Method for constructing network propagation influence index model

Abstract

The invention belongs to the technical field of network propagation evaluation, and particularly relates to a network propagation influence index model construction method, which comprises the steps of constructing an INC index system and performing correlation analysis, constructing a calculation model of six first-level indexes, and constructing a propagation influence index total score model. The method has the advantages that the propagation influence index can integrate data of a plurality of platforms such as social media platforms, news websites, search engines and the like. The cross-platform integration capability can more comprehensively reflect the propagation effect of brands or information in the whole network range, and avoids the one-sidedness of single channel evaluation.

Description

Technical Field

[0001] This application belongs to the field of network communication evaluation technology, specifically involving a method for constructing a network communication influence index model. Background Technology

[0002] With the deepening development of the all-media era, communication media and methods have become increasingly diverse and multi-dimensional, significantly enhancing the speed and reach of communication. How to determine the scope of information dissemination's impact and accurately assess the effectiveness of brand information communication has become a focal point for governments, businesses, and media. However, the rapid development of new media has greatly increased the difficulty of assessment. Traditional media's external information dissemination method was singular, simply "media → audience," with a highly controllable process. Today, however, everyone is a content creator and producer, meaning "media → opinion leaders → the public → forming communication influence." Therefore, a more scientific and comprehensive communication influence index model is needed to accurately quantify the overall effect of brand communication in the digital age. Summary of the Invention

[0003] This application uses the INC index to set up multiple primary and secondary indicators, including reach, popularity, online attention, market activity, and brand reputation. These dimensions together constitute a comprehensive evaluation of the communication effect. This multi-dimensional evaluation system can more comprehensively reflect the complexity and diversity of brand communication, thus providing a more accurate and comprehensive assessment of communication influence. The technical solution is as follows: A method for constructing a network dissemination influence index model includes the following steps: S1. Construct the INC index indicator system and conduct correlation analysis; The INC index indicator system is constructed, setting six primary indicators: breadth of dissemination, popularity of dissemination, online attention, presentation on proprietary platforms, market activity, and brand reputation; the breadth of dissemination is calculated separately for each indicator. Popularity of dissemination Online attention Presentation level on proprietary platforms Market activity Brand reputation The Pearson correlation coefficients among the secondary indicators under these six primary indicators; S2. Construct a calculation model for six primary indicators; S3. Construct the total score model of the Communication Influence Index: The total score of the INC index is first calculated based on the original data to limit the values ​​of each sub-index to a reasonable range. Then, it is calculated by addition and finally logarithmic compression is performed to obtain the theoretical limit value of the total score.

[0004] Preferably, the secondary indicator of dissemination breadth is The third-level indicators include the amount of information on Weibo. WeChat information volume Total number of clients ,website ,forum ,video and digital newspaper ;

[0005]

[0006] in, To spread the score, For the logarithmic coefficient, , This is a constant term.

[0007] Equal to the total volume That is, the sum of the information content of the seven tertiary indicators.

[0008] Preferred, popularity The secondary indicators include forwarding rate. and comment rate Among them, the forwarding rate Including Weibo information volume Weibo reposts Comment rate include and the total number of comments across the entire network Forwarding rate and comment rate The calculation formula is as follows: ; ; The formula for the forwarding rate score in the spread popularity is as follows:

[0009] in, To earn points for forwarding, For forwarding rate, It's about the amount of information on Weibo. It is related to forwarding weight. , For constant terms; The formula for the comment rate score is as follows: ; in, To score the comments, For comment rate, It is a comment-related weight. It's the overall online buzz. , For constant terms, The popularity score is calculated by summing the forwarding score and the commenting score, with a value range of [0,5]. The formula is as follows:

[0010] in, To determine the spread of popularity score, For coefficient terms, This is a constant term.

[0011] Preferably, online attention includes six secondary indicators and seven tertiary indicators; the secondary indicators are Baidu search... TikTok search Headline Search Popularity Black Cat Complaints and the increase in comments The third-level indicators include Baidu Index. TikTok search Headline Search Popularity Black Cat Complaints Ctrip's review volume this month And Ctrip's review volume last month ; The formula for calculating the secondary indicators of online attention is as follows: ; i It is 4-9. It is a constant and is related to the industry sector being evaluated. The formula for online attention is as follows: ; Industry search coefficient and the search coefficient for each evaluation unit. , For search weight, i It is 4-9.

