A news periodical intelligent typesetting design method and system

Abstract

The present invention discloses a news periodical intelligent typesetting design method and system, which relates to the field of periodical intelligent typesetting. The main scheme is: collecting multiple types of data of news periodicals, detecting image copyright and watermarks, classifying data, and calculating news comprehensive content nature index CNI. R and heat index H R , and determine the standard index M of the text size R , calculate the area of ​​the text area A R and the image area A S , and according to the main color saturation C R To select the layout and color scheme, adjust the image area, and finally preview the content, format and visual effects of the typesetting results; this program helps to improve the quality, standardization, attractiveness and reader satisfaction of news periodicals.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent typesetting of journals, and in particular to an intelligent typesetting design method and system for news journals. Background Art

[0002] With the rapid advancement of the information age, news periodicals need to reasonably and effectively layout the massive and diverse news information within the limited page space; on the one hand, they need to highlight important and popular news to attract readers' attention and enable them to quickly obtain key information; on the other hand, they also need to balance the different types and importance of news to avoid important news being ignored or minor news occupying too much space; At the same time, readers' expectations and requirements for reading experience are also constantly improving. They are no longer satisfied with just obtaining news content itself, but have higher standards for the layout aesthetics, visual comfort, and reading convenience and readability of news periodicals. For example, the size and style of fonts, color matching, the ratio and layout of pictures and texts, etc., will affect readers' reading experience and acceptance of news. In recent years, the rapid development of digital technology has brought new opportunities and challenges to the typesetting design of news periodicals; the emergence of big data technology has made it possible to collect, analyze and mine massive amounts of news data, which can provide in-depth understanding of readers' reading behavior, interest preferences and market trends, and provide valuable data support and decision-making basis for news typesetting; However, although these advanced technologies bring great potential for the intelligent typesetting of news periodicals, there are still problems such as low typesetting efficiency, low quality, unattractive content and difficulty in meeting readers' needs. Summary of the invention

[0003] 1. Technical issues to be resolved In view of the deficiencies in the prior art, the present invention provides a method and system for intelligent typesetting design of news periodicals, which collects multiple types of data of news periodicals, detects copyrights and watermarks of images, classifies data, calculates relevant indicators to determine the font size, calculates the area of ​​text and image regions, and the saturation of the main color tone to select the layout and color scheme, adjusts the image area, and finally previews and checks the content, format and visual effects of the typesetting results, thereby solving the problems of low typesetting efficiency, low quality, unattractive content and difficulty in meeting readers' needs.

[0004] (II) Technical solution To achieve the above objectives, the present invention is implemented through the following technical solutions: a news periodical intelligent typesetting design method, comprising the following steps: Collect content data, image data and reading preference data of news periodicals; Calculate the image watermark detection value BPi And copyright detection value BT i , check whether there are watermarks or copyright issues; Classify content data and image data based on news topic categories; Based on the content data of different news topic categories, calculate the corresponding content nature index CNI R ; Calculate the heat index H based on reading preference data R ; Based on the comprehensive content nature index CNI R and heat index H R , calculate the font size standard index M R ; Preset font size standard threshold set E, according to the text font size standard index M R and the font size standard threshold set E, determining the font size of the corresponding content data; Based on content data of different news topic categories, topic heat index H R And font size standard index M R , calculate the text area A of different content data in the journal R ; Through the topic heat index H R and text area A R Calculate the main color saturation C R ; By the main color saturation C R Choose a layout and color scheme; According to the total area S of the journal and the area A of the text area R Calculate the total area A of the image region S , and calculate the area A of each subject category image i ; Preview the typeset news journal and check the preview results for content, format and visual effects.

[0005] In the preferred embodiment of the intelligent typesetting design method for a news periodical, the method for checking whether there is a watermark and whether there is a copyright problem is as follows: The feature vector F of each image in the image data is extracted by SIFT image feature extraction algorithm i , extract the watermark feature vector F of the picture in the watermark database through the SIFT image feature extraction algorithm (w,i) And the copyright feature vector F of the picture in the copyright database (k,i) , through the feature vector F of the picture i and watermark feature vector F (w,i) , calculate the watermark detection value BP i , the calculation formula is: , Among them, F irepresents the feature vector of the i-th picture, i is the serial number of pictures of different subject categories, and its value is 1, 2, 3...n, where n is the total number of pictures; F (w,i) Represents the watermark feature vector of the i-th image; Set the watermark detection threshold BP1, if BP i >BP1, there is a watermark in the image, and the matching watermark-free image is retrieved; if BP i ≤BP1, there is no watermark in the image, and the image is included in the image list for subsequent processing and typesetting; Through the feature vector F of the image to be detected i and the copyright feature vector F of the image in the copyright database (k,i) , calculate the copyright detection value BT i , the calculation formula is: , Among them, F (k,i) Represents the copyright feature vector of the i-th image; Set the copyright detection threshold BT2. If BT i > BT2, there is a copyright issue in the image. Stop using the image immediately and obtain a new image that has no copyright issue and matches the image. If BT i ≤BT2, there is no copyright issue in the image and the image is marked as legal for use.

[0006] In the preferred embodiment of the intelligent typesetting design method for news periodicals mentioned above: news subject categories include at least policies and regulations, technological innovation, market dynamics, corporate dynamics, project progress, power field, and popular science knowledge.

[0007] In the preferred embodiment of the intelligent typesetting design method for a news journal, based on the text content of different news subject categories, a comprehensive content professional index PI is obtained. R , Importance Index II R and timeliness index TI R , calculate the content nature index CNI R , the formula based on is:

[0008] Among them, ω1 is the comprehensive professional index PI R The weight coefficient is 0.3-0.4; ω2 is the importance index II R The weight coefficient is 0.3-0.4; ω3 is the timeliness index TI R The weight coefficient is 0.2-0.3, and ω1+ω2+ω3=1.

