Method for measuring urban space vlog phenomenon

Abstract

The application discloses a kind of urban space net red phenomenon determination method, it is related to urban space technical field, including the following steps: collection short video check-in data, obtain the online traffic and offline traffic of multiple net red check-in points;Based on the ratio of offline traffic and online traffic, obtain the traffic conversion rate of multiple net red check-in points in the preset research range;The surrounding area of net red check-in point is researched, and the influence factor of urban space net red phenomenon is determined and corresponding data is obtained;Based on the traffic conversion rate of net red check-in point and each influence factor, linear regression model is established respectively, and multiple regression analysis equation is obtained;The weight of each influence factor is obtained by using principal component regression analysis method, and weight analysis result is obtained.The application can analyze the correlation between urban net red phenomenon and influence factor, help to develop urban space potential, and help to promote the development of urban space net red phenomenon by optimizing space quality targetedly.

Description

Technical Field

[0001] This invention relates to the field of urban space technology, and more specifically to a method for measuring the phenomenon of urban spaces becoming "internet famous". Background Technology

[0002] The information technology revolution, represented by the internet, has permeated every corner of today's society, changing the public's production, lifestyle, thinking, and behavior. These changes in lifestyle will further lead to changes in urban physical spaces. In the internet environment, the way information is carried has transformed; online media has become an important means for the public to understand the world, and traditional urban public spaces have gained richer ways of being perceived.

[0003] Short videos, as a rich and widely accepted information carrier, have enriched and made diverse information more intuitive, optimized the way information is delivered, attracted a large number of users, and given rise to a new form of recreation: short video check-ins. The public has transformed from a mere recipient of information to an integrated role of information creation, dissemination, and sharing. The visual consumption information of urban spaces is continuously disseminated through online platforms. After browsing related content online, the public conducts offline check-ins and continues to share on online platforms. In this cycle, the overall popularity of urban spaces and the number of offline visitors continuously increase. This process is considered the phenomenon of urban spaces becoming "internet famous."

[0004] As a visual consumption space that receives high attention from the mobile internet, "Instagrammable" spots provide a new perspective for urban space research. Identifying the influencing factors of the phenomenon of "Instagrammable" urban spaces can help develop the potential of urban waterfront spaces, and optimizing spatial quality can help promote the development of the "Instagrammable" phenomenon in urban spaces. However, there is currently very little research in this area, and it is still impossible to accurately judge the degree of "Instagrammable" urban spaces.

[0005] Therefore, determining the measurement indicators and influencing factors of the phenomenon of urban space becoming a "celebrity city" and accurately measuring this phenomenon based on big data from mobile internet platforms are technical problems that urgently need to be solved by those skilled in the art. Summary of the Invention

[0006] In view of this, the present invention provides a method for measuring the phenomenon of urban space becoming a "celebrity city," which can analyze the correlation between the phenomenon and influencing factors, help to develop the potential of urban space, and optimize spatial quality in a targeted manner to promote the development of the phenomenon of urban space becoming a "celebrity city."

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] A method for measuring the phenomenon of urban spaces becoming "Instagrammable" includes the following steps:

[0009] Collect short video check-in data to obtain online and offline traffic from multiple popular check-in locations;

[0010] The conversion rate of traffic to multiple popular online check-in spots within a pre-defined research scope was obtained based on the ratio of offline traffic to online traffic.

[0011] The study investigates the areas surrounding popular online photo spots to identify the influencing factors of urban space becoming a "Instagrammable" phenomenon and obtains relevant data.

[0012] Linear regression models were established based on the traffic conversion rate of popular online check-in spots and various influencing factors, and multiple regression analysis equations were obtained.

[0013] Principal component regression analysis was used to obtain the weights of each influencing factor, and the weight analysis results were obtained.

[0014] The technical effects achieved by the above technical solution are as follows: using traffic conversion rate as a metric, the relationship between the phenomenon of urban internet celebrity and influencing factors can be studied. Based on the weight analysis results, the influencing factors that have a significant impact on traffic conversion rate can be identified, which helps to optimize spatial quality in a targeted manner and promote the development of the phenomenon of urban internet celebrity.

[0015] Optionally, online traffic is the number of people who view the check-in video. A check-in video being viewed by a user is considered as one view.