[0012] Preferably, the self-owned platform presentation dimension involves six secondary indicators and twelve tertiary indicators; the secondary indicators include the number of keywords included. The amount of information collected Publication volume on WeChat official accounts WeChat Official Account Headline Reading Count Number of video works and video likes The third-level indicators include Baidu PC. Baidu Mobile Sogou PC Sogou Mobile 360PC 360 Mobile Shenma January Collection Publication volume on WeChat official accounts WeChat Official Account Headline Reading Count Number of video works and video likes ; Number of keywords included It includes seven third-level indicators, including Baidu PC. Baidu Mobile Sogou PC Sogou Mobile 360PC 360 Mobile Shenma First, perform a weighted summation. , j =10,...,15; Information carrying capacity It is the product of contrast strength and impact. ; Contrast intensity is the magnitude of the differences among various evaluation values ​​of the same indicator, calculated in the form of standard deviation: i =10,...,15; Conflict The degree of correlation between different indicators , It is an indicator i With indicators j The correlation coefficient between them;

[0013] After dimensionless processing of the six secondary indicators, their weights are determined according to industry type. The presentation formula of the proprietary platform is as follows: ; in, To increase the weight of included keywords, Weighting based on the amount of information included. For the relevant weight of the official account, For video account related weight, , , , , This is a constant term.

[0014] Preferably, market activity includes the secondary indicator of JD.com brand product market share. Top-selling products on Tmall and offline sales share The third-level indicator includes the volume of products sold on JD.com. Top-selling products on Tmall Single item sold on Tmall and offline sales ; Sales volume was standardized using the reciprocal method to calculate market share, and then a function was fitted. The formula for calculating the market share of top products on Tmall is as follows: ; in, It refers to the top-selling products in the category on Tmall. This refers to the sales volume of a single brand's top-selling products on Tmall. This refers to the number of units sold for each product of the brand on the platform. This refers to the number of units sold for each product category on the platform. JD.com's brand product market share The calculation formula is as follows: ; Offline sales share The calculation formula is as follows:

[0015] in, This refers to the sales volume of this product category on JD.com. This refers to the brand's sales volume on JD.com. It represents the offline sales revenue of the product category. It refers to the brand's offline sales revenue; ; in, , , These are the weighting coefficients, This is a constant term.

[0016] Preferably, the brand reputation dimension includes two secondary indicators and five tertiary indicators; the secondary indicators include the overall score. and positive review rate The third-level indicators include the overall score. Number of reviews this month Number of positive reviews this month Number of reviews last month Compared to last month's positive reviews ; Overall rating The value range is [0, 5], and the positive review rate is... The value range is [0,1]. The formula for calculating brand reputation is as follows: .

[0017] Preferably, based on experience, the industry is divided into four categories: consumer brands, scenic spot brands, catering and hotels, and others, and the original value of INC is introduced. This is the unlog-standardized INC score, calculated using the following formula, which varies depending on the industry: Consumer Brands The overall score model for the communication influence index is as follows: ; ; in, The original value coefficient of INC. It is a constant; Scenic spot brand category The overall score model for the communication influence index is as follows: ; in i =4, 5, 6, 7, 9; Therefore, , It is a constant; Catering and Hotel The overall score model for the communication influence index is as follows: ; in i =4, 5, 6, 7, Brand reputation coefficient; , It is a constant; The overall score model for other brand communication influence indices is as follows: ; in i =4, 5, 6, 7; , It is a constant.

[0018] Compared with the prior art, the beneficial effects of this application are as follows: First, the index has low distortion. The complexity of data often leads to the distortion of the compiled index when reflecting comparisons and changes in things. The index model fully considers the impact of the size of the same type of data on the index. For example, the comment rate in the popularity index is not simply the ratio of the number of comments to the total amount of information, but also takes into account the impact of comment activity and the total amount of information on the index, thus preventing the index from being distorted.

[0019] Another unique feature is that the communication influence index may have different evaluation dimensions and weights across different industries and sectors. For example, market activity may be more important for the consumer goods industry, while brand reputation may be more crucial for the hotel and catering industry. This customized evaluation method makes the index more closely aligned with actual needs, reflecting its uniqueness.

