Method and system for measuring influence of multiple subjects in social media

Abstract

The invention discloses a method and system for measuring the influence of multiple subjects in social media, and the method comprises the steps: obtaining an influence index attribute value of a preset user subject based on the social media, and carrying out the calculation to obtain the basic influence of the user subject; according to the collaborative relationship strength between the user subjects, calculating to obtain multi-subject comprehensive basic influence; according to the user behavior, the corresponding preset weight coefficient and the nonlinear effect, the interaction influence of the user main body is obtained through calculation; according to the maximum saturation value of the interaction influence and a preset weight coefficient, calculating to obtain the multi-subject interaction influence of the user subject; and according to the multi-subject comprehensive basic influence, the multi-subject interaction influence and the corresponding weight coefficients, calculating to obtain the multi-subject comprehensive influence of the user subject. Through the technical scheme of the invention, the accurate evaluation of the multi-subject comprehensive influence is realized, and the timeliness of the user interaction behavior and the decreasing effect of the multi-subject interaction influence are dynamically reflected.

Description

Technical Field

[0001] The present invention relates to the field of information dissemination technology, and in particular to a method for measuring the influence of multiple subjects in social media and a system for measuring the influence of multiple subjects in social media. Background Art

[0002] With the rapid development and widespread popularity of social media, it has become an important platform for the study of information dissemination, user interaction, and social behavior. In social networks, user influence is not only a key driving force for information dissemination, but also a core factor in understanding network dynamics and optimizing dissemination strategies. However, traditional user influence research focuses on a single subject and ignores the problem of comprehensive influence measurement under the joint action of multiple subjects.

[0003] A large proportion of shared content in current social networks is jointly owned by multiple subjects. For example, a group photo involves multiple users, or an advertising post contains multiple brand logos. In a multi-subject scenario, the breadth and depth of information dissemination depends not only on the influence of a single user, but also on the characteristics, relevance, and overall synergy of multiple subjects. Therefore, how to scientifically quantify the comprehensive influence of multiple subjects and reveal their driving mechanism for information dissemination has become an important direction of current research.

[0004] In recent years, with the widespread application of social networks and the diversification of information dissemination methods, the study of user influence has gradually become an important direction in social network analysis.

[0005] In the study of user influence, existing technologies often rely too much on a single dimension, such as social relationships or communication paths, and lack comprehensive consideration of multi-dimensional characteristics of users, resulting in insufficient accuracy and comprehensiveness in influence measurement. The measurement method based on deep learning has low computational efficiency in large-scale social networks and is difficult to cope with real-time requirements and huge amounts of data. In addition, existing user influence measurement models usually do not fully consider the context of multiple subjects and mainly focus on the influence evaluation of a single user. This leads to limitations in these methods when dealing with multi-subject interactions, and they cannot fully reflect the interactions of multiple subjects in information dissemination and their comprehensive impact on the overall dissemination effect. Summary of the invention

[0006] In response to the above problems, the present invention provides a method and system for measuring the influence of multiple subjects in social media, which introduces multi-dimensional factors and fully considers the complex interactive effects among multiple subjects, introduces synergy relationship parameters to quantify the synergy between users, and combines basic influence with interactive influence to form a multi-subject comprehensive influence measurement model, thereby achieving an accurate assessment of the comprehensive influence of multiple subjects. By combining the time attenuation function and the nonlinear adjustment coefficient, it dynamically reflects the timeliness of user interactive behavior and its decreasing effect on the interactive influence of multiple subjects.

[0007] To achieve the above object, the present invention provides a method for measuring the influence of multiple subjects in social media, comprising:

[0008] Acquire the influence index attribute value of a preset user subject based on social media, and calculate the basic influence of the user subject according to the influence index attribute value and a preset weight coefficient;

[0009] According to the strength of the collaborative relationship between the user subjects, the comprehensive basic influence of multiple subjects is calculated based on the basic influence of the user subjects;

[0010] Calculate the interactive influence of the user subject according to the user behavior of the user subject and the corresponding preset weight coefficient and nonlinear effect;

[0011] The multi-subject interactive influence of the user subject is calculated according to the maximum saturation value of the interactive influence and the preset weight coefficient of the user subject;

[0012] The multi-subject comprehensive influence of the user subject is calculated based on the multi-subject comprehensive basic influence, the multi-subject interactive influence and the corresponding weight coefficients.

