A user credibility assessment method and apparatus

Abstract

A user credibility evaluation method and device, comprising: obtaining attitude evaluation information of a target user according to a plurality of microblogs published by the target user; wherein the target user is a user to be evaluated for credibility, and the attitude evaluation information is used to evaluate the degree of positive attitude of the plurality of microblogs published by the target user; obtaining topic labels involved in the plurality of microblogs published by the target user, and performing the following operations on each topic label obtained: obtaining the activity, attention, mention and popularity of the target user under the topic label, respectively; and obtaining influence information according to the activity, attention, mention and popularity; and obtaining the credibility of the target user under each topic label according to the influence information of the target user under each topic label and the attitude evaluation information of the target user. The embodiments of the present application comprehensively and accurately realize the evaluation of user credibility, thereby realizing the identification of false users.

Description

TECHNICAL FIELD

[0001] The present application relates to information processing technology, in particular to a user credibility evaluation method and device. BACKGROUND

[0002] With the rapid development of Internet technology, microblog has become the main way of communication for people. However, some institutions and individuals are driven by economic interests and register many false accounts in microblog to carry out commercial marketing, false hype, spread false information and disseminate bad information.

[0003] In the prior art, the research on user credibility in microblog is limited to identifying zombie users in microblog.

[0004] However, it is difficult to identify false users such as "live powder" and "manual powder" in microblog, so the related technology lacks an effective mechanism for comprehensive and accurate evaluation of user credibility. SUMMARY

[0005] The present application provides a user credibility evaluation method and device, which can comprehensively and accurately evaluate the credibility of users, thereby realizing the identification of false users.

[0006] In one aspect, the present application provides a user credibility evaluation method, comprising:

[0007] Obtaining attitude evaluation information of a target user from multiple microblogs published by the target user; wherein the target user is a user to be evaluated for credibility, and the attitude evaluation information is used to evaluate the degree of positive attitude of the multiple microblogs published by the target user;

[0008] Obtaining topic labels involved in the multiple microblogs published by the target user, and performing the following operations on each obtained topic label: obtaining the activity of the target user under the topic label, obtaining the attention of the microblog published by the target user under the topic label, obtaining the mention of the microblog published by the target user under the topic label, and obtaining the popularity of the microblog published by the target user under the topic label; obtaining influence information of the target user under the topic label according to the activity of the target user under the topic label, the attention of the microblog published by the target user, the mention of the microblog published by the target user, and the popularity of the microblog published by the target user;

[0009] Obtaining the credibility of the target user under each topic label according to the influence information of the target user under each topic label and the attitude evaluation information of the target user.

[0010] In another aspect, the present application provides a user credibility evaluation device, comprising a memory and a processor, wherein the memory is used to save an executable program;

[0011] The processor is configured to read and execute the executable program to implement the user credibility evaluation method as described above.

[0012] Compared with the related art, the embodiment of the application first acquires attitude evaluation information of a target user according to multiple micro blogs published by the target user, then classifies the multiple micro blogs according to topic labels, comprehensively measures influence information of the target user under each topic label from different dimensions such as activity, attention, mention and popularity, and finally evaluates credibility of the target user under each topic label based on the influence information under each topic label and the attitude evaluation information of the target user. Therefore, the evaluation of the user credibility is comprehensively and accurately realized, and the identification of the false user is realized.

[0013] Other features and advantages of the present application will be set forth in the following description, and in part will become apparent to those skilled in the art upon examination of the following or can be learned by practice of the present application. Other advantages of the present application can be realized and attained by means of the instrumentalities and combinations particularly pointed out in the description and appended claims.

[0014] Other features and advantages of the present application will be set forth in the following description, and in part will become apparent to those skilled in the art upon examination of the following or can be learned by practice of the present application. Other advantages of the present application can be realized and attained by means of the instrumentalities and combinations particularly pointed out in the description and appended claims. BRIEF DESCRIPTION OF DRAWINGS

[0015] The accompanying drawings are included to provide a further understanding of the present application, and constitute a part of the specification, illustrate embodiments of the present application and are used to explain the technical solutions of the present application, and do not constitute a limitation on the technical solutions of the present application.

[0016] Figure 1 A flowchart of a user credibility evaluation method according to an embodiment of the application is shown in FIG. 1;

[0017] Figure 2 A process framework of a user credibility evaluation according to an embodiment of the application is shown in FIG. 2. DETAILED DESCRIPTION

[0018] The present application describes a number of embodiments, but the description is exemplary rather than limiting and it will be apparent to those of ordinary skill in the art that numerous more embodiments and implementations are possible within the scope of the embodiments described in the present application. Although a number of possible combinations of features have been set forth in the appended figures and discussed above, many other combinations of the disclosed features are possible. Unless specifically intended otherwise, any feature or element of any embodiment can be used in combination with any other feature or element of any other embodiment, or in replacement of any other feature or element in any other embodiment.