[0020] Furthermore, the Communication Influence Index integrates data from multiple social media platforms, news websites, search engines, and other platforms. This cross-platform integration capability can more comprehensively reflect the dissemination effect of a brand or message across the entire internet, avoiding the one-sidedness of evaluation from a single channel. Attached Figure Description

[0021] Figure 1 A curve was fitted to show the extent of propagation.

[0022] Figure 2 A curve was fitted to represent market activity.

[0023] Figure 3 Screenshot from the brand's big data platform.

[0024] Figure 4 This is a flowchart of the application process. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] A method for constructing a network dissemination influence index model includes the following steps: S1. Construct the INC index indicator system and conduct correlation analysis; The INC index indicator system is constructed, setting six primary indicators: breadth of dissemination, popularity of dissemination, online attention, presentation on proprietary platforms, market activity, and brand reputation; the breadth of dissemination is calculated separately for each indicator. Popularity of dissemination Online attention Presentation level on proprietary platforms Market activity Brand reputation The Pearson correlation coefficients among the secondary indicators under these six primary indicators; S2. Construct a calculation model for six primary indicators; S3. Construct the total score model of the Communication Influence Index: The total score of the INC index is first calculated based on the original data to limit the values ​​of each sub-index to a reasonable range. Then, it is calculated by addition and finally logarithmic compression is performed to obtain the theoretical limit value of the total score.

[0027] First, the principles for constructing indicators: Scientific Principle: The information dissemination influence index system should be based on scientific methods and data. The selected indicators should be representative and able to objectively and truthfully reflect the brand's communication influence. At the same time, the authenticity, accuracy, and completeness of the data must be guaranteed to ensure the objectivity and accuracy of the index results, avoiding subjective assumptions and methods that are difficult to quantify.

[0028] Systematic principle: The online communication influence index involves multiple industry sectors and multiple evaluation dimensions. Therefore, the construction of the model system needs to comprehensively consider the relationship between various elements and be able to comprehensively reflect the brand's communication influence.

[0029] Hierarchical principle: The constructed evaluation index system should have a certain hierarchical structure, determine which subsystems the evaluation system includes, and which specific measurement indicators are included under each subsystem. At the same time, each indicator should be independent of the others, targeted, and avoid duplication of dimensions.

[0030] Feasibility principle: When constructing an effective communication influence index system, the ease of data collection should be considered when selecting indicator dimensions to ensure that the index results can provide effective reference and guidance for improving information dissemination.

[0031] Second, the selection of primary indicators: Communication power: Communication power is short for media communication power, referring to the strength of a media outlet and its ability to collect information, report news, and influence society. Specifically, it can include the scope of dissemination, speed of dissemination, duration of dissemination, and coverage of the target audience, with communication effect being the main indicator of media communication power.

[0032] Impact: The degree of influence on society when information is delivered to an audience through certain communication methods in order to achieve a specific communication effect. Impact is a quantitative assessment of the effectiveness of news communication, which can be measured through a comprehensive evaluation of the information release entity, the content released, the audience, the communication channels, and audience feedback.

[0033] The indicators are constructed based on six dimensions: After defining the concept and measurement methods of communication influence, Yidian Think Tank selected evaluation indicators based on big data theory, brand management theory, communication theory, SEO theory, etc., according to the principles of feasibility and hierarchy, and selected a total of 6 dimensions.

[0034] (a) Breadth of dissemination Dissemination breadth is a crucial indicator for measuring the reach and channel coverage of brand information online. It comprehensively reflects the quantity of information disseminated across the entire internet and the breadth of channel coverage. Specifically, it is constructed as follows: ① Dissemination breadth is primarily reflected in the total volume of information across the entire internet, i.e., the volume of the information's impact. This includes original and forwarded information about the brand or content on all online dissemination channels (such as news websites, social media, forums, etc.). ② Dissemination breadth also considers which channels this information is disseminated through. If relevant information can be widely disseminated across multiple platforms, its dissemination breadth is relatively high. Based on platform attributes, it is divided into seven dissemination channels: Weibo, WeChat, mobile apps, websites, forums, video platforms, and digital newspapers.

[0035] (ii) Popularity of dissemination The term "spread popularity" refers to the extent and enthusiasm with which a brand is widely disseminated and discussed on social media, online platforms, and other channels within a monitored period. This metric is an important standard for measuring the influence and popularity of information dissemination. Specifically, it is constructed as follows: ① For social media popularity, it is measured by the number of reposts, likes, views, and comments. ② For platforms outside of social media, such as mobile apps and websites, it is measured by the number of comments and views.

