A news event exposure amount evaluation method and system based on internet dissemination

Abstract

This invention relates to a method and system for evaluating the exposure of news events based on internet dissemination, comprising the following steps: S1: Automatically acquiring media content data related to the target news event from multiple internet platforms and preprocessing it; S2: Based on the preprocessed data, employing a multi-dimensional fusion weighted model to comprehensively calculate the overall exposure of the news event, integrating search engine data, social media interaction data, and sampled data from closed platforms; S3: After calculation, the exposure results are stored in time-series format and displayed through a visual interface. This invention more comprehensively reflects the true breadth of news dissemination and social influence in the complex internet ecosystem, possessing strong practicality and scalability.

Description

Technical Field

[0001] This invention relates to the field of intelligent statistics, and in particular to a method and system for evaluating the exposure of news events based on internet dissemination. Background Technology

[0002] In the internet age, the speed and reach of news events have expanded rapidly. Traditional assessment methods based on clicks or views on a single platform are no longer sufficient to fully reflect their social impact, and a complete method and system for calculating exposure is lacking. Currently, the industry often measures news popularity by statistically analyzing access data from single platforms (such as news websites and social media), which has the following shortcomings: First, it only focuses on direct access, ignoring secondary dissemination behaviors such as forwarding and citation across multiple platforms, resulting in incomplete exposure statistics; second, it lacks in-depth analysis of the dissemination path, as the influence and exposure vary across different platforms, meaning the number of times related news is seen by the public differs; third, inconsistent data standards across different platforms make cross-platform integrated calculations difficult, affecting the objectivity and accuracy of the assessment results. These shortcomings prevent existing technologies from scientifically and comprehensively quantifying the true exposure level of news events. Summary of the Invention

[0003] To address the aforementioned issues, the present invention aims to provide a method and system for evaluating the exposure of news events based on internet dissemination. This method comprehensively reflects the true breadth of news dissemination and its social impact within the complex internet ecosystem, and possesses strong practicality and scalability.

[0004] To achieve the above objectives, the present invention adopts the following technical solution:

[0005] A method for evaluating the exposure of news events based on internet dissemination includes the following steps:

[0006] S1: Automatically acquire media content data related to the target news event from multiple internet platforms and preprocess it;

[0007] S2: Based on the preprocessed data, a multi-dimensional fusion weighted model is adopted to calculate the overall exposure of news events by integrating search engine data, social media interaction data, and closed platform sampling data.

[0008] S3: After the calculation is completed, the exposure results are stored in time series form and displayed through a visualization interface.

[0009] Furthermore, media content data related to the target news event is automatically obtained from multiple internet platforms, as detailed below:

[0010] Users can configure the list of media sites to be monitored through the management interface; the collection account supports the configuration of multiple legally authorized third-party platform API access accounts to achieve compliant collection of content protected by permissions;

[0011] It also sets core keywords, extended thesaurus of synonyms, event aliases, related people, places, and organization names for the target news event, forming a multi-dimensional search strategy. The keywords support Boolean logic combination and fuzzy matching algorithms.

[0012] The task scheduling mechanism based on cron expressions supports the automatic start of data scraping tasks on a minute, hour, or day basis. After the data scraping task is executed, the raw data is cleaned, deduplicated, and formatted and then stored in a structured database.

[0013] Further preprocessing is performed as follows:

[0014] The collected text content, including web page text, social media posts, and short video titles and descriptions, is processed by word segmentation, stop word filtering, and entity recognition to extract structured fields such as news topic, publication time, publishing media, author, number of views, number of likes, number of comments, and number of reposts.

[0015] Clustering of reports from different sources based on semantic similarity models determines whether they point to the same news event, thus avoiding duplicate counting.

[0016] Based on a pre-set media classification database, each report is automatically categorized as a central-level authoritative media outlet, a provincial-level mainstream media outlet, a commercial portal website, a self-media account, or a social media user.

[0017] Furthermore, a multi-dimensional weighted fusion model is adopted, integrating search engine data, social media interaction data, and sampled data from closed platforms to calculate the overall exposure of news events, as detailed below:

[0018] S21: Send an HTTP request to the search engines, using a unified keyword combination to query the number of related reports for the target news event. The query platforms include Baidu, Sogou, and Bing Chinese version. To reduce noise interference, the query syntax uses exact matching and time range limitation.