[0016] Offline traffic refers to the number of people who participate in offline check-ins; a user uploading a POI video is considered one check-in.

[0017] Optionally, collect short video check-in data, specifically including the following steps:

[0018] Using "city + name of short video check-in point" as keywords, search for short videos with corresponding tags on short video platforms, and paste the corresponding web page links into a spreadsheet to form a list of videos to be crawled;

[0019] The system uses Python to loop through the webpage links in the list of videos to be crawled, calls short video platform plugins and ffmpeg plugins to batch download the videos corresponding to the webpage links, and uses the videos to collect short video-related information, including the number of viewers, the number of people who checked in, the influence of the publisher, and the attention of the check-in points.

[0020] Optionally, the method further includes:

[0021] After obtaining the short video check-in data, the data is processed to identify and clean up duplicate data, product advertisements and marketing advertisements that are irrelevant to the check-in points.

[0022] Optionally, the specific method for confirming duplicate data is as follows:

[0023] The feature information of the first video and the second video to be verified is extracted respectively. The feature information includes image information and audio information obtained by taking screenshots of the first video and the second video at preset time intervals.

[0024] Calculate the image similarity between the image information of the first video and the image information of the second video, as well as the audio similarity between the audio information of the first video and the audio information of the second video; weight the image similarity and audio similarity to obtain the feature similarity between the feature information of the first video and the feature information of the second video.

[0025] When the feature similarity exceeds the first preset threshold, key frames of the first video and the second video are extracted respectively, and the similarity of the key frames is calculated.

[0026] When the similarity of keyframes exceeds a second preset threshold, the first video and the second video are determined to be duplicate data.

[0027] Optionally, the influencing factors of the phenomenon of urban spaces becoming "Instagrammable" are divided into online and offline dimensions. The offline dimension selects environmental and spatial factors surrounding the check-in points as evaluation factors, while the online dimension's impact on public spaces comes from information received and fed back by online users; specifically,

[0028] Environmental factors mainly consider the functional units that can affect the check-in activities, including factors that attract online celebrities, functional competitiveness, transportation accessibility, distance from residential areas, and the completeness of service facilities;

[0029] Spatial factors mainly consider the public's perception of the environment. Semantic recognition is used to analyze photos taken by tourists around the check-in points, and building ratio, green view rate, water accessibility, and landscape uniformity are selected as evaluation factors.

[0030] The influencing factors for the online dimension were selected as nighttime check-in rate, online interaction intensity, network influence, functional usability, and consumption intensity.

[0031] Optionally, obtain the multiple regression analysis equation, which specifically includes the following steps:

[0032] The data corresponding to each influencing factor are standardized, and the variance contribution rate of each principal component is obtained by using principal component regression analysis.

[0033] The variance contribution rates of each principal component are summed sequentially to obtain the explanatory rates of different numbers of principal components for explaining combinations of influencing factors.

[0034] Select a corresponding number of principal components based on a preset threshold, and obtain the component matrix of each influencing factor in the selected principal components and the eigenvalues ​​of the selected principal components.

[0035] Based on the component matrix and the eigenvalues ​​of the selected principal components, the coefficients of each influencing factor in the linear combination of the selected principal components and in the comprehensive score model are obtained, and a multiple regression analysis equation is established.

[0036] Optionally, the variance contribution rate and explained rate of each principal component can be obtained, specifically including the following steps:

[0037] Using X1, X2, ..., X respectively n Let X represent the vector composed of all the data collected for each impact factor. The n-dimensional vector formed by these n impact factors is X = (X1, X2, ..., X...). n ) T T represents the transpose of a vector;

[0038] Calculate the covariance matrix R:

[0039]

[0040] In the formula: R ij Indicates the impact factor X i With X j The correlation coefficients, i,j=1,2,...,n, and R ij =R ji The calculation formula is as follows:

[0041]

[0042] Solving the characteristic equation |λE-R|=0 using the covariance matrix R yields the eigenvalues ​​λ. i and corresponding to the eigenvalue λ i eigenvector e i ,Require Where e ij Represents vector e i The j-th component;

[0043] Calculate the variance contribution rate of the principal components:

[0044]

[0045] The formula for calculating the explanatory power of different numbers of principal components for the combination of influencing factors is as follows:

[0046]

[0047] Before calculating the explanatory rate, the variance contribution rates of each principal component are sorted by size, Q. p Let p be the explanatory power of the p principal components in explaining the combination of influencing factors.

