaluation index, set the evaluation information that meets the dynamic credibility threshold as reliable evaluation data, obtain a reliable evaluation data set, and share user evaluation information of the target platform.

[0015] The beneficial effects of the present invention are as follows: by performing feature clustering analysis on false evaluation information of a target platform, a first abnormal evaluation index is determined; then, based on big data, feature clustering analysis is performed on false evaluation information of multiple similar platforms of the target platform to determine multiple associated abnormal evaluation indexes; further, an abnormal evaluation index is determined based on the fusion of the first abnormal evaluation index and multiple associated abnormal evaluation indexes; then, data credibility analysis is performed on the evaluation information set of the target platform according to the abnormal evaluation index, and the evaluation information that meets the dynamic credibility threshold is set as reliable evaluation data to obtain a reliable evaluation data set, and user evaluation information of the target platform is shared; that is, by performing clustering analysis on the false evaluation features of the target platform and multiple similar platforms, false positive reviews and malicious negative reviews can be more accurately identified, unreliable evaluation data can be filtered out, and true evaluation information can be ensured to be fed back to customers, thereby reducing the negative impact of false evaluations on the platform's reputation, optimizing user experience, and improving customer satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic diagram of a process flow of a user information sharing method based on big data provided by the present invention; Figure 2 A schematic diagram of the structure of a user information sharing system based on big data provided by the present invention.

[0017] In the accompanying drawings, the components represented by the reference numerals are described as follows: A first abnormal evaluation index determination module 11, a related abnormal evaluation index determination module 12, an abnormal evaluation index fusion module 13, and a data credibility analysis module 14. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0019] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0020] In the description of the present invention, the term "for example" is used to mean "used as an example, illustration or explanation". Any embodiment described as "for example" in the present invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any technician in the field to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes will not be elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed in the present invention.

[0021] Embodiment 1, as Figure 1 As shown, the embodiment of the present invention provides a user information sharing method based on big data, which specifically includes the following steps: S100: Perform feature clustering analysis on the false evaluation information of the target platform to determine a first abnormal evaluation index.

[0022] Furthermore, step S100 of the present invention further includes: S110: Obtain a historical false evaluation information set of the target platform, wherein the false evaluation information includes false positive evaluation data and malicious negative evaluation data.

[0023] Specifically, all historical user evaluation data is obtained from the target platform, including evaluation information of all products, services, merchants or service providers; then, based on the existing data of the platform or manual annotation, the historical false evaluation information set of the target platform is obtained, where false evaluation information includes false positive evaluation data and malicious negative evaluation data. False positive evaluation data usually manifests as excessive praise, evaluation without actual experience, or obviously unreasonable positive evaluation (for example, the evaluation content is too consistent, the rating is extremely positive, etc.); malicious negative evaluation data manifests as negative evaluation without actual use experience, extreme and malicious negative evaluation, or obvious signs of malicious attack by competitors. By obtaining the historical false evaluation information set of the target platform and classifying it into false positive evaluation data and malicious negative evaluation data, this provides basic data support for subsequent false evaluation identification, feature analysis and credibility assessment.

[0024] S120: Perform feature extraction and cluster analysis on the historical false positive review data set according to predetermined evaluation feature attributes to obtain a first false positive review index.

[0025] Furthermore, step S120 of the present invention further includes: S121: Obtain predetermined evaluation feature attributes, wherein the predetermined evaluation feature attributes include at least the number of evaluation words, keywords, modal particles, number of pictures and picture quality; S122: According to the predetermined evaluation feature attributes, perform feature extraction on multiple historical false positive review data in the historical false positive review data set respectively to obtain multiple evaluation feature data sets; S123: Perform feature clustering on the multiple evaluation feature data sets using a K-means clustering algorithm to obtain the first false positive review indicator.

[0026] Specifically, first, the predetermined evaluation feature attributes are obtained, wherein the predetermined evaluation feature attributes include at least the number of evaluation words, keywords, modal particles, the number of pictures and the quality of pictures. The number of evaluation words refers to the number of characters in each evaluation. False evaluations may have certain regularities in the number of words (such as too short or too long), which is inconsistent with the natural evaluation habits of real users. Keywords refer to specific words contained in the evaluation, which can reveal the authenticity of the evaluation. For example, frequently appearing keywords such as "good", "perfect", and "very satisfied" may be characteristics of false positive reviews. Modal particles refer to words or phrases used in the evaluation to express emotions, such as modal particles such as "very" and "extremely". The frequency and intensity of use of these modal particles may be more extreme in false evaluations. The number of pictures refers to the number of pictures attached to the evaluation. False positive reviews may often contain a large number of irrelevant pictures, or a small number of but irrelevant pictures. Picture quality refers to the clarity and authenticity of the pictures attached to the evaluation. Low-quality pictures or pictures that do not match the content of the evaluation may be characteristics of false evaluations.