[0013] In the above technical solution, preferably, the process of obtaining the influence index attribute value of a preset user subject based on social media and calculating the basic influence of the user subject according to the influence index attribute value and a preset weight coefficient includes:

[0014] Based on the social media, the number of fans and the number of followers of the user subject are obtained, and the user subject u is calculated using a logarithmic function according to the number of fans and the number of followers. i Basic influence The calculation formula is:

[0015]

[0016] Among them, α and β are weight coefficients, F i Represents the user subject u i Number of fans; G i Represents user ui Number of followers.

[0017] In the above technical solution, preferably, the multi-subject comprehensive basic influence is calculated based on the basic influence of the user subjects according to the strength of the collaborative relationship between the user subjects, and the specific process includes:

[0018] Define the user combination U in the social media = {u1,u2,…,u n};

[0019] The Jaccard similarity coefficient is used to calculate the user subject u i and u j The strength of the synergistic relationship between ij , the calculation formula is:

[0020]

[0021] in, and Represents user subject u i and user principal u j A collection of commonly participated groups, interactive activities, or commonly-interested content;

[0022] According to the basic influence of the user subject and the strength of the synergy relationship between the user subjects, the multi-subject comprehensive basic influence of the user combination is calculated, and the calculation formula is:

[0023]

[0024] Among them, γ represents the weight coefficient of the synergistic effect.

[0025] In the above technical solution, preferably, the interactive influence of the user subject is calculated according to the user behavior of the user subject and the corresponding preset weight coefficient and nonlinear effect, and the specific process includes:

[0026] Obtain parameter values ​​of the user subject's like behavior, comment behavior, and forwarding behavior;

[0027] According to the weight coefficient of each user behavior and the corresponding nonlinear effect, the single user subject u is calculated i The interactive influence The formula is:

[0028]

[0029] Among them, ζ, η, θ represent the weight coefficients of likes, comments and forwarding behaviors. Represents user u i At time t ′ The number of likes for the published content, Represents user u i At time t ′ Number of comments on the post, Represents user u i At time t ′ The number of reposts of the published content, δ,ε,∈ is the nonlinear adjustment coefficient, δ,ε,∈<1 means that the relationship between user behavior and influence is decreasing. represents the nonlinear effect of likes, comments, and reposts, It is expressed as a time decay function, which means that the interactive influence of user behavior decays over time. λ is the time decay coefficient.

[0030] In the above technical solution, preferably, the multi-subject interactive influence of the user subject is calculated based on the maximum saturation value of the interactive influence and the preset weight coefficient of the user subject, and the specific process includes:

[0031] According to the maximum saturation value C of the interactive influence, and the adjustment parameter k of the interactive influence growth rate and saturation degree, the multi-subject interactive influence of the user subject is calculated The calculation formula is:

[0032]

[0033] Among them, the maximum saturation value C of the interactive influence represents the upper limit of the interactive influence, w i Represents user u i The weight coefficient of .

[0034] In the above technical solution, preferably, the multi-subject comprehensive influence of the user subject is calculated according to the multi-subject comprehensive basic influence and the multi-subject interactive influence and the corresponding weight coefficients, and the specific process includes:

[0035] Combined with the comprehensive basic influence of the multi-subject and the multi-agent interactive influence Calculate the comprehensive influence of multiple subjects I U (t), the calculation formula is:

[0036]

[0037] Among them, ξ is the weight coefficient of the comprehensive basic influence of multiple subjects, It is the weighted contribution of the interactive influence of multiple agents.

