A delivery analysis method and system based on content characteristics and communication effects

Abstract

The present invention relates to the technical field of content analysis, and specifically discloses a delivery analysis method and system based on content characteristics and communication effects, the method comprising: obtaining multiple delivery contents based on target keywords; conducting multi-dimensional delivery effect analysis to generate a comprehensive delivery effect score; conducting content characteristic analysis on multiple delivery contents to generate a content quality score; obtaining the communication data of multiple delivery contents, building a communication network based on the communication data, inputting the communication network into a pagerank model, and outputting a communication effect score; generating a delivery analysis report based on the comprehensive delivery effect score, content quality score, and communication effect score. By analyzing multi-dimensional indicators such as delivery volume, category, time distribution, and reinvestment ratio, a comprehensive delivery effect score is generated based on a standardization and weight allocation mechanism, which can evaluate the overall performance of the delivery strategy, provide delivery effect feedback, and provide systematic decision support to improve delivery efficiency and return rate.

Description

Technical Field

[0001] The present invention relates to the technical field of content analysis, and in particular to a delivery analysis method and system based on content characteristics and communication effects. Background Art

[0002] With the rapid development of digital marketing and advertising, advertisers and content publishers are paying more and more attention to the effect analysis of advertising, content quality evaluation and the mining of communication influence. For example, social platforms such as WeChat official accounts have become important channels for the dissemination of advertising content, and the demand for analysis of delivery strategies for these platforms is particularly important. However, the existing technology lacks multi-dimensional comprehensive evaluation in the analysis of delivery effect. The current delivery effect analysis focuses on the data statistics of a single dimension such as the number of readings and the number of likes, while ignoring the comprehensive influence of multiple dimensions such as the delivery time period, category, and re-delivery ratio. In addition, the weight distribution and evaluation methods between these dimensions lack standardization, making it difficult to accurately measure the overall effect of the delivery content. At the same time, the content quality analysis method has limitations. In terms of content analysis, most existing technologies only focus on whether the article is original, but lack in-depth analysis of the attractiveness of the article title, the rationality of the content structure, and the interactive behavior of readers. This limitation makes it difficult for advertisers to accurately evaluate the content quality and thus optimize the delivery strategy. The depth of the analysis of the communication effect after delivery is insufficient. For the evaluation of the communication effect, most current technologies only focus on the direct dissemination range of the article (such as the number of forwardings, the distribution of the reading population, etc.), and lack a global analysis of the dissemination path. In addition, existing technologies fail to fully explore the mutual influence between communication nodes in communication network modeling, nor do they provide an effective scoring model to quantify the communication effect, making it difficult to provide advertisers with comprehensive communication optimization suggestions. In summary, although existing analysis tools can achieve partial analysis, they lack the ability to integrate these analysis results into one, making it difficult to generate a systematic delivery analysis report to provide a reliable basis for optimization strategies. Summary of the invention

[0003] In view of this, the present invention proposes a delivery analysis method and system based on content characteristics and communication effects, which can effectively analyze the delivery content from multiple dimensions to obtain effect analysis and suggestions for delivery. The present invention provides the following technical solutions:

[0004] A delivery analysis method based on content characteristics and communication effects, the method comprising: obtaining a plurality of delivery contents based on set target keywords; performing multi-dimensional delivery effect analysis on the plurality of delivery contents and generating a comprehensive delivery effect score; performing content characteristic analysis on the plurality of delivery contents and generating a content quality score; obtaining communication data of the plurality of delivery contents and constructing a communication network based on the communication data, inputting the communication network into a pagerank model and outputting a communication effect score; generating a delivery analysis report according to the comprehensive delivery effect score, the content quality score and the communication effect score.

[0005] Optionally, the method for obtaining multiple delivery contents based on set target keywords includes: obtaining the set target keywords and performing synonym expansion through natural language processing to obtain a keyword set; obtaining multiple content sources and their API interfaces, and obtaining content data through the API interface; obtaining delivery content data from content data through direct matching, semantic matching, context matching and priority matching of keywords; and performing content filtering and deduplication on the delivery content data to obtain multiple delivery contents.

[0006] Optionally, the method for performing multi-dimensional delivery effect analysis on the plurality of delivery contents and generating a comprehensive delivery effect score includes: identifying basic delivery index data, spatiotemporal index data and industry category data of the delivery contents; performing delivery volume and reinvestment ratio analysis on the delivery contents from the basic delivery index dimension to obtain a basic delivery index score; performing delivery effect analysis on the delivery contents on weekdays and weekends from the spatiotemporal index dimension to obtain a spatiotemporal index delivery score; analyzing the delivery contents from the industry category dimension to obtain an industry category delivery score; and generating a comprehensive delivery effect score through the basic delivery index score, the spatiotemporal index delivery score and the industry category delivery score.

