A method for identifying illegal network communities and related equipment

Abstract

The embodiment of the present application provides a method for identifying illegal network communities and related devices for safeguarding the rights and interests of live broadcast platforms. The method includes: determining the initial suspicion level of a first community; generating a first comment vector and a second comment vector based on the comment texts published by a first user and a second user in the first community; obtaining a first synergy coefficient of the first user and the second user in the first community; obtaining the total synergy effect of the first user and the second user based on the first synergy coefficient, and obtaining a second synergy coefficient of the first user and the second user in the community set based on the total synergy effect; determining a similar user set and a non-similar user set based on the second synergy coefficient; constructing an optimization function to determine the vector representation of users in the first community; obtaining the user closeness of the first community based on the user's vector representation; when the user closeness is less than a first threshold, determining the first community as an illegal network community.

Description

Technical Field

[0001] The present application relates to the field of information processing technology, and in particular to a method for identifying illegal network communities and related equipment. Background Art

[0002] On the live broadcast platform, the popularity of the live broadcast room reflects the influence of the anchor in the live broadcast room, which can affect the display position of the live broadcast room, that is, the degree to which resources can be obtained. Therefore, in order to obtain better resources, the anchor will use various methods to increase the popularity of his live broadcast room. Among them, the number of forum comments is an important way to increase the popularity of the live broadcast room.

[0003] During the development of live broadcast platforms, some black industry groups often register many platform accounts in batches to achieve private purposes, and use these accounts to post comments in live broadcast rooms, forum comments, bullet comments, and popularity, etc., in an attempt to increase the popularity of a certain live broadcast room, making the resources of the entire live broadcast platform unequal, and profit from it. For example, some anchors increase the popularity of their live broadcast rooms by posting comments on forums in Douyu Fish Bar, in order to obtain more platform resources, which directly damages the rights and interests of the live broadcast platform and other anchors.

[0004] In the prior art, the method of identifying Yuba comments usually starts from the comment text itself, that is, the text itself does not comply with the relevant regulations of the forum, such as the text about violence or politics, etc. However, when Yuba comments are used to increase popularity, the text itself is likely to be normal. Therefore, in order to better protect the rights and interests of the live broadcast platform, how to identify illegal Yuba that is suspected of brushing forum comments from a behavioral perspective is currently a top priority. Summary of the invention

[0005] The embodiments of the present application provide a method for identifying illegal online communities and related devices, which can identify illegal communities to protect the rights and interests of live broadcast platforms.

[0006] A first aspect of an embodiment of the present application provides a method for recommending a live broadcast room, including: determining an initial suspicion level of a first community, where the first community is any community in a community set; generating a first comment vector based on a comment text posted by a first user in the first community, and generating a second comment vector based on a comment text posted by a second user in the first community, where the first user and the second user are any users who have posted comments in the first community; obtaining a first synergy coefficient of the first user and the second user in the first community based on the first comment vector, the second comment vector, and the initial suspicion level; obtaining a total synergy effect of the first user and the second user based on the first synergy coefficient, and obtaining a second synergy coefficient of the first user and the second user in the community set based on the total synergy effect; determining a similar user set and a dissimilar user set based on the second synergy coefficient; constructing an optimization function based on the similar user set and the dissimilar user set to determine a vector representation of users in the first community; obtaining a user closeness of the first community based on the vector representation of the users; and determining that the first community is an illegal network community when the user closeness is less than a first threshold.

[0007] Optionally, determining the initial suspicion level of the first community includes: obtaining the initial suspicion level by the following formula: Among them, the s p is used to represent the initial suspicion level, p is used to represent the first community, r(p) is used to represent the number of people who post in the first community, θ is a normalization constant between 0 and 1, P represents the community set, It is used to represent the maximum number of posters in a single community in the community set P, and q is used to represent the community corresponding to the maximum number of posters.

[0008] Optionally, obtaining a first synergy coefficient of the first user i and the second user j in the first community according to the first comment vector, the second comment vector and the initial suspicion level includes: when a time interval between a posting time of the first user in the first community and a posting time of the second user in the first community is less than a minimum time interval, obtaining the first synergy coefficient according to the following formula: Wherein, col(i,j,p) is used to represent the first collaborative parameter, i is used to represent the first user, j is used to represent the second user, α and β are weight coefficients greater than 0, and satisfy α+β=1, α<β, and is used to indicate the time when the first user posted in the first community, is used to indicate the posting time of the second user in the first community, the τ tIt is used to indicate the minimum time interval. is used to represent the first comment vector, the is used to represent the second review vector.