[0038] The present invention also proposes a system for measuring the influence of multiple subjects in social media, which uses the method for measuring the influence of multiple subjects in social media disclosed in any one of the above technical solutions, including:

[0039] A basic indicator influence calculation module is used to obtain the influence indicator attribute value of a preset user subject based on social media, and calculate the basic influence of the user subject according to the influence indicator attribute value and a preset weight coefficient;

[0040] A multi-subject basic influence calculation module is used to calculate the multi-subject comprehensive basic influence based on the basic influence of the user subjects according to the collaborative relationship strength between the user subjects;

[0041] A single-subject interaction influence calculation module is used to calculate the interaction influence of the user subject according to the user behavior of the user subject and the corresponding preset weight coefficient and nonlinear effect;

[0042] A multi-agent interaction influence calculation module, used to calculate the multi-agent interaction influence of the user subject according to the maximum saturation value of the interaction influence and the preset weight coefficient of the user subject;

[0043] The multi-subject comprehensive influence calculation module is used to calculate the multi-subject comprehensive influence of the user subject according to the multi-subject comprehensive basic influence and the multi-subject interactive influence and the corresponding weight coefficients.

[0044] In the above technical solution, preferably, the basic indicator influence calculation module is specifically used to:

[0045] Based on the social media, the number of fans and the number of followers of the user subject are obtained, and the user subject u is calculated using a logarithmic function according to the number of fans and the number of followers. i Basic influence The calculation formula is:

[0046]

[0047] Among them, α and β are weight coefficients, F i Represents the user subject u i Number of fans; G i Represents user u i Number of followers.

[0048] In the above technical solution, preferably, the multi-agent basic influence calculation module is specifically used for:

[0049] Define the user combination U in the social media = {u1,u2,…,u n};

[0050] The Jaccard similarity coefficient is used to calculate the user subject u i and u j The strength of the synergistic relationship between ij, the calculation formula is:

[0051]

[0052] in, and Represents user subject u i and user principal u j A collection of commonly participated groups, interactive activities, or commonly-interested content;

[0053] According to the basic influence of the user subject and the strength of the synergy relationship between the user subjects, the multi-subject comprehensive basic influence of the user combination is calculated, and the calculation formula is:

[0054]

[0055] Among them, γ represents the weight coefficient of the synergistic effect.

[0056] In the above technical solution, preferably, the single-agent interaction impact calculation module is specifically used to:

[0057] Obtain parameter values ​​of the user subject's like behavior, comment behavior, and forwarding behavior;

[0058] According to the weight coefficient of each user behavior and the corresponding nonlinear effect, the single user subject u is calculated i The interactive influence The formula is:

[0059]

[0060] Among them, ζ, η, θ represent the weight coefficients of likes, comments and forwarding behaviors. Represents user u i At time t ′ The number of likes for the published content, Represents user u i At time t ′ Number of comments on the post, Represents user u i At time t ′ The number of reposts of the published content, δ,ε,∈ is the nonlinear adjustment coefficient, δ,ε,∈<1 means that the relationship between user behavior and influence is decreasing. represents the nonlinear effect of likes, comments, and reposts, It is expressed as a time decay function, which means that the interactive influence of user behavior decays over time. λ is the time decay coefficient;

[0061] The multi-agent interaction impact calculation module is specifically used for:

[0062] According to the maximum saturation value C of the interactive influence, and the adjustment parameter k of the interactive influence growth rate and saturation degree, the multi-subject interactive influence of the user subject is calculated The calculation formula is:

[0063]

[0064] Among them, the maximum saturation value C of the interactive influence represents the upper limit of the interactive influence, w i Represents user u i The weight coefficient of .

[0065] Compared with the prior art, the beneficial effects of the present invention are as follows: by introducing multi-dimensional factors and fully considering the complex interactive effects among multiple subjects, introducing collaborative relationship parameters to quantify the collaborative effects among users, and integrating basic influence with interactive influence to form a multi-subject comprehensive influence measurement model, an accurate evaluation of the comprehensive influence of multiple subjects is achieved, and by combining the time attenuation function and the nonlinear adjustment coefficient, the timeliness of user interactive behavior and its decreasing effect on the interactive influence of multiple subjects are dynamically reflected. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 A schematic diagram of a flow chart of a method for measuring the influence of multiple subjects in social media disclosed in an embodiment of the present invention;

[0067] Figure 2 A schematic diagram of modules of a system for measuring the influence of multiple subjects in social media disclosed in an embodiment of the present invention.