[0048] Optionally, the coefficients of each influencing factor in the selected principal component linear combination and in the comprehensive score model are obtained, specifically including the following steps:

[0049] The corresponding component matrix in the principal components is:

[0050] Z = XE (5);

[0051] In the formula: Let R be the orthonormal eigenvalue matrix corresponding to the q non-negative eigenvalues ​​in the covariance matrix R. It is a matrix composed of q principal components;

[0052] Obtain the standardized values ​​of the component matrix and the square roots of the eigenvalues ​​of the selected principal components. Then, use the ratio of the standardized value of each selected principal component to the square root of its corresponding eigenvalue as the coefficient 'a' of each influencing factor in the linear combination of the selected principal components. i ;

[0053] The established multiple regression equation is expressed as follows:

[0054]

[0055] in:

[0056]

[0057] In the formula: α i β is a component of the first coefficient. i As a component of the second coefficient, Here, Y represents the coefficients of each influencing factor in the comprehensive scoring model, and Y represents the traffic conversion rate.

[0058] Optionally, obtain the weights of each influencing factor, specifically:

[0059]

[0060] As can be seen from the above technical solution, compared with the prior art, the present invention discloses a method for measuring the phenomenon of urban space becoming a "celebrity spot," which has the following beneficial effects:

[0061] (1) This invention uses traffic conversion rate as a metric to study the relationship between the phenomenon of urban internet celebrity and influencing factors. Based on the weight analysis results, we can understand the influencing factors that have a significant impact on traffic conversion rate, which helps to optimize spatial quality in a targeted manner and promote the development of the phenomenon of urban internet celebrity.

[0062] (2) This invention uses “city + name of short video check-in point” as keywords to collect videos of popular check-in points and obtain relevant information. Compared with the existing technology that uses the number of viewers or check-in points as the measurement index, this technical solution uses traffic conversion rate as the measurement index, which can better reflect the degree of the phenomenon of urban space becoming popular. In the process of confirming duplicate data, the similarity of the video is determined based on the image information and the sound information, and further determined whether it is a duplicate video based on the extracted key frames, which can improve the accuracy of video similarity judgment and improve the efficiency of video detection.

[0063] (3) This invention uses principal component analysis to obtain the weight analysis results of each influencing factor. By studying the short video check-in data, it improves the limitation of evaluating urban space by area. It can gain a deeper understanding of a certain point or a certain type of point in urban space, further explore urban spaces with high potential, cultivate distinctive functions, and contribute to the high-quality development of urban space. Attached Figure Description

[0064] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0065] Figure 1 A flowchart for measuring the phenomenon of urban spaces becoming "Instagrammable";

[0066] Figure 2 A schematic diagram illustrating the phenomenon of urban spaces becoming "Instagrammable" or "viral."

[0067] Figure 3 A flowchart for establishing the multiple regression analysis equation. Detailed Implementation

[0068] 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.

[0069] In the mobile internet era, technologies such as remote work and virtual reality tours have reduced the public's dependence on urban spaces. The ways in which the public participates in urban interaction have become more diversified. Urban spaces with strong visual consumption attributes are constantly being shared by social media, gaining significant attention on internet platforms. Some online users, in order to attract traffic, choose to film and post short videos of widely popular urban spaces to "drive traffic." While users are "driving traffic" for themselves, short videos on social media platforms also present the visual consumption attributes of urban spaces to other online users from different perspectives, thereby attracting another group of people to "check in" and "drive traffic" offline, forming a virtuous cycle. Based on the above analysis, and combined with existing research definitions of "internet-famous spaces," "internet-famous destinations," "internet-famous cities," and "internet-famous villages," this embodiment argues that the virtuous cycle of the continuous dissemination of visual consumption information of urban spaces through online platforms, attracting the public to check in offline and continue sharing the space's content, constitutes a phenomenon of urban space becoming "internet-famous." This phenomenon continuously enhances the vitality of urban spaces and increases the public's enthusiasm for participating in the evaluation of urban spaces, which is conducive to the high-quality development of urban spaces.