[0007] Optionally, the method for generating a comprehensive score of delivery effect through the basic delivery index score, the spatiotemporal index delivery score and the industry category delivery score includes: performing data cleaning on the basic delivery index score, the spatiotemporal index delivery score and the industry category delivery score; performing min-max normalization on the cleaned scores of each dimension to normalize the scores to the interval [0, 1]; and calculating the comprehensive score of the normalized scores of each dimension using the weighted average method, and the formula is: ;in, is the dimension count, For the comprehensive rating, For the The rating of the dimension, For the The weight of the dimension score.

[0008] Optionally, the method of performing content characteristic analysis on the plurality of delivered contents and generating content quality scores includes: performing multi-dimensional content characteristic analysis on the delivered contents in terms of originality, graphic structure, external link references, title characteristics and content richness; wherein the originality characteristic analysis includes: performing content comparison with an existing data set based on a content similarity algorithm, and the originality score is: ,in, is the repetition rate, and its value range is [0,1]. Among them, the analysis of the graphic structure characteristics includes: calculating the ratio of the number of pictures to the number of paragraphs, and setting an ideal ratio range, and deducting points from the content that does not meet the ideal ratio range based on the basic total score to obtain a graphic structure score. Among them, the analysis of the external link reference characteristics includes: checking the number of external links and the credibility of their sources, taking the product of the number of external links on a single platform and the credibility of the platform as the external link score of the platform, and adding up multiple external link scores of a single content as the external link reference score of the content. Among them, the title feature analysis includes: using a machine learning model to analyze the attractiveness of the title, scoring based on the title feature construction of historically highly disseminated content and obtaining a title feature score. Among them, the content richness feature analysis includes: calculating the vocabulary diversity and logical hierarchy of the content through natural language processing and obtaining a content richness score; calculating a comprehensive score based on the originality score, graphic structure score, external link reference score, title feature score and content richness score to generate a content quality score.

[0009] Optionally, the method of obtaining the propagation data of the plurality of delivered contents, constructing a propagation network based on the propagation data, inputting the propagation network into a pagerank model, and outputting a propagation effect score comprises: the propagation data comprises reading data, sharing data, comment interaction data, and external citation data; performing data cleaning and data aggregation on the propagation data, and generating a propagation characteristic matrix:

[0010] , where R ij To deliver content To the delivery content The propagation association weight between , ; Take each delivered content as a network node, and the edge weight between network nodes is the communication influence between delivered content. The calculation formula of edge weight is: ,in, To deliver content Associated content The normalized value of the number of shares. To deliver content Content being delivered The normalized value of the number of citations, To deliver content Interaction data and delivery content The normalized value of the number of associations, , , is the weight coefficient, and + + =1; construct a directed graph through network nodes and edge weights, and combine the propagation characteristic matrix to construct a propagation network; construct a transfer matrix P according to the edge weights of the propagation network: ,in, For network nodes To network node The propagation weight between is the sum of the out-degree weights of network node i; the communication influence score of each delivered content is calculated iteratively :

[0011] ,

[0012] in, is the damping coefficient, For Node The communication influence For Node To Node The transition probability, For Node The sum of out-degree weights; iterative calculation to Converge to output the final communication effect score of each delivered content.

[0013] Optionally, the method for generating a delivery analysis report based on the comprehensive delivery effect score, content quality score and communication effect score includes: assigning adjustable weight coefficients to the comprehensive delivery effect score, content quality score and communication effect score respectively, and calculating the product of the comprehensive delivery effect score, content quality score and communication effect score with the corresponding weight coefficients and adding them to obtain a comprehensive score for each delivery content; obtaining the data of delivery effect analysis, content quality analysis and communication effect analysis in a single delivery content; ranking each delivery content according to the comprehensive score through a visualization tool, and loading the data of delivery effect analysis, content quality analysis and communication effect analysis of the corresponding delivery content in turn according to the ranking to generate a delivery analysis report.

[0014] The present invention further discloses a delivery analysis system based on content characteristics and communication effects, including: a data acquisition module, used to acquire multiple delivery contents based on set target keywords; a scoring module, used to perform multi-dimensional delivery effect analysis on the multiple delivery contents, and generate a comprehensive delivery effect score; also used to perform content characteristic analysis on the multiple delivery contents, and generate a content quality score; also used to acquire the communication data of the multiple delivery contents, and build a communication network based on the communication data, input the communication network into a pagerank model, and output a communication effect score; an analysis report generation module, used to generate a delivery analysis report according to the comprehensive delivery effect score, the content quality score and the communication effect score.

[0015] The present invention further discloses a computer-readable storage medium, wherein the storage medium stores a computer program, and the computer program implements the above method when executed by a processor.

[0016] The present invention further discloses an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above method when executing the program.