[0027] Next, according to the predetermined evaluation feature attributes, feature extraction is performed on multiple historical false positive reviews in the historical false positive review data set, including extracting the number of evaluation words, extracting keywords, extracting modal particles, extracting the number of pictures and extracting the quality of pictures. For example, the content of each evaluation is counted and the number of words is recorded; the evaluation content is processed using natural language processing (NLP) technology (such as word segmentation, part-of-speech tagging, etc.) to extract high-frequency keywords therein; modal particles (such as "extremely", "very", "perfect", etc.) in the evaluation are extracted through regular expressions or sentiment analysis tools; the number of pictures attached to each evaluation is counted according to the picture field in the evaluation data; the quality of the pictures attached to each evaluation is evaluated through image processing technology (such as clarity detection, content analysis, etc.), and factors such as the resolution, clarity, and relevance of the pictures to the evaluation content are evaluated to generate a picture quality score; for each historical false positive review data, the corresponding features are extracted according to the above steps, and an evaluation feature data set containing all predetermined feature attributes is generated to obtain multiple evaluation feature data sets. By extracting predetermined evaluation feature attributes such as the number of evaluation words, keywords, modal particles, number of pictures and picture quality from historical false positive review data sets, we can construct an evaluation data set containing multidimensional features. This data set provides basic data for subsequent false review identification, cluster analysis and credibility assessment, helping to accurately identify false reviews.

[0028] K-means clustering is an unsupervised learning method that divides data into K clusters, each of which contains samples with similar features. Then, the K-means clustering algorithm is used to perform feature clustering on the multiple evaluation feature data sets. First, K initial centroids are randomly selected; then, for each data point, its distance from all centroids is calculated and assigned to the nearest centroid; the centroid of each cluster is further updated, and the new centroid is the mean of all data points in the cluster; then the above steps are repeated until the cluster assignment no longer changes or the maximum number of iterations is reached; after clustering, the feature value of each cluster can represent some common patterns of false positive reviews, and features that frequently appear in false positive reviews are extracted from the clustering results, such as certain specific word ranges, keyword frequencies, the use of specific modal particles, excessive number of pictures, etc. If the centroid of a cluster indicates that the number of evaluation words is generally longer and the modal particles are used more frequently, this may indicate that the cluster belongs to a typical false positive review. Next, by analyzing the clustering results, we can determine which features appear more frequently in false positive reviews, and set the high-frequency features as the first false positive review indicator, for example, frequently appearing positive words such as "perfect", "absolutely good", "most satisfactory", etc.; low-quality pictures or irrelevant pictures may be signs of false positive reviews, etc. These indicators will be used for subsequent false review identification and filtering.

[0029] S130: performing feature extraction and cluster analysis on a historical malicious negative review data set according to predetermined evaluation feature attributes to obtain a first malicious negative review index; S140: using the first false positive review index and the first malicious negative review index as the first abnormal evaluation index.

[0030] Specifically, according to the predetermined evaluation feature attributes, feature extraction is performed on the historical malicious negative review data set, that is, relevant feature data is extracted from the historical malicious negative review data set, and a feature data set is generated, including extracting the number of evaluation words, extracting keywords (using natural language processing technology to analyze the keywords in each negative review, especially negative sentiment words such as "bad", "terrible", "disappointed", etc.), using regular expressions or sentiment analysis tools to extract modal particles in the evaluation, such as "worst", "never again", etc.; then the K-means clustering algorithm is used to perform feature clustering on the feature data set to obtain the first malicious negative review index, that is, the most representative malicious negative review features, which will be used for subsequent malicious negative review identification and filtering. Finally, the first false positive review index and the first malicious negative review index are used as the first abnormal evaluation index.

[0031] By performing feature extraction and K-means cluster analysis on the historical malicious negative review data set, we can identify the typical characteristics of malicious negative reviews, such as too short review words, frequently appearing negative keywords, extreme modal particles, abnormal number of pictures and low-quality pictures. The first malicious negative review indicator obtained through cluster analysis will help to identify and filter malicious negative reviews in the future and improve the credibility of platform evaluations.

[0032] S200: Based on big data, feature clustering analysis is performed on false evaluation information of multiple similar platforms of the target platform to determine multiple associated abnormal evaluation indicators.