[0019] The overall search engine index E is defined as follows:

[0020] E=max(B,S,G)+0.3* (B+S+G- max(B,S,G));

[0021] Where B represents the number of results returned by Baidu search; S represents the number of results returned by Sogou search; and G represents the number of results returned by Bing search.

[0022] S22: For each search result in E, by identifying the domain name and matching the media attribute database, calculate the proportion of traditional authoritative media and new media, and split E into two parts: RN and RO.

[0023] S23: Define the enhancement coefficient C1 to determine whether the target news event has been reported by authoritative central-level media.

[0024] S24: Extract the top N most popular blog posts related to this news event from new media sampling platforms and extract their interaction data: number of likes. Number of reposts Number of comments Calculate the average number of interactions per blog post:

[0025] ;

[0026] Define the social media activity coefficient C2 = max(I,5), meaning that if the average interaction volume is lower than 5, then take the base value of 5 to ensure basic activity.

[0027] S25: For closed content ecosystem platforms that cannot be effectively indexed by search engines, use API interfaces or compliant crawling methods to directly collect their publicly available data, collect all articles / micro-headlines matching keywords, obtain the number of views and publications for each piece of content, and define its equivalent exposure T as:

[0028] Among them, read j M represents the actual number of views for the j-th article; M represents the total number of matched articles.

[0029] Based on the data from the above dimensions, this invention defines the total exposure of a news event as:

[0030] Total = RO*C1 + RN*C2 + T.

[0031] Furthermore, the visual interface displays information including exposure trend charts, pie charts showing the distribution of media sources, analysis of the contribution ratio of trending platforms, a list of key reports and links, supports exporting PDF reports or API interfaces for third-party systems to use, and sets a threshold alarm mechanism to automatically trigger public opinion warnings when the exposure of a news event surges in a short period of time.

[0032] An internet-based news event exposure assessment system includes a data acquisition module, a data processing and analysis module, an exposure calculation engine module, and a result output and visualization module. The data acquisition module automatically acquires media content data related to the target news event from multiple internet sources. The data processing and analysis module preprocesses the acquired data. The exposure calculation engine module, based on the preprocessed data, uses a multi-dimensional fusion weighted model to integrate search engine data, social media interaction data, and closed platform sampling data to calculate the overall exposure of the news event. After the result output and visualization module completes the calculation, it stores the exposure results in time-series format and displays them through a visualization interface.

[0033] Furthermore, the data acquisition module is as follows: Users can configure a list of media sites to be monitored through the management interface; the acquisition account supports configuring multiple legally authorized third-party platform API access accounts to achieve compliant acquisition of content protected by permissions; and the core keywords, synonyms, event aliases, related people, locations, and organization names of target news events can be set to form a multi-dimensional search strategy. Keywords support Boolean logic combination and fuzzy matching algorithms; the task scheduling mechanism based on cron expressions supports automatic start of data crawling tasks on a minute, hour, or day cycle. After the acquisition task is executed, the raw data is cleaned, deduplicated, and formatted before being stored in a structured database.

[0034] Further, the data processing and analysis module performs the following: It segments, filters stop words, and identifies entities in the collected text content, including webpage text, social media posts, and short video titles and descriptions, extracting structured fields such as news topic, publication time, publishing media, author, number of views, number of likes, number of comments, and number of reposts; it clusters reports from different sources based on a semantic similarity model to determine whether they refer to the same news event, avoiding duplicate counting; and it automatically labels each report as published by a pre-defined media classification database, categorizing it as a central-level authoritative media outlet, provincial mainstream media, commercial portal website, self-media account, or social media user.

[0035] Furthermore, the exposure calculation engine module integrates search engine data, social media interaction data, and sampled data from closed platforms to calculate the overall exposure of news events, as detailed below:

[0036] S21: Send an HTTP request to the search engines, using a unified keyword combination to query the number of related reports for the target news event. The query platforms include Baidu, Sogou, and Bing Chinese version. To reduce noise interference, the query syntax uses exact matching and time range limitation.

[0037] The overall search engine index E is defined as follows:

[0038] E=max(B,S,G)+0.3* (B+S+G- max(B,S,G));

[0039] Where B represents the number of results returned by Baidu search; S represents the number of results returned by Sogou search; and G represents the number of results returned by Bing search.