[0009] Optionally, obtaining a total synergy effect of the first user and the second user according to the first synergy coefficient, and obtaining a second synergy coefficient of the first user and the second user in the community set according to the total synergy effect includes: obtaining the second synergy coefficient by the following formula:

[0010] Wherein, σ(i,j) is used to represent the total synergistic effect, P i The community set P is used to represent the community set where the first user posts comments. j It is used to represent the community set where the second user posts comments. i ∩P j | is used to indicate the number of communities where the first user and the second user jointly post comments, |P i ∪P j | is used to represent the number of communities where the first user and the second user post comments, P i ∩P j The community set used to represent the comments jointly posted by the first user and the second user, wherein p is the P i ∩P j For any community in , the φ(i,j) is used to represent the second synergy coefficient.

[0011] Optionally, determining a similar user set and a dissimilar user set based on the second collaboration coefficient includes: when the second collaboration coefficient of the first user and the second user is greater than a second threshold, determining that the first user and the second user are a similar user pair to obtain the similar user set; when the second collaboration coefficient of the first user and the second user is less than a third threshold, determining that the first user and the second user are a dissimilar user pair to obtain the dissimilar user set.

[0012] Optionally, constructing an optimization function according to the similar user set and the non-similar user set includes: Among them, the D p Used to represent the similar user set, D n is used to represent the non-similar user set, ν c , ν a is the vector representation of the user pairs in the set, ν c ' is used to represent ν c The transpose of .

[0013] Optionally, obtaining the user closeness of the first community according to the vector representation of the user includes: Wherein, spam(p) is used to represent the user closeness, R(p) is used to represent the set of users who have posted comments in the first community, i is any user in R(p), and |R(p)| is used to represent the number of people who have posted in the first community. is the vector representation of the user i.

[0014] According to a second aspect of an embodiment of the present application, there is provided an identification device, comprising: a calculation unit, configured to determine an initial suspicion level of a first community, wherein the first community is any community in a community set; a generation unit, configured to generate a first comment vector according to a comment text posted by a first user i in the first community, and to generate a second comment vector according to a comment text posted by a second user j in the first community, wherein the first user i and the second user j are any users who have posted comments in the first community; the calculation unit is further configured to obtain a first synergy coefficient of the first user i and the second user j in the first community according to the first comment vector, the second comment vector and the initial suspicion level; the calculation unit is further configured to obtain a total synergy effect of the first user and the second user according to the first synergy coefficient, and obtain a second synergy coefficient of the first user and the second user in the community set according to the total synergy effect; a determination unit, configured to determine a similar user set and a dissimilar user set according to the second synergy coefficient; the determination unit is further configured to construct an optimization function according to the similar user set and the dissimilar user set to determine a vector representation of users in the first community; the calculation unit is further configured to obtain a user closeness level of the first community according to the vector representation of the users; the determination unit is further configured to determine that the first community is an illegal network community when the user closeness level is less than a first threshold.

[0015] In summary, it can be seen that in the embodiment provided by the present application, the initial suspicion level of the first community is determined, and the first community is any community in the community set; a first comment vector is generated according to the comment text published by the first user in the first community, and a second comment vector is generated according to the comment text published by the second user in the first community, and the first user and the second user are any users who have posted comments in the first community; a first synergy coefficient of the first user and the second user in the first community is obtained according to the first comment vector, the second comment vector and the initial suspicion level; the total synergy effect of the first user and the second user is obtained according to the first synergy coefficient, and the second synergy coefficient of the first user and the second user in the community set is obtained according to the total synergy effect; a similar user set and a non-similar user set are determined according to the second synergy coefficient; an optimization function is constructed according to the similar user set and the non-similar user set to determine the vector representation of the users in the first community; the user closeness of the first community is obtained according to the vector representation of the users; when the user closeness is less than the first threshold, the first community is determined to be an illegal network community. Through the embodiment of the present application, illegal communities can be accurately identified to safeguard the rights and interests of live broadcast platforms. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A flowchart of a possible method for identifying illegal network communities provided in an embodiment of the present application;

[0017] Figure 2 A schematic diagram of a possible identification device provided in an embodiment of the present application;

[0018] Figure 3 A schematic diagram of the hardware structure of a possible identification device provided in an embodiment of the present application;

[0019] Figure 4 A schematic diagram of a possible electronic device provided in an embodiment of the present application;

[0020] Figure 5 A schematic diagram of an embodiment of a possible computer-readable storage medium provided for an embodiment of the present application. DETAILED DESCRIPTION

[0021] The embodiments of the present application provide a method for identifying illegal online communities and related devices, which can identify illegal communities to protect the rights and interests of live broadcast platforms.