[0070] To study the phenomenon of urban spaces becoming "internet celebrity" sites, this invention discloses a method for measuring this phenomenon, such as... Figure 1 As shown, it includes the following steps:

[0071] Collect short video check-in data to obtain online and offline traffic from multiple popular check-in locations;

[0072] The conversion rate of traffic to multiple popular online check-in spots within a pre-defined research scope was obtained based on the ratio of offline traffic to online traffic.

[0073] The study investigates the areas surrounding popular online photo spots to identify the influencing factors of urban space becoming a "Instagrammable" phenomenon and obtains relevant data.

[0074] Linear regression models were established based on the traffic conversion rate of popular online check-in spots and various influencing factors, and multiple regression analysis equations were obtained.

[0075] Principal component regression analysis was used to obtain the weights of each influencing factor, and the weight analysis results were obtained.

[0076] The phenomenon of urban spaces becoming "internet sensations" represents a high degree of integration between cyberspace and physical space. Analysis must consider both the influence of online factors and the relevance of offline participants to online platforms. Therefore, measuring the phenomenon of urban spaces becoming "internet sensations" should include three parts: 1. Measuring online traffic to urban spaces; 2. Measuring offline traffic generated by online traffic; 3. Calculating traffic conversion rates and the factors influencing these rates. Among these, for example... Figure 2As shown, online traffic originates in cyberspace and is mainly generated by the public participating in online activities on social platforms; offline traffic originates in physical spaces and is generated by the public participating in offline activities.

[0077] Analysis and screening of social media platforms generating online traffic revealed that Douyin (TikTok), a platform that has emerged in recent years, features online-offline integration, and its short video check-in data is suitable for measuring the phenomenon of urban spaces becoming "internet celebrity" destinations. Short videos taken by the public on Douyin can carry POI (Point of Interest) information, accurately recording the geographical location where the video was filmed. The POI address is highly relevant to the content displayed in the video, effectively attracting users to save the address and subsequently check in at that location. As a carrier of urban spatial content, these short check-in videos, through various indicators such as the number of check-ins, the number of views on the videos, the influence of the video publisher, and the popularity of the check-in point, can relatively accurately reflect the influence and attractiveness of a city or a specific area.

[0078] Browsing short videos showcasing popular POIs (Points of Interest) can give the public some understanding of urban spaces, and some of these people choose to visit the trending spots featured in the videos. When people visit these spots, offline traffic is generated, and some of them continue to upload videos, creating a cyclical phenomenon of POI popularity. On Douyin (TikTok), uploading a POI video is considered a check-in, and a video being viewed by another user is considered a view. Douyin uses algorithms to recommend videos to users who may be interested in or about to visit the location; therefore, the number of check-ins and views for the same location on Douyin is highly correlated. Based on this analysis, this embodiment measures the phenomenon of urban space POI popularity by combining online traffic generated from users browsing videos on Douyin with offline traffic generated from users visiting these spots. The measurement formula is as follows:

[0079]

[0080] In the formula, P offline P represents the offline traffic generated by the check-in activity. online C represents the online traffic generated by the check-in activity, and C represents the ratio of traffic converted into offline traffic. See Table 1 for the measurement table.

[0081] Table 1. Measurement Table of Urban Waterfront Space Popularization

[0082]

[0083] Furthermore, the collection of short video check-in data includes the following steps:

[0084] Using "city + name of short video check-in point" as keywords, search for short videos with corresponding tags on short video platforms, and paste the corresponding web page links into a spreadsheet to form a list of videos to be crawled;

[0085] The program uses Python to loop through the webpage links in the list of videos to be crawled, calls short video platform plugins and ffmpeg plugins to batch download the videos corresponding to the webpage links, and uses the videos to collect short video-related information, including the number of viewers, the number of people who checked in, the influence of the publisher, and the attention of the check-in point.

[0086] Furthermore, to ensure the relevance of the short video data to the research content and to exclude irrelevant data: after obtaining the short video check-in data, the short video check-in data was processed to identify and clean up duplicate data, product advertisements and marketing advertisements unrelated to the check-in points that were initially collected.