[0009] In the preferred solution of the above intelligent typesetting design method for a news journal: Calculate the popularity index H R The method is as follows: The reading preference data includes the view count P R , the comment count C R , the share count G R and the like count L R ; Calculate the popularity index H R based on the following formula:

[0010] where α is the weight coefficient of the view count P R , with a value range of 0.3 - 0.5; β is the weight coefficient of the comment count C R , with a value range of 0.2 - 0.3; γ is the weight coefficient of the share count G R , with a value range of 0.1 - 0.2; δ is the weight coefficient of the like count L R , with a value range of 0.1 - 0.2; and α + β + γ + δ = 1.

[0011] In the preferred solution of the above intelligent typesetting design method for a news journal: Calculate the font size standard index M R The formula is as follows:

[0012] where V is the basic font size coefficient; a and b are the index coefficients of the content nature index CNI R and the popularity index H R , with a value range of [0, 2]; c and d are the parameters for adjusting the content nature index CNI R and the popularity index H R , with a value range of [0, 5]; e is the parameter for the combined effect of the content nature index CNI R and the popularity index H R , with a value range of [0, 5]; θ is the adjustment coefficient of the font size, with a value range of [0.5, 1.5]; Preset the font size threshold set E, which includes the first font size threshold E1, the second font size threshold E2, and the third font size threshold E3, and E1 < E2 < E3; Compare the font size standard index M R with the font size threshold set E to determine the selected font size: When M R < E1, select the first font size scheme; When E1 ≤ M R < E2, select the second font size scheme; When E2 < M R ≤ E3, select the third font size scheme; When M R >E3, select font size option 4.

[0013] In the preferred embodiment of the intelligent typesetting design method for a news periodical, the area A of different text regions is calculated. R The method is: Get the corresponding word count W based on the content data R , according to the topic heat index H R , font size standard index M R Calculate the area of ​​the text area A R , the formula based on is:

[0014] Among them, A R It is the area of ​​the text content of the Rth news topic category, R is the serial number of the news topic category, ranging from 1 to 7; g is the exponential coefficient, ranging from 0.5 to 2; k is a constant.

[0015] In the preferred embodiment of the intelligent typesetting design method for a news journal, based on the topic heat index H R and text area A R Calculate the main color saturation C R , the formula based on is: , Among them, k1 is the topic popularity to the main color saturation C R The influence degree is 0.01-0.1; k2 is the layout area A R Saturation of main color C R The degree of influence is 0.001-0.01; Preset the first threshold C of the main color saturation R1 and the second threshold C of the main hue saturation R2 , and C R1 <C R2 : When C R <C R1 When using the layout and color scheme 1; When C R1 ≤C R <C R2 When using the layout and color scheme 2; When C R ≥C R2 When using the layout and color scheme three.

[0016] In the preferred embodiment of the intelligent typesetting design method for a news periodical, the total area A of the image region is calculated. S The method is:

[0017] Calculate the area A of each subject category image i , the calculation formula is:

[0018] Among them, r i It is the proportion of pictures of different subject categories in the total picture area, and the value is 0-1.

[0019] The present invention also discloses a news periodical intelligent typesetting design system, which is used to implement the above-mentioned news periodical intelligent typesetting design method: Data collection module, used to collect content data, image data and reading preference data of news periodicals; Image adjustment module, used to calculate the image watermark detection value BP i And copyright detection value BT i , check whether there are watermarks or copyright issues; A classification processing module, used to classify content data and image data based on news topic categories; The text preprocessing module is used to calculate the corresponding content nature index CNI based on the content data of different news topic categories. R ; Calculate the heat index H based on reading preference data R ; Based on the comprehensive content nature index CNI R and heat index H R , calculate the font size standard index M R ; Preset font size standard threshold set E, according to the text font size standard index M R and the font size standard threshold set E, determining the font size of the corresponding content data; The layout planning module is used to plan the content data of different news topic categories and the topic heat index H R And font size standard index M R , calculate the text area A of different content data in the journal R ; Through the topic heat index H R and text area A R Calculate the main color saturation C R ; By the main color saturation C R Choose a layout and color scheme; The layout optimization module is used to optimize the layout according to the total area S of the journal and the area A of the text area. R Calculate the total area A of the image region S , and calculate the area A of each subject category image i ; The preview and adjustment module is used to preview the news periodicals after typeset, and to check the content, format and visual effects of the preview results.

[0020] (III) Beneficial effects The present invention provides a news periodical intelligent typesetting design method and system, which has the following beneficial effects: (1) Collecting content data, image data, and reading preference data of news periodicals can comprehensively obtain all kinds of news-related information and provide a rich data foundation for subsequent typesetting design; (2) Calculate the image watermark detection value BP i And copyright detection value BT i , strictly check whether there are watermarks and copyright issues, ensure that the images used are legal and compliant, avoid potential legal disputes, and maintain the good reputation of news journals; (3) Classify content data and image data based on news topic categories, so that content with similar topics can be processed and displayed in a centralized manner, improving the logic and orderliness of typesetting; (4) Calculate the content nature index CNI based on content data of different news topic categories R , calculate the heat index H based on the reading preference data R , and then synthesize the font size standard index M R , and preset a font size standard threshold set E to determine the font size of the corresponding content data; such a process can accurately determine the appropriate font size based on factors such as the importance and popularity of the news, so that key information can be highlighted and readers can more easily get the key points; (5) Content data and topic popularity index H based on different news topic categories R And font size standard index M R , calculate the text area A of different content data in the journal R , through the topic heat index H R and text area A R Calculate the main color saturation C R , and then through the main color saturation C R Select the layout and color scheme; this series of steps makes the layout of the page more reasonable, important news can get more display space and more eye-catching colors, enhance visual appeal, and also reduce readers' reading fatigue and improve reading experience; (6) Based on the total journal area S and the text area A R Calculate the total area A of the image region S , and calculate the area A of each subject category image pThis ensures that text and pictures can be properly distributed within the limited journal pages, and there will be neither too many pictures and insufficient text information, nor too dense text and too few pictures, thus achieving optimal use of resources. (7) Preview the news periodical after typeset printing, and conduct content check, format check and visual effect check on the preview results; in terms of content, it can ensure the accuracy and completeness of information; format check ensures the standardization and consistency of fonts, paragraphs, titles, etc.; visual effect check optimizes color matching, image quality and page layout, thereby comprehensively improving the quality of news periodicals. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 A schematic diagram of the steps of a method and system for intelligent typesetting design of news periodicals according to the present invention; DETAILED DESCRIPTION