[0022] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices. The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments.

[0023] The recommendation method for the live broadcast room is described below from the perspective of a recommendation device for the live broadcast room. The recommendation device for the live broadcast room may be a server or a service unit in the server.

[0024] See also Figure 1 , which is a flow chart of a possible method for identifying illegal network communities provided in an embodiment of the present application, specifically comprising:

[0025] 101. Determine the initial level of suspicion of the first community;

[0026] In order to facilitate a better understanding of the embodiments of the present application, the present application will take Douyu's Fish Bar as an example to illustrate the application of the present application in actual scenarios. It should be noted that the present application can also be applied to other scenarios, such as Weibo Square, Baidu Tieba, etc., which are not specifically limited here. When it is necessary to identify whether any community in the community set, such as the first community, is maliciously brushing forum comments to obtain more live broadcast resources, such as identifying whether a certain fish bar has the behavior of brushing comments, it is necessary to first determine the initial suspicion of the fish bar. Among them, the initial suspicion of the fish bar is estimated based on the number of people posting in the fish bar. In order to increase the popularity, multiple users must be used to post. Therefore, the more people post, the greater the initial suspicion in theory. In this embodiment, the number of posters in the fish bar is compared with the maximum number of posters in all fish bars. The smaller the difference, the greater the initial suspicion. Specifically, the initial suspicion of the first community, i.e., the fish bar, can be obtained by the following formula:

[0027]

[0028] Among them, s p is used to indicate the initial suspicion level, p is used to indicate the first community, r(p) is used to indicate the number of people posting in the first community, and P indicates the community set. It is used to represent the maximum number of postings in a single community in the community set P, and q is used to represent the community corresponding to the maximum number of postings. θ is a normalized constant between 0 and 1, and in practice it can be taken as 0.4. The factors that affect the value of this constant are: if you want the initial suspicion to be more discriminatory and have a greater impact in the subsequent first synergy coefficient, then increase the value of θ; otherwise, decrease it.

[0029] 102. Generate a first comment vector according to a comment text posted by a first user in the first community, and generate a second comment vector according to a comment text posted by a second user in the first community;

[0030] Randomly select users who post comments in the first community, i.e., the fish bar, including the first user and the second user, and generate a first comment vector from the comment text posted by the first user in the fish bar, and generate a second comment vector from the comment text posted by the second user in the fish bar. Among them, the method of generating comment vectors from comment texts can be to generate word vectors for the comment texts using methods such as word2vecc or glove, which are not specifically limited here, and then add the word vectors in the comment texts and calculate the average value to obtain the comment vector. It should be noted that the method of generating word vectors from comment texts is a common method in the prior art, and this application will not be specifically summarized.

[0031] 103. Obtaining a first synergy coefficient between the first user and the second user in the first community according to the first comment vector, the second comment vector, and the initial suspicion level;

[0032] After obtaining the first comment vector and the second comment vector, the first synergy coefficient of the first user and the second user in the fish bar is obtained. Assuming that the first user posted in the fish bar at The first comment vector generated by the published comment text is The second user posted on the fish bar at The first comment vector generated by the published comment text is And set the minimum time interval τ t ,

[0033] When the first user posts in the first community The time when the second user posted in the first community The time interval is no greater than the minimum time interval τ t ,Right now When , the first synergy coefficient is obtained according to the following formula:

[0034]

[0035] Where col(i,j,p) is used to represent the first coordination parameter, i is used to represent the first user, j is used to represent the second user, s p It is used to indicate the initial suspicion level of the first community, i.e., the fish bar. α and β are weight coefficients greater than 0, and satisfy α+β=1, α<β. The factors affecting the two weight coefficients α and β are: if the users identified in recent business as posting large amounts of comments in a short period of time, then α will be increased; if the contents of the comments identified in recent business are similar, then α will be reduced.

[0036] when , col(i,j,p)=0.

[0037] It is understandable that in practical applications, it can be considered that only two users’ behaviors in the same fish bar occur at a time less than the minimum time interval τ t Synergistic effect is considered to exist only when t is a constant, with a value greater than 0, affecting the minimum time interval τ t The factors for determining the value are: If the accuracy of recognition needs to be guaranteed, then τ t To set it smaller, if more users need to be identified, τ t Set it bigger.