[0087] Furthermore, the specific method for confirming duplicate data is as follows:

[0088] The feature information of the first video and the second video to be verified is extracted respectively. The feature information includes image information and audio information obtained by taking screenshots of the first video and the second video at preset time intervals.

[0089] Calculate the image similarity between the image information of the first video and the image information of the second video, as well as the audio similarity between the audio information of the first video and the audio information of the second video; weight the image similarity and audio similarity to obtain the feature similarity between the feature information of the first video and the feature information of the second video.

[0090] When the feature similarity exceeds the first preset threshold, key frames of the first video and the second video are extracted respectively, and the similarity of the key frames is calculated.

[0091] When the similarity of keyframes exceeds a second preset threshold, the first video and the second video are determined to be duplicate data.

[0092] This technical solution transforms video comparison into image and audio comparison, reducing the complexity of data processing and improving the efficiency of video detection. At the same time, it determines the similarity of videos based on image and audio information, and further determines whether they are duplicate videos based on extracted keyframes, thereby improving the accuracy of video similarity judgment and saving processing time.

[0093] In the mobile internet era, online information has become an important basis for the public's choice of recreational spaces in cities, while physical spaces remain a primary source of online information. Therefore, this embodiment selects the online and offline dimensions of urban spaces as influencing factors for the phenomenon of urban spaces becoming "internet famous."

[0094] The offline influencing factors mainly consider people's feelings and ability to choose recreational activities, selecting the surrounding environment and spatial factors of check-in points as evaluation factors. Environmental factors mainly consider functional units that can influence check-in activities, including factors related to the gathering of online celebrities, functional competitiveness, transportation accessibility, distance from residential areas, and the completeness of service facilities. Spatial factors mainly consider the public's perception of the environment, using semantic recognition to analyze photos taken by tourists around check-in points, selecting building ratio, green view ratio, water accessibility, and landscape evenness as evaluation factors. Taking into account the research data and the suitability of the indicators, as shown in Table 2, this embodiment establishes nine factors in two dimensions—environmental factors and spatial factors—as offline influencing factors for the phenomenon of urban waterfront space becoming a popular tourist destination.

[0095] Table 2 Summary of Offline Influencing Factors of the "Internet Celebrity" Phenomenon

[0096]

[0097] According to literature review, the impact of online factors on public spaces mainly stems from information received and fed back by online users. Online users learn about spatial characteristics through images and videos, understand functional features through e-commerce platforms, and further understand the space by recording interactions such as likes, favorites, and comments, as well as check-in times. Therefore, as shown in Table 3, the online factors contributing to the phenomenon of urban waterfront spaces becoming "Instagrammable" include nighttime check-in rate, online interaction intensity, online influence, functional practicality, and consumption intensity.

[0098] Table 3. Summary of Online Influencing Factors of the "Internet Celebrity" Phenomenon

[0099]

[0100]

[0101] like Figure 3 As shown, the multivariate regression equation is obtained, specifically through the following steps:

[0102] The data corresponding to each influencing factor are standardized, and the variance contribution rate of each principal component is obtained by using principal component regression analysis.

[0103] The variance contribution rates of each principal component are summed sequentially to obtain the explanatory rates of different numbers of principal components for explaining combinations of influencing factors.

[0104] Select a corresponding number of principal components based on a preset threshold, and obtain the component matrix of each influencing factor in the selected principal components and the eigenvalues ​​of the selected principal components.

[0105] Based on the component matrix and the eigenvalues ​​of the selected principal components, the coefficients of each influencing factor in the linear combination of the selected principal components and in the comprehensive score model are obtained, and a multiple regression analysis equation is established.

[0106] Furthermore, the variance contribution rate and explained rate of each principal component are obtained, specifically including the following steps:

[0107] Using X1, X2, ..., X respectively n Let X represent the vector composed of all the data collected for each impact factor. The n-dimensional vector formed by these n impact factors is X = (X1, X2, ..., X...). n ) T T represents the transpose of a vector;

[0108] Calculate the covariance matrix R:

[0109]

[0110] In the formula: R ij Indicates the impact factor X i With X j The correlation coefficients, i,j=1,2,...,n, and R ij =R ji The calculation formula is as follows:

[0111]

[0112] Solving the characteristic equation |λE-R|=0 using the covariance matrix R yields the eigenvalues ​​λ. i and corresponding to the eigenvalue λ i eigenvector e i ,Require Where e ij Represents vector e i The j-th component;

[0113] Calculate the variance contribution rate of the principal components:

[0114]

[0115] The formula for calculating the explanatory power of different numbers of principal components for the combination of influencing factors is as follows:

[0116]

[0117] Before calculating the explanatory rate, the variance contribution rates of each principal component are sorted by size, Q. p Let p be the explanatory power of the p principal components in explaining the combination of influencing factors.