[0033] Furthermore, step S200 of the present invention further includes: S210: Taking the platform attribute characteristics of the target platform as constraints, performing associated retrieval based on big data, determining multiple similar platforms, and obtaining multiple similar historical false evaluation information sets of the multiple similar platforms, wherein the similar historical false evaluation information includes similar false positive evaluation data and similar malicious negative evaluation data; S220: According to predetermined evaluation feature attributes, performing feature extraction and cluster analysis on multiple similar historical false positive evaluation data sets and similar historical malicious negative evaluation data sets, respectively, to obtain multiple associated abnormal evaluation indicators.

[0034] Specifically, first, it is necessary to clarify the platform attribute characteristics of the target platform, including but not limited to platform type, user activity, product or service category, user group characteristics, etc.; then, with the platform attribute characteristics of the target platform as a constraint, perform association retrieval based on big data to determine multiple similar platforms, that is, through big data analysis and association retrieval technology, find similar platforms with similar attributes to the target platform, and calculate the similarity between the various attributes of the target platform and the attributes of other platforms to identify similar platforms with high similarity. Then, for each similar platform, obtain the historical false evaluation information set of the platform, and obtain multiple similar historical false evaluation information sets of the multiple similar platforms, where similar historical false evaluation information includes similar false positive evaluation data and similar malicious negative evaluation data.

[0035] Next, according to the predetermined evaluation feature attributes, feature extraction and cluster analysis are performed on multiple similar historical false positive review data sets and similar historical malicious negative review data sets, where the feature extraction and clustering methods are the same as the method for obtaining the first false positive review indicator mentioned above, and are not explained in detail here. Multiple associated abnormal evaluation indicators are obtained, which can help distinguish the characteristics of false positive reviews and malicious negative reviews.

[0036] S300: Determine an abnormality evaluation index by fusing the first abnormality evaluation index and a plurality of associated abnormality evaluation indexes.

[0037] Furthermore, step S300 of the present invention further includes: S310: Based on the platform attribute characteristics, respectively analyze the correlation between the target platform and multiple similar platforms to determine the correlations of multiple platforms; S320: Calculate the ratio of each platform correlation to the sum of the correlations of multiple platforms, set it as the indicator fusion weight, and obtain multiple indicator fusion weights; S330: Perform weighted fusion on the multiple associated anomaly evaluation indicators according to the multiple indicator fusion weights to determine the first associated anomaly evaluation indicator; S340: Perform secondary weighted fusion on the first anomaly evaluation indicator and the first associated anomaly evaluation indicator to obtain the anomaly evaluation indicator, wherein the fusion weight of the first anomaly evaluation indicator is 0.7, and the fusion weight of the first associated anomaly evaluation indicator is 0.3.

[0038] Specifically, first, based on the platform attribute characteristics (platform type, user activity, commodity or service category, user group characteristics), the target platform and multiple similar platforms are analyzed for association, and the similarity or relationship between each platform is calculated. For example, by calculating the cosine similarity of the attribute characteristics between the platforms, the association between the target platform and the similar platform is obtained, and the association between multiple platforms is determined. Then, the ratio of the association of each platform to the sum of the associations of multiple platforms is calculated, set as the indicator fusion weight, and multiple indicator fusion weights are obtained. Further, the multiple associated abnormal evaluation indicators are weighted fused according to the multiple indicator fusion weights, that is, according to the calculated indicator fusion weight of each platform, the multiple associated abnormal evaluation indicators are weighted calculated to obtain the first associated abnormal evaluation indicator. The fusion weight of the first abnormal evaluation indicator and the fusion weight of the first associated abnormal evaluation indicator are obtained, wherein the fusion weight of the first abnormal evaluation indicator is 0.7, and the fusion weight of the first associated abnormal evaluation indicator is 0.3; then, the first abnormal evaluation indicator and the first associated abnormal evaluation indicator are weighted fused twice to obtain the abnormal evaluation indicator. By weighted fusion of the first abnormal evaluation index and the first associated abnormal evaluation index, the final abnormal evaluation index can be obtained, which comprehensively considers the characteristics and historical evaluation data of the target platform and multiple similar platforms, and can more accurately identify and filter false evaluations or malicious negative reviews. Through the correlation analysis of platform attribute characteristics, combined with the clustering results of false evaluation features of multiple platforms, a more accurate abnormal evaluation index can be obtained by weighted fusion.