[0038] The principle of the first synergy coefficient formula is: the similarity of two users in the same fish bar can be measured from two perspectives. One perspective is that users who use scripts to increase traffic often operate in a short period of time. Therefore, if the time interval between the two user behaviors is The shorter the time interval, the higher the synergy in terms of behavioral synchronization. In order to make the synergy similarity value of this angle between 0 and 1, the transformation is performed based on the time interval. Another perspective is the synchronization of the content in the Fish Bar. Users who post comments usually use some scripts to generate comments, so the comments will have a large similarity in format and word composition. Therefore, the cosine distance is directly used. The value is also between 0 and 1. The weighted coefficient is then used to sum the collaborative similarities of the two angles to obtain the final collaborative similarity. In order to further reduce the false recognition rate, the collaborative similarity needs to be multiplied by the initial suspicion of the fish bar when obtaining the final first collaborative coefficient. Since the greater the initial suspicion of the fish bar, the more suspicious the collaborative behavior between the two users is, so this can effectively avoid user misidentification caused by behavioral coincidence.

[0039] 104. Obtain a total synergy effect of the first user and the second user according to the first synergy coefficient, and obtain a second synergy coefficient of the first user and the second user in the community set according to the total synergy effect;

[0040] After obtaining the first synergy coefficient, the total synergy effect of the first user and the second user is obtained according to the first synergy coefficient, and then the second synergy coefficient of the first user and the second user in the community set is obtained according to the total synergy effect. Specifically, it can be obtained by the following formula:

[0041]

[0042]

[0043] Among them, σ(i,j) is used to represent the total synergy effect, φ(i,j) is used to represent the second synergy coefficient, P i It is used to represent the community set where the first user posts a comment, P j Used to represent the community collection where the second user posts a comment, |P i ∩P j | is used to indicate the number of communities where the first user and the second user jointly post comments, |P i ∪P j | is used to represent the number of communities where the first user and the second user post comments, P i ∩P j It is used to represent the community set where the first user and the second user jointly post comments. p is P i ∩P j Any community in .

[0044] The above formula for calculating the second synergy coefficient by calculating the total synergy effect can be: The total synergy effect σ(i,j) is measured by two parts, one of which is It represents the ratio of the number of fish bars where the first user i and the second user j post together to the number of their union. The larger the ratio, the higher the ratio of posts posted by the two users, and the stronger the behavioral coordination of the two users. The other part is It is the synergy coefficient of all the posts of the first user and the second user. The larger the value, the higher the degree of synchronization of the two users in certain specific fish bars. The product form is used because the dimensions of these two types of influence are inconsistent and cannot be directly added.

[0045] The second synergy coefficient φ(i,j) is transformed based on the total synergy effect σ(i,j). When σ(i,j) is 0, the second synergy coefficient is also 0. When the total synergy effect σ(i,j) increases, the second synergy coefficient will continue to increase and approach 1. It is easy to find that the value of the second synergy coefficient is between 0 and 1. The reason why the exponential form is used in the transformation is that the influence of the total synergy effect on the second synergy coefficient is nonlinear. When the total synergy effect is relatively small, its increase will cause a more significant increase in the second synergy coefficient. When the total synergy effect reaches a certain level, the effect of further increase on the second synergy coefficient is also negligible.

[0046] In order to better understand how to obtain the second synergy coefficient in the embodiment of the present application, an example is given below:

[0047] Assume that users A and B post in Fish Bar a as follows:

[0048] User A posted at 10:14 on 20200101 [The content of the anchor’s live broadcast today is very exciting]

[0049] User B posted at 10:15 on 20200101 [The anchor’s live broadcast today was very good]

[0050] Take the minimum time interval τ t And the weight coefficients α and β are: τ t =5,α=0.5,β=0.5;

[0051] In addition, the number of people posting in Fish Bar a is 100, and the maximum number of people posting is 500, so the first synergy coefficient of two users in the specific Fish Bar a is obtained:

[0052]

[0053] |P i ∩P j |=1,|P i ∪P j |=2;

[0054] The total synergistic effect can be obtained:

[0055] Then the second synergy coefficient is obtained as:

[0056]

[0057] 105. Determine a similar user set and a dissimilar user set according to the second coordination coefficient;

[0058] After determining the second synergy coefficient of the first user and the second user in all fish bars, determine the similar user set and the dissimilar user set in the first community according to the second synergy coefficient. Specifically, when the second synergy coefficient of the first user and the second user is greater than the second threshold, the first user and the second user are determined to be a similar user pair to obtain the similar user set; when the second synergy coefficient of the first user and the second user is less than the third threshold, the first user and the second user are determined to be a dissimilar user pair to obtain the dissimilar user set. That is, for the similar user set, take the similar user pairs whose second synergy coefficient is greater than the second threshold ω; for the dissimilar user set, take the similar user pairs whose synergy coefficient is less than the third threshold τ. The factors affecting the value of the threshold are: if it is necessary to ensure that the user vectors of obvious brushing behavior are as close as possible, then the second threshold ω can be increased and the third threshold τ can be decreased; if some suspected users who are not very obvious and users with obvious brushing behavior can be made close, then the second threshold ω can be decreased and the third threshold τ can be increased.