[0036] (III) Online Attention Online attention is a crucial indicator of brand influence in the digital age, primarily reflecting the level of search interest in brand information on search platforms. In the INC index evaluation system, online attention is a key metric for measuring brand awareness and audience interest. Specifically: ① Data sources for online attention mainly include search big data platforms such as Baidu, 360, Douyin, and Toutiao. These platforms record and analyze netizens' search behavior, comprehensively reflecting netizens' active search intentions and interest in brand information. ② The number of complaints on the Black Cat Complaint Platform has a negative impact on netizens' attention levels to some extent. ③ For the scenic spot sector, higher ticket purchase reviews indicate higher attention.

[0037] (iv) Presentation on proprietary platforms The visibility of a brand's own platforms is a crucial indicator for evaluating the brand's capabilities and effectiveness in building its own communication platforms. It primarily reflects the visibility and information display effectiveness of a brand's own platforms (such as its official website and official social media accounts) in search engines. Specifically, it is structured as follows: ① Official Website Visibility: Based on SEO theory, this indicator measures a brand's independent communication strength in cyberspace by evaluating the visibility of its official website on search platforms. This indicator not only focuses on the platform's own construction quality, such as page design, content update frequency, and information richness, but also on factors such as the platform's inclusion in search engines, its ranking position, and user experience. ② WeChat Official Account Visibility: Measured by the frequency of article publications, readership, number of headlines, and headline readership. ③ Video Account Visibility: Measured by the number of videos published and viewed.

[0038] (v) Market activity Market activity is a crucial indicator reflecting the frequency of a brand's transactions and the richness of its product offerings. It embodies consumer interest and enthusiasm for a specific brand or product and is a key dimension in the quantitative evaluation system of brand communication effectiveness. Specifically, it is constructed as follows: ① For online sales platforms, brand sales volume, the number of paid orders for top-selling products in each category, and the number of brands included in the rankings measure online activity. ② For offline markets, sales volume and sales revenue reflect brand activity.

[0039] (vi) Brand reputation Brand reputation reflects a brand's positive image and level of trust among consumers; it represents the degree of goodwill and trust people have towards a particular brand in the market. It includes not only consumer evaluations of the brand's product quality, performance, and price, but also recognition of the brand's image, culture, and social responsibility. Specific indicator construction: For consumer food and beverage brands, ratings and positive review rates on Dianping (a Chinese review platform) are used for measurement. Higher ratings and higher positive review rates indicate better performance in terms of ingredients, taste, service, and value for money.

[0040] Third, indicator selection based on correlation analysis: To reduce redundancy among the indicator systems, correlation analysis was performed on the indicator data.

[0041] First, subjectively remove indicators such as online sales figures that are difficult to obtain and micro-indices with inconsistent time ranges, and then calculate the breadth of dissemination separately. Popularity of dissemination Online attention Presentation level on proprietary platforms Market activity Brand reputation The Pearson correlation coefficients between the secondary indicators under these six primary indicators are shown in Table 1-6.

[0042] Table 1. Results of Correlation Analysis of Dissemination Breadth Indicators

[0043] Note: **At the 0.01 level (two-tailed), the correlation is significant.

[0044] The so-called "overall online presence" refers to the total amount of information a brand generates across various online platforms, including social media, forums, comment sections, and news websites.

[0045] In terms of reach, since the total volume of voice is the sum of information from seven channels, it is significantly correlated with each dimension. Therefore, only the total volume of voice indicator is retained.

[0046] ; In the dimension of virality, the indicator data needs to be processed again. Weibo is representative of social media repost volume, therefore Weibo data is selected.

[0047] ; ; in, For forwarding rate, For comment rate, For Weibo reposts, For the amount of information on Weibo, This represents the total number of comments across the entire internet.

[0048] Table 2. Correlation Analysis Results of Dissemination Breadth Indicators

[0049] According to the data in Table 2, the correlation coefficient between the repost ratio and the comment ratio is 0.416, which is acceptable. Therefore, both indicators are retained.

[0050] In terms of online attention, the number of travel reviews also needs to be processed: ; in, To increase the number of comments, This is the number of reviews on Ctrip this month. This is the number of reviews on Ctrip last month.