[0040] S22: For each search result in E, by identifying the domain name and matching the media attribute database, calculate the proportion of traditional authoritative media and new media, and split E into two parts: RN and RO.

[0041] S23: Define the enhancement coefficient C1 to determine whether the target news event has been reported by authoritative central-level media.

[0042] S24: Extract the top N most popular blog posts related to this news event from new media sampling platforms and extract their interaction data: number of likes. Number of reposts Number of comments Calculate the average number of interactions per blog post:

[0043] ;

[0044] Define the social media activity coefficient C2 = max(I,5), meaning that if the average interaction volume is lower than 5, then take the base value of 5 to ensure basic activity.

[0045] S25: For closed content ecosystem platforms that cannot be effectively indexed by search engines, use API interfaces or compliant crawling methods to directly collect their publicly available data, collect all articles / micro-headlines matching the keywords, obtain the number of views and publications for each piece of content, and define its equivalent exposure T as:

[0046] Among them, read j M represents the actual number of views for the j-th article; M represents the total number of matched articles.

[0047] Based on the data from the above dimensions, this invention defines the total exposure of a news event as:

[0048] Total = RO*C1 + RN*C2 + T.

[0049] Furthermore, the results output and visualization module displays exposure trend charts, media source distribution pie charts, hot platform contribution ratio analysis, key report lists and link jumps. It supports exporting PDF reports or API interfaces for third-party systems to call, and sets a threshold alarm mechanism to automatically trigger public opinion warnings when the exposure of a news event surges in a short period of time.

[0050] The present invention has the following beneficial effects:

[0051] This invention comprehensively acquires information on the dissemination of news events across multiple channels through a cross-platform data collection and fusion model, effectively overcoming the shortcomings of incomplete statistics from a single platform. By integrating search engine data and social media interaction data, it constructs a multi-source, heterogeneous, and hierarchical quantitative model for news event exposure. Compared to traditional methods of statistically analyzing single readership or search volume, this solution more comprehensively reflects the true breadth of news dissemination and social influence in the complex internet ecosystem. It has strong practicality and scalability, and is applicable to scenarios such as government public opinion monitoring, corporate brand communication evaluation, and media influence analysis. Attached Figure Description

[0052] Figure 1 This is a system architecture diagram of the present invention. Detailed Implementation

[0053] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments:

[0054] refer to Figure 1 In this embodiment, a method for evaluating the exposure of news events based on internet dissemination is provided, including the following steps:

[0055] S1: Automatically acquire media content data related to the target news event from multiple internet platforms and preprocess it;

[0056] S2: Based on the preprocessed data, a multi-dimensional fusion weighted model is adopted to calculate the overall exposure of news events by integrating search engine data, social media interaction data, and closed platform sampling data.

[0057] S3: After the calculation is completed, the exposure results are stored in time series form and displayed through a visualization interface.

[0058] In this embodiment, media content data related to the target news event is automatically obtained from multiple internet sources, as detailed below:

[0059] Users can configure the list of media sites to be monitored through the management interface, including but not limited to Weibo, WeChat official accounts, Toutiao, Kuaishou, Douyin, Zhihu, Bilibili, Sohu News, NetEase News, Tencent News, and other mainstream new media and social platforms. The collection account supports the configuration of multiple legally authorized third-party platform API access accounts (such as Weibo developer accounts, WeChat developer accounts, Toutiao accounts, etc.) to achieve compliant collection of content protected by permissions.

[0060] It also sets core keywords, extended thesaurus of synonyms, event aliases, related people, places, and organization names for the target news event, forming a multi-dimensional search strategy. Keywords support Boolean logic combination (AND / OR / NOT) and fuzzy matching algorithms (such as edit distance and pinyin similarity).

[0061] The task scheduling mechanism based on cron expressions supports the automatic start of data scraping tasks on a minute, hour, or day basis. After the data scraping task is executed, the raw data is cleaned, deduplicated, and formatted and then stored in a structured database.

[0062] In this embodiment, the preprocessing is as follows:

[0063] The collected text content, including web page text, social media posts, and short video titles and descriptions, is processed by word segmentation, stop word filtering, and new entity recognition (NER) to extract structured fields such as news topic, publication time, publishing media, author, number of views, number of likes, number of comments, and number of reposts.

[0064] Clustering of reports from different sources based on semantic similarity models (such as BERT-Siamese Network or Sentence-BERT) determines whether they point to the same news event and avoids double counting.