[0039] S400: Performing data credibility analysis on the evaluation information set of the target platform according to the abnormal evaluation index, setting the evaluation information that meets the dynamic credibility threshold as reliable evaluation data, obtaining a reliable evaluation data set, and sharing user evaluation information of the target platform.

[0040] Furthermore, step S400 of the present invention further includes: S410: Obtain the abnormal evaluation index, wherein the abnormal evaluation index includes multiple abnormal evaluation attribute data; S420: Randomly select first evaluation information from the evaluation information set, perform similarity analysis on the first evaluation information according to the multiple abnormal evaluation attribute data, obtain a first similarity set, and calculate the first similarity coefficient by average; S430: Subtract the first similarity coefficient from 1 to obtain a first data credibility coefficient, and sequentially analyze to obtain multiple data credibility coefficients of multiple evaluation information in the evaluation information set.

[0041] Specifically, the abnormal evaluation index is obtained, wherein the abnormal evaluation index includes a plurality of abnormal evaluation attribute data, including the number of evaluation words, the frequency of occurrence of keywords, the frequency of use of modal particles, the number and quality of pictures, etc.; then, any one evaluation information is randomly selected from the evaluation information set as the first evaluation information, and the first evaluation information is respectively subjected to similarity analysis according to the plurality of abnormal evaluation attribute data, that is, the feature data of the first evaluation information and other evaluations in the evaluation information set are compared, and the similarity is calculated, such as the similarity between the evaluation information is calculated by the cosine similarity of the feature data, and then the similarity values ​​of the first evaluation information and other evaluation information are formed into a first similarity set, that is, the similarity scores of the first evaluation information and all other evaluation information are included; further, all similarity values ​​in the first similarity set are averaged to obtain a first similarity coefficient, wherein a high similarity means that the evaluation information is highly similar to other evaluation information, and may be a false evaluation, so the credibility is low; a low similarity means that the evaluation information is relatively independent and has a high credibility. Then, 1 is subtracted from the first similarity coefficient, and the difference between the two is taken as the first data credibility coefficient, which represents the credibility of the evaluation information and can be used as a basis for screening true evaluations and false evaluations. Then, the same method for calculating the first data credibility coefficient is used to calculate the data credibility coefficients of the multiple evaluation information in the evaluation information set in turn to obtain multiple data credibility coefficients of the multiple evaluation information.

[0042] Furthermore, step S400 of the present invention further includes: S440: Based on big data, obtain the sequence of the number of false evaluation information and the sequence of the proportion of false evaluation information on the target platform within a preset historical time zone, as well as multiple sequences of the number of similar false evaluation information and multiple sequences of the proportion of similar false evaluation information on multiple similar platforms within a preset time zone.

[0043] Specifically, configure the preset historical time zone, that is, select an appropriate time range (for example, the past week, month, quarter, etc.) according to the needs, and this time interval will be used as the basis for analyzing false evaluation information. Next, based on big data, obtain the number sequence of false evaluation information and the proportion sequence of false evaluation information in the preset historical time zone of the target platform, that is, count the number of false evaluations of the target platform in each time period (for example, every hour, every day, every week, etc.) in the preset time zone to form a false evaluation number sequence; calculate the proportion of false evaluations of the target platform in each time period in the preset time zone. On the other hand, obtain multiple similar false evaluation information number sequences and multiple similar false evaluation information proportion sequences of multiple similar platforms in the preset time zone. By obtaining the historical false evaluation information of the target platform and multiple similar platforms, and analyzing their quantity and proportion sequences, we can clearly understand the performance and changing trends of false evaluations on each platform, and provide effective support for further optimizing the false evaluation filtering algorithm and improving the credibility of platform data.

[0044] S450: pre-training an evaluation trend analysis plug-in, and inputting the false evaluation information quantity sequence, false evaluation information ratio sequence, multiple platform associations, multiple similar false evaluation information quantity sequences and multiple similar false evaluation information ratio sequences into the evaluation trend analysis plug-in, and outputting an evaluation trend coefficient.

[0045] Further, step S450 of the present invention further includes: S451: Based on big data, collect sample training data sets, wherein the sample training data includes a sample false evaluation information quantity sequence, a sample false evaluation information ratio sequence, a sample platform correlation set, a sample similar false evaluation information quantity sequence set, a sample similar false evaluation information ratio sequence set and a sample evaluation trend coefficient, and the sample evaluation trend coefficient can be obtained by manual labeling; S452: Use the sample training data set to perform supervised training and testing on the BP neural network until the loss function converges to obtain a trained evaluation trend analysis plug-in.