[0051] Table 3. Correlation Analysis Results of Online Attention Indicators

[0052] According to the data in Table 3, the 360 ​​trend of online attention is highly correlated with Baidu, Douyin search, and Toutiao search, and its search engine attributes overlap with Baidu. Therefore, it can be deleted, while other indicators are retained.

[0053] Table 4. Correlation Analysis Results of Presentation Indicators on Owned Platforms

[0054] Table 4 shows that, regarding visibility on proprietary platforms, the correlation between the number of WeChat official account articles, the number of WeChat official account headline views, the number of video works and video likes, and monthly indexing with other indicators is not very high. However, the correlation between keyword data related to Baidu PC and Baidu Mobile indexing is relatively high. Therefore, we retain the indicators of WeChat official account article volume, WeChat official account headline views, the number of video works and video likes, and monthly indexing, and summarize the indexed keyword volume. .

[0055] ; in, It refers to the total number of keywords included. As weight.

[0056] Table 5. Correlation Analysis Results of Market Activity Indicators

[0057] According to the data in Table 5, the correlation between offline sales volume and offline sales revenue is relatively high. Considering that the correlation between offline sales revenue and other indicators is low, offline sales volume is deleted and other indicators are retained.

[0058] Table 6. Correlation Analysis Results of Brand Reputation Indicators

[0059] Regarding brand reputation, the quantitative indicators of evaluation were processed, and consumer reviews were used. Positive feedback rate Measurement for the hotel and catering sector.

[0060] ; The index system selected after correlation analysis is shown in Table 7.

[0061]

[0062]

[0063]

[0064] Rationality test of the indicator system The rationality of an indicator system can be tested by its information contribution rate, which measures how much information the final indicator system retains compared to the original data. This is usually achieved by comparing the standard deviations of the two, as the standard deviation reflects the amount of information. If the final system can retain approximately 90% of the information from the original data (based on standard deviation assessment), then the indicator system is considered reasonable.

[0065] Calculations show that the information contribution rate of the INC index is 90.56%, indicating that the indicator system after correlation analysis can represent 90.56% of the information, and the 21 secondary indicators can represent most of the initial indicators. This demonstrates that the constructed communication influence index system is reasonable and effective.

[0066] Fourth, the online communication influence index model: (1) Set keywords: By monitoring over 200,000 dissemination platforms and hundreds of millions of dissemination sites on the information big data platform, we collect data on the overall dissemination breadth, popularity, and content aggregation of brand-related information, as well as the dissemination popularity of individual pieces of information; by monitoring the search big data platform, we collect data on netizens' attention and event popularity.

[0067] Keyword settings during information collection should adhere to the following principles: The principle of uniqueness: that is, excluding irrelevant content.

[0068] Comprehensiveness principle: On the basis of uniqueness, ensure the comprehensiveness of data and prevent relevant information from being excluded.

[0069] Regional limitation principle. Region + brand, or limit the region in the frog's eyes.

[0070] Word splitting and expansion principle: To ensure comprehensiveness, words that are not usually expressed continuously should be split. For example, Dongying Kehong Chemical Co., Ltd. can be split into: Dongying + Kehong Chemical.

[0071] Large word contraction method: Product brands cannot be split, such as "Joyoung Soymilk Maker".

[0072] (2) Construction of the Dissemination Influence Index Dimension Model: Dissemination reach: The breadth of dissemination dimension involves one secondary indicator (total online buzz) and seven tertiary indicators. These seven dissemination channels are equally important and therefore have equal weight. Dimensional scores are calculated using a five-level system and obtained through the Delphi method.

[0073] The consultation invited experts from various industries, including communications, branding, culture and tourism, and finance. During the process, experts were required to evaluate the volume of different information disseminations and provide scores based on their industry experience, judgment criteria, and relevance to their research directions. These scores served as support for the scientific validity of the quantification of communication volume. The more familiar the experts were with the indicators, the more scientific their methods, and the closer their research directions, the more reliable their quantification of communication volume would be.

[0074] The scoring rule is a maximum score of 5 points. Each expert scores based on the overall online buzz, and the weighted average is taken.

[0075] The propagation extent corresponding to different sound levels was obtained by weighted averaging, and curve fitting was performed based on this. The result conforms to a logarithmic function. =0.983, indicating that the model can explain 98.3% of the variation, and the F test is significant.