[0065] Based on a pre-set media classification database, each report is automatically categorized as a central-level authoritative media outlet (such as People's Daily Online, Xinhua Net, and CCTV.com), a provincial mainstream media outlet, a commercial portal website, a self-media account, or a social media user.

[0066] In this embodiment, a multi-dimensional fusion weighted model is adopted, which integrates search engine data, social media interaction data, and closed platform sampling data to calculate the overall exposure of news events, as detailed below:

[0067] S21: Send an HTTP request to the search engines, using a unified keyword combination to query the number of related reports for the target news event. The query platforms include Baidu, Sogou, and Bing Chinese version. To reduce noise interference, the query syntax uses exact matching and time range limitation.

[0068] The overall search engine index E is defined as follows:

[0069] E=max(B,S,G)+0.3* (B+S+G- max(B,S,G));

[0070] Where B represents the number of results returned by Baidu search; S represents the number of results returned by Sogou search; and G represents the number of results returned by Bing search.

[0071] S22: For each search result in E, by identifying the domain name and matching the media attribute database, calculate the proportion of traditional authoritative media and new media, and split E into two parts: RN and RO.

[0072] RN: The proportion of new media reports in E.

[0073] RO: The proportion of traditional media coverage in E.

[0074] Based on empirical modeling, new media has a fast dissemination speed, strong interactivity, and more exposure, accounting for about 60%; traditional media has strong credibility but a longer dissemination chain and lower interactivity, accounting for about 40%.

[0075] Therefore, E is split into two parts: RN = 60%E, RO = 40%E.

[0076] S23: Define the enhancement coefficient C1, which is used to determine whether the target news event has been reported by authoritative central-level media (specifically People's Daily Online, Xinhua Net, and CCTV Net);

[0077] This judgment is made by matching the normalized report dataset in the data acquisition module;

[0078] The enhancement coefficient C1 is defined as follows: If none of the three media outlets report on it, then C1 = 5; if only one media outlet reports on it, then C1 = 10; if two media outlets report on it, then C1 = 20; if all three media outlets report on it, then C1 = 30. This coefficient reflects the authority of central-level media and the secondary reprint effect it triggers. Being reported by central-level media will significantly increase the exposure potential of an event.

[0079] S24: Extract the top N most popular blog posts (N is configurable, e.g., N=50) related to this news event from new media sampling platforms, and extract their interaction data: number of likes. Number of reposts Number of comments Calculate the average number of interactions per blog post:

[0080] ;

[0081] Define the social media activity coefficient C2 = max(I,5), meaning that if the average interaction volume is lower than 5, then take the base value of 5 to ensure basic activity.

[0082] S25: For closed content ecosystem platforms that cannot be effectively indexed by search engines (such as Toutiao, Douyin, Kuaishou, and WeChat Official Accounts), use API interfaces or compliant crawling methods to directly collect their publicly available data. Taking Toutiao as an example, collect all articles / micro-blogs matching keywords, obtain the number of views and publications for each piece of content, and define its equivalent exposure T as:

[0083] Among them, read j M represents the actual number of views for the j-th article; M represents the total number of matched articles.

[0084] Based on the data from the above dimensions, this invention defines the total exposure of a news event as:

[0085] Total = RO*C1 + RN*C2 + T.

[0086] In this embodiment, the visualization interface includes an exposure trend chart (by hourly / daily granularity), a pie chart of media source distribution, analysis of the contribution ratio of hot platforms, a list of key reports and link jumps, supports exporting PDF reports or API interfaces for third-party systems to call, and sets a threshold alarm mechanism. When the exposure of a news event surges in a short period of time (such as a growth rate exceeding 200% / hour), a public opinion warning is automatically triggered.

[0087] An internet-based news event exposure assessment system includes a data acquisition module, a data processing and analysis module, an exposure calculation engine module, and a result output and visualization module. The data acquisition module automatically acquires media content data related to the target news event from multiple internet sources. The data processing and analysis module preprocesses the acquired data. The exposure calculation engine module, based on the preprocessed data, uses a multi-dimensional fusion weighted model to integrate search engine data, social media interaction data, and closed platform sampling data to calculate the overall exposure of the news event. After the result output and visualization module completes the calculation, it stores the exposure results in time-series format and displays them through a visualization interface.