[0059] 106. Construct an optimization function based on the similar user set and the dissimilar user set to determine a vector representation of users in the first community;

[0060] After distinguishing the similar user set and the dissimilar user set by the second synergy coefficient, an optimization function is constructed according to the similar user set and the dissimilar user set to determine the vector representation of the users in the first community. Specifically, the constructed optimization function can be expressed as follows:

[0061]

[0062] Among them, D p Used to represent similar user sets, D n Used to represent a set of non-similar users, ν c , ν a is the vector representation of the user pairs in the set, ν c ' is used to represent ν c The transpose of .

[0063] The principle of constructing the above optimization function is: in order to comprehensively consider the impact of similar user sets and non-similar user sets on the vector representation of users in the first community, the constructed optimization function must include the vector representation of user pairs corresponding to the two sets. In the embodiment of the present application, the operation properties of logarithms can be used to solve the optimal representation of user vectors. Since the multiplication of two logarithms cannot be solved, the addition of two logarithms is used. In addition, for similar users, if the user vectors are learned very appropriately, the vectors between similar users will be closer, which is reflected in the optimization function as vector dot product - v' c v a The smaller the result, the The larger the value, the more appropriate the vector learning is; on the contrary, for two dissimilar user vectors, we hope The smaller the The better, so the overall optimization function is the larger the better. Therefore, based on the above principle, when the optimization function is the largest, it is the optimal representation of the user vector, that is, it is determined as the vector representation of the user in the first community.

[0064] 107. Obtain the user closeness of the first community according to the vector representation of the user;

[0065] 108. When the user closeness is less than a first threshold, determining that the first community is an illegal network community.

[0066] After the users are represented by vectors, the user closeness of the first community is obtained according to the user's vector representation, which can be obtained by the following formula:

[0067]

[0068] Among them, spam(p) is used to indicate the closeness of users, R(p) is used to indicate the set of users who have posted comments in the first community, i is any user in R(p), |R(p)| is used to indicate the number of people who have posted in the first community, is the vector representation of user i.

[0069] The principle of the above formula for obtaining user closeness is: It represents the overall center vector of the first community, i.e. the fish bar, which is the average value of all user vectors in the fish bar. It represents the distance between user i and the overall center vector of the fish bar. It represents the sum of the squares of the distances between user i and the center vector of the entire fish bar. Therefore, the actual meaning of spam(p) is the average of the sum of the squares of the distances between the users in the fish bar and the entire fish bar. If the average is smaller, it means that the users in the fish bar are closer, and the suspicion of the comments in the fish bar is higher.

[0070] Therefore, when the user closeness spam(p) is less than the first threshold, it means that the users in the first community are very close, and it is inferred that the comments in the first community are highly suspected of being spam, so the first community is determined to be an illegal network community. On the contrary, when the user closeness spam(p) is greater than the first threshold, it means that the users in the first community have a normal relationship, and the first community can be determined to be a legal network community.

[0071] In order to better understand how to determine whether the first community is an illegal network community in the embodiment of the present application, an example is given below:

[0072] Assume that there are three users A1, B1, and C1 posting in Fish Bar B, and their vector representations are:

[0073] A1=(1,2,1), B1=(2,3,1), C1=(1,4,2), then the average vector of all users in Fish Bar B is (1.33,3,1.33)

[0074] So the user closeness of Yuba B is:

[0075]

[0076] If the first threshold is set to 1, then the user closeness of Fish Bar B, 1.11, is greater than the first threshold, so Fish Bar B is considered to be an illegal online community.

[0077] In the embodiments of the present application, based on the principle that in order to achieve the purpose of posting without being discovered, the anchor must use multiple accounts to manipulate the posts. Therefore, the coordination of behaviors between user accounts in the fish bar that has not been posted is low, and vice versa. A method for identifying illegal online communities is provided, which accurately identifies illegal online communities and protects the rights and interests of the live broadcast platform.

[0078] The above describes the method for identifying illegal network communities in the embodiment of the present application. The following describes the identification device in the embodiment of the present application.