[0068] In the figure, the corresponding relationship between each component and the reference numeral is as follows:

[0069] 1. Basic indicator influence calculation module, 2. Multi-subject basic influence calculation module, 3. Single-subject interactive influence calculation module, 4. Multi-subject interactive influence calculation module, 5. Multi-subject comprehensive influence calculation module. DETAILED DESCRIPTION

[0070] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0071] The present invention is further described in detail below in conjunction with the accompanying drawings:

[0072] like Figure 1As shown, a method for measuring the influence of multiple subjects in social media provided by the present invention includes:

[0073] Based on social media, the influence index attribute value of the preset user subject is obtained, and the basic influence of the user subject is calculated according to the influence index attribute value and the preset weight coefficient;

[0074] According to the strength of the collaborative relationship between user subjects, the comprehensive basic influence of multiple subjects is calculated based on the basic influence of user subjects;

[0075] The interactive influence of the user subject is calculated based on the user behavior of the user subject and the corresponding preset weight coefficient and nonlinear effect;

[0076] According to the maximum saturation value of the interactive influence and the preset weight coefficient of the user subject, the multi-subject interactive influence of the user subject is calculated;

[0077] According to the comprehensive basic influence of multiple subjects, the interactive influence of multiple subjects and the corresponding weight coefficients, the comprehensive influence of multiple subjects of the user subject is calculated.

[0078] In this implementation, by introducing multi-dimensional factors and fully considering the complex interactive effects among multiple subjects, synergy relationship parameters are introduced to quantify the synergy between users, and the basic influence and interactive influence are integrated to form a multi-subject comprehensive influence measurement model, thereby achieving an accurate assessment of the comprehensive influence of multiple subjects. By combining the time attenuation function and the nonlinear adjustment coefficient, the timeliness of user interactive behavior and its diminishing effect on the interactive influence of multiple subjects are dynamically reflected.

[0079] Specifically, in order to accurately evaluate the comprehensive influence of multiple users in the process of information dissemination, firstly, based on the social network structure, the basic influence of users is defined in combination with the number of fans and the number of followers of users, and the synergy between users is quantified by introducing synergy relationship parameters, thereby obtaining the comprehensive basic influence of multiple subjects. Secondly, taking into account the interactive behaviors of users (such as likes, comments, and forwarding), the present invention further defines the interactive influence, and reflects the timeliness of the interactive behaviors of multiple subjects and their decreasing effect on influence through the time decay function and the nonlinear adjustment coefficient. Finally, the basic influence and the interactive influence are combined and fused using the weight coefficient to finally form a more comprehensive and accurate multi-subject comprehensive influence measurement model.

[0080] In the above implementation, preferably, the influence index attribute value of a preset user subject is obtained based on social media, and the basic influence of the user subject is calculated according to the influence index attribute value and the preset weight coefficient. The specific process includes:

[0081] Based on social media, the number of fans and followers of the user subject is obtained. According to the number of fans and followers, the logarithmic function is used to calculate the user subject u i Basic influence The calculation formula is:

[0082]

[0083] Among them, α and β are weight coefficients, F i Represents the user subject u i Number of fans; G i Represents user u i Number of followers.

[0084] Specifically, the structure of social networks is the basis for the diffusion and spread of influence. In social networks, the number of fans of a user usually reflects his or her popularity, while the number of followers represents his or her influence in the network. Therefore, the number of fans and the number of followers can be used as important indicators to measure the influence of a user on social networks.

[0085] The logarithmic function is used to eliminate the order of magnitude difference between the number of fans and the number of followers, and to avoid imbalance in influence evaluation caused by excessive values. In addition, the logarithmic transformation makes the contribution of the growth of the number of fans and the number of followers to the influence gradually decrease, which is in line with the characteristics of diminishing marginal effects in reality.

[0086] In the above implementation, in the multi-subject influence model, the influence of multiple user bases is not simply summed, but the structural synergy between multiple subjects needs to be considered. For example, when multiple opinion leaders (Influencers) promote the same product at the same time, their combined influence far exceeds the sum of the influence of a single opinion leader.