[0017] According to the technical solution of the present invention, by analyzing multi-dimensional indicators such as delivery volume, category, time period distribution, re-delivery ratio, and generating a comprehensive delivery effect score based on standardization and weight distribution mechanism, the overall performance of the delivery strategy can be comprehensively evaluated, and accurate delivery effect feedback can be provided to advertisers. At the same time, multi-dimensional content characteristic analysis methods such as originality, title attractiveness, and content structure rationality are introduced to generate content quality scores, which can effectively identify high-quality content features and help optimize content creation and delivery strategies. By constructing a dissemination network through dissemination data, and combining the PageRank model to quantify the importance of dissemination nodes and the dissemination effect of content, this solution can accurately evaluate the dissemination influence of content and provide advertisers with a scientific basis for optimizing dissemination paths. The delivery effect score, content quality score, and dissemination effect score are integrated to generate a comprehensive delivery analysis report containing delivery strategy optimization suggestions, providing advertisers with systematic decision-making support and improving delivery efficiency and return rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] For the purpose of illustration and not limitation, the present invention is now described in conjunction with the embodiments of the present invention and the accompanying drawings, in which:

[0019] Figure 1 It is a flowchart of a delivery analysis method based on content characteristics and communication effects in an embodiment of the present invention;

[0020] Figure 2 It is a structural diagram of a delivery analysis system based on content characteristics and communication effects in an embodiment of the present invention;

[0021] Figure 3 It is a schematic diagram of the structure of an electronic device in an embodiment of the present invention. DETAILED DESCRIPTION

[0022] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the implementation mode of the present application will be clearly and completely described below in conjunction with the drawings in the implementation mode of the present application. Obviously, the described implementation mode is only a part of the implementation mode of the present application, not all the implementation modes. Based on the implementation mode in the present application, all other implementation modes obtained by ordinary technicians in the field without creative work should fall within the scope of protection of the present application.

[0023] It should be noted that the terms "first", "second", etc. in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present application described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0024] In the present application, the terms "upper", "lower", "left", "right", "front", "back", "top", "bottom", "inner", "outer", "middle", "vertical", "horizontal", "lateral", "longitudinal" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings. These terms are mainly used to better describe the present application and its implementation methods, and are not used to limit the indicated devices, elements or components to have a specific orientation, or to be constructed and operated in a specific orientation.

[0025] In addition, some of the above terms may be used to express other meanings in addition to indicating orientation or positional relationship. For example, the term "on" may also be used to express a certain dependency or connection relationship in some cases. For those of ordinary skill in the art, the specific meanings of these terms in this application can be understood according to the specific circumstances. In addition, the term "plurality" should mean two or more than two.

[0026] It should be noted that, in the absence of conflict, the embodiments of the present application and the features in the embodiments can be combined with each other. The embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0027] refer to Figure 1This embodiment discloses a delivery analysis method based on content characteristics and communication effects, the method comprising:

[0028] S100: Acquire multiple delivery contents based on the set target keywords. Among them, the keywords are the basic data for searching the historical delivery records. The keywords can be divided into keywords in various fields and keywords at various levels, such as industry, product, brand, core keywords and extended keywords, etc., to ensure wide coverage of the search target. Further, after setting the target keywords, the target keywords are subjected to synonym expansion, synonym matching and keyword context analysis through natural language processing technology to generate a more comprehensive keyword set. The sources of delivery content include but are not limited to publicly available content platforms such as WeChat public accounts, Weibo, short video platforms (such as Douyin, Xiaohongshu), and news portals. After obtaining the platform information, the target content is obtained by calling a public API (such as the crawler interface of the WeChat public account platform, the query interface of the article aggregation platform) or customizing the crawler. In order to avoid repeated searches of historical data every time a search is performed, the delivery content search task is run on a daily basis in this embodiment, and the new content released on the day is obtained through keyword matching, and a comprehensive database is constructed based on this.

[0029] For keyword matching, after constructing a keyword set, the delivery content data is obtained from the content data through direct keyword matching, semantic matching, context matching and priority matching. The semantic matching is based on a semantic model (such as a BERT or GPT model) to analyze the semantic similarity between the content and the target keyword, and the threshold screens out highly relevant content. The context matching is to combine the context environment (such as the position of the keyword in the paragraph and its importance), and further optimize the matching accuracy through models such as TF-IDF and BM25. In one or more embodiments, the priority of core keywords and extended keywords can also be set, and the content matching the core keywords is preferentially screened; for extended keywords, the context weight and semantic relationship are combined for screening. After obtaining the target delivery content, the delivery content data is generated by using hash values ​​and content features (such as titles, text summaries) to generate a unique identifier, filter duplicate delivery content, and perform content filtering and deduplication. At the same time, the classifier is used to filter content that is not related to the delivery target (such as advertising content, paragraphs without substantive information, etc.) to obtain the final multiple delivery content. This embodiment exemplifies an example of searching for delivery content. For example, assuming the target keyword is "smart hardware", call the content interface of the WeChat public account and the news portal platform to obtain all articles published on the day. Filter out articles that contain "smart hardware" in the title or text through a direct matching strategy. Use a semantic model to further analyze the semantic relevance of the article (such as "smart devices" and "IoT hardware" are also included in the match). De-duplicate, filter noise, and perform multi-condition screening on the screened content to ultimately obtain a delivery content set that meets the requirements.