[0022] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] See also Figure 1 The present invention provides a news periodical intelligent typesetting design method, comprising the following steps: Step 1: Collect content data, image data and reading preference data of news periodicals; Step 101: Collect content data of news journals, including: news title, news text, author information, release time, news source, news classification label, and keywords; Step 102: The collected image data includes: the file format and pixel information of the image, a simple text description of the image content, the source of the image and the photographer; Step 103: Collect reading preference data of news periodicals, including: number of clicks, browsing time, number of favorites, comment content, number of shares, number of search keywords, and topic categories of interest; It should be noted that when collecting content data from news journals, it is necessary to ensure that it covers everything comprehensively and accurately, including but not limited to news of various types and fields, while paying attention to the timeliness and authority of the data; during the collection process, it is necessary to follow the principles of legality and compliance and not to infringe on copyright and privacy.

[0024] Collecting content data can fully understand the diversity and coverage of news, explore hot topics and trends, and establish a news database; collecting image data can enhance visualization effects, intuitively convey information, supplement the shortcomings of text, and improve visual quality; collecting reading preference data can accurately push news, provide personalized services, understand market demand and reader feedback, optimize topic planning, and adjust advertising strategies to increase revenue.

[0025] Step 2: Calculate the image watermark detection value BP i And copyright detection value BT i , check whether there are watermarks or copyright issues; Step 201: Extract the feature vector F of each image in the image data using the SIFT image feature extraction algorithm i , extract the watermark feature vector F of the picture in the watermark database through the SIFT image feature extraction algorithm (w,i) And the copyright feature vector F of the picture in the copyright database (k,i) , through the feature vector F of the picture i and watermark feature vector F (w,i) , calculate the watermark detection value BP i , the calculation formula is:

[0026] Among them, F i represents the feature vector of the i-th picture, i is the serial number of pictures of different subject categories, and its value is 1, 2, 3...n, where n is the total number of pictures; F (w,i) Represents the watermark feature vector of the i-th image; Step 202: Based on a large amount of historical experimental data and experience, set the watermark detection threshold BP1. i >BP1, there is a watermark in the image, and the matching watermark-free image is retrieved; if BP i ≤BP1, there is no watermark in the image, and the image is included in the image list for subsequent processing and typesetting; The working principle of this formula is: the watermark detection value is obtained by calculating the inner product of the image feature vector and the watermark feature vector and dividing it by the product of the modulus lengths of the two vectors; Molecular part It represents the inner product of the image feature vector and the watermark feature vector, reflecting the similarity between the two in each component; Denominator It is the product of the moduli of the two vectors; The function of the whole formula is to normalize the similarity between the image feature vector and the watermark feature vector, so that the watermark detection value BP iIt can take values ​​between 0 and 1. The closer the value is to 1, the higher the similarity between the features of the image and the watermark, and the greater the possibility of the presence of a watermark. The closer the value is to 0, the lower the similarity between the two, and the smaller the possibility of the presence of a watermark.

[0027] Step 203: Using the feature vector F of the image to be detected i and the copyright feature vector F of the image in the copyright database (k,i) , calculate the copyright detection value BT i , the calculation formula is: , Among them, F (k,i) Represents the copyright feature vector of the i-th image; Step 204: Perform detection calculations on a large number of images with and without copyright issues, observe the distribution of detection values, and thus set a copyright detection threshold BT2. i > BT2, there is a copyright issue in the image. Stop using the image immediately and obtain a new image that has no copyright issue and matches the image. If BT i ≤BT2, there is no copyright issue in the image and the image is marked as legal for use.

[0028] It should be noted that the SIFT image feature extraction algorithm is an algorithm used in the field of image processing, especially in computer vision and image recognition. It is an existing mature computing model. It is used to detect and describe local features in images. These features remain unchanged for image scaling, rotation, and even brightness changes. Through the SIFT image feature extraction algorithm, images can be converted into feature vectors. This vector can usually reflect the color, texture, shape and other information of the image, and is used for image recognition, classification, matching and other tasks.

[0029] It should be noted that the watermark database is a database specifically used to store various types of watermark information. These watermark information are usually embedded in multimedia content, such as images, audio, video, and documents, to identify the ownership and integrity of the content or provide other tracking information, and can obtain licenses from media platforms, copyright registration agencies, and copyright trading platforms. The copyright database is a database that stores copyright information. The copyright database is used to verify whether the content used has been legally authorized and to ensure that no copyright is infringed during the typesetting process. For example, copyright registration agencies, copyright trading platforms, and copyright management systems (CMS) have built-in copyright database functions.

[0030] It should be noted that the working principle of this formula is: by calculating the inner product of the feature vector of the image to be detected and the copyright feature vector of the image in the copyright database, and dividing by the product of the module lengths of the two vectors, the copyright detection value is obtained; Molecular part Represents the inner product of the feature vector of the image to be detected and the feature vector of the image in the copyright database, reflecting the similarity between them in each component; the denominator is the product of the moduli of these two vectors; The whole formula aims to normalize the similarity between the feature vector of the image to be detected and the copyright feature vector of the image in the copyright database, so that the copyright detection value BT i It can take values ​​between 0 and 1. The closer the value is to 1, the higher the feature similarity between the image to be detected and the image in the copyright database, and the greater the possibility of copyright issues. The closer the value is to 0, the lower the similarity, and the less likely there is a copyright issue. This method helps ensure that images used by news journals comply with legal regulations, avoid legal action and fines for copyright infringement, and maintain the good reputation and legal status of news organizations. It can also promote the news industry's respect for and protection of intellectual property rights and promote the formation of a healthy and standardized image use environment throughout the industry.