[0088] In this embodiment, the data acquisition module is specifically as follows: Users can configure a list of media sites to be monitored through the management interface, including but not limited to Weibo, WeChat official accounts, Toutiao, Kuaishou, Douyin, Zhihu, Bilibili, Sohu News, NetEase News, and Tencent News, which are mainstream new media and social platforms; the acquisition account supports the configuration of multiple legally authorized third-party platform API access accounts (such as Weibo developer accounts, WeChat developer accounts, Toutiao accounts, etc.) to achieve compliant acquisition of content protected by permissions; and the core keywords, synonym extended thesaurus, event aliases, related people, locations, and organization names of the target news event are set to form a multi-dimensional search strategy. Keywords support Boolean logic combination (AND / OR / NOT) and fuzzy matching algorithms (such as edit distance and pinyin approximation); the task scheduling mechanism based on cron expressions supports automatic start of data crawling tasks on a minute, hour, or day cycle. After the acquisition task is executed, the raw data is cleaned, deduplicated, and formatted and then stored in a structured database.

[0089] In this embodiment, the data processing and analysis module specifically performs the following: It performs word segmentation, stop word filtering, and entity recognition (NER) on the collected text content, including webpage text, social media posts, and short video titles and descriptions, extracting structured fields such as news topic, publication time, publishing media, author, number of views, number of likes, number of comments, and number of reposts; it clusters reports from different sources based on semantic similarity models (such as BERT-Siamese Network or Sentence-BERT) to determine whether they point to the same news event, avoiding duplicate counting; and it automatically labels each report as published by a pre-set media classification database, categorizing it as a central-level authoritative media (such as People's Daily Online, Xinhua Net, and CCTV.com), provincial mainstream media, commercial portal websites, self-media accounts, or social media users.

[0090] In this embodiment, the exposure calculation engine module integrates search engine data, social media interaction data, and closed platform sampling data to calculate the overall exposure of a news event, as detailed below:

[0091] S21: Send an HTTP request to the search engines, using a unified keyword combination to query the number of related reports for the target news event. The query platforms include Baidu, Sogou, and Bing Chinese version. To reduce noise interference, the query syntax uses exact matching and time range limitation.

[0092] The overall search engine index E is defined as follows:

[0093] E=max(B,S,G)+0.3* (B+S+G- max(B,S,G));

[0094] Where B represents the number of results returned by Baidu search; S represents the number of results returned by Sogou search; and G represents the number of results returned by Bing search.

[0095] S22: For each search result in E, by identifying the domain name and matching the media attribute database, calculate the proportion of traditional authoritative media and new media, and split E into two parts: RN and RO.

[0096] RN: The proportion of new media reports in E.

[0097] RO: The proportion of traditional media coverage in E.

[0098] Based on empirical modeling, new media has a fast dissemination speed, strong interactivity, and more exposure, accounting for about 60%; traditional media has strong credibility but a longer dissemination chain and lower interactivity, accounting for about 40%.

[0099] Therefore, E is split into two parts: RN = 60%E, RO = 40%E.

[0100] S23: Define the enhancement coefficient C1, which is used to determine whether the target news event has been reported by authoritative central-level media (specifically People's Daily Online, Xinhua Net, and CCTV Net);

[0101] This judgment is made by matching the normalized report dataset in the data acquisition module;

[0102] The enhancement coefficient C1 is defined as follows: If none of the three media outlets report on it, then C1 = 5; if only one media outlet reports on it, then C1 = 10; if two media outlets report on it, then C1 = 20; if all three media outlets report on it, then C1 = 30. This coefficient reflects the authority of central-level media and the secondary reprint effect it triggers. Being reported by central-level media will significantly increase the exposure potential of an event.

[0103] S24: Extract the top N most popular blog posts (N is configurable, e.g., N=50) related to this news event from new media sampling platforms, and extract their interaction data: number of likes. Number of reposts Number of comments Calculate the average number of interactions per blog post:

[0104] ;

[0105] Define the social media activity coefficient C2 = max(I,5), meaning that if the average interaction volume is lower than 5, then take the base value of 5 to ensure basic activity.