[0030] S200: Perform multi-dimensional delivery effect analysis on the plurality of delivery contents, and generate a comprehensive delivery effect score. Specifically, the method includes:

[0031] S210: Identify basic delivery index data, spatiotemporal index data, and industry category data of delivery content;

[0032] S220: Analyze the delivery volume and re-delivery ratio of the delivered content from the basic delivery index dimension to obtain the basic delivery index score. Delivery volume refers to the actual number of times the delivered content is actually released, and re-delivery ratio refers to the ratio of the same content being delivered again in different channels or time periods to evaluate its popularity.

[0033] S230: Analyze the delivery effect of the delivered content on weekdays and weekends from the dimension of time and space indicators to obtain the delivery score of time and space indicators. Specifically, analyze the difference in delivery effect of content in different time periods (weekdays and weekends), analyze the distribution characteristics of content delivery in a day (such as morning, noon and evening), explore the peak effect of time periods, and combine the impact of page layout analysis (such as headline, second or bottom position) on the effect.

[0034] S240: Analyze the delivered content from the perspective of industry categories to obtain industry category delivery scores, so as to classify the delivered content by industry or category (such as electronic products, education and training, etc.) and compare industry delivery trends and effect differences.

[0035] S250: Generate a comprehensive score for delivery effect through the basic delivery index score, spatiotemporal indicator delivery score and industry category delivery score.

[0036] Specifically, it includes: data cleaning of basic delivery index scores, spatiotemporal indicator delivery scores, and industry category delivery scores; min-max normalization of the cleaned scores of each dimension to normalize the scores to the [0, 1] interval. The normalization formula is:

[0037] , where X is the value of the original data, i.e. the actual value of a certain score (such as the amount of delivery, the value of the communication effect, etc.), min : The minimum value of the score among all samples, X max : The maximum value of the score value in all samples, X′: The standardized value. After the formula conversion, the original value is mapped to the interval [0, 1] to eliminate the influence of the indicator dimension or value range, which is convenient for the comparison and comprehensive calculation of different indicators.

[0038] The normalized scores of each dimension are used to calculate the comprehensive score of each delivery content using the weighted average method. The formula is:

[0039] ;

[0040] in, is the dimension count, For the comprehensive rating, For the The rating of the dimension, For the The weight of the dimension score.

[0041] Based on this, you can also use visualization tools such as bar charts, pie charts, radar charts, etc. to show the performance of each delivery content in different dimensions. Based on the distribution of comprehensive scores, you can identify delivery methods with outstanding or poor results. At the same time, you can convert the scoring results into actionable optimization plans, for example, generate horizontal (between different contents) and vertical (in different time periods) comparison reports of delivery effects. Combined with the scoring results, recommend efficient delivery time periods, suitable delivery locations, and high-quality content templates to users.

[0042] S300: Analyze the content characteristics of the plurality of delivered contents and generate a content quality score. In order to achieve a comprehensive analysis of the inherent characteristics of the delivered contents, this embodiment will evaluate the content quality from multiple perspectives, specifically:

[0043] Conduct a multi-dimensional content characteristic analysis of the originality, graphic structure, external link references, title characteristics and content richness of the delivered content;

[0044] The originality characteristic analysis includes: content comparison with existing data sets based on content similarity algorithms, that is, judging whether the content is original, and determining originality by similarity comparison with historical databases or content on the Internet. The originality score is: ,in, is the repetition rate, and its value range is [0,1].

[0045] The analysis of the graphic structure characteristics includes: calculating the ratio of the number of pictures to the number of paragraphs, and setting an ideal ratio range. For example, the ideal ratio range is 1:3-1:5. Points are deducted from the content that does not meet the ideal ratio range based on the basic total score to obtain a graphic structure score.

[0046] The analysis of external link citation characteristics includes: checking the number of external links and the credibility of their sources, taking the product of the number of external links on a single platform and the credibility of the platform as the external link score of the platform, and adding up multiple external link scores of a single content as the external link citation score of the content.

[0047] The title feature analysis includes: using a machine learning model to analyze the attractiveness of the title, constructing a title feature based on historically highly disseminated content, and obtaining a title feature score.