[0031] Step 3: Classify content data and image data based on news topic categories; Step 301: Classify the acquired news content data and image data, where the subject categories at least include policies and regulations, technological innovation, market dynamics, enterprise dynamics, project progress, power field, and popular science knowledge; It should be noted that news topics may include politics, economy, culture, technology, entertainment, sports, etc.; as for content data, they are classified into corresponding categories according to the core topics of the news reports.

[0032] Step 4: Based on the content data of different news topic categories, calculate the corresponding content nature index CNI R ; Calculate the heat index H based on reading preference data R ; Based on the comprehensive content nature index CNI R and heat index H R , calculate the font size standard index M R ; Preset font size standard threshold set E, according to the text font size standard index M R and the font size standard threshold set E, determining the font size of the corresponding content data; Step 401: By establishing a professional vocabulary database, conducting vocabulary analysis on the text in the news periodical, and counting the number of times the professional vocabulary appears to count the number of professional vocabulary N in the news periodical. PR ; By searching for references in the news journal text, we can count the number of cited authoritative documents N aR ; By analyzing the paragraphs of the news journal text, the number of commentary analysis paragraphs N is counted dRand count; It should be noted that the number of review and analysis paragraphs N dR The paragraphs of the content data in the news periodicals are analyzed by vocabulary, such as paragraphs containing commentary and analysis vocabulary such as "think, advocate, viewpoint, analysis, evaluation, judgment", and the number of paragraphs containing such commentary and analysis vocabulary is counted, and then the number of commentary and analysis paragraphs N is obtained. dR .

[0033] Step 402: Based on the number N of professional words in the news periodical PR 、Number of cited authoritative literature aR , Number of review and analysis paragraphs N dR , then calculate the comprehensive professional index PI of the content R The formula is:

[0034] Among them, N total is the total vocabulary or total number of paragraphs; w p1 is the number of professional words N PR The weight coefficient is 0.3-0.4; w p2 N is the number of cited authoritative literature aR The weight coefficient is 0.3-0.4; w p3 The number of paragraphs for review analysis is N dR The weight coefficient is 0.2-0.3; and w p1 +w p2 +w p3 =1; It should be noted that the principle of this formula is: the number of professional words in news journals N PR It reflects the use of terminology in specific professional fields in news; more professional vocabulary means that the content involves specific professional fields more deeply and has a higher degree of professionalism; divide it by the total vocabulary or total number of paragraphs and other related totals N total Standardization is performed so that news content of different sizes can be compared on the same scale; a weight coefficient w is given to it. p1 , indicating that the importance of professional vocabulary in the overall professional evaluation is within a certain range; the number of cited authoritative literature N aR It reflects the credibility of the news content and the authority of the basis; citing authoritative literature can increase the reliability and professionalism of the news; similarly, dividing by the total vocabulary or total number of paragraphs and other related totals N total Standardize and give weight coefficient w p2 , indicating that citing authoritative literature also plays an important role in professional evaluation; the number of review and analysis paragraphs N dRIt measures the degree of discussion and analysis of the issue in the news; the number and total amount of commentary and analysis usually means a more comprehensive and in-depth understanding of the topic, thereby improving the professionalism of the content; after standardization, the weight coefficient w p3 It reflects the relative importance of the number of review and analysis paragraphs in the professional evaluation; p1 +w p2 +w p3 =1 ensures that the weight distribution of each factor in the comprehensive professional evaluation is reasonable, so that the final calculated comprehensive content professional index can more accurately reflect the overall professional level of the news periodical content; Step 403: By monitoring and analyzing the data of the news release system, the interval T between the news release time and the current time is obtained. intervalR and news update frequency F updateR ; Based on the interval T between news release time and current time intervalR and news update frequency F updateR , calculate the content importance index II R The formula is:

[0035] Among them, T maxR is the maximum time interval threshold set; w t1 is the interval T between the news release time and the current time intervalR The weight coefficient is 0.4-0.6; w t2 The news update frequency F updateR The weight coefficient is 0.4-0.5; and w t1 +w t2 =1; It should be noted that the operating principle of this formula is that this formula comprehensively calculates the importance index of news content from two aspects: timeliness and dynamic update of news; on the one hand, considering the interval T between the news release time and the current time intervalR ; As time goes by, the freshness of news will gradually decrease; Used in formula To reflect the impact of this factor on importance; when T intervalR When it is close to 0, that is, the news has just been released, the value of this formula is close to 1, and the contribution to the importance index is large; intervalR Close to the set maximum time interval threshold T maxR When , the value of this formula is close to 0, and the contribution to the importance index is small; the weight coefficient w t1This indicates that the time interval has a certain degree of importance in the importance assessment. The range of this coefficient also reflects that although timeliness is important, it is not the only determining factor. On the other hand, the news update frequency F updateR It reflects the attention level and dynamic changes of news; if a piece of news is updated frequently, it means it is a hot topic or has high importance; the weight coefficient w t2 represents the relative importance of news update frequency in importance assessment; similarly, w t2 The value range of also reflects that the update frequency is also an important aspect of importance assessment, but it is not the only dominant factor; t1 +w t2 =1, which ensures that the weight distribution of the two factors of time interval and update frequency in the comprehensive calculation of the importance index is reasonable, so that the final content importance index can more accurately reflect the comprehensive importance of news in terms of timeliness and dynamic update; Step 404: Obtain the social attention of the news by monitoring the news's spreading speed, number of reposts, number of comments, and user sentiment on social media. socialR ; Obtain the policy influence of news based on the frequency and depth of mentions of policy analysis institutions, government departments, expert opinions, and policy-related content in news reports I policyR ; Obtain the economic impact of news by analyzing financial market reactions, changes in consumer behavior, and fluctuations in corporate performance I economyR ; Step 405: Based on the social attention of the news A socialR 、Policy Influence I policyR 、Economic Influence I economyR , calculate the news timeliness index TI R The formula is:

[0036] Among them, w i1 The social attention of the news socialR The weight coefficient is 0.3-0.5; w i2 For policy influence policyR The weight coefficient is 0.2-0.4; w i3 Economic Influence I economyR The weight coefficient is 0.2-0.3; and w i1 +w i2 +w i3 =1; It should be noted that the operating principle of this formula is that, first, the social attention to the news A socialR 、Policy Influence I policyR 、Economic Influence IeconomyR Quantify; Social attention A socialR By counting the number of times news is shared, commented on, and liked, weight coefficients can be assigned to the number of times news is shared, commented on, and liked, such as A socialR =a1*number of news shares + a2*number of comments + a3*number of likes, a1 can be 0.1~0.4, a2 can be 0.2~0.5, a3 can be 0.2~0.4, and then the social attention A is obtained. socialR Quantitative value of policy influence I policyR By counting the number of times the public mentions the policy online and the percentage of public opinion survey responses, we can assign weight coefficients to the number of times the public mentions the policy online and the percentage of public opinion survey responses, such as the policy influence I policyR =b1*number of times the policy is mentioned online+b2*the proportion of public opinion survey responses. b1 can be 0.3~0.6, and b2 can be 0.4~0.7, thus obtaining the policy influence I policyR Quantitative value of economic influence I economyR By counting the change value of the relevant stocks after the news is released and the change value of the consumer confidence index in the public opinion survey, a weight coefficient can be assigned to the change value of the relevant stocks and the change value of the consumer confidence index, such as the economic influence I economyR =c1*change in the relevant stock value+c2*change in the consumer confidence index, c1 can be 0.2~0.6, c2 can be 0.4~0.8, and then the economic impact I is obtained. economyR The quantitative value of the weight coefficient w can be obtained through the above method. i1 、w i2 and w i3 Used to indicate the importance of different aspects in evaluating news timeliness; these weight coefficients are determined based on actual conditions, expert opinions or historical data analysis; in this formula, the range of weight coefficients gives a rough guide, that is, the weight coefficient of social attention is higher, and the weight coefficients of policy influence and economic influence are relatively lower but also different; after determining the quantitative value and weight coefficient of each indicator, multiply the quantitative value of each indicator by the corresponding weight coefficient, and then add all the products to get the news timeliness index TI R The final value of the news. The higher the value, the more timely the news is, that is, the news has a higher importance or influence in terms of social attention, policy influence and economic influence. Step 406: Comprehensive professional index PI based on content R , Importance Index II R and timeliness index TI R , calculate the content nature index CNIR , the formula based on is:

[0037] Among them, ω1 is the comprehensive professional index PI R The weight coefficient is 0.3-0.4; ω2 is the importance index II R The weight coefficient is 0.3-0.4; ω3 is the timeliness index TI R The weight coefficient is 0.2-0.3, and ω1+ω2+ω3=1; The principle of this formula is: by comprehensive professional index PI R , Importance Index II R and timeliness index TI R Assign certain weight coefficients ω1, ω2 and ω3 respectively, and calculate the content nature index CNI by weighted summation R ; Step 407: Obtain news reading preference data including page views P R 、Number of comments C R , share G R And the number of likes L R ; Calculate the heat index H by reading preference data R The formula is as follows:

[0038] Among them, α is the pageview volume P R The weight coefficient is 0.3-0.5; β is the number of comments C R The weight coefficient is 0.2-0.3; γ is the number of shares G R The weight coefficient is 0.1-0.2; δ is the number of likes L R The weight coefficient is 0.1-0.2; and α+β+γ+δ=1; The principle of this formula is: Views P R The weight coefficient α ranges from 0.3 to 0.5, reflecting the importance of page views in measuring news popularity. A higher page view usually means that more people pay attention to the news, so it contributes more to the popularity. The number of comments C R The weight coefficient β is 0.2-0.3, indicating that the number of comments can reflect the readers' participation in the news and the enthusiasm of discussion to a certain extent; the number of shares G R The weight coefficient γ is 0.1-0.2, which means that sharing behavior represents the reader’s willingness to actively spread news, and can also reflect the attractiveness and value of news; the number of likes L RThe weight coefficient δ is 0.1-0.2, and the number of likes reflects the readers' initial recognition and liking of the news. By assigning different weights and making the sum of the weight coefficients 1, a reasonable balance and comprehensive consideration of these four factors is achieved, thereby comprehensively evaluating the popularity of the news. Step 408: Calculate the font size standard index M R The formula is:

[0039] Among them, V is the basic font size coefficient; a and b are the content nature index CNI R and heat index H R The exponential coefficient is [0,2]; c and d are the content nature index CNI R and heat index H R The parameter is [0,5]; e is the content nature index CNI R and heat index H R The parameter that works together takes the value [0,5]; θ is the adjustment coefficient of the font size, takes the value [0.5,1.5]; The principle of this formula is: the basic font size coefficient V is a fixed value, which is the font size standard index M R The calculation of sets an initial benchmark size; the exponential coefficients a and b act on the content nature index CNI respectively R and heat index H R ; When the value of a is large, it means that CNI R Index M for the final font size standard R The influence of b is greater, and vice versa. Similarly, the value of b determines H R To M R The influence of CNI is strong or weak; their value range is between [0,2], and CNI can be flexibly adjusted according to specific application scenarios and needs. R and V relative importance; parameters c and d are respectively related to CNI R and H R Multiply them together to further fine-tune the influence of each of these two factors; if the value of c is large, then CNI R The influence on the calculation results will be relatively stronger; if the value of d is large, then H R The effect will be more significant; parameter e is used to adjust CNI R and H R The effect of synergy; its existence makes CNI R and H RThe combined effect can be fine-tuned according to specific circumstances to achieve more accurate calculation results that meet actual requirements; finally, the adjustment coefficient θ of the font size is an overall scaling factor; when the value of θ is close to 1.5, the calculation result will be enlarged, resulting in the final font size standard index M R increasing, corresponding to a larger font size; when the value of θ is close to 0.5, the calculation result will be reduced, and the font size standard index M R decreases, corresponding to a smaller font size; in summary, by cleverly combining and adjusting various parameters, this formula can comprehensively consider various factors such as the nature and popularity of the content, thus providing an accurate and flexible calculation method for determining the appropriate font size standard, enabling more effective highlighting of key and popular information when presenting content; Step 409: Preset a set of font size thresholds E according to experience, data analysis, user requirements, and the overall requirements of page layout and design. The set of font size thresholds E includes a first font size threshold E1, a second font size threshold E2, and a third font size threshold E3, and E1 < E2 < E3; Compare the font size standard index M R with the set of font size thresholds E to determine the selected font size: When M R < E1, select font size option one; When E1 ≤ M R < E2, select font size option two; When E2 < M R ≤ E3, select font size option three; When M R > E3, select font size option four.