[0106] S25: For closed content ecosystem platforms that cannot be effectively indexed by search engines (such as Toutiao, Douyin, Kuaishou, and WeChat Official Accounts), use API interfaces or compliant crawling methods to directly collect their publicly available data. Taking Toutiao as an example, collect all articles / micro-blogs matching keywords, obtain the number of views and publications for each piece of content, and define its equivalent exposure T as:

[0107] Among them, read j M represents the actual number of views for the j-th article; M represents the total number of matched articles.

[0108] Based on the data from the above dimensions, this invention defines the total exposure of a news event as:

[0109] Total = RO*C1 + RN*C2 + T.

[0110] In this embodiment, the results output and visualization module displays an exposure trend chart (by hourly / daily granularity), a pie chart showing the distribution of media sources, an analysis of the contribution ratio of trending platforms, a list of key reports, and link jumps. It supports exporting PDF reports or API interfaces for third-party systems to use, and sets a threshold alarm mechanism. When the exposure of a news event surges in a short period of time (e.g., the growth rate exceeds 200% / hour), a public opinion warning is automatically triggered.

[0111] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0112] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0113] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0114] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0115] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. An Internet-based news event exposure evaluation method, characterized in that, Includes the following steps: S1: Automatically acquire media content data related to the target news event from multiple internet platforms and preprocess it; S2: Based on the preprocessed data, a multi-dimensional fusion weighted model is adopted to calculate the overall exposure of news events by integrating search engine data, social media interaction data, and closed platform sampling data. S3: After the calculation is completed, the exposure results are stored in time series form and displayed through a visualization interface; The method employs a multi-dimensional weighted fusion model, integrating search engine data, social media interaction data, and sampled data from closed platforms to calculate the overall exposure of news events, as detailed below: S21: Send an HTTP request to the search engines, using a unified keyword combination to query the number of related reports for the target news event. The query platforms include Baidu, Sogou, and Bing Chinese version. To reduce noise interference, the query syntax uses exact matching and time range limitation. The overall search engine index E is defined as follows: E=max(B,S,G)+0.3 (B+S+G- max(B,S,G)); Where B represents the number of results returned by Baidu search; S represents the number of results returned by Sogou search; and G represents the number of results returned by Bing search. S22: For each search result in E, by identifying the domain name and matching the media attribute database, calculate the proportion of traditional authoritative media and new media, and split E into two parts: RN and RO. S23: Define the enhancement coefficient C1 to determine whether the target news event has been reported by authoritative central-level media. S24: Extract the top N hot blog posts related to the news event from the new media sampling platform, and extract the interaction data: number of likes , number of forwards , number of comments , and calculate the average total number of interactions per blog post: ; Define the social media activity coefficient C2 = max(I,5), meaning that if the average interaction volume is lower than 5, then take the base value of 5 to ensure basic activity. S25: For closed content ecosystem platforms that cannot be effectively indexed by search engines, use API interfaces or compliant crawling methods to directly collect their publicly available data, collect all articles / micro-headlines matching keywords, obtain the number of views and publications for each piece of content, and define its equivalent exposure T as: ; wherein read j is the actual reading volume of the jth article; M is the total number of matching articles; Based on the data from the above dimensions, the total exposure of a news event is defined as: Total=RO C1+RN C2+T。 2. The method of claim 1, wherein, The automatic acquisition of media content data related to the target news event from multiple internet sources is as follows: Users can configure the list of media sites to be monitored through the management interface; the collection account supports the configuration of multiple legally authorized third-party platform API access accounts to achieve compliant collection of content protected by permissions; It also sets core keywords, extended thesaurus of synonyms, event aliases, related people, places, and organization names for the target news event, forming a multi-dimensional search strategy. The keywords support Boolean logic combination and fuzzy matching algorithms. The task scheduling mechanism based on cron expressions supports the automatic start of data scraping tasks on a minute, hour, or day basis. After the data scraping task is executed, the raw data is cleaned, deduplicated, and formatted and then stored in a structured database. 3.The Internet-based news event exposure evaluation method according to claim 2, characterized in that, The preprocessing is as follows: The collected webpage text, social media posts, and short video titles and descriptions are processed by word segmentation, stop word filtering, and entity recognition to extract structured fields such as news topic, publication time, publishing media, author, number of views, number of likes, number of comments, and number of reposts. Clustering of reports from different sources based on semantic similarity models determines whether they point to the same news event, thus avoiding duplicate counting. Based on a pre-set media classification database, each report is automatically categorized as a central-level authoritative media outlet, a provincial-level mainstream media outlet, a commercial portal website, a self-media account, or a social media user.