[0048] The content richness characteristic analysis includes: calculating the vocabulary diversity and logical hierarchy of the content through natural language processing and obtaining a content richness score.

[0049] A comprehensive score is calculated based on the originality score, the image and text structure score, the external link citation score, the title feature score and the content richness score to generate a content quality score. Specifically, weights are assigned to the scores of different dimensions and adjusted according to the importance of the content features. For example, the weight of originality is 40%, the weight of image and text structure is 20%, the weight of external link citation is 20%, the weight of title features is 10%, and the weight of content richness is 10%. Based on this, the calculation formula for the comprehensive score corresponding to each delivered content is: ,in, For a single dimension rating, is the corresponding weight.

[0050] S400: Acquire the propagation data of the multiple delivered contents, build a propagation network based on the propagation data, input the propagation network into the pagerank model, and output a propagation effect score. Specifically including:

[0051] The dissemination data includes reading data, sharing data, comment interaction data and external citation data;

[0052] The propagation data is cleaned and aggregated, and a propagation characteristic matrix is ​​generated:

[0053] , where R ij To deliver content To the delivery content The propagation association weight between , ;

[0054] Each delivered content is regarded as a network node. The edge weight between network nodes is the communication influence between delivered content. The calculation formula of edge weight is: ,in, To deliver content Associated content The normalized value of the number of shares. To deliver content Content being delivered The normalized value of the number of citations, To deliver content Interaction data and delivery content The normalized value of the number of associations, , , is the weight coefficient, and + + =1;

[0055] Construct a directed graph through network nodes and edge weights, and combine it with the propagation characteristic matrix to construct a propagation network;

[0056] Construct the transfer matrix P according to the edge weights of the propagation network: ,in, For network nodes To network node The propagation weight between is the sum of out-degree weights of network node i;

[0057] Iteratively calculate the communication influence score of each delivered content :

[0058] ,

[0059] in, is the damping coefficient, For Node The communication influence For Node To Node The transition probability, For Node The sum of out-degree weights; iterative calculation to Converge to output the final communication effect score of each delivery content, that is, .

[0060] After obtaining the communication effect score, The value is sorted to identify the content with the strongest communication effect. At the same time, the main communication paths and key nodes are displayed based on visualization tools. It can also analyze which content occupies a core position in the communication network and is suitable as the focus of subsequent delivery. For content with weaker communication effects, the network interruption points or communication path shortcomings are analyzed to make optimization suggestions.

[0061] S500: Generate a delivery analysis report based on the delivery effect comprehensive score, content quality score and communication effect score. That is, the delivery effect comprehensive score, content quality score and communication effect score calculated above are respectively assigned adjustable weight coefficients, and the product of the delivery effect comprehensive score, content quality score and communication effect score and the corresponding weight coefficients is calculated and added to obtain the comprehensive score of each delivery content. The calculation formula of the comprehensive score in this embodiment is:

[0062] ,in, is an adjustable weight parameter, and .

[0063] Extract the data of delivery effect analysis, content quality analysis and communication effect analysis from a single delivery content. Delivery effect analysis data include delivery volume distribution (time period, category, region), delivery effect comparison between weekdays and weekends, the impact of delivery position on conversion effect, statistical re-delivery ratio of delivery content, and analysis of common characteristics of high re-delivery content (such as title attractiveness and originality). Content quality analysis data include originality distribution analysis: charts show the ratio of original articles to non-original articles. Give scores and rankings to content with high originality. Comparison of single-image and multi-image effects: analyze the differences in reading, interaction and communication between single-image and multi-image. Content structure and attractiveness analysis: analyze the attractiveness characteristics of content based on title attractiveness score and paragraph structure. Communication effect analysis data include: Communication network analysis visualization: visualize the communication network to display the main communication paths and key nodes. Highlight the communication center nodes and the content with the greatest impact on communication. Communication effect ranking: list the delivery content with the greatest communication influence based on the communication effect score. Communication timeliness analysis: analyze the change trend of communication data over time and identify the peak time of communication. Through visualization tools, each delivery content is ranked according to the comprehensive score, and the delivery effect analysis, content quality analysis and communication effect analysis data of the corresponding delivery content are loaded in sequence according to the ranking to generate a delivery analysis report.

[0064] In summary, the delivery analysis method based on content characteristics and communication effects has been completed. Through the analysis of multi-dimensional indicators such as delivery volume, category, time distribution, and re-delivery ratio, and based on the standardization and weight distribution mechanism, a comprehensive delivery effect score is generated, which can comprehensively evaluate the overall performance of the delivery strategy and provide advertisers with accurate delivery effect feedback. At the same time, multi-dimensional content characteristic analysis methods such as originality, title attractiveness, and content structure rationality are introduced to generate content quality scores, which can effectively identify high-quality content characteristics and help optimize content creation and delivery strategies. By constructing a communication network through communication data and combining the PageRank model to quantify the importance of communication nodes and the communication effect of content, this solution can accurately evaluate the communication influence of content and provide advertisers with a scientific basis for optimizing the communication path. The delivery effect score, content quality score, and communication effect score are integrated to generate a comprehensive delivery analysis report containing delivery strategy optimization suggestions, providing advertisers with systematic decision support and improving delivery efficiency and return rate.