[0040] It should be noted that font size option one is the smallest font size option, suitable for situations with low content density or where space needs to be saved, and the font size range is 6pt - 8pt; selecting font size option two is a medium-sized font size option, suitable for most regular text content, and the font size range is 10pt - 14pt; selecting font size option three is a larger font size option, suitable for content that needs to be emphasized or highlighted, and the font size range is 16pt - 20pt; font size option four is the largest font size option, usually used for titles, slogans, or particularly important content, and the font size range is 24pt - 28pt; This method can conduct precise professional evaluations for different news topics, fully consider reader feedback and participation levels, highlight key and popular news, optimize the reading experience, enable readers to obtain key information more quickly, improve the rationality and aesthetics of page layout, adapt to diverse needs, and moreover, the entire process is based on objective data, reducing subjective biases and making the determination of font size more scientific and accurate.

[0041] Step 5: Based on the content data of different news topic categories and the topic heat index H R And font size standard index M R , calculate the text area A of different content data in the journal R ; Through the topic heat index H R and text area A R Calculate the main color saturation C R ; By the main color saturation C R Choose a layout and color scheme; Step 501: Obtain the corresponding word count W based on the content data R , according to the topic heat index H R , font size standard index M R Calculate the area of ​​the text area A R , the formula based on is:

[0042] Among them, A R is the area of ​​the text content of the Rth news topic category, R is the serial number of the news topic category, ranging from 1 to 7; g is the exponential coefficient, ranging from 0.5 to 2; k is a constant; The principle of this formula is: This formula is calculated by taking the topic heat index H R , font size standard index M R And the number of words W R Combined to calculate the text area A R ; Topic heat index H R It reflects the degree of attention and interest in the news topic. The higher the popularity, the larger the area may need to be allocated to display the relevant text. The font size standard index M R The function of is to consider the influence of font size on text display size and space occupancy, and the exponential coefficient g is used to adjust the degree of its influence; the number of words W R pass This part is involved in the calculation, where the log function performs a nonlinear transformation on the number of words, and the constant k is used to further adjust the effect of this transformation; adding 1 is to avoid the situation where the log function is undefined when the number of words is 0; in summary, this formula aims to reasonably calculate the required text area based on the popularity of the news topic, the font size standard and the number of words, so as to achieve more effective information display and layout arrangement; Step 502: Based on the topic heat index H R and text area A R Calculate the main color saturation C R , the formula based on is:

[0043] Among them, k1 is the topic popularity to the main color saturation C R The influence degree is 0.01-0.1; k2 is the layout area A R Saturation of main color C R The degree of influence is 0.001-0.01; The principle of this formula is: Topic heat index H R Multiplied by the coefficient k1, this means that the higher the topic popularity, the greater the contribution to the saturation of the main color tone; a higher k1 value reflects that the topic popularity has a certain influence on determining the saturation of the main color tone, but this influence is not absolutely dominant, but plays a role to a certain extent; the text area A R The square root of is multiplied by the coefficient k2; the square root treatment may be to moderately adjust the impact of the area to avoid excessive impact of the direct numerical value of the area on the result; the value of k2 is small, indicating that the influence of the text area on the saturation of the main color is relatively weak, but it is still a factor that needs to be considered; in summary, this formula determines the final saturation of the main color by considering the different degrees of topic popularity and text area area; in this way, the saturation of the main color can more reasonably reflect the importance and display needs of the news content, that is, when the topic popularity is high or the text area is large, the saturation of the main color is increased accordingly to achieve a more prominent and attractive visual effect; when the popularity is low or the area is small, the saturation of the main color is appropriately reduced to maintain the overall balance and coordination; Step 503: Preset the first threshold C of the main color saturation based on empirical data, user visual preference research, tests on the effects of different saturation colors, and the degree of coordination with the overall layout design style. R1 and the second threshold C of the main hue saturation R2 , and C R1 <C R2 : When C R <C R1 When using a topic, choose layout and color scheme 1; layout and color scheme 1 is: use relatively elegant and low-key colors, and the main color saturation is low; for example, light blue, light green or light gray as the main color, which gives people a fresh and soothing feeling as a whole. It is suitable for situations where the topic is not very popular and the text area is small, and will not cause a strong visual impact on readers; When C R1 ≤C R <C R2When the page is in the middle, choose the second layout and color scheme; the second layout and color scheme is: the colors may be relatively rich and bright, but still maintain a certain degree of moderation; for example, choose the main color with medium saturation such as orange, lavender or yellow-green, which can attract the reader's attention without being too glaring, and is suitable for the situation where the main color saturation is at a medium level; When C R ≥C R2 When using the third layout and color scheme, the third layout and color scheme is to use highly saturated, bright and eye-catching colors; choose bright red, bright yellow or royal blue as the main color, highlight important and popular content with a strong visual effect, match the high topic popularity and large text area, and attract readers' attention to the greatest extent; Based on content data of different news topic categories, topic heat index H R And font size standard index M R To calculate the area of ​​the text area, we can achieve accurate quantification and layout planning of news content; this can ensure that important and popular news get more sufficient display space, highlight the key points, and improve the efficiency of information transmission; through the topic heat index H R and text area A R Calculate the main color saturation C R , so that the color of the layout can match the popularity and display needs of the news; popular news can attract readers' attention through brighter and more saturated main colors, enhance visual impact, and improve the attractiveness and readability of the news.