[0079] See also Figure 2 In one embodiment of the identification device in the embodiment of the present application, the recommendation device of the live broadcast room includes:

[0080] A calculation unit 201 is used to determine an initial suspicion level of a first community, where the first community is any community in the community set;

[0081] A generating unit 202 is configured to generate a first comment vector according to a comment text posted by a first user i in the first community, and to generate a second comment vector according to a comment text posted by a second user j in the first community, wherein the first user i and the second user j are any user who has posted a comment in the first community;

[0082] The calculation unit 201 is further configured to obtain a first synergy coefficient of the first user i and the second user j in the first community according to the first comment vector, the second comment vector and the initial suspicion level;

[0083] The calculation unit 201 is further configured to obtain a total synergy effect of the first user and the second user according to the first synergy coefficient, and obtain a second synergy coefficient of the first user and the second user in the community set according to the total synergy effect;

[0084] A determination unit 203, configured to determine a similar user set and a dissimilar user set according to the second cooperation coefficient;

[0085] The determining unit 203 is further configured to construct an optimization function according to the similar user set and the non-similar user set to determine a vector representation of users in the first community;

[0086] The calculation unit 201 is further used to obtain the user closeness of the first community according to the vector representation of the user;

[0087] The determining unit 203 is further configured to determine that the first community is an illegal network community when the user closeness is less than a first threshold.

[0088] Optionally, the calculation unit 201 is specifically configured to obtain the initial suspicion level by using the following formula:

[0089]

[0090] Among them, the s p is used to represent the initial suspicion level, p is used to represent the first community, r(p) is used to represent the number of people who post in the first community, θ is a normalization constant between 0 and 1, P represents the community set, It is used to represent the maximum number of posters in a single community in the community set P, and q is used to represent the community corresponding to the maximum number of posters.

[0091] Optionally, the calculation unit 201 is specifically used for:

[0092] When the time interval between the posting time of the first user in the first community and the posting time of the second user in the first community is less than the minimum time interval, the first synergy coefficient is obtained according to the following formula:

[0093]

[0094] Wherein, col(i,j,p) is used to represent the first collaborative parameter, i is used to represent the first user, j is used to represent the second user, α and β are weight coefficients greater than 0, and satisfy α+β=1, α<β, and is used to indicate the time when the first user posted in the first community, is used to indicate the posting time of the second user in the first community, the τ t It is used to indicate the minimum time interval. is used to represent the first comment vector, the is used to represent the second review vector.

[0095] Optionally, the calculation unit 201 is specifically used to obtain the second synergy coefficient by the following formula:

[0096]

[0097]

[0098] Wherein, σ(i,j) is used to represent the total synergistic effect, P i The community set P is used to represent the community set where the first user posts comments. j It is used to represent the community set where the second user posts comments. i ∩P j | is used to indicate the number of communities where the first user and the second user jointly post comments, |P i ∪P j | is used to represent the number of communities where the first user and the second user post comments, P i ∩P j The community set used to represent the comments jointly posted by the first user and the second user, wherein p is the P i ∩P j For any community in , the φ(i,j) is used to represent the second synergy coefficient.

[0099] Optionally, the determination unit 203 is specifically used to: when the second synergy coefficient of the first user and the second user is greater than a second threshold, determine that the first user and the second user are a similar user pair to obtain the similar user set; when the second synergy coefficient of the first user and the second user is less than a third threshold, determine that the first user and the second user are a dissimilar user pair to obtain the dissimilar user set.

[0100] Optionally, the determining unit 203 is specifically configured to:

[0101]

[0102] Among them, the D p Used to represent the similar user set, D n is used to represent the non-similar user set, ν c , ν a is the vector representation of the user pairs in the set, ν c ' is used to represent ν c The transpose of .

[0103] Optionally, the calculation unit 201 is specifically configured to obtain the user closeness by:

[0104]

[0105] Wherein, spam(p) is used to represent the user closeness, R(p) is used to represent the set of users who have posted comments in the first community, i is any user in R(p), and |R(p)| is used to represent the number of people who have posted in the first community. is the vector representation of the user i.

[0106] above Figure 2 The recognition device in the embodiment of the present application is described from the perspective of modular functional entities. The recognition device in the embodiment of the present application is described in detail from the perspective of hardware processing. Figure 3 , an embodiment of the recommendation device 300 of the live broadcast room in the embodiment of the present application includes:

[0107] An input device 301, an output device 302, a processor 303 and a memory 304 (the number of the processor 303 can be one or more, Figure 3 In some embodiments of the present application, the input device 301, the output device 502, the processor 303 and the memory 304 may be connected via a bus or other means, wherein: Figure 3 The example of connecting through bus is taken in the following.