[0118] Furthermore, the coefficients of each influencing factor in the selected principal component linear combination and in the comprehensive score model are obtained, specifically including the following steps:

[0119] The corresponding component matrix in the principal components is:

[0120] Z = XE (5);

[0121] In the formula: Let R be the orthonormal eigenvalue matrix corresponding to the q non-negative eigenvalues ​​in the covariance matrix R. It is a matrix composed of q principal components;

[0122] Obtain the standardized values ​​of the component matrix and the square roots of the eigenvalues ​​of the selected principal components. Then, use the ratio of the standardized value of each selected principal component to the square root of its corresponding eigenvalue as the coefficient 'a' of each influencing factor in the linear combination of the selected principal components. i ;

[0123] The established multiple regression equation is expressed as follows:

[0124]

[0125] in:

[0126]

[0127] In the formula: α i β is a component of the first coefficient. i As a component of the second coefficient, Here, Y represents the coefficients of each influencing factor in the comprehensive scoring model, and Y represents the traffic conversion rate.

[0128] Furthermore, the weights of each influencing factor are obtained, specifically:

[0129]

[0130] The phenomenon of urban spaces becoming "Instagrammable" reflects the public's subjective perception of urban spaces through short video platforms. As the number of people engaging with these spaces increases and the content becomes more in-depth, urban spaces gain public recognition and gradually shed their physical constraints. Short video check-in data overcomes the limitations of evaluating urban spaces on a surface-by-surface basis, allowing for deeper understanding of specific points or types of points within an urban space. For urban spaces where the "Instagrammable" phenomenon is in its early stages, identifying relevant influencing factors for similar check-in points enables targeted optimization of spatial quality and the cultivation of distinctive functions, contributing to the high-quality development of urban spaces.

[0131] The various embodiments described in this specification are presented in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for measuring the phenomenon of urban spaces becoming "internet celebrity" sites, characterized in that, Includes the following steps: Collect short video check-in data to obtain online and offline traffic from multiple popular check-in locations; The conversion rate of traffic to multiple popular online check-in spots within a pre-defined research scope was obtained based on the ratio of offline traffic to online traffic. The study investigates the areas surrounding popular online photo spots to identify the influencing factors of urban space becoming a "Instagrammable" phenomenon and obtains relevant data. Linear regression models were established based on the traffic conversion rate of popular online check-in spots and various influencing factors, and multiple regression analysis equations were obtained. Principal component regression analysis was used to obtain the weights of each influencing factor, and the weight analysis results were obtained. Online traffic refers to the number of people who view the check-in videos; a check-in video being viewed by a user is considered one view. Offline traffic refers to the number of people who participate in offline check-ins; a user uploading a POI video is considered one check-in. The influencing factors of the phenomenon of urban spaces becoming "Instagrammable" are divided into online and offline dimensions. The offline dimension uses environmental and spatial factors surrounding the check-in points as evaluation factors, while the online dimension's impact on public spaces stems from information received and fed back by online users; specifically... Environmental factors that can affect the check-in activity include factors such as the gathering of online celebrities, functional competitiveness, transportation accessibility, distance from residential areas, and the completeness of service facilities. Spatial factors take into account the public's perception of the environment. Semantic recognition is used to analyze photos taken by tourists around the check-in points, and building ratio, green view rate, water accessibility, and landscape uniformity are selected as evaluation factors. The influencing factors for the online dimension were selected as nighttime check-in rate, online interaction intensity, network influence, functional usability, and consumption intensity.