[0046] Specifically, firstly, based on big data, sample training data sets are collected, where the sample training data include sample false evaluation information quantity sequence, sample false evaluation information proportion sequence, sample platform correlation set, sample similar false evaluation information quantity sequence set, sample similar false evaluation information proportion sequence set and sample evaluation trend coefficient. The evaluation trend coefficient represents the changing trend of false evaluations of the target platform and similar platforms within a preset time interval, which can be determined through manual labeling. The labeler determines the degree of increase, decrease or stability of the trend by checking the fluctuation trend of the platform's false evaluation quantity sequence and proportion sequence. For example, if the number of false evaluations on a platform is on an upward trend, the labeling trend coefficient is 1.1. If it is on a downward trend, it can be labeled as 0.95.

[0047] Then, an evaluation trend analysis plug-in is constructed based on the BP neural network. The evaluation trend analysis plug-in is a BP neural network model that can be iteratively optimized in machine learning. The evaluation trend analysis plug-in includes an input layer, multiple hidden layers, and an output layer, wherein the input data of the input layer are a sample false evaluation information quantity sequence, a sample false evaluation information ratio sequence, a sample platform association set, a sample similar false evaluation information quantity sequence set, and a sample similar false evaluation information ratio sequence set, and the output data is a sample evaluation trend coefficient. Then, the sample training data set is divided into a training set and a verification set according to a predetermined ratio, for example, the training set ratio is 85%, and the verification set ratio is 15%; then, the training set and the verification set are used to supervise the training and verification of the evaluation trend analysis plug-in; during the training process, first, the input data in the training set is propagated through each layer of the neural network, and the output of each node is calculated to finally obtain the predicted evaluation trend coefficient; then, the loss function is used to calculate the gap between the predicted value and the true value; then, the gradient of the loss function for each weight is calculated, and the weights and biases in the neural network are updated by the gradient descent method; it is continuously iterated until the loss function converges. Each iteration optimizes the weights through forward propagation and back propagation, and ultimately makes the loss function as small as possible. After each round of training, the model is verified using the validation set, and the error between the predicted results on the validation set and the actual results is calculated. By comparing the errors on the training set and the validation set, the training progress of the model can be monitored to avoid overfitting. Get the evaluation trend analysis plug-in after the training is completed.

[0048] Then, the false evaluation information quantity sequence, false evaluation information proportion sequence, multiple platform associations, multiple similar false evaluation information quantity sequences and multiple similar false evaluation information proportion sequences are input into the trained evaluation trend analysis plug-in to output the evaluation trend coefficient.

[0049] S460: Optimize and adjust the predetermined credibility threshold according to the evaluation trend coefficient to obtain a dynamic credibility threshold; S470: Judge the multiple data credibility coefficients according to the dynamic credibility threshold, set the evaluation information corresponding to the data credibility coefficient greater than the dynamic credibility threshold as reliable evaluation data, and obtain a reliable evaluation data set.

[0050] Specifically, according to the evaluation trend coefficient, the credibility threshold is dynamically adjusted so that when the number of false evaluations increases, the threshold is raised to reduce the impact of false evaluations; when the number of false evaluations decreases, the threshold is lowered to more accurately screen out true evaluations, that is, the evaluation trend coefficient is multiplied by the predetermined credibility threshold to obtain a dynamic credibility threshold. The multiple data credibility coefficients are further judged according to the dynamic credibility threshold. If the data credibility coefficient is greater than the dynamic credibility threshold, it indicates that the credibility of the evaluation information is high and can be considered as reliable evaluation data; if the data credibility coefficient is less than or equal to the dynamic credibility threshold, it indicates that the credibility of the evaluation information is low and can be considered as false evaluation data and needs to be filtered out; then the evaluation information corresponding to the data credibility coefficient greater than the dynamic credibility threshold is set as reliable evaluation data to obtain a reliable evaluation data set.

[0051] By optimizing and adjusting the credibility threshold based on the evaluation trend coefficient and screening out reliable evaluation data according to the dynamic credibility threshold, the platform can more accurately identify and filter out false evaluations and improve the quality and accuracy of evaluation data; ultimately, the platform can provide users with real and reliable evaluation information, enhance user trust, and thus improve the overall reputation of the platform and user experience.