[0087] Therefore, preferably, according to the strength of the collaborative relationship between the user subjects, the comprehensive basic influence of multiple subjects is calculated based on the basic influence of the user subjects. The specific process includes:

[0088] Define the user combination U in social media = {u1,u2,…,u n};

[0089] The Jaccard similarity coefficient is used to calculate the user subject u i and u j The strength of the synergistic relationship between ij , the calculation formula is:

[0090]

[0091] in, and Represents user subject u i and user principal u jA collection of commonly participated groups, interactive activities, or commonly-interested content;

[0092] According to the basic influence of the user subject and the strength of the synergy relationship between the user subjects, the multi-subject comprehensive basic influence of the user combination is calculated. The calculation formula is:

[0093]

[0094] Among them, γ represents the weight coefficient of the synergistic effect.

[0095] In the above implementation, preferably, the interactive influence of the user subject is calculated according to the user behavior of the user subject and the corresponding preset weight coefficient and nonlinear effect. The specific process includes:

[0096] Get the parameter values ​​of the user's like, comment, and forward behavior;

[0097] According to the weight coefficient of each user behavior and the corresponding nonlinear effect, the single user subject u is calculated i The interactive influence The formula is:

[0098]

[0099] Among them, ζ, η, θ represent the weight coefficients of likes, comments and forwarding behaviors. Represents user u i At time t ′ The number of likes for the published content, Represents user u i At time t ′ Number of comments on the post, Represents user u i At time t ′ The number of reposts of the published content,δ,ε,∈ is the nonlinear adjustment coefficient,δ,ε,∈<1 means that the relationship between user behavior and influence is decreasing, represents the nonlinear effect of likes, comments, and reposts, It is expressed as a time decay function, which means that the interactive influence of user behavior decays over time. λ is the time decay coefficient.

[0100] Specifically, user behavior (likes, comments, and reposts) can reflect the user's activity, interactivity, and content dissemination ability in social networks, and is a direct indicator for measuring the user's social network influence. Among them, the interactive influence generated by user behavior in social networks will decay over time, so the time decay function is used to adjust the interactive influence to make it more in line with the real scene.

[0101] In addition, in real social networks, the influence of user interaction is not constant, but gradually decreases as the interaction increases. In particular, when a piece of content has received a large number of likes, the marginal contribution of each additional like to the influence will decrease. Therefore, the present invention uses exponential weighting to adjust the effect of interaction on user influence.

[0102] In the above implementation, preferably, the multi-subject interactive influence of the user subject is calculated according to the maximum saturation value of the interactive influence and the preset weight coefficient of the user subject, and the specific process includes:

[0103] According to the maximum saturation value C of the interactive influence, and the adjustment parameter k of the interactive influence growth rate and saturation degree, the multi-subject interactive influence of the user subject is calculated The calculation formula is:

[0104]

[0105] Among them, the maximum saturation value C of the interactive influence represents the upper limit of the interactive influence, w i Represents user u i The weight coefficient of .

[0106] Specifically, in real social networks, users’ attention and influence have their upper limits. When the interactions of multiple high-influence users are accumulated, the combined influence will not grow linearly, but tend to a certain stable level, reflecting the phenomenon of audience attention saturation and information overload.

[0107] In the above implementation, preferably, the multi-subject comprehensive influence of the user subject is calculated according to the multi-subject comprehensive basic influence and the multi-subject interactive influence and the corresponding weight coefficients. The specific process includes:

[0108] Combined with the comprehensive basic influence of multiple subjects Interactive influence of multiple subjects Calculate the comprehensive influence of multiple subjects I U (t), the calculation formula is:

[0109]

[0110] Among them, ξ is the weight coefficient of the comprehensive basic influence of multiple subjects, It is the weighted contribution of the interactive influence of multiple agents.

[0111] Specifically, in social networks, a user's influence depends not only on his position in the network structure (such as the number of fans and followers), but also on his activity and interactive behavior on the platform (such as likes, comments, and reposts). Therefore, in order to comprehensively measure the comprehensive influence of multiple subjects, the present invention combines basic influence and interactive influence to obtain the comprehensive influence of multiple subjects.

[0112] In summary, by combining the basic influence of social networks and the interactive influence of user behaviors, the influence of multiple subjects can be accurately quantified, overcoming the limitation of existing research that only focuses on the influence of a single user. The synergistic relationship strength parameter is introduced, and the Jaccard similarity coefficient is used to quantify the synergistic effect between users, effectively capturing the nonlinear synergistic effect of the basic influence of multiple subjects. In addition, the time attenuation function and the nonlinear adjustment coefficient are combined to dynamically evaluate the timeliness and decreasing effect of user interactive behaviors, thereby improving the adaptability and responsiveness of the model. In general, the present invention realizes the accurate and dynamic evaluation of the comprehensive influence of multiple subjects, significantly enhancing the comprehensiveness and practicality of influence measurement.