[0076] The breadth of dissemination is shown in the following formula: ; in, To spread the score, For the logarithmic coefficient, , This is a constant term. The fitted image is shown in Figure 1.

[0077] Popularity: The popularity dimension involves two secondary indicators and four tertiary indicators. To ensure the rationality of the dimension evaluation and reduce the influence of interfering factors, the index innovatively uses dynamic weights. (Weibo repost rate) The value range is [0,1], and the forwarding volume Compared with original hair volume There is a positive correlation. Therefore, the formula for the forwarding score in the spread popularity is as follows: ; in, To earn points for forwarding, For forwarding rate, It's about the amount of information on Weibo. It is related to forwarding weight. , This is a constant term, related to the industry sector being evaluated.

[0078] Compared to Weibo reposts, comments come from various platforms across the internet, with a comment rate of [missing information]. Affected by the total amount of information on the entire network The impact is greater. Therefore, the comment score is: ; in, To score the comments, For comment rate, It is a comment-related weight. It's the overall online buzz. , This is a constant term, related to the industry sector being evaluated.

[0079] The popularity score is calculated by summing the forwarding score and the commenting score, with a value range of [0,5]. The formula is as follows: ; in, To spread the buzz, For coefficient terms, This is a constant term.

[0080] Online attention: The online attention dimension involves 6 secondary indicators and 7 tertiary indicators, with data sourced from different big data platforms. To avoid interference from indicator data of different dimensions and magnitudes on the evaluation results, the evaluation indicators were logarithmically transformed, which also eliminated model heteroscedasticity and reduced collinearity to some extent.

[0081]

[0082] i is 4-9 It is a constant and is related to the industry sector being evaluated.

[0083] Depending on the industry, the applicable indicators fall into three categories:

[0084] The Delphi method was used to assess the importance of search metrics for each industry through scoring by multiple industry experts. Simultaneously, based on the correlation between search keywords and industries / evaluation units, the industry search coefficient is determined. and the search coefficient for each evaluation unit. .

[0085] The formula for online attention is as follows: ; Specifically, online attention for consumer brands includes the Black Cat Complaint indicator, which is deducted as a negative value (i=4, 5, 6, 7, 8); for scenic spot brands, it includes the comment growth indicator (i=4, 5, 6, 7, 9); and for other brands, i=4, 5, 6, 7. Search weight varies depending on the brand type. There are some differences.

[0086] Presentation level on proprietary platforms: The self-owned platform presentation dimension involves 6 secondary indicators and 12 tertiary indicators, including the number of keywords included. It contains 7 tertiary indicators, which are first weighted and summed. Considering the correlation and variability among the indicators, the CRITIC weighting method is used for calculation.

[0087] The formula for the CRITIC weighting method is: , Information carrying capacity It is the product of contrast strength and impact. .

[0088] Contrast intensity is the magnitude of the differences among various evaluation values ​​of the same indicator, calculated in the form of standard deviation:

[0089] Conflict The degree of correlation between different indicators , It is the correlation coefficient between indicator i and indicator j.

[0090] Table 7 Weight Calculation Results

[0091] therefore, ; After dimensionless processing of the six obtained Level 2 indicators, their weights are determined according to industry type. The presentation formula of the proprietary platform is as follows: ; in, To increase the weight of included keywords, Weighting based on the amount of information included. For the relevant weight of the official account, For video account related weight, , , , , This is a constant term.

[0092] Market activity: The market activity dimension involves four secondary indicators and four tertiary indicators. First, platform data is filtered and aggregated to obtain category data and brand data. To ensure the data exhibits consistent trends, sales volume is standardized using the reciprocal method to calculate market share, and then a function is fitted.

[0093] Regarding the market share of a certain brand's top-selling products on Tmall, ; in, It refers to the top-selling products in the category on Tmall. This refers to the sales volume of a single brand's top-selling products on Tmall. This refers to the number of units sold for each product of the brand on the platform. This refers to the number of units sold for each product category on the platform.

[0094] Taking some home appliance brands as examples, the relationship between product categories and brands is shown in the table below.

[0095]

[0096] Choosing the power function, the adjusted R-squared is 0.963, indicating that the independent variables can explain 96.3% of the variation in the dependent variable. The p-value of the F-value is less than 0.05, indicating that the combination of explanatory variables in the model has a significant impact on the explained variable, and all coefficients are also significant.