[0065] refer to Figure 2 This embodiment further discloses a delivery analysis system based on content characteristics and communication effects, including:

[0066] The data acquisition module 21 is used to obtain multiple delivery contents based on the set target keywords; specifically: obtain the set target keywords and perform synonym expansion through natural language processing to obtain a keyword set; obtain multiple content sources and their API interfaces, and obtain content data through the API interface; obtain delivery content data from content data through direct matching, semantic matching, context matching and priority matching of keywords; perform content filtering and deduplication on the delivery content data to obtain multiple delivery contents.

[0067] Scoring module 22 is used to perform multi-dimensional delivery effect analysis on the multiple delivery contents and generate a comprehensive delivery effect score; specifically, identify the basic delivery index data, spatiotemporal index data and industry category data of the delivery content; analyze the delivery volume and reinvestment ratio of the delivery content from the basic delivery index dimension to obtain the basic delivery index score; analyze the delivery effect of the delivery content on weekdays and weekends from the spatiotemporal index dimension to obtain the spatiotemporal index delivery score; analyze the delivery content from the industry category dimension to obtain the industry category delivery score; generate a comprehensive delivery effect score through the basic delivery index score, spatiotemporal index delivery score and industry category delivery score. It is also used to clean the basic delivery index score, spatiotemporal index delivery score and industry category delivery score; perform min-max normalization on the cleaned scores of each dimension to normalize the score to the [0, 1] interval; calculate the comprehensive score of the normalized scores of each dimension using the weighted average method, and the formula is:

[0068] ;

[0069] Where i is the dimension count, For the comprehensive rating, is the score of the i-th dimension, is the weight of the score of the i-th dimension. It is also used to analyze the content characteristics of the multiple delivered contents and generate content quality scores; it is also used to analyze the multi-dimensional content characteristics of the delivered contents in terms of originality, graphic structure, external link references, title characteristics and content richness; wherein the originality characteristic analysis includes: comparing the content with the existing data set based on the content similarity algorithm, and the originality score is: ,in, is the repetition rate, and its value range is [0,1]; wherein, the analysis of the graphic structure characteristics includes: calculating the ratio of the number of pictures to the number of paragraphs, and setting an ideal ratio range, and deducting points from the content that does not meet the ideal ratio range based on the basic total score to obtain a graphic structure score; wherein the analysis of the external link reference characteristics includes: checking the number of external links and the credibility of their sources, taking the product of the number of external links on a single platform and the credibility of the platform as the external link score of the platform, and adding up multiple external link scores of a single content as the external link reference score of the content; wherein the title feature analysis includes: using a machine learning model to analyze the attractiveness of the title, scoring based on the title feature construction of historically highly disseminated content and obtaining a title feature score; wherein the content richness feature analysis includes: calculating the vocabulary diversity and logical hierarchy of the content through natural language processing and obtaining a content richness score; and calculating a comprehensive score based on the originality score, graphic structure score, external link reference score, title feature score and content richness score to generate a content quality score. It is also used to obtain the propagation data of the multiple delivered contents, build a propagation network based on the propagation data, input the propagation network into the pagerank model, and output a propagation effect score; it is also used to, the propagation data includes reading data, sharing data, comment interaction data and external citation data; perform data cleaning and data aggregation on the propagation data, and generate a propagation characteristic matrix:

[0070] , where R ij To deliver content To the delivery content The propagation association weight between , ; Take each delivered content as a network node, and the edge weight between network nodes is the communication influence between delivered content. The calculation formula of edge weight is: ,in, To deliver content Associated content The normalized value of the number of shares. To deliver content Content being delivered The normalized value of the number of citations, To deliver content Interaction data and delivery content The normalized value of the number of associations, , , is the weight coefficient, and + + =1; construct a directed graph through network nodes and edge weights, and combine the propagation characteristic matrix to construct a propagation network; construct a transfer matrix P according to the edge weights of the propagation network: ,in, For network nodes To network node The propagation weight between For network nodes The sum of the out-degree weights; the communication influence score of each delivered content is calculated iteratively :

[0071] ,

[0072] in, is the damping coefficient, For Node The communication influence For Node To Node The transition probability, For Node The sum of out-degree weights; iterative calculation to Converge to output the final communication effect score of each delivered content.