[0113] like Figure 2 As shown, the present invention also proposes a system for measuring the influence of multiple subjects in social media, which applies the method for measuring the influence of multiple subjects in social media disclosed in any one of the above embodiments, including:

[0114] The basic indicator influence calculation module 1 is used to obtain the influence indicator attribute value of the preset user subject based on social media, and calculate the basic influence of the user subject according to the influence indicator attribute value and the preset weight coefficient;

[0115] The multi-subject basic influence calculation module 2 is used to calculate the multi-subject comprehensive basic influence based on the basic influence of the user subjects according to the collaborative relationship strength between the user subjects;

[0116] The single subject interaction influence calculation module 3 is used to calculate the interaction influence of the user subject according to the user behavior of the user subject and the corresponding preset weight coefficient and nonlinear effect;

[0117] The multi-subject interaction influence calculation module 4 is used to calculate the multi-subject interaction influence of the user subject according to the maximum saturation value of the interaction influence and the preset weight coefficient of the user subject;

[0118] The multi-subject comprehensive influence calculation module 5 is used to calculate the multi-subject comprehensive influence of the user subject according to the multi-subject comprehensive basic influence and the multi-subject interactive influence and the corresponding weight coefficients.

[0119] In this implementation, by introducing multi-dimensional factors and fully considering the complex interactive effects among multiple subjects, synergy relationship parameters are introduced to quantify the synergy between users, and the basic influence and interactive influence are integrated to form a multi-subject comprehensive influence measurement model, thereby achieving an accurate assessment of the comprehensive influence of multiple subjects. By combining the time attenuation function and the nonlinear adjustment coefficient, the timeliness of user interactive behavior and its diminishing effect on the interactive influence of multiple subjects are dynamically reflected.

[0120] In the above implementation, preferably, the basic indicator influence calculation module 1 is specifically used for:

[0121] Based on social media, the number of fans and followers of the user subject is obtained. According to the number of fans and followers, the logarithmic function is used to calculate the user subject u i Basic influence The calculation formula is:

[0122]

[0123] Among them, α and β are weight coefficients, F i Represents user subject u i Number of fans; G i Represents user u i Number of followers.

[0124] In the above implementation, preferably, the multi-agent basic impact calculation module 2 is specifically used for:

[0125] Define the user combination U in social media = {u1,u2,…,u n};

[0126] The Jaccard similarity coefficient is used to calculate the user subject u i and u j The strength of the synergistic relationship between ij , the calculation formula is:

[0127]

[0128] in, and Represents user subject u i and user principal u j A collection of commonly participated groups, interactive activities, or commonly-interested content;

[0129] According to the basic influence of the user subject and the strength of the synergy relationship between the user subjects, the multi-subject comprehensive basic influence of the user combination is calculated. The calculation formula is:

[0130]

[0131] Among them, γ represents the weight coefficient of the synergistic effect.

[0132] In the above implementation, preferably, the single-agent interaction influence calculation module 3 is specifically used for:

[0133] Get the parameter values ​​of the user's like, comment, and forward behavior;

[0134] According to the weight coefficient of each user behavior and the corresponding nonlinear effect, the single user subject u is calculated i The interactive influence The formula is:

[0135]

[0136] Among them, ζ, η, θ represent the weight coefficients of likes, comments and forwarding behaviors. Represents user u i At time t ′ The number of likes for the published content, Represents user u i At time t ′ Number of comments on the post, Represents user u i At time t ′ The number of reposts of the published content,δ,ε,∈ is the nonlinear adjustment coefficient,δ,ε,∈<1 means that the relationship between user behavior and influence is decreasing, represents the nonlinear effect of likes, comments, and reposts, It is expressed as a time decay function, which means that the interactive influence of user behavior decays over time. λ is the time decay coefficient;

[0137] The multi-agent interaction influence calculation module 4 is specifically used for:

[0138] According to the maximum saturation value C of the interactive influence, and the adjustment parameter k of the interactive influence growth rate and saturation degree, the multi-subject interactive influence of the user subject is calculated The calculation formula is:

[0139]

[0140] Among them, the maximum saturation value C of the interactive influence represents the upper limit of the interactive influence, w i Represents user u i The weight coefficient of .