[0052] The user information sharing method based on big data provided by the embodiment of the present invention has at least the following technical effects: By performing feature clustering analysis on the false evaluation information of the target platform, a first abnormal evaluation index is determined; then, based on big data, feature clustering analysis is performed on the false evaluation information of multiple similar platforms of the target platform to determine multiple related abnormal evaluation indicators; further, the abnormal evaluation indicator is determined based on the fusion of the first abnormal evaluation indicator and multiple related abnormal evaluation indicators; then, data credibility analysis is performed on the evaluation information set of the target platform according to the abnormal evaluation indicator, and the evaluation information that meets the dynamic credibility threshold is set as reliable evaluation data to obtain a reliable evaluation data set, and user evaluation information of the target platform is shared; that is, by performing clustering analysis on the false evaluation features of the target platform and multiple similar platforms, false positive reviews and malicious negative reviews can be identified more accurately, unreliable evaluation data can be filtered out, and true evaluation information can be ensured to be fed back to customers, thereby reducing the negative impact of false evaluations on the platform's reputation, optimizing user experience, and improving customer satisfaction.

[0053] Embodiment 2, as Figure 2As shown, based on the same inventive concept as the user information sharing method based on big data provided in Example 1, an embodiment of the present invention also provides a user information sharing system based on big data, including: a first abnormal evaluation index determination module 11, used to perform feature clustering analysis on the false evaluation information of the target platform, and determine a first abnormal evaluation index; an associated abnormal evaluation index determination module 12, used to perform feature clustering analysis on the false evaluation information of multiple similar platforms of the target platform based on big data, and determine multiple associated abnormal evaluation indicators; an abnormal evaluation index fusion module 13, used to determine the abnormal evaluation index based on the fusion of the first abnormal evaluation index and multiple associated abnormal evaluation indicators; a data trust analysis module 14, used to perform data trust analysis on the evaluation information set of the target platform according to the abnormal evaluation index, set the evaluation information that meets the dynamic credibility threshold as reliable evaluation data, obtain a reliable evaluation data set, and share user evaluation information of the target platform.

[0054] Furthermore, the user information sharing system based on big data is also used to: obtain a historical false evaluation information set of a target platform, wherein the false evaluation information includes false positive evaluation data and malicious negative evaluation data; perform feature extraction and cluster analysis on the historical false positive evaluation data set according to predetermined evaluation feature attributes to obtain a first false positive evaluation index; perform feature extraction and cluster analysis on the historical malicious negative evaluation data set according to predetermined evaluation feature attributes to obtain a first malicious negative evaluation index; and use the first false positive evaluation index and the first malicious negative evaluation index as the first abnormal evaluation index.

[0055] Furthermore, the user information sharing system based on big data is also used to: obtain predetermined evaluation feature attributes, wherein the predetermined evaluation feature attributes include at least the number of evaluation words, keywords, modal particles, number of pictures and picture quality; according to the predetermined evaluation feature attributes, perform feature extraction on multiple historical false positive review data in the historical false positive review data set to obtain multiple evaluation feature data sets; perform feature clustering on the multiple evaluation feature data sets using the K-means clustering algorithm to obtain the first false positive review indicator.

[0056] Furthermore, the user information sharing system based on big data is also used to: use the platform attribute characteristics of the target platform as constraints, perform associated retrieval based on big data, determine multiple similar platforms, and obtain multiple similar historical false evaluation information sets of the multiple similar platforms, wherein the similar historical false evaluation information includes similar false positive review data and similar malicious negative review data; perform feature extraction and cluster analysis on multiple similar historical false positive review data sets and similar historical malicious negative review data sets according to predetermined evaluation feature attributes, and obtain multiple associated abnormal evaluation indicators.

[0057] Furthermore, the user information sharing system based on big data is also used to: based on platform attribute characteristics, respectively analyze the correlation between the target platform and multiple similar platforms to determine the correlations of multiple platforms; respectively calculate the ratio of the correlation of each platform to the sum of the correlations of multiple platforms, set it as the indicator fusion weight, and obtain multiple indicator fusion weights; perform weighted fusion on the multiple associated abnormality evaluation indicators according to the multiple indicator fusion weights to determine a first associated abnormality evaluation indicator; perform secondary weighted fusion on the first abnormality evaluation indicator and the first associated abnormality evaluation indicator to obtain the abnormality evaluation indicator, wherein the fusion weight of the first abnormality evaluation indicator is 0.7, and the fusion weight of the first associated abnormality evaluation indicator is 0.3.

[0058] Furthermore, the user information sharing system based on big data is also used to: obtain the abnormal evaluation index, wherein the abnormal evaluation index includes multiple abnormal evaluation attribute data; randomly select first evaluation information from the evaluation information set, and perform similarity analysis on the first evaluation information according to the multiple abnormal evaluation attribute data to obtain a first similarity set, and calculate the first similarity coefficient by average; subtract the first similarity coefficient from 1 to obtain a first data credibility coefficient, and analyze in turn to obtain multiple data credibility coefficients of multiple evaluation information in the evaluation information set.