[0073] The analysis report generation module 23 is used to generate a delivery analysis report based on the delivery effect comprehensive score, content quality score and communication effect score. It is also used to assign adjustable weight coefficients to the delivery effect comprehensive score, content quality score and communication effect score respectively, and calculate the product of the delivery effect comprehensive score, content quality score and communication effect score and the corresponding weight coefficients and add them to obtain the comprehensive score of each delivery content; extract the delivery effect analysis, content quality analysis and communication effect analysis data in a single delivery content; rank each delivery content according to the comprehensive score through a visualization tool, and load the delivery effect analysis, content quality analysis and communication effect analysis data of the corresponding delivery content in turn according to the ranking to generate a delivery analysis report.

[0074] This embodiment further discloses a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the above-mentioned delivery analysis method based on content characteristics and communication effects.

[0075] This embodiment further discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned delivery analysis method based on content characteristics and communication effects when executing the program.

[0076] Figure 3 A schematic diagram of the physical structure of an electronic device provided in an embodiment of the present invention, such as Figure 3As shown, the electronic device 50 includes: a processor 501 (processor), a memory 502 (memory) and a bus 503;

[0077] The processor 501 and the memory 502 communicate with each other via the bus 503 ; the processor 501 is used to call program instructions in the memory 502 to execute the methods provided by the above-mentioned method implementation methods.

[0078] This embodiment provides a non-transitory computer-readable storage medium, which stores computer instructions. The computer instructions enable a computer to execute the methods provided by the above-mentioned method embodiments.

[0079] A person skilled in the art can understand that all or part of the steps for implementing the above-mentioned method implementation method can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium, which, when executed, executes the steps of the above-mentioned method implementation method; and the aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk, etc., various storage media that can store program codes.

[0080] The device implementation described above is merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, i.e., they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present implementation scheme. Those of ordinary skill in the art may understand and implement it without creative effort.

[0081] Through the description of the above implementation modes, those skilled in the art can clearly understand that each implementation mode can be implemented by means of software plus a necessary general hardware platform, or of course by hardware. Based on such an understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of each implementation mode or some parts of the implementation mode.

[0082] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may occur depending on design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A delivery analysis method based on content characteristics and communication effects, characterized in that: The method comprises: Get multiple delivery contents based on the set target keywords; Performing a multi-dimensional delivery effect analysis on the plurality of delivery contents, and generating a comprehensive delivery effect score; Performing content characteristic analysis on the plurality of delivered contents and generating content quality scores; Acquire the propagation data of the multiple delivered contents, build a propagation network based on the propagation data, input the propagation network into the pagerank model, and output a propagation effect score, including: the propagation data includes reading data, sharing data, comment interaction data and external citation data; perform data cleaning and data aggregation on the propagation data, and generate a propagation characteristic matrix: , where R ij To deliver content To the delivery content The propagation association weight between , ; Take each delivered content as a network node, and the edge weight between network nodes is the communication influence between delivered content. The calculation formula of edge weight is: ,in, To deliver content The normalized value of the number of shares associated with the delivery content j, To deliver content Content being delivered The normalized value of the number of citations, To deliver content Interaction data and delivery content The normalized value of the number of associations, , , is the weight coefficient, and + + =1; construct a directed graph through network nodes and edge weights, and combine the propagation characteristic matrix to construct a propagation network; construct a transfer matrix P according to the edge weights of the propagation network: ,in, From network node i to network node The propagation weight between For network nodes The sum of out-degree weights of ; Iteratively calculate the communication influence score of each delivered content : ， in, is the damping coefficient, For Node The communication influence For Node To Node The transition probability, For Node The sum of out-degree weights; iterative calculation to Converge to output the final communication effect score of each delivered content; Generate a delivery analysis report based on the comprehensive delivery effect score, content quality score and communication effect score.

2. The delivery analysis method based on content characteristics and communication effects according to claim 1 is characterized in that: The method for acquiring multiple delivery contents based on the set target keywords includes: Obtain the set target keywords and perform synonym expansion through natural language processing to obtain a keyword set; Obtain multiple content sources and their API interfaces, and obtain content data through the API interfaces; Obtain delivery content data from content data through direct keyword matching, semantic matching, context matching, and priority matching; Content filtering and deduplication are performed on the delivered content data to obtain multiple delivered contents.

3. The delivery analysis method based on content characteristics and communication effects according to claim 1 is characterized in that: The method of performing multi-dimensional delivery effect analysis on the plurality of delivery contents and generating a comprehensive delivery effect score includes: Identify the basic delivery index data, spatiotemporal index data and industry category data of the delivered content; Analyze the delivery volume and re-delivery ratio of the delivered content from the perspective of basic delivery indicators to obtain the basic delivery indicator score; Analyze the delivery effects of the delivered content on weekdays and weekends from the perspective of time and space indicators to obtain the time and space indicator delivery scores; Analyze the delivered content from the perspective of industry categories to obtain industry category delivery scores; A comprehensive score of delivery effect is generated through the basic delivery index score, spatiotemporal index delivery score and industry category delivery score.