2. The method for measuring the phenomenon of urban spatial "internet celebrity" status according to claim 1, characterized in that, Collecting short video check-in data includes the following steps: Using "city + name of short video check-in point" as keywords, search for short videos with corresponding tags on short video platforms, and paste the corresponding web page links into a spreadsheet to form a list of videos to be crawled; The system uses Python to loop through the webpage links in the list of videos to be crawled, calls short video platform plugins and ffmpeg plugins to batch download the videos corresponding to the webpage links, and uses the videos to collect short video-related information, including the number of viewers, the number of people who checked in, the influence of the publisher, and the attention of the check-in points.

3. The method for measuring the phenomenon of urban space becoming a "celebrity spot" according to claim 1, characterized in that, The method further includes: After obtaining the short video check-in data, the data is processed to identify and clean up duplicate data, product advertisements and marketing advertisements that are irrelevant to the check-in points.

4. The method for measuring the phenomenon of urban space becoming a "celebrity spot" according to claim 3, characterized in that, The specific method for confirming duplicate data is as follows: The feature information of the first video and the second video to be verified is extracted respectively. The feature information includes image information and audio information obtained by taking screenshots of the first video and the second video at preset time intervals. Calculate the image similarity between the image information of the first video and the image information of the second video, as well as the audio similarity between the audio information of the first video and the audio information of the second video; weight the image similarity and audio similarity to obtain the feature similarity between the feature information of the first video and the feature information of the second video. When the feature similarity exceeds the first preset threshold, key frames of the first video and the second video are extracted respectively, and the similarity of the key frames is calculated. When the similarity of keyframes exceeds a second preset threshold, the first video and the second video are determined to be duplicate data.

5. The method for measuring the phenomenon of urban spatial "internet celebrity" status according to claim 1, characterized in that, Obtaining the multiple regression equation involves the following steps: The data corresponding to each influencing factor are standardized, and the variance contribution rate of each principal component is obtained by using principal component regression analysis. The variance contribution rates of each principal component are summed sequentially to obtain the explanatory rates of different numbers of principal components for explaining combinations of influencing factors. Select a corresponding number of principal components based on a preset threshold, and obtain the component matrix of each influencing factor in the selected principal components and the eigenvalues ​​of the selected principal components. Based on the component matrix and the eigenvalues ​​of the selected principal components, the coefficients of each influencing factor in the linear combination of the selected principal components and in the comprehensive score model are obtained, and a multiple regression analysis equation is established.

6. The method for measuring the phenomenon of urban spatial "internet celebrity" status according to claim 5, characterized in that, The variance contribution rate and explained rate of each principal component are obtained, which includes the following steps: Use respectively This represents a vector composed of all the data collected for each influencing factor. n Composed of several influencing factors n dimensional vector is T represents the transpose of a vector; Calculate the covariance matrix R : （1）； In the formula: Indicating the impact factor and The correlation coefficient, ,and The calculation formula is as follows: （2）； Based on the covariance matrix R For the characteristic equation Solve the problem to obtain the eigenvalues. and corresponding to eigenvalues eigenvectors ,Require ,in Representing vectors The j One component; Calculate the variance contribution rate of the principal components: （3）； The formula for calculating the explanatory power of different numbers of principal components for the combination of influencing factors is as follows: （4）； Before calculating the explanatory rate, the variance contribution rates of each principal component are sorted by size. for p The principal components explain the explanatory power of the combination of influencing factors.

7. The method for measuring the phenomenon of urban spatial "internet celebrity" status according to claim 6, characterized in that, Obtaining the coefficients of each influencing factor in the selected principal component linear combination and in the comprehensive score model involves the following steps: The corresponding component matrix in the principal components is: （5）； In the formula: Covariance matrix R In q The orthonormal eigenma matrix corresponding to each non-negative eigenvalue. for q A matrix composed of principal components; Obtain the standardized values ​​of the component matrix and the square roots of the eigenvalues ​​of the selected principal components. Then, use the ratio of the standardized value of each selected principal component to the square root of its corresponding eigenvalue as the coefficient of each influencing factor in the linear combination of the selected principal components. ; The established multiple regression equation is expressed as follows: （6）； in: （7）； In the formula: As a component of the first coefficient, As a component of the second coefficient, These are the coefficients of each influencing factor in the comprehensive scoring model. Y Traffic conversion rate.

8. The method for measuring the phenomenon of urban space becoming a "celebrity spot" according to claim 7, characterized in that, The weights of each influencing factor are obtained as follows: （8）。

Images