[0019] The present application includes and contemplates combinations of features and elements known to those of ordinary skill in the art. The embodiments, features and elements disclosed herein can also be combined with any conventional features or elements to form a unique application of the presently claimed application that is not specifically disclosed. Any feature or element of any embodiment can also be combined with features or elements from other applications to form another unique application of the presently claimed application that is not specifically disclosed. Thus, it should be understood that any feature shown and / or discussed in the present application can be used alone or in any suitable combination. Accordingly, the embodiments are not to be restricted, except as by the appended claims and their equivalents. Furthermore, various modifications and changes can be made within the scope of the appended claims.

[0020] Furthermore, in describing representative embodiments, the specification can have presented the method and / or process as a particular sequence of steps. However, to the extent that the method or process depends on the particular order of steps, this description should not be construed as limiting unless specifically so stated. Other steps can be performed in between described steps without departing from the spirit of the application. For instance, it is possible that data was processed by additional steps not mentioned in the above description of the method and / or process. Thus, the specific order of steps recited in the specification is not an inherent part of the embodiments. The methods and processes described herein are also not limited to the specific recited order of steps unless specifically stated as such. Furthermore, the claims can not be limited to the specific recited steps of the method and / or process, as the skilled artisan can readily appreciate that the order of steps can vary and still remain within the spirit and scope of the application.

[0021] With the rapid development of Internet technology, microblog has become the main way of communication and exchange for people, and the users in microblog also show an explosive growth trend. As the main information acquisition channel for people, microblog not only facilitates the communication between people, but also brings great convenience to people's life. However, some institutions and individuals are driven by economic interests and register a large number of false accounts on the microblog platform to carry out commercial marketing, false hype, spread false information and spread bad information, etc. Due to the characteristics of microblog that the information data volume is huge and the transmission speed is extremely fast, it has become a breeding ground for false information and bad information. The large number of false users and various false information in microblog not only affect people's acquisition of normal information, but also pose a threat to social stability.

[0022] The existing microblog platform, users can publish their own views on such platform at any time and anywhere, and can also see the views, opinions and attitudes published by other users. If the credibility of the user cannot be accurately calculated and properly constrained according to the credibility, the unverified remarks published by the user with low credibility may ferment into shocking false information in a few minutes. However, the research on the credibility of users in microblog is currently mainly focused on content mining, theme discovery and user influence evaluation, and the proposed solutions are mostly limited to identifying zombie fans in microblog. For example, some researchers use the idea of text classification to establish a feature library for registered users of microblog, and combine artificial neural network (ANN) and support vector machine (SVM) and other methods to realize automatic classification of users, so as to identify zombie fans. However, there are relatively few relevant solutions on how to identify false users in microblog and how to calculate the credibility of users according to the user features in microblog.

[0023] In summary, most of the current user credibility calculation methods only perform binary classification processing on users, and less reasonably quantify and evaluate the user credibility by taking into account multiple feature dimensions. Although the proposed implementation technology can identify zombie fans in microblog, it is difficult to identify false users such as "live fans" and "artificial fans" in microblog. Therefore, it is an urgent task to establish an effective mechanism for judging and measuring the credibility of users in microblog according to various key feature dimensions of users, so as to more comprehensively and accurately evaluate the credibility of users.

[0024] Therefore, the embodiments of the present application provide a user credibility evaluation method, as shown in Figure 1 , comprising:

[0025] In step 101, attitude evaluation information of a target user is obtained according to a plurality of microblogs published by the target user; the target user is a user to be evaluated for credibility, and the attitude evaluation information is used to evaluate the degree of positive attitude of the plurality of microblogs published by the target user.

[0026] In step 102, topic labels involved in the plurality of microblogs published by the target user are obtained, and each obtained topic label is operated as follows: the activity of the target user under the topic label is obtained, the attention degree of the microblog published by the target user under the topic label is obtained, the mentioned degree of the microblog published by the target user under the topic label is obtained, and the popularity of the microblog published by the target user under the topic label is obtained; and influence information of the target user under the topic label is obtained according to the activity of the target user under the topic label, the attention degree of the microblog published by the target user under the topic label, the mentioned degree of the microblog published by the target user under the topic label, and the popularity of the microblog published by the target user under the topic label.