[0097] Both JD.com and offline stores use the same calculation method, and both are modeled using logarithmic functions. In particular, some brands include multiple sub-categories, and in the calculation, a weighted average is performed using the monthly sales of each category as the weight.

[0098] ; ; in, This refers to the sales volume of this product category on JD.com. This refers to the brand's sales volume on JD.com. It represents the offline sales revenue of the product category. It refers to the brand's offline sales revenue.

[0099] The final market activity formula is: ; in, , , These are the weighting coefficients for Tmall, JD.com, and offline sales, respectively. This is a constant term.

[0100] Brand reputation: Brand reputation involves 2 secondary indicators and 5 tertiary indicators. Overall score. The value range is [0, 5], and the positive review rate is... The value range is [0,1], and the two indicators are ultimately weighted at 0.8 and 1 respectively to be included in the brand reputation score.

[0101] .

[0102] Fifth, the construction of the overall score model for the communication influence index: The INC index total score is first calculated based on the original data to limit the values ​​of each sub-index to a stable and reasonable range. Then, it is calculated by addition (the volume is still compressed into a standardized volume by taking the square root). Finally, it is logarithmically compressed. The theoretical limit of the total score is 1000.

[0103] INC indices cover different dimensions in different industries

[0104] Introducing the original value of INC This is the unlog-standardized INC score, calculated using the following formula, which varies depending on the industry: 5.3.1 Consumer Brands ; Therefore, ； in, The original value coefficient of INC. constant 5.3.2 Scenic Area Brand ; Where i = 4, 5, 6, 7, 9; Therefore ， ； in, It is a constant.

[0105] 5.3.3 Catering and Hotel Brands ; Where i = 4, 5, 6, 7, This represents the brand reputation coefficient.

[0106] Therefore, , It is a constant.

[0107] 5.3.4 Other Brands ; Where i = 4, 5, 6, 7 Therefore, , It is a constant.

[0108] VI. Application of the Online Communication Influence Index To make the evaluation results more intuitive and easier to understand, each of the six dimensions is divided into a five-level rating system, with the following meanings:

[0109] Since the official release of the Internet Communication Influence Index in 2020, the monthly ranking has been updated for four consecutive years, covering more than 30 industry sectors and more than 20,000 brand units, and its scientific nature, objectivity and accuracy have been fully verified.

[0110] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for constructing a network dissemination influence index model, characterized in that, Includes the following steps: S1. Construct the INC index indicator system and conduct correlation analysis; The INC index indicator system is constructed, setting six primary indicators: breadth of dissemination, popularity of dissemination, online attention, presentation on proprietary platforms, market activity, and brand reputation; the breadth of dissemination is calculated separately for each indicator. Popularity of dissemination Online attention Presentation level on proprietary platforms Market activity Brand reputation The Pearson correlation coefficients among the secondary indicators under these six primary indicators; S2. Construct a calculation model for six primary indicators; S3. Construct the total score model of the Communication Influence Index: The total score of the INC index is first calculated based on the original data to limit the values ​​of each sub-index to a reasonable range. Then, it is calculated by addition and finally logarithmic compression is performed to obtain the theoretical limit value of the total score.

2. The method for constructing a network communication influence index model according to claim 1, characterized in that, The secondary indicator of the breadth of dissemination is The third-level indicators include the amount of information on Weibo. WeChat information volume Total number of clients ,website ,forum ,video and digital newspaper ; ； ； in, To spread the score, For the logarithmic coefficient, , For constant terms; Equal to the total volume That is, the sum of the information content of the seven tertiary indicators.

3. The method for constructing a network dissemination influence index model according to claim 1, characterized in that, Popularity The secondary indicators include forwarding rate. and comment rate Among them, the forwarding rate Including Weibo information volume Weibo reposts Comment rate include and the total number of comments across the entire network Forwarding rate and comment rate The calculation formula is as follows: ； ； The formula for the forwarding rate score in the spread popularity is as follows: ； in, To earn points for forwarding, For forwarding rate, It's about the amount of information on Weibo. It is related to forwarding weight. , For constant terms; The formula for the comment rate score is as follows: ； in, To score the comments, For comment rate, It is a comment-related weight. It's the overall online buzz. , For constant terms, The popularity score is calculated by summing the forwarding score and the commenting score, with a value range of [0,5]. The formula is as follows: ； in, To determine the spread of popularity score, For coefficient terms, This is a constant term.