[0044] Step 6: Based on the total journal area S and the text area A R Calculate the total area A of the image region S , and calculate the area A of each subject category image i ; Step 601: Calculate the total area A of the image region S The method is:

[0045] Step 602: By analyzing the readers' reading behavior data, we can understand the readers' clicks, reading time, and number of comments on news on different topics, so as to judge the attention and importance of each topic, and then determine the proportion of pictures of different topic categories in the total picture area. i value; Step 603: Calculate the area A of each subject category image i , the calculation formula is:

[0046] Among them, r i The value range is 0-1; It should be noted that if a news topic is very important and needs more pictures to assist in explaining and attracting readers' attention, then the r corresponding to the topic should be i If a topic is widely concerned in the current period, in order to meet the needs of readers and increase the visual appeal of the topic, its r value can be appropriately increased. i value; The total area of ​​the picture area is calculated based on the total area of ​​the journal and the area of ​​the text area, and the area of ​​the pictures in each subject category is further calculated; this can reasonably allocate layout resources and avoid an imbalance in the layout of text and pictures; at the same time, accurate calculation of the picture area for different subject categories can highlight key topics, enhance visual appeal and information communication effects; in addition, it can improve the overall coordination and aesthetics of the layout, meet readers' reading needs and visual expectations, and improve the production efficiency and quality of the journal, reduce arbitrariness and uncertainty in the design, and ensure the stability and professionalism of the layout design.

[0047] Step 7: Preview the news journal after typeset, and check the content, format and visual effect of the preview result; Step 701: Check the content of the preview effect, including: Information accuracy: Ensure that the text, data and other contents presented are accurate and contain no erroneous or misleading information; Completeness: Check whether the key information is complete and no important parts are omitted; Logical coherence: whether the narration and organization of the content are logically clear, the transitions between the parts are natural, and easy to understand; Language standardization: whether the language expression is standard, grammatically correct, and free of typos, grammatical errors, etc. Step 702: Perform format check on the preview effect, including: Font consistency: whether the font type, size, color, etc. are consistent throughout the document; Paragraph layout: whether the paragraph indentation, line spacing, alignment, etc. meet the design requirements and are neat and beautiful; Title hierarchy: whether the formats of titles at all levels are clearly distinguished and the hierarchy is clear; Chart annotation: whether the chart number, title, annotation, etc. are accurate and clear; Step 703: Perform visual effect check on the preview effect, including: Color matching: whether the color selection is coordinated, whether the main color and auxiliary color match comfortably and not dazzling; Page layout: whether the distribution of page elements is balanced, whether the blank space is reasonable, and whether it is not too crowded or empty; Overall style: Check whether the overall visual style is unified and consistent with the expected design concept and the aesthetic taste of the target audience; This method can ensure that the content is accurate, complete, and logically clear, thereby enhancing credibility; it can ensure that the format is consistent, enhance the reading experience, and reflect rigorous and high quality; it can make the visual effect attractive and comfortable, reduce reader fatigue, improve the quality and competitiveness of the journal, meet reader needs, and establish a brand image.

[0048] On the other hand, the present invention also discloses a news periodical intelligent typesetting design system, which is used to implement the above-mentioned news periodical intelligent typesetting design method, comprising: Data collection module, used to collect content data, image data and reading preference data of news periodicals; Image adjustment module, used to calculate the image watermark detection value BP i And copyright detection value BT i , check whether there are watermarks or copyright issues; A classification processing module, used to classify content data and image data based on news topic categories; The text preprocessing module is used to calculate the corresponding content nature index CNI based on the content data of different news topic categories. R ; Calculate the heat index H based on reading preference data R ; Based on the comprehensive content nature index CNI R and heat index H R , calculate the font size standard index M R ; Preset font size standard threshold set E, according to the text font size standard index M R and the font size standard threshold set E, determining the font size of the corresponding content data; The layout planning module is used to plan the content data of different news topic categories and the topic heat index H R And font size standard index M R , calculate the text area A of different content data in the journal R ; Through the topic heat index H R and text area A R Calculate the main color saturation C R ; By the main color saturation C R Choose a layout and color scheme; The layout optimization module is used to optimize the layout according to the total area S of the journal and the area A of the text area. R Calculate the total area A of the image region S , and calculate the area A of each subject category image i ; The preview and adjustment module is used to preview the news periodicals after typeset, and to check the content, format and visual effects of the preview results.

[0049] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented using software, the above embodiments may be implemented in whole or in part in the form of a computer program product. A person of ordinary skill in the art may appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein may be implemented in electronic hardware or in combination with computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution.

[0050] 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, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0051] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A method for intelligent typesetting design of news periodicals, characterized by: The following steps are involved: Collect content data, image data and reading preference data of news periodicals; Calculate the image watermark detection value BP i And copyright detection value BT i , check whether there are watermarks or copyright issues; Among them, the image watermark detection value BP i And copyright detection value BT i The calculation method is as follows: Extract the feature vector F of each image in the image data through the SIFT image feature extraction algorithm i , extract the watermark feature vector F of the picture in the watermark database through the SIFT image feature extraction algorithm (w,i) And the copyright feature vector F of the picture in the copyright database (k,i) , through the feature vector F of the picture i and watermark feature vector F (w,i) Calculate the watermark detection value BP i , the calculation formula is: , where F i represents the feature vector of the i-th picture, i is the serial number of pictures of different subject categories, and its value is 1, 2, 3...n, where n is the total number of pictures; F (w,i) Represents the watermark feature vector of the i-th image; Through the feature vector F of the image to be detected i and the copyright feature vector F of the image in the copyright database (k,i) , calculate the copyright detection value BT i , the calculation formula is: , where F (k,i) Represents the copyright feature vector of the i-th image; Classify content data and image data based on news topic categories; Based on the content data of different news topic categories, calculate the corresponding content nature index CNI R ; Calculate the heat index H based on reading preference data R ; Based on the comprehensive content nature index CNI R and heat index H R , calculate the font size standard index M R ; Preset font size standard threshold set E, according to the text font size standard index M R and the font size standard threshold set E, determining the font size of the corresponding content data; Based on content data of different news topic categories, topic heat index H R And font size standard index M R , calculate the text area A of different content data in the journal R ; Through the topic heat index H R and text area A R Calculate the main color saturation C R ; By the main color saturation C R Choose a layout and color scheme; According to the total area S of the journal and the area A of the text area R Calculate the total area A of the image region S , and calculate the area A of each subject category image i ; Preview the typeset news journal and check the preview results for content, format and visual effects.