[0027] In step 103, credibility of the target user under each topic label is obtained according to the influence information of the target user under each topic label and the attitude evaluation information of the target user.

[0028] For example, the number of the plurality of microblogs published by the target user can be a pre-set number.

[0029] The user credibility evaluation method provided by the embodiment of the present application first obtains the attitude evaluation information of the target user according to the plurality of microblogs published by the target user, then classifies the plurality of microblogs according to topic labels, comprehensively measures the influence information of the target user under each topic label from the activity, the attention degree, the mentioned degree, and the popularity, and finally evaluates the credibility of the target user under each topic label based on the influence information under each topic label and the attitude evaluation information of the target user. Therefore, the evaluation of the credibility of the user is comprehensive and accurate, and the identification of the false user is realized.

[0030] The user credibility evaluation method provided by the embodiment of the present application can calculate the credibility of the user in the microblog. The method first calculates the historical attitude of the user and the influence information of the user according to the multi-dimensional features of the user, then comprehensively considers the historical attitude of the user and the influence information of the user to form a unified index of the credibility of the user, and more comprehensively and accurately represents the credibility of the user. That is, the user credibility evaluation method provided by the embodiment of the present application mainly includes four steps, which are as follows:

[0031] In step 1, the historical attitude of the user is calculated according to the content of the published blog of the user.

[0032] In step 2, the current activity of the user is calculated.

[0033] Step 3, calculate the user attention, user mentioned degree and user popularity, and then calculate the user influence by using the calculated results;

[0034] Step 4, according to the user historical attitude, user activity and user influence, the user credibility is calculated.

[0035] Compared with the prior art, the user credibility evaluation method provided by the embodiment has the following beneficial effects:

[0036] 1. The method provides a user credibility calculation method for users in microblog, reasonably quantifies and evaluates the user credibility, represents the credibility of a user, and solves the problem that the existing method can only rigidly classify users;

[0037] 2. According to the calculated user credibility, the method can further identify false users in microblog, and is not limited to identifying zombie fans in microblog;

[0038] 3. The user credibility evaluation method provided by the method can be applied to most users in microblog, and has high universality.

[0039] In an exemplary example, the attitude evaluation information of the target user is obtained according to a plurality of microblogs published by the target user, including:

[0040] First, for each microblog in the plurality of microblogs published by the target user, the following operations are performed: the frequency of occurrence of positive words and the frequency of occurrence of negative words in the microblog are counted by using a pre-set positive word set and a pre-set negative word set, and the attitude information of the microblog is determined according to the frequency of occurrence of positive words and the frequency of occurrence of negative words in the microblog;

[0041] Second, the attitude evaluation information of the target user is calculated according to the attitude information of all microblogs of the target user and the pre-set attitude weight corresponding to each attitude information.

[0042] Wherein, the attitude information of each microblog can include positive microblog, negative microblog and neutral microblog. Since a microblog can contain both positive words and negative words, first, the frequency of occurrence of positive words C + (W i ) and the frequency of occurrence of negative words C - (W i ) in a microblog are counted by using a pre-set positive word set and a pre-set negative word set, and then the attitude information of the microblog (W i ) is represented according to the following formula:

[0043]

[0044] wherein C + (W i ) represents the frequency of appearance of positive words in the microblog W i (W - ) represents the frequency of appearance of negative words in the microblog W i i represents the positive microblog, W + represents the positive microblog, W - represents the negative microblog, and W 0 represents the neutral microblog. In the process of calculating the attitude evaluation information of the target user, the weight of the positive microblog can be set as 2, the weight of the neutral microblog can be set as 1, and the weight of the negative microblog can be set as 0.

[0045] The attitude weight sum of all microblogs published by the target user is calculated according to the attitude information of all microblogs of the target user and the attitude weight corresponding to each kind of attitude information set in advance, and the calculated attitude weight sum is directly taken as the attitude evaluation information of the target user.

[0046] Alternatively, the attitude weight sum of all microblogs published by the target user is calculated according to the attitude information of all microblogs of the target user and the attitude weight corresponding to each kind of attitude information set in advance, and then the attitude evaluation information of the target user is obtained after normalization processing. The specific calculation process can be shown in the following formula:

[0047]

[0048] wherein represents the number of positive microblogs published by the target user u i represents the number of neutral microblogs published by the target user u i represents the number of negative microblogs published by the target user u i

[0049] In an exemplary instance, the obtaining of the activity of the target user under the topic label comprises:

[0050] First, the number of microblogs published by the target user under the topic label is obtained, and the number of all microblogs published under the topic label is obtained;

[0051] Second, the activity of the target user under the topic label is calculated according to the number of microblogs published by the target user under the topic label and the number of all microblogs published under the topic label.