[0108] Wherein, by calling the operation instruction stored in the memory 304, the processor 303 is used to perform the following steps:

[0109] Determining an initial suspicion level of a first community, where the first community is any community in the community set;

[0110] Generate a first comment vector according to a comment text posted by a first user in the first community, and generate a second comment vector according to a comment text posted by a second user in the first community, wherein the first user and the second user are any user who posted a comment in the first community;

[0111] Obtaining a first synergy coefficient between the first user and the second user in the first community according to the first comment vector, the second comment vector, and the initial suspicion level;

[0112] Obtaining a total synergy effect between the first user and the second user according to the first synergy coefficient, and obtaining a second synergy coefficient between the first user and the second user in the community set according to the total synergy effect;

[0113] Determine a similar user set and a dissimilar user set according to the second cooperation coefficient;

[0114] constructing an optimization function according to the similar user set and the non-similar user set to determine vector representations of users in the first community;

[0115] Obtaining the user closeness of the first community according to the vector representation of the user;

[0116] When the user closeness is less than a first threshold, the first community is determined to be an illegal network community.

[0117] By calling the operation instructions stored in the memory 304, the processor 303 is also used to execute Figure 1 Any method in the corresponding embodiment.

[0118] See also Figure 4 , Figure 4 A schematic diagram of an embodiment of an electronic device provided in an embodiment of the present application.

[0119] like Figure 4 As shown, an embodiment of the present application provides an electronic device, including a memory 410, a processor 420, and a computer program 411 stored in the memory 420 and executable on the processor 420. When the processor 420 executes the computer program 411, the following steps are implemented:

[0120] Determining an initial suspicion level of a first community, where the first community is any community in the community set;

[0121] Generate a first comment vector according to a comment text posted by a first user in the first community, and generate a second comment vector according to a comment text posted by a second user in the first community, wherein the first user and the second user are any user who posted a comment in the first community;

[0122] Obtaining a first synergy coefficient between the first user and the second user in the first community according to the first comment vector, the second comment vector, and the initial suspicion level;

[0123] Obtaining a total synergy effect between the first user and the second user according to the first synergy coefficient, and obtaining a second synergy coefficient between the first user and the second user in the community set according to the total synergy effect;

[0124] Determine a similar user set and a dissimilar user set according to the second cooperation coefficient;

[0125] constructing an optimization function according to the similar user set and the non-similar user set to determine vector representations of users in the first community;

[0126] Obtaining the user closeness of the first community according to the vector representation of the user;

[0127] When the user closeness is less than a first threshold, the first community is determined to be an illegal network community.

[0128] In the specific implementation process, when the processor 420 executes the computer program 411, it can achieve Figure 1 Any implementation manner in the corresponding embodiments.

[0129] Since the electronic device introduced in this embodiment is a device used to implement an identification device in the embodiment of the present application, based on the method introduced in the embodiment of the present application, technical personnel in this field can understand the specific implementation of the electronic device of this embodiment and its various variations. Therefore, how the electronic device implements the method in the embodiment of the present application is not introduced in detail here. As long as the equipment used by technical personnel in this field to implement the method in the embodiment of the present application is within the scope of protection of this application.

[0130] See also Figure 5 , Figure 5 A schematic diagram of an embodiment of a computer-readable storage medium provided in an embodiment of the present application.

[0131] like Figure 5 As shown, this embodiment provides a computer-readable storage medium 500, on which a computer program 511 is stored. When the computer program 511 is executed by a processor, the following steps are implemented:

[0132] Determining an initial suspicion level of a first community, where the first community is any community in the community set;

[0133] Generate a first comment vector according to a comment text posted by a first user in the first community, and generate a second comment vector according to a comment text posted by a second user in the first community, wherein the first user and the second user are any user who posted a comment in the first community;

[0134] Obtaining a first synergy coefficient between the first user and the second user in the first community according to the first comment vector, the second comment vector, and the initial suspicion level;

[0135] Obtaining a total synergy effect between the first user and the second user according to the first synergy coefficient, and obtaining a second synergy coefficient between the first user and the second user in the community set according to the total synergy effect;

[0136] Determine a similar user set and a dissimilar user set according to the second cooperation coefficient;

[0137] constructing an optimization function according to the similar user set and the non-similar user set to determine vector representations of users in the first community;

[0138] Obtaining the user closeness of the first community according to the vector representation of the user;

[0139] When the user closeness is less than a first threshold, the first community is determined to be an illegal network community.

[0140] In the specific implementation process, when the computer program 511 is executed by the processor, it can achieve Figure 1 Any implementation manner in the corresponding embodiments.

[0141] It should be noted that in the above embodiments, the description of each embodiment has its own emphasis, and for parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0142] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0143] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0144] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0145] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0146] The present application also provides a computer program product, which includes computer software instructions. When the computer software instructions are executed on a processing device, the processing device is caused to execute the following Figure 1 The process in the method for designing a digital platform for a wind farm in the corresponding embodiment.