[0059] Furthermore, the user information sharing system based on big data is also used to: based on big data, obtain a sequence of the number of false evaluation information and a sequence of the proportion of false evaluation information on a target platform within a preset historical time zone, as well as multiple sequences of the number of similar false evaluation information and multiple sequences of the proportion of similar false evaluation information on multiple similar platforms within a preset time zone; pre-train an evaluation trend analysis plug-in, and input the sequence of the number of false evaluation information, the sequence of the proportion of false evaluation information, the correlation of multiple platforms, the sequence of the number of multiple similar false evaluation information and the sequence of the proportion of multiple similar false evaluation information into the evaluation trend analysis plug-in, and output an evaluation trend coefficient; optimize and adjust a predetermined credibility threshold according to the evaluation trend coefficient to obtain a dynamic credibility threshold; judge the multiple data credibility coefficients according to the dynamic credibility threshold, set the evaluation information corresponding to the data credibility coefficient greater than the dynamic credibility threshold as reliable evaluation data, and obtain a reliable evaluation data set.

[0060] Furthermore, the user information sharing system based on big data is also used for: collecting sample training data sets based on big data, wherein the sample training data includes a sample false evaluation information quantity sequence, a sample false evaluation information proportion sequence, a sample platform correlation set, a sample similar false evaluation information quantity sequence set, a sample similar false evaluation information proportion sequence set and a sample evaluation trend coefficient, and the sample evaluation trend coefficient can be obtained by manual labeling; using the sample training data set to perform supervised training and testing on the BP neural network until the loss function converges, thereby obtaining a trained evaluation trend analysis plug-in.

[0061] Although preferred embodiments of the present invention have been described, additional changes and modifications may occur to these embodiments once those skilled in the art understand the basic inventive concepts.

[0062] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention belong to the scope of the present invention and its equivalent technologies, the present invention is also intended to include these changes and variations.

Claims

1. A user information sharing method based on big data, characterized in that: Methods include: Perform feature clustering analysis on the false evaluation information of the target platform to determine the first abnormal evaluation index; Based on big data, feature clustering analysis is performed on the false evaluation information of multiple similar platforms of the target platform to determine multiple associated abnormal evaluation indicators; Determine an abnormality evaluation index by fusing the first abnormality evaluation index and a plurality of associated abnormality evaluation indexes; According to the abnormal evaluation index, the evaluation information set of the target platform is subjected to data credibility analysis, and the evaluation information that meets the dynamic credibility threshold is set as reliable evaluation data to obtain a reliable evaluation data set, and the user evaluation information of the target platform is shared.

2. According to the user information sharing method based on big data in claim 1, it is characterized in that: Perform feature cluster analysis on the false evaluation information of the target platform to determine the first abnormal evaluation index, including: Obtain a set of historical false evaluation information of the target platform, wherein the false evaluation information includes false positive evaluation data and malicious negative evaluation data; According to the predetermined evaluation feature attributes, feature extraction and cluster analysis are performed on the historical false positive review data set to obtain a first false positive review index; According to the predetermined evaluation feature attributes, feature extraction and cluster analysis are performed on the historical malicious negative review data set to obtain a first malicious negative review index; The first false positive review indicator and the first malicious negative review indicator are used as the first abnormal evaluation indicator.

3. A user information sharing method based on big data according to claim 2, characterized in that: According to the predetermined evaluation feature attributes, feature extraction and cluster analysis are performed on the historical false positive review data set to obtain the first false positive review index, including: Obtaining predetermined evaluation feature attributes, wherein the predetermined evaluation feature attributes at least include the number of evaluation words, keywords, modal particles, the number of pictures, and the quality of the pictures; According to the predetermined evaluation feature attributes, feature extraction is performed on multiple historical false positive review data in the historical false positive review data set to obtain multiple evaluation feature data sets; The plurality of evaluation feature data sets are clustered using a K-means clustering algorithm to obtain the first false positive rating indicator.

4. The user information sharing method based on big data according to claim 2 is characterized in that: Based on big data, we conduct feature clustering analysis on the false evaluation information of multiple similar platforms of the target platform and determine multiple associated abnormal evaluation indicators, including: Taking the platform attribute characteristics of the target platform as a constraint, performing association retrieval based on big data, determining multiple similar platforms, and obtaining multiple similar historical false evaluation information sets of the multiple similar platforms, wherein the similar historical false evaluation information includes similar false positive evaluation data and similar malicious negative evaluation data; According to the predetermined evaluation feature attributes, feature extraction and cluster analysis are performed on multiple similar historical false positive review data sets and similar historical malicious negative review data sets to obtain multiple associated abnormal evaluation indicators.