[0052] ​​​​For example, the user activity of a user under a topic label is calculated by collecting the number of microblogs published by the user under the topic label. Assuming that the set of all users participating in a topic and publishing a number of microblogs greater than or equal to 1 under a topic label t∈T is U, and the set of all microblogs published under the topic is W t , the set of microblogs published by a user u i ∈U under the topic label t is The activity A i (u t ) of the user u i under the topic label t is calculated as follows:

[0053]

[0054] In an example, the attention degree of the microblog published by the target user under the topic label is obtained by:

[0055] First, the number of fans of the target user is obtained, and the number of fans of the user with the largest number of fans under the topic label is obtained.

[0056] Second, the attention degree of the microblog published by the target user under the topic label is calculated according to the number of fans of the target user and the number of fans of the user with the largest number of fans under the topic label.

[0057] For example, the calculation process of the attention degree can be implemented by logarithm. In a given topic label t∈T, the attention degree F t (u i ) of the microblog published by the target user under the topic label is calculated according to the number of fans num follow (u i ) of the target user, and the calculation method is as follows:

[0058]

[0059] Wherein, the numerator represents the logarithmic value of the number of fans of the target user u i , and the denominator represents the logarithmic value of the number of fans of the user with the largest number of fans under the topic label t.

[0060] Logarithmic calculation can reduce differentiation and improve evaluation accuracy.

[0061] In an example, the mentioned degree of the microblog published by the target user under the topic label is obtained by:

[0062] First, the number of mentions of the microblog published by the target user under the topic label is obtained, and the number of mentions of the microblog published by the user with the largest number of mentions under the topic label is obtained.

[0063] Secondly, the mentioned degree of the microblog published by the target user under the topic label is calculated according to the mentioned times of the microblog published by the target user under the topic label and the mentioned times of the microblog published by the user with the most mentioned times under the topic label.

[0064] Exemplarily, the calculation process of the mentioned degree can be realized by logarithm. In the given topic label t∈T, the mentioned degree M t (u i ) of the microblog published by the target user under the topic label is calculated according to the mentioned times num mention (u i ) of the microblog published by the target user under the topic label, and the calculation method is as follows:

[0065] M t (u i ) = log(num mention (u i )) / max(log(num mention (U,t)))

[0066] The numerator represents the logarithm value of the mentioned times of the target user u i under the topic label t, and the denominator represents the logarithm value of the mentioned times of the user with the most mentioned times under the topic label t.

[0067] In an exemplary instance, the obtaining of the popularity of the microblog published by the target user under the topic label comprises:

[0068] Firstly, the liked degree of the microblog published by the target user under the topic label is obtained, the forwarded degree of the microblog published by the target user under the topic label is obtained, and the collected degree of the microblog published by the target user under the topic label is obtained.

[0069] Secondly, the popularity of the microblog published by the target user under the topic label is calculated according to the liked degree, the forwarded degree and the collected degree of the microblog published by the target user under the topic label.

[0070] The popularity of the microblog published by the target user under the topic label is composed of the liked degree, the forwarded degree and the collected degree of the microblog published by the target user under the topic label. The liked degree, the forwarded degree and the collected degree of the microblog published by the target user under the topic label are respectively calculated according to the total number of the liked times of the microblog published by the target user under the topic label, the total number of the forwarded times of the microblog published by the target user under the topic label and the total number of the collected times of the microblog published by the target user under the topic label.

[0071] In an exemplary instance, the obtaining the like degree of the microblog published by the target user under the topic label comprises:

[0072] First, the total number of likes of the microblog published by the target user under the topic label is obtained, and the total number of likes of the microblog published by the user with the largest total number of likes among the microblogs published under the topic label is obtained;

[0073] Second, the like degree of the microblog published by the target user under the topic label is calculated according to the total number of likes of the microblog published by the target user under the topic label and the total number of likes of the microblog published by the user with the largest total number of likes among the microblogs published under the topic label.

[0074] Exemplarily, the calculation process of the like degree can be implemented by logarithm. In a given topic t∈T, the like degree L t (u i of the microblog published by the target user under the topic label is calculated according to the total number of likes of the microblog published by the target user under the topic label, and the calculation manner is as follows:

[0075]

[0076] The numerator represents the logarithmic value of the total number of likes of the microblog published by the target user u i under the topic label, and the denominator represents the logarithmic value of the total number of likes of the microblog published by the user with the largest total number of likes among the microblogs published under the topic label.