4. The delivery analysis method based on content characteristics and communication effects according to claim 3 is characterized in that: The method for generating a comprehensive score of delivery effect through the basic delivery index score, the spatiotemporal index delivery score and the industry category delivery score includes: Carry out data cleaning on basic delivery index scores, spatiotemporal indicator delivery scores, and industry category delivery scores; Perform min-max standardization on the cleaned scores of each dimension to normalize the scores to the interval [0, 1]; The normalized scores of each dimension are used to calculate the comprehensive score using the weighted average method. The formula is: ； in, is the dimension count, For the comprehensive rating, For the The rating of the dimension, For the The weight of the dimension score.

5. The delivery analysis method based on content characteristics and communication effects according to claim 1 is characterized in that: The method of analyzing the content characteristics of the plurality of delivered contents and generating content quality scores includes: Conduct a multi-dimensional content characteristic analysis of the originality, graphic structure, external link references, title characteristics and content richness of the delivered content; The originality characteristic analysis includes: content comparison with existing data sets based on content similarity algorithms, and the originality score is: ,in, is the repetition rate, the value range is [0,1]; The analysis of the characteristics of the graphic structure includes: calculating the ratio of the number of pictures to the number of paragraphs, setting an ideal ratio range, and deducting points from the content that does not meet the ideal ratio range based on the basic total score to obtain a graphic structure score; The analysis of external link citation characteristics includes: checking the number of external links and the credibility of their sources, taking the product of the number of external links on a single platform and the credibility of the platform as the external link score of the platform, and adding up multiple external link scores of a single delivery content as the external link citation score of the delivery content; The title feature analysis includes: using machine learning models to analyze the attractiveness of titles, constructing and scoring title feature scores based on the title features of historically highly disseminated content, and obtaining title feature scores; The content richness feature analysis includes: calculating the vocabulary diversity and logical hierarchy of the content through natural language processing and obtaining a content richness score; A comprehensive score is calculated based on the originality score, image and text structure score, external link citation score, title feature score and content richness score to generate a content quality score.

6. The delivery analysis method based on content characteristics and communication effects according to claim 1 is characterized in that: The method for generating a delivery analysis report based on the delivery effect comprehensive score, content quality score and communication effect score includes: Assign adjustable weight coefficients to the comprehensive score of delivery effect, content quality score and communication effect score respectively, and calculate the product of the comprehensive score of delivery effect, content quality score and communication effect score and the corresponding weight coefficients and add them up to obtain the comprehensive score of each delivery content; Extract data from single delivery content for delivery effect analysis, content quality analysis, and communication effect analysis; Through visualization tools, each delivery content is ranked according to the comprehensive score, and the delivery effect analysis, content quality analysis and communication effect analysis data of the corresponding delivery content are loaded in sequence according to the ranking to generate a delivery analysis report.

7. A delivery analysis system based on content characteristics and communication effects, characterized in that: include: A data acquisition module is used to acquire multiple delivery contents based on the set target keywords; The scoring module is used to perform multi-dimensional delivery effect analysis on the plurality of delivery contents and generate a comprehensive delivery effect score; and is also used to: Performing content characteristic analysis on the plurality of delivered contents and generating content quality scores; and further used for, Acquire the propagation data of the multiple delivered contents, build a propagation network based on the propagation data, input the propagation network into the pagerank model, and output a propagation effect score, including: the propagation data includes reading data, sharing data, comment interaction data and external citation data; perform data cleaning and data aggregation on the propagation data, and generate a propagation characteristic matrix: , where R ij To deliver content To the delivery content The propagation association weight between , ; Take each delivered content as a network node, and the edge weight between network nodes is the communication influence between delivered content. The calculation formula of edge weight is: ,in, To deliver content The normalized value of the number of shares associated with the delivery content j, To deliver content Content being delivered The normalized value of the number of citations, To deliver content Interaction data and delivery content The normalized value of the number of associations, , , is the weight coefficient, and + + =1; construct a directed graph through network nodes and edge weights, and combine the propagation characteristic matrix to construct a propagation network; construct a transfer matrix P according to the edge weights of the propagation network: ,in, From network node i to network node The propagation weight between For network nodes The sum of out-degree weights of ; Iteratively calculate the communication influence score of each delivered content : ， in, is the damping coefficient, For Node The communication influence For Node To Node The transition probability, For Node The sum of out-degree weights; iterative calculation to Converge to output the final communication effect score of each delivered content; The analysis report generation module is used to generate a delivery analysis report based on the comprehensive delivery effect score, content quality score and communication effect score.

8. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the method described in any one of claims 1 to 6 is implemented.

Images