[0141] According to the system for measuring the influence of multiple subjects in social media disclosed in the above-mentioned implementation mode, the functions to be implemented by each module thereof correspond to the steps in the method for measuring the influence of multiple subjects in social media disclosed in the above-mentioned implementation mode. During the implementation process, operations are performed with reference to the above-mentioned implementation mode, which will not be repeated here.

[0142] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for measuring the influence of multiple subjects in social media, characterized in that: include: Acquire the influence index attribute value of a preset user subject based on social media, and calculate the basic influence of the user subject according to the influence index attribute value and a preset weight coefficient; According to the strength of the collaborative relationship between the user subjects, the comprehensive basic influence of multiple subjects is calculated based on the basic influence of the user subjects; Calculate the interactive influence of the user subject according to the user behavior of the user subject and the corresponding preset weight coefficient and nonlinear effect; The multi-subject interactive influence of the user subject is calculated according to the maximum saturation value of the interactive influence and the preset weight coefficient of the user subject; The multi-subject comprehensive influence of the user subject is calculated based on the multi-subject comprehensive basic influence, the multi-subject interactive influence and the corresponding weight coefficients.

2. The method for measuring the influence of multiple subjects in social media according to claim 1, characterized in that: The process of obtaining the influence index attribute value of a preset user subject based on social media and calculating the basic influence of the user subject according to the influence index attribute value and a preset weight coefficient includes: Based on the social media, the number of fans and the number of followers of the user subject are obtained, and the user subject u is calculated using a logarithmic function according to the number of fans and the number of followers. i Basic influence The calculation formula is: Among them, α and β are weight coefficients, F i Represents the user subject u i Number of fans; G i Represents user u i Number of followers.

3. The method for measuring the influence of multiple subjects in social media according to claim 1, characterized in that: The multi-subject comprehensive basic influence is calculated based on the basic influence of the user subjects according to the strength of the collaborative relationship between the user subjects. The specific process includes: Define the user combination U in the social media = {u1,u2,…,u n }; The Jaccard similarity coefficient is used to calculate the user subject u i and u j The strength of the synergistic relationship between ij , the calculation formula is: in, and Represents user subject u i and user principal u j A collection of commonly participated groups, interactive activities, or commonly-interested content; According to the basic influence of the user subject and the strength of the synergy relationship between the user subjects, the multi-subject comprehensive basic influence of the user combination is calculated, and the calculation formula is: Among them, γ represents the weight coefficient of the synergistic effect.

4. The method for measuring the influence of multiple subjects in social media according to claim 1, characterized in that: The interactive influence of the user subject is calculated based on the user behavior of the user subject and the corresponding preset weight coefficient and nonlinear effect. The specific process includes: Obtain parameter values ​​of the user subject's like behavior, comment behavior, and forwarding behavior; According to the weight coefficient of each user behavior and the corresponding nonlinear effect, the single user subject u is calculated i The interactive influence The formula is: Among them, ζ, η, θ represent the weight coefficients of likes, comments and forwarding behaviors. Represents user u i At time t ′ The number of likes for the published content, Represents user u i At time t ′ Number of comments on the post, Represents user u i At time t ′ The number of reposts of the published content,δ,ε,∈ is the nonlinear adjustment coefficient,δ,ε,∈<1 means that the relationship between user behavior and influence is decreasing, represents the nonlinear effect of likes, comments, and reposts, It is expressed as a time decay function, which means that the interactive influence of user behavior decays over time. λ is the time decay coefficient.

5. The method for measuring the influence of multiple subjects in social media according to claim 1, characterized in that: The multi-subject interactive influence of the user subject is calculated based on the maximum saturation value of the interactive influence and the preset weight coefficient of the user subject. The specific process includes: According to the maximum saturation value C of the interactive influence, and the adjustment parameter k of the interactive influence growth rate and saturation degree, the multi-subject interactive influence of the user subject is calculated The calculation formula is: Among them, the maximum saturation value C of the interactive influence represents the upper limit of the interactive influence, w i Represents user u i The weight coefficient of .