[0077] In an exemplary instance, the obtaining the retweet degree of the microblog published by the target user under the topic label comprises:

[0078] First, the total number of retweets of the microblog published by the target user under the topic label and the total number of retweets of the microblog published by the user with the largest total number of retweets among the microblogs published under the topic label are obtained;

[0079] Second, the retweet degree of the microblog published by the target user under the topic label is calculated according to the total number of retweets of the microblog published by the target user under the topic label and the total number of retweets of the microblog published by the user with the largest total number of retweets among the microblogs published under the topic label.

[0080] Exemplarily, the calculation process of the retweet degree can be implemented by logarithm. In a given topic t∈T, the retweet degree R t (u iThe total number of retweets of a user's Weibo posts under this hashtag is calculated as follows:

[0081]

[0082] The molecule represents the target user u i The logarithm of the total number of reposts of Weibo posts under hashtag t, where the denominator represents the logarithm of the total number of reposts of the user with the most Weibo posts under hashtag t under that hashtag.

[0083] In one exemplary instance, obtaining the number of times a Weibo post by the target user under the topic tag has been saved includes:

[0084] First, obtain the total number of times the target user's Weibo posts under this topic tag were collected, and the total number of times the user whose Weibo posts under this topic tag had the highest total number of collections was collected.

[0085] Secondly, the number of times the target user's Weibo posts under this topic tag are collected is calculated based on the total number of times they are collected, and the total number of times the user with the most collected Weibo posts under this topic tag is collected.

[0086] For example, the calculation of the number of times a post is favorited can be implemented using logarithms. Given a topic tag t∈T, the number of times a Weibo post by a target user under that topic tag is favorited is Ψt(u i The total number of times a Weibo post by a target user under this hashtag is saved or liked is calculated as follows:

[0087]

[0088] The molecule represents the target user u i The sum of the total number of times Weibo posts under hashtag t were favorited is logarithmic. The denominator represents the logarithmic sum of the total number of times Weibo posts under hashtag t were favorited by the user with the most favorites.

[0089] In one exemplary instance, calculating the popularity of a target user's Weibo posts under a specific hashtag based on the number of likes, reposts, and favorites includes:

[0090] The popularity of the target user's Weibo posts under this topic tag is calculated based on the number of likes, reposts, and favorites, as well as the pre-set weights for likes, reposts, and favorites.

[0091] For example, the weights for likes, reposts, and favorites can each be set to 1. The target user's likes, reposts, and favorites are calculated based on the number of likes, reposts, and favorites of their Weibo posts under this hashtag, along with the pre-set weights for likes, reposts, and favorites. i Popularity of Weibo posts under this hashtag Θ t (u i The calculation method is as follows:

[0092]

[0093] L t (u i R represents the number of likes received by Weibo posts published by the target user under this hashtag. t (u i Ψ represents the number of times a Weibo post by a target user under a specific hashtag is retweeted. t (u i This refers to the number of times a Weibo post by a target user under a specific hashtag is saved or collected.

[0094] In one exemplary instance, obtaining the target user's influence information under the topic tag based on the target user's activity level, the attention level of their posted Weibo posts, the mention level of their posted Weibo posts, and the popularity of their posted Weibo posts includes:

[0095] First, the target user's influence information under the topic tag is calculated based on the target user's activity level, the attention level of the Weibo posts, the mention level of the Weibo posts, the popularity level of the Weibo posts, and the pre-set activity level weight, attention level weight, mention level weight, and popularity level weight.

[0096] For example, the activity weight, attention weight, mention weight, and popularity weight can each be set to 1. Based on the target user's activity level under this topic tag, the attention level of their posted Weibo posts, the mention rate of their posted Weibo posts, the popularity of their posted Weibo posts, and the pre-set activity weight, attention weight, mention weight, and popularity weight, the target user u is calculated. i Influence information under this hashtag I t (u i The calculation formula is:

[0097]

[0098] The embodiment of the present application also provides a user credibility evaluation method, the overall framework of which is shown in the figure, and the method comprises the following steps: Figure 2

[0099] Step 1, calculating the historical attitude of the user according to the content of the latest 100 blog posts of the user. Specifically, the embodiment of the step comprises two specific sub-steps:

[0100] Sub-step 1-1, calculating the attitude attribute of the historical microblog published by the user in the microblog. Since positive words and negative words may appear in a microblog, first, the number of positive words C + (W i ) and the number of negative words C - (W i ) in a microblog are counted. Then, the attitude attribute of the microblog is expressed according to the following formula:

[0101]

[0102] Wherein C + (W i ) represents the number of positive words in the microblog W i , C - (W i ) represents the number of negative words in the microblog W i , W + represents positive microblog, W - represents negative microblog, and W 0 represents neutral microblog.