4. The method for constructing a network dissemination influence index model according to claim 1, characterized in that, Online attention includes six secondary indicators and seven tertiary indicators; the secondary indicators are Baidu search... TikTok search Headline Search Popularity Black Cat Complaints and the increase in comments The third-level indicators include Baidu Index. TikTok search Headline Search Popularity Black Cat Complaints Ctrip's review volume this month And Ctrip's review volume last month ; The formula for calculating the secondary indicators of online attention is as follows: ; i It is 4-9. It is a constant and is related to the industry sector being evaluated. The formula for online attention is as follows: ； Industry search coefficient and the search coefficient for each evaluation unit. , For search weight, i It is 4-9.

5. The method for constructing a network dissemination influence index model according to claim 1, characterized in that, The self-owned platform presentation dimension involves six secondary indicators and twelve tertiary indicators; the secondary indicators include the number of keywords included. The amount of information collected Publication volume on WeChat official accounts WeChat Official Account Headline Reading Count Number of video works and video likes The third-level indicators include Baidu PC. Baidu Mobile Sogou PC Sogou Mobile 360PC 360 Mobile Shenma January Collection Publication volume on WeChat official accounts WeChat Official Account Headline Reading Count Number of video works and video likes ; Number of keywords included It includes seven third-level indicators, including Baidu PC. Baidu Mobile Sogou PC Sogou Mobile 360PC 360 Mobile Shenma First, perform a weighted summation. ， j =10,...,15; Information carrying capacity It is the product of contrast strength and impact. ; Contrast intensity is the magnitude of the differences among various evaluation values ​​of the same indicator, calculated in the form of standard deviation: i =10,...,15; Conflict The degree of correlation between different indicators , It is an indicator i With indicators j The correlation coefficient between them; ； After dimensionless processing of the six secondary indicators, their weights are determined according to industry type. The presentation formula of the proprietary platform is as follows: ； in, To increase the weight of included keywords, Weighting based on the amount of information included. For the relevant weight of the official account, For video account related weight, , , , , This is a constant term.

6. The method for constructing a network dissemination influence index model according to claim 1, characterized in that, Market activity includes the secondary indicator of JD.com's brand product market share. Top-selling products on Tmall and offline sales share The third-level indicator includes the volume of products sold on JD.com. Top-selling products on Tmall Single item sold on Tmall and offline sales ; Sales volume was standardized using the reciprocal method to calculate market share, and then a function was fitted. The formula for calculating the market share of top products on Tmall is as follows: ； in, It refers to the top-selling products in the category on Tmall. This refers to the sales volume of a single brand's top-selling products on Tmall. This refers to the number of units sold for each product of the brand on the platform. This refers to the number of units sold for each product category on the platform. JD.com's brand product market share The calculation formula is as follows: ； Offline sales share The calculation formula is as follows: ； in, This refers to the sales volume of this product category on JD.com. This refers to the brand's sales volume on JD.com. It represents the offline sales revenue of the product category. It refers to the brand's offline sales revenue; ； in, , , These are the weighting coefficients, This is a constant term.

7. The method for constructing a network dissemination influence index model according to claim 1, characterized in that, Brand reputation includes two secondary indicators and five tertiary indicators; the secondary indicators include the overall score. and positive review rate ; The third-level indicators include the overall score. Number of reviews this month Number of positive reviews this month Number of reviews last month Compared to last month's positive reviews ; Overall rating The value range is [0, 5], and the positive review rate is... The value range is [0,1]. The formula for calculating brand reputation is as follows: 。 8. The method for constructing a network dissemination influence index model according to claim 1, characterized in that, Based on experience, the industry is divided into four categories: consumer brands, scenic spot brands, catering and hotels, and others. The original value of INC is then introduced. This is the unlog-standardized INC score, calculated using the following formula, which varies depending on the industry: Consumer Brands The overall score model for the communication influence index is as follows: ； ； in, The original value coefficient of INC. It is a constant; Scenic spot brand category The overall score model for the communication influence index is as follows: ； in i =4, 5, 6, 7, 9; Therefore, , It is a constant; Catering and Hotel The overall score model for the communication influence index is as follows: ； in i =4, 5, 6, 7, Brand reputation coefficient; , It is a constant; The overall score model for other brand communication influence indices is as follows: ； in i =4, 5, 6, 7; , It is a constant.

Images