6. The method for measuring the influence of multiple subjects in social media according to claim 1, characterized in that: The multi-subject comprehensive influence of the user subject is calculated based on the multi-subject comprehensive basic influence, the multi-subject interactive influence and the corresponding weight coefficients. The specific process includes: Combined with the comprehensive basic influence of the multi-subject and the multi-agent interactive influence Calculate the comprehensive influence of multiple subjects I U (t), the calculation formula is: Among them, ξ is the weight coefficient of the comprehensive basic influence of multiple subjects, It is the weighted contribution of the interactive influence of multiple agents.

7. A system for measuring the influence of multiple subjects in social media, characterized in that: The method for measuring the influence of multiple subjects in social media as claimed in any one of claims 1 to 6 comprises: A basic indicator influence calculation module is used to obtain the influence indicator attribute value of a preset user subject based on social media, and calculate the basic influence of the user subject according to the influence indicator attribute value and a preset weight coefficient; A multi-subject basic influence calculation module is used to calculate the multi-subject comprehensive basic influence based on the basic influence of the user subjects according to the collaborative relationship strength between the user subjects; A single-subject interaction influence calculation module is used to calculate the interaction influence of the user subject according to the user behavior of the user subject and the corresponding preset weight coefficient and nonlinear effect; A multi-agent interaction influence calculation module, used to calculate the multi-agent interaction influence of the user subject according to the maximum saturation value of the interaction influence and the preset weight coefficient of the user subject; The multi-subject comprehensive influence calculation module is used to calculate the multi-subject comprehensive influence of the user subject according to the multi-subject comprehensive basic influence and the multi-subject interactive influence and the corresponding weight coefficients.

8. The system for measuring the influence of multiple subjects in social media according to claim 7, characterized in that: The basic indicator influence calculation module is specifically used for: Based on the social media, the number of fans and the number of followers of the user subject are obtained, and the user subject u is calculated using a logarithmic function according to the number of fans and the number of followers. i Basic influence The calculation formula is: Among them, α and β are weight coefficients, F i Represents the user subject u i Number of fans; G i Represents user u i Number of followers.

9. The system for measuring the influence of multiple subjects in social media according to claim 7, characterized in that: The multi-agent basic impact calculation module is specifically used for: Define the user combination U in the social media = {u1,u2,…,u n }; The Jaccard similarity coefficient is used to calculate the user subject u i and u j The strength of the synergistic relationship between ij , the calculation formula is: in, and Represents user subject u i and user principal u j A collection of commonly participated groups, interactive activities, or commonly-interested content; According to the basic influence of the user subject and the strength of the synergy relationship between the user subjects, the multi-subject comprehensive basic influence of the user combination is calculated, and the calculation formula is: Among them, γ represents the weight coefficient of the synergistic effect.

10. The system for measuring the influence of multiple subjects in social media according to claim 7, characterized in that: The single-agent interaction impact calculation module is specifically used for: Obtain parameter values ​​of the user subject's like behavior, comment behavior, and forwarding behavior; According to the weight coefficient of each user behavior and the corresponding nonlinear effect, the single user subject u is calculated i The interactive influence The formula is: Among them, ζ, η, θ represent the weight coefficients of likes, comments and forwarding behaviors. Represents user u i At time t ′ The number of likes for the published content, Represents user u i At time t ′ Number of comments on the post, Represents user u i At time t ′ The number of reposts of the published content, δ,ε,∈ is the nonlinear adjustment coefficient, δ,ε,∈<1 means that the relationship between user behavior and influence is decreasing. represents the nonlinear effect of likes, comments, and reposts, It is expressed as a time decay function, which means that the interactive influence of user behavior decays over time. λ is the time decay coefficient; The multi-agent interaction impact calculation module is specifically used for: According to the maximum saturation value C of the interactive influence, and the adjustment parameter k of the interactive influence growth rate and saturation degree, the multi-subject interactive influence of the user subject is calculated The calculation formula is: Among them, the maximum saturation value C of the interactive influence represents the upper limit of the interactive influence, w i Represents user u i The weight coefficient of .

Images