[0103] Sub-step 1-2, calculating the historical attitude of the user according to the attitude attribute of the historical microblog of the user. For the user u i ∈U, the historical attitude of the user Δu i can be obtained by analyzing the latest 100 historical microblogs of the user, and is expressed as:

[0104]

[0105] Wherein represents the number of positive microblogs published by the user u i , represents the number of neutral microblogs published by the user u i , represents the number of negative microblogs published by the user u i .

[0106] Step 2, calculating the activity degree of the user. Specifically, the implementation process of the step is as follows: ​

[0107] The user activity under a topic is calculated by collecting the number of microblogs published by the user under a certain topic label. Assume that the set of all users participating in a topic and publishing a number of microblogs greater than or equal to 1 under a topic label t∈T is U, and the set of all published microblogs under the topic is W t , the set of microblogs published by the user u i ∈U under the topic t is The calculation method of the activity of the user u i under the topic t is as follows:

[0108]

[0109] Where A t (u i ) represents the activity of the user.

[0110] Step 3, calculate the user attention, user mentioned and user popularity, and then calculate the user influence using the calculated results. Specifically, the implementation process of this step is divided into three sub-steps:

[0111] Sub-step 3-1, calculate the user attention. In a given topic t∈T, the attention of the user F t (u i ) is calculated according to the number of fans num follow (u i ) of the user, and the calculation method is as follows:

[0112]

[0113] The numerator represents the logarithmic value of the number of fans of the user u i , and the denominator represents the logarithmic value of the number of fans of the user with the most fans in the topic t.

[0114] Sub-step 3-2, calculate the user mentioned. In a given topic t∈T, the mentioned degree M t (u i ) of the user is calculated according to the number of @ num mention (u i ) of the user under the topic, and the calculation method is as follows:

[0115] M t (u i )=log(num mention (u i )) / max(log(num mention (U,t)))

[0116] The numerator represents the number of fans of the user u iThe logarithm value of the number of likes of the microblog of the user u in the topic t, and the denominator represents the logarithm value of the number of likes of the microblog of the user with the most likes in the topic t.

[0117] Sub-step 3-3, calculating the user popularity. The user popularity is composed of three parts, i.e., the user microblog like degree, the user microblog forwarding degree and the user microblog collection degree. The user microblog like degree, the user microblog forwarding degree and the user microblog collection degree are calculated according to the number of likes, the number of forwarding and the number of collection of the microblog of the user respectively.

[0118] In a given topic t∈T, the user microblog like degree L t (u i ) is calculated according to the total number of likes of the microblog of the user in the topic, and the calculation method is as follows:

[0119]

[0120] The numerator represents the logarithm value of the total number of likes of the microblog of the user u i in the topic t, and the denominator represents the logarithm value of the total number of likes of the microblog of the user with the most likes in the topic t.

[0121] In a given topic t∈T, the user microblog forwarding degree R t (u i ) is calculated according to the total number of forwarding of the microblog of the user in the topic, and the calculation method is as follows:

[0122]

[0123] The numerator represents the logarithm value of the total number of forwarding of the microblog of the user u i in the topic t, and the denominator represents the logarithm value of the total number of forwarding of the microblog of the user with the most forwarding in the topic t.

[0124] In a given topic t∈T, the user microblog collection degree Ψ t (u i ) is calculated according to the total number of collection of the microblog of the user in the topic, and the calculation method is as follows:

[0125]

[0126] The numerator represents the logarithm value of the total number of collection of the microblog of the user u i in the topic t, and the denominator represents the logarithm value of the total number of forwarding of the microblog of the user with the most forwarding in the topic t.

[0127] The user microblog like degree, the user microblog forwarding degree and the user microblog collection degree can be calculated by the above formula, and the user popularity Θ can be obtained by combining the three.t (u i

[0128]

[0129] The user influence can be obtained according to the calculated user activity, user attention, user mention and user popularity, and the calculation formula is as follows:

[0130]

[0131] wherein I t (u i ) represents the influence of the user u i .