[0147] The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from a website site, a computer, a server, or a data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (digital subscriber line, DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, server, or data center. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a server or a data center that includes one or more available media integrated. The available medium may be a magnetic medium, (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid state disk (SSD)), etc.

[0148] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0149] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0150] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0151] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0152] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), disk or optical disk and other media that can store program codes.

[0153] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for identifying illegal network communities, characterized in that: include: Determining an initial suspicion level of a first community, where the first community is any community in the community set; Generate a first comment vector according to a comment text posted by a first user in the first community, and generate a second comment vector according to a comment text posted by a second user in the first community, wherein the first user and the second user are any user who posted a comment in the first community; Obtaining a first synergy coefficient between the first user and the second user in the first community according to the first comment vector, the second comment vector, and the initial suspicion level; Obtaining a total synergy effect between the first user and the second user according to the first synergy coefficient, and obtaining a second synergy coefficient between the first user and the second user in the community set according to the total synergy effect; Determine a similar user set and a dissimilar user set according to the second cooperation coefficient; constructing an optimization function according to the similar user set and the non-similar user set to determine vector representations of users in the first community; Obtaining the user closeness of the first community according to the vector representation of the user; When the user closeness is less than a first threshold, determining that the first community is an illegal network community; Determining the initial suspicion level of the first community includes: obtaining the initial suspicion level by the following formula: Among them, the s p is used to represent the initial suspicion level, p is used to represent the first community, r(p) is used to represent the number of people who post in the first community, θ is a normalization constant between 0 and 1, P represents the community set, is used to represent the maximum number of posters in a single community in the community set P, and q is used to represent the community corresponding to the maximum number of posters; The obtaining of a first synergy coefficient between the first user i and the second user j in the first community according to the first comment vector, the second comment vector and the initial suspicion level comprises: When the time interval between the posting time of the first user in the first community and the posting time of the second user in the first community is less than the minimum time interval, the first synergy coefficient is obtained according to the following formula: Wherein, col(i,j,p) is used to represent the first synergy coefficient, i is used to represent the first user, j is used to represent the second user, α and β are weight coefficients greater than 0, and satisfy a+β=1, a<β, and is used to indicate the time when the first user posted in the first community, is used to indicate the time when the second user posted in the first community. t It is used to indicate the minimum time interval. is used to represent the first comment vector, the Used to represent the second comment vector; The obtaining of the total synergy effect of the first user and the second user according to the first synergy coefficient, and obtaining the second synergy coefficient of the first user and the second user in the community set according to the total synergy effect includes: obtaining the second synergy coefficient by the following formula: Wherein, σ(i,j) is used to represent the total synergistic effect, P i The community set P is used to represent the community set where the first user posts comments. j It is used to represent the community set where the second user posts comments. i ∩P j | is used to indicate the number of communities where the first user and the second user jointly post comments, |P i ∪P j | is used to represent the number of communities where the first user and the second user post comments, P i ∩P j The community set used to represent the comments jointly posted by the first user and the second user, wherein p is the P i ∩P j For any community in , the φ(i,j) is used to represent the second synergy coefficient; The obtaining the user closeness of the first community according to the vector representation of the user comprises: Wherein, spam(p) is used to represent the user closeness, R(p) is used to represent the set of users who have posted comments in the first community, i is any user in R(p), and |R(p)| is used to represent the number of people who have posted in the first community. is the vector representation of the user i.

2. The method according to claim 1, characterized in that Determining a similar user set and a dissimilar user set according to the second cooperation coefficient comprises: When the second synergy coefficient between the first user and the second user is greater than a second threshold, determining that the first user and the second user are a similar user pair to obtain the similar user set; When the second synergy coefficient between the first user and the second user is less than a third threshold, it is determined that the first user and the second user are a non-similar user pair to obtain the non-similar user set.

3. The method according to claim 2, characterized in that The constructing of an optimization function according to the similar user set and the non-similar user set comprises: Among them, the D p Used to represent the similar user set, D n is used to represent the non-similar user set, v c 、V a is the vector representation of the user pairs in the set, V c ' is used to represent V c The transpose of .

4. An electronic device, comprising a memory and a processor, characterized in that: The processor is used to implement the steps of the method for identifying an illegal network community as claimed in any one of claims 1 to 3 when executing the computer management program stored in the memory.

5. A computer-readable storage medium having a computer management program stored thereon, characterized in that: When the computer management program is executed by a processor, the steps of the method for identifying an illegal network community as claimed in any one of claims 1 to 3 are implemented.

Images