2. The intelligent typesetting design method for news periodicals according to claim 1, characterized in that: The method to check whether there is a watermark and no copyright issues is: Set the watermark detection threshold BP1, if BP i >BP1, there is a watermark in the image, and the matching watermark-free image is retrieved; if BP i ≤BP1, there is no watermark in the image, and the image is included in the image list for subsequent processing and typesetting; Set the copyright detection threshold BT2. If BT i > BT2, there is a copyright issue in the image. Stop using the image immediately and obtain a new image that has no copyright issue and matches the image. If BT i ≤BT2, there is no copyright issue in the image and the image is marked as legal for use.

3. The intelligent typesetting design method for news periodicals according to claim 2, characterized in that: News topic categories include at least policies and regulations, technological innovation, market trends, corporate trends, project progress, power sector, and popular science knowledge.

4. The intelligent typesetting design method for news periodicals according to claim 3 is characterized by: Based on the text content of different news topic categories, obtain the comprehensive content professionalism index PI R , Importance Index II R and timeliness index TI R , calculate the content nature index CNI R , the formula based on is: ， Among them, ω1 is the comprehensive professional index PI R The weight coefficient is 0.3-0.4; ω2 is the importance index II R The weight coefficient is 0.3-0.4; ω3 is the timeliness index TI R The weight coefficient is 0.2-0.3, and ω1+ω2+ω3=1.

5. The intelligent typesetting design method for news periodicals according to claim 4, characterized in that: Calculate the heat index H R The method is as follows: Reading preference data includes pageviews P R 、Number of comments C R , share G R And the number of likes L R ; Calculate the heat index H by reading preference data R The formula is as follows: , Among them, α is the pageview volume P R The weight coefficient is 0.3-0.5; β is the number of comments C R The weight coefficient is 0.2-0.3; γ is the number of shares G R The weight coefficient is 0.1-0.2; δ is the number of likes L R The weight coefficient is 0.1-0.2; and α+β+γ+δ=1.

6. The intelligent typesetting design method for news periodicals according to claim 5, characterized in that: Calculate the font size standard index M R The formula is: ， Among them, V is the basic font size coefficient; a and b are the content nature index CNI R and heat index H R The exponential coefficient is [0,2]; c and d are the content nature index CNI R and heat index H R The parameter is [0,5]; e is the content nature index CNI R and heat index H R The parameter that works together takes the value [0,5]; θ is the adjustment coefficient of the font size, takes the value [0.5,1.5]; A font size threshold set E is preset, and the font size threshold set E includes a font size threshold 1 E1, a font size threshold 2 E2, and a font size threshold 3 E3, and E1<E2<E3; The font size standard index M R Compare with the font size threshold set E to determine the selected font size: When M R <is less than E1, select font size option one; When E1 ≤ M R <When E2, select font size option two; When E2 <M R When ≤E3, select font size plan 3; When M R >E3, select font size option 4.

7. The intelligent typesetting design method for news periodicals according to claim 6, characterized in that: Calculate the area A of different text areas R The method is: Get the corresponding word count W based on the content data R , according to the topic heat index H R , font size standard index M R Calculate the area of ​​the text area A R , the formula based on is: , Among them, A R It is the area of ​​the text content of the Rth news topic category, R is the serial number of the news topic category, ranging from 1 to 7; g is the exponential coefficient, ranging from 0.5 to 2; k is a constant.

8. The intelligent typesetting design method for a news periodical according to claim 7, characterized in that: Based on the topic heat index H R and text area A R Calculate the main color saturation C R , the formula based on is: , Among them, k1 is the topic popularity to the main color saturation C R The influence degree is 0.01-0.1; k2 is the layout area A R Saturation of main color C R The degree of influence is 0.001-0.01; Preset the first threshold C of the main color saturation R1 and the second threshold C of the main hue saturation R2 , and C R1 <C R2 : When C R <C R1 When using the layout and color scheme 1; When C R1 ≤C R <C R2 When using the layout and color scheme 2; When C R ≥C R2 When using the layout and color scheme three.

9. The intelligent typesetting design method for news periodicals according to claim 8, characterized in that: Calculate the total area A of the image region S The method is: ， Calculate the area A of each subject category image i , the calculation formula is: ， Among them, r i It is the proportion of pictures of different subject categories in the total picture area, and the value is 0-1.

10. An intelligent typesetting design system for news periodicals, characterized by: Data collection module, used to collect content data, image data and reading preference data of news periodicals; Image adjustment module, used to calculate the image watermark detection value BP i And copyright detection value BT i , check whether there are watermarks or copyright issues; A classification processing module, used to classify content data and image data based on news topic categories; The text preprocessing module is used to calculate the corresponding content nature index CNI based on the content data of different news topic categories. R ; Calculate the heat index H based on reading preference data R ; Based on the comprehensive content nature index CNI R and heat index H R , calculate the font size standard index M R ; Preset font size standard threshold set E, according to the text font size standard index M R and the font size standard threshold set E, determining the font size of the corresponding content data; The layout planning module is used to plan the content data of different news topic categories and the topic heat index H R And font size standard index M R , calculate the text area A of different content data in the journal R ; Through the topic heat index H R and text area A R Calculate the main color saturation C R ; By the main color saturation C R Choose a layout and color scheme; The layout optimization module is used to optimize the layout according to the total area S of the journal and the area A of the text area. R Calculate the total area A of the image region S , and calculate the area A of each subject category image i ; The preview and adjustment module is used to preview the news periodicals after typeset, and to check the content, format and visual effects of the preview results.

Images