[0132] Step 4, calculating the user credibility. Specifically, the implementation process is as follows:

[0133] The user credibility is calculated by combining the user historical attitude and the user influence, and the calculation method is as follows:

[0134] Υ t (u i ) = Δu i × I t (u i )

[0135] wherein Υ t (u i ) represents the user credibility of the user u i , Δu i represents the user historical attitude, and I t (u i ) represents the user influence.

[0136] The embodiment of the application further provides a user credibility evaluation device, comprising a memory and a processor, the memory is used for saving an executable program;

[0137] The processor is used for reading and executing the executable program, so as to realize the user credibility evaluation method as described in any one of the above embodiments.

[0138] The user credibility evaluation device provided by the embodiment of the application first acquires the attitude evaluation information of a target user according to multiple micro blogs published by the target user, then classifies the multiple micro blogs according to topic labels, comprehensively measures the influence information of the target user under each topic label from the activity, attention, mention and popularity, and finally evaluates the credibility of the target user under each topic label based on the influence information under each topic label and the attitude evaluation information of the target user. Therefore, the evaluation of the user credibility is realized comprehensively and accurately, and the identification of the false user is realized ​

[0139] Those of ordinary skill in the art will realize and understand that all or some of the steps in the methods disclosed above and the functional modules / units in the systems and devices can be implemented as software, firmware, hardware, and appropriate combinations thereof. In hardware implementation, the division between the functional modules / units mentioned in the above description does not necessarily correspond to the division of physical components; for example, one physical component can have multiple functions, or one function or step can be performed by several physical components in cooperation. Some or all of the components can be implemented as software executed by a processor, such as a digital signal processor or a microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software can be distributed on computer-readable media, which can include computer storage media (or non-transitory media) and communication media (or transitory media). As is well known to those of ordinary skill in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can be accessed by a computer. Furthermore, it is common and well understood by those of ordinary skill in the art that communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and can include any information delivery media.

Claims

1. A user trustworthiness assessment method, characterized by, Comprise: According to the target user published a plurality of microblog obtain the target user's attitude evaluation information; wherein, the target user is to be carried out credibility evaluation user, the attitude evaluation information is used to evaluate the positive attitude degree of the target user publishing a plurality of microblog; Obtain the topic label involved in the plurality of microblog published by the target user, and each topic label obtained is operated as follows: respectively obtaining the activity of the target user under the topic label, obtaining the attention degree of the microblog published by the target user under the topic label, obtaining the mentioned degree of the microblog published by the target user under the topic label, obtaining the popularity of the microblog published by the target user under the topic label; According to the activity of the target user under the topic label, the attention degree of the microblog published, the mentioned degree of the microblog published, the popularity of the microblog published, the influence information of the target user under the topic label is obtained; According to the influence information of the target user under each topic label and the attitude evaluation information of the target user, the credibility of the target user under each topic label is obtained; Wherein, the attitude evaluation information of the target user is obtained according to the analysis of the attitude attribute of the plurality of microblog of the target user, combined with the weight of each type of attitude preset; The activity of the target user under the topic label is determined according to the ratio of the number of microblog published by the target user under the topic label to the number of all microblog published under the topic label; The attention degree of the target user under the topic label is determined according to the ratio of the logarithmic value of the number of fans of the target user to the logarithmic value of the number of fans of the user with the most number of fans under the topic label; The mentioned degree of the target user under the topic label is determined according to the ratio of the logarithmic value of the number of mentions of the microblog published by the target user under the topic label to the logarithmic value of the number of mentions of the microblog published by the user with the most number of mentions under the topic label; The popularity of the target user under the topic label is obtained by comprehensively calculating the like degree, the forwarding degree and the collection degree of the microblog published by the target user under the topic label; The like degree is determined according to the ratio of the logarithmic value of the total number of likes of the microblog published by the target user under the topic label to the logarithmic value of the total number of likes of the microblog published by the user with the most total number of likes under the topic label; The forwarding degree is determined according to the ratio of the logarithmic value of the total number of forwarding of the microblog published by the target user under the topic label to the logarithmic value of the total number of forwarding of the microblog published by the user with the most total number of forwarding under the topic label; The collection degree is determined according to the ratio of the logarithmic value of the total number of collections of the microblog published by the target user under the topic label to the logarithmic value of the total number of collections of the microblog published by the user with the most total number of collections under the topic label.