5. A user information sharing method based on big data according to claim 4, characterized in that: Determining an abnormality evaluation index by fusing the first abnormality evaluation index and a plurality of associated abnormality evaluation indexes includes: Based on the platform attribute characteristics, the target platform and multiple similar platforms are analyzed for correlation, and the correlation between multiple platforms is determined; Calculate the ratio of each platform's correlation degree to the sum of multiple platform's correlation degrees respectively, set it as the indicator fusion weight, and obtain multiple indicator fusion weights; Performing weighted fusion on the multiple associated abnormality evaluation indicators according to the multiple indicator fusion weights to determine a first associated abnormality evaluation indicator; The first abnormality evaluation index and the first associated abnormality evaluation index are subjected to secondary weighted fusion to obtain the abnormality evaluation index, wherein the fusion weight of the first abnormality evaluation index is 0.7, and the fusion weight of the first associated abnormality evaluation index is 0.

3.

6. A user information sharing method based on big data according to claim 5, characterized in that: Performing data credibility analysis on the evaluation information set of the target platform according to the abnormal evaluation index includes: Acquiring the abnormal evaluation index, wherein the abnormal evaluation index includes a plurality of abnormal evaluation attribute data; Randomly selecting first evaluation information from the evaluation information set, performing similarity analysis on the first evaluation information according to the plurality of abnormal evaluation attribute data to obtain a first similarity set, and calculating the mean to obtain a first similarity coefficient; The first similarity coefficient is subtracted from 1 to obtain a first data credibility coefficient, and multiple data credibility coefficients of multiple evaluation information in the evaluation information set are analyzed in sequence.

7. A user information sharing method based on big data according to claim 6, characterized in that: The evaluation information that meets the dynamic credibility threshold is set as reliable evaluation data, and a reliable evaluation data set is obtained, including: Based on big data, obtain the number sequence of false evaluation information and the ratio sequence of false evaluation information of the target platform in a preset historical time zone, as well as multiple number sequences of similar false evaluation information and multiple ratio sequences of similar false evaluation information of multiple similar platforms in a preset time zone; Pre-training an evaluation trend analysis plug-in, and inputting the false evaluation information quantity sequence, the false evaluation information ratio sequence, multiple platform associations, multiple similar false evaluation information quantity sequences, and multiple similar false evaluation information ratio sequences into the evaluation trend analysis plug-in, and outputting an evaluation trend coefficient; Optimizing and adjusting the predetermined credibility threshold according to the evaluation trend coefficient to obtain a dynamic credibility threshold; The multiple data credibility coefficients are judged according to the dynamic credibility threshold, and the evaluation information corresponding to the data credibility coefficient greater than the dynamic credibility threshold is set as reliable evaluation data to obtain a reliable evaluation data set.

8. The user information sharing method based on big data according to claim 7 is characterized in that: Pre-training evaluation trend analysis plug-in, including: Based on big data, a sample training data set is collected, wherein the sample training data includes a sample false evaluation information quantity sequence, a sample false evaluation information ratio sequence, a sample platform association degree set, a sample similar false evaluation information quantity sequence set, a sample similar false evaluation information ratio sequence set, and a sample evaluation trend coefficient, wherein the sample evaluation trend coefficient can be obtained by manual annotation; The sample training data set is used to perform supervised training and testing on the BP neural network until the loss function converges, thereby obtaining a trained evaluation trend analysis plug-in.

9. A user information sharing system based on big data, characterized in that: The steps for implementing the user information sharing method based on big data as described in any one of claims 1 to 8 include: A first abnormal evaluation index determination module, used to perform feature cluster analysis on the false evaluation information of the target platform to determine a first abnormal evaluation index; The module for determining the associated abnormal evaluation index is used to perform feature clustering analysis on the false evaluation information of multiple similar platforms of the target platform based on big data, and determine multiple associated abnormal evaluation indicators; An abnormality evaluation index fusion module, used to determine an abnormality evaluation index by fusing the first abnormality evaluation index and a plurality of associated abnormality evaluation indicators; The data credibility analysis module is used to perform data credibility analysis on the evaluation information set of the target platform according to the abnormal evaluation index, set the evaluation information that meets the dynamic credibility threshold as reliable evaluation data, obtain a reliable evaluation data set, and share user evaluation information of the target platform.

Images