4. The method of claim 1, wherein, The visualization interface includes an exposure trend chart, a pie chart of media source distribution, an analysis of the contribution ratio of hot platforms, a list of key reports and links, supports exporting PDF reports or API interfaces for third-party systems to call, and sets a threshold alarm mechanism to automatically trigger a public opinion warning when the exposure of a news event surges in a short period of time.

5. An Internet-based news event exposure evaluation system, characterized in that, It includes a data acquisition module, a data processing and analysis module, an exposure calculation engine module, and a result output and visualization module; the data acquisition module automatically acquires media content data related to the target news event from multiple Internet sources; the data processing and analysis module preprocesses the acquired data; The exposure calculation engine module uses a multi-dimensional fusion weighted model based on preprocessed data to calculate the overall exposure of news events by integrating search engine data, social media interaction data, and closed platform sampling data. After the results are output and the visualization model is calculated, the exposure results are stored in time series form and displayed through a visualization interface. The exposure calculation engine module integrates search engine data, social media interaction data, and closed platform sampling data to calculate the overall exposure of news events, as detailed below: S21: Send an HTTP request to the search engines, using a unified keyword combination to query the number of related reports for the target news event. The query platforms include Baidu, Sogou, and Bing Chinese version. To reduce noise interference, the query syntax uses exact matching and time range limitation. The overall search engine index E is defined as follows: E = max(B, S, G) + 0.3 (B + S + G - max(B, S, G)); Where B represents the number of results returned by Baidu search; S represents the number of results returned by Sogou search; and G represents the number of results returned by Bing search. S22: For each search result in E, by identifying the domain name and matching the media attribute database, calculate the proportion of traditional authoritative media and new media, and split E into two parts: RN and RO. S23: Define the enhancement coefficient C1 to determine whether the target news event has been reported by authoritative central-level media. S24: Extract the top N most popular blog posts related to this news event from new media sampling platforms and extract their interaction data: number of likes. Number of reposts Number of comments Calculate the average number of interactions per blog post: ; Define the social media activity coefficient C2 = max(I,5), meaning that if the average interaction volume is lower than 5, then take the base value of 5 to ensure basic activity. S25: For closed content ecosystem platforms that cannot be effectively indexed by search engines, use API interfaces or compliant crawling methods to directly collect their publicly available data, collect all articles / micro-headlines matching the keywords, obtain the number of views and publications for each piece of content, and define its equivalent exposure T as: ; wherein read j is the actual reading volume of the jth article; M is the total number of matching articles; Based on the data from the above dimensions, the total exposure of a news event is defined as: Total=RO C1+RN C2+T。 6. The Internet-based news event exposure evaluation system according to claim 5, wherein, The data acquisition module is as follows: users can configure a list of media sites to be monitored through the management interface; the acquisition account supports the configuration of multiple legally authorized third-party platform API access accounts to achieve compliant acquisition of content protected by permissions; It sets core keywords, synonyms, event aliases, related people, locations, and organization names for target news events to form a multi-dimensional search strategy. Keywords support Boolean logic combination and fuzzy matching algorithms. Based on the cron expression-based task scheduling mechanism, it supports automatic data crawling tasks to be started on a minute, hour, or day basis. After the data collection task is executed, the raw data is cleaned, deduplicated, and formatted and then stored in a structured database.

7. The system according to claim 5, wherein, The data processing and analysis module specifically performs word segmentation, stop word filtering, and entity recognition on the collected web page text, social media posts, short video titles, and description text content, and extracts structured fields such as news topic, publication time, publishing media, author, number of views, number of likes, number of comments, and number of reposts. Based on a semantic similarity model, reports from different sources are clustered to determine whether they point to the same news event, thus avoiding duplicate counting. According to a pre-set media classification database, each report is automatically labeled as a category published by a central-level authoritative media, a provincial mainstream media, a commercial portal website, a self-media account, or a social media user. 8.The Internet-based news event exposure evaluation system according to claim 5, wherein, The results output and visualization module displays exposure trend charts, media source distribution pie charts, hot platform contribution ratio analysis, key report lists and link jumps. It supports exporting PDF reports or API interfaces for third-party systems to call, and sets a threshold alarm mechanism to automatically trigger public opinion warnings when the exposure of a news event surges in a short period of time.

Images