2. The method of claim 1, wherein, The target user's attitude evaluation information is obtained according to the analysis of the attitude attribute of the plurality of microblog of the target user, combined with the weight of each type of attitude preset; For each of the multiple microblogs published by the target user, the following operations are performed: the frequency of positive words and the frequency of negative words in the microblog are counted using a pre-set set of positive words and a set of negative words, and the attitude information of the microblog is determined based on the frequency of positive words and the frequency of negative words in the microblog. The target user's attitude evaluation information is calculated based on the attitude information of all the target user's Weibo posts and the pre-set attitude weights corresponding to each type of attitude information.

3. The method of claim 1, wherein, The process of obtaining the target user's activity level under the topic tag includes: Get the number of Weibo posts published by the target user under this topic tag, and get the total number of Weibo posts published under this topic tag; The activity level of the target user under this topic tag is calculated based on the number of Weibo posts published by the target user under this topic tag and the total number of Weibo posts published under this topic tag.

4. The method of claim 1, wherein, The process of obtaining the attention level of the Weibo posts published by the target user under this topic tag includes: Obtain the number of followers of the target user, and obtain the number of followers of the user with the most followers under this topic tag; The level of attention received by the target user's Weibo posts under this hashtag is calculated based on the number of followers of the target user and the number of followers of the user with the most followers under this hashtag.

5. The method of claim 4, wherein, The process of obtaining the mention rate of the Weibo posts published by the target user under this topic tag includes: Get the number of times the target user's Weibo posts under this topic tag are mentioned, and get the number of times the Weibo posts of the user with the most mentions under this topic tag are mentioned; The mention rate of the target user's posts under this hashtag is calculated based on the number of times the target user's posts are mentioned and the number of times the posts by the user who has been mentioned the most among the posts under this hashtag.

6. The method of claim 5, wherein, The process of obtaining the popularity of the Weibo posts published by the target user under this topic tag includes: Get the number of likes on the Weibo posts published by the target user under the topic tag, get the number of reposts on the Weibo posts published by the target user under the topic tag, and get the number of favorites on the Weibo posts published by the target user under the topic tag. The popularity of the target user's Weibo posts under this hashtag is calculated based on the number of likes, reposts, and favorites.

7. The method of claim 6, wherein, The process of obtaining the number of likes received by the target user's Weibo posts under this topic tag includes: Get the total number of likes on the Weibo posts published by the target user under this topic tag, and get the total number of likes on the Weibo posts published by the user with the most likes under this topic tag. The likes score of the target user's Weibo posts under this topic tag is calculated based on the total number of likes received by the target user's Weibo posts under this topic tag, and the total number of likes received by the user with the most likes among all Weibo posts under this topic tag. The obtaining of the forwarding degree of the microblog published by the target user under the topic label comprises: obtaining the total number of times of forwarding of the microblog published by the target user under the topic label, and the total number of times of forwarding of the microblog published by the user with the maximum total number of times of forwarding under the topic label; calculating the forwarding degree of the microblog published by the target user under the topic label according to the total number of times of forwarding of the microblog published by the target user under the topic label and the total number of times of forwarding of the microblog published by the user with the maximum total number of times of forwarding under the topic label; The obtaining of the collection degree of the microblog published by the target user under the topic label comprises: obtaining the total number of times of collection of the microblog published by the target user under the topic label, and the total number of times of collection of the microblog published by the user with the maximum total number of times of collection under the topic label; calculating the collection degree of the microblog published by the target user under the topic label according to the total number of times of collection of the microblog published by the target user under the topic label and the total number of times of collection of the microblog published by the user with the maximum total number of times of collection under the topic label.

8. The method of claim 6, wherein, The calculation of the popularity of the microblog published by the target user under the topic label according to the like degree, the forwarding degree and the collection degree of the microblog published by the target user under the topic label comprises: calculating the popularity of the microblog published by the target user under the topic label according to the like degree, the forwarding degree and the collection degree of the microblog published by the target user under the topic label, and the pre-set like degree weight, forwarding degree weight and collection degree weight; The obtaining of the influence information of the target user under the topic label according to the activity degree, the attention degree, the mentioned degree and the popularity of the microblog published by the target user under the topic label comprises: calculating the influence information of the target user under the topic label according to the activity degree, the attention degree, the mentioned degree and the popularity of the microblog published by the target user under the topic label, and the pre-set activity degree weight, attention degree weight, mentioned degree weight and popularity weight.

9. The method of claim 6, wherein, The calculation process of the attention degree, the mentioned degree, the like degree, the forwarding degree and the collection degree adopts logarithm.

10. A user trustworthiness evaluation apparatus characterized by comprising: comprise: a memory and a processor, the memory is used to save an executable program; the processor is used to read and execute the executable program to realize the user credibility evaluation method in any one of claims 1-9.

Images