Integrated media management system

Abstract

The invention discloses a convergent media management system, which belongs to the technical field of Internet, and comprises a target population news data acquisition module used for distinguishing news browsing populations in a preset area, acquiring browsing contents of the news browsing populations, and marking corresponding populations according to the browsing contents; the target news data browsing fusion calculation unit is used for performing per capita browsing number calculation on the news publishing platforms in the preset area by taking each natural day as a collection time period, and sequencing the news publishing platforms in the preset area from front to back by taking each month as a statistical period based on the calculated per capita browsing number; according to the method and the device, the browsing behaviors of the browsing crowds in the preset area and the browsing time period in which the browsing crowds are most concentrated are obtained, and the browsing mode is obtained in the time period, so that the browsing experience of the browsers is improved from a more humanized perspective, thinning and fragmentation of news pushing are avoided, and the news transmission effect is guaranteed.

Description

Technical Field

[0001] The present invention relates to the field of Internet technology, and in particular to a media convergence management system. Background Art

[0002] The integrated media management system is a product of the combination of modern information technology and the media field. It can realize the integration, processing and distribution of multimedia content, and has efficient content management, user interaction and data processing capabilities. With the continuous development of information technology and the demand of human beings for information data, the media methods for information dissemination according to media attributes have been diversified. Among them, "integrated media" has been widely used among the general public. It can realize a variety of media forms. It can recommend relevant content to users for selection by calculating the similarity between content feature vectors and user preference vectors according to the characteristics and historical preferences of users. In order to enable "integrated media" to serve the public efficiently, the prior art application number CN202310360246.9 provides a content management system for integrated media, and its technical content includes content acquisition module, original content import module, initial content storage module, content integrated management module and dissemination push module; wherein, the content acquisition module, original content import module, initial content storage module, content integrated management module and dissemination push module are connected in sequence. When searching for content, this technical application can quickly find the required information according to people's needs, and can avoid the interference of some related content, so as to improve the efficiency of search.

[0003] Another technical application number is CN202211193527.1, which provides a broadcasting integrated media information collection, editing and publishing management system based on a cloud platform. Its technical content includes a target news manuscript news category identification module, a target news manuscript preferred publishing platform screening module, a target news manuscript adaptation publishing form analysis module, a management database, a target news manuscript template editing and processing module, a news finished product preferred publishing time point determination module, a news finished product adaptation background music selection module and a news finished product publishing terminal. This technical application identifies the news category of the target news manuscript and then selects the preferred publishing platform corresponding to the target news manuscript, thereby realizing the targeted release of the collected and edited news, effectively making up for the shortcomings of the current widespread release method for the collected and edited news, not only greatly improving the vertical dissemination of news releases, but also avoiding a large amount of ineffective waste of publishing costs and improving the utilization rate of news release funds.

[0004] The media content in today's society is becoming more and more entertainment-oriented, making it difficult for today's integrated media to conduct basic data analysis and public opinion analysis based on functions and applications, and unable to provide scientific data support for operational decisions, resulting in the shallowness and fragmentation of news, which is not conducive to ensuring the effectiveness of news dissemination. Summary of the invention

[0005] In view of the above problems existing in the existing Internet technology field, the present invention is proposed.

[0006] Therefore, one of the objects of the present invention is to provide a media convergence management system, which can obtain the browsing behavior of the browsing population in the preset area by calculating the average number of views per capita for news released on the news release platform in a preset area with each natural day as the collection period, and then push the corresponding news categories to the browsing population in the preset area in a targeted manner, and can also collect the browsing time periods with the most concentrated browsing population in the preset area, and obtain the browsing methods during these time periods, so as to improve the browsing experience of the viewers from a more humane perspective, avoid the shallowness and fragmentation of news push, and ensure the dissemination effect of news.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0008] The present invention provides a media integration management system, comprising:

[0009] The target population news data acquisition module is used to distinguish the news browsing population in a preset area, obtain their browsing content, and mark the corresponding population according to the browsing content;

[0010] The target news data browsing fusion calculation unit is used to calculate the average number of views per capita for the news publishing platform in the preset area with each natural day as the collection period, and based on the calculated average number of views per capita, the news publishing platform in the preset area is sorted in order with each month as a statistical period. The target news data browsing fusion calculation unit includes a news data feature collection module, a pageview fluctuation analysis module and a determination module;

[0011] The news data feature collection module is used to collect news categories released by the corresponding news platform on each natural day, and the news categories include news, communication news, commentary news and feature news;

[0012] The pageview fluctuation analysis module responds to the news categories collected by the news data feature collection module and published on the corresponding news platform, and is used to analyze the fluctuation changes in the number of daily pageviews of different news categories, and generate a data set based on one of the statistical cycles;

[0013] At the same time, the fluctuation of the number of daily viewers of different news categories is divided into high fluctuation, medium fluctuation and low fluctuation.

[0014] The determination module is used to limit the frequency of news categories released by the news platform in the future period by using the daily browsing data of every 100 people as reference data, wherein, in the reference data, if the daily browsing fluctuation number of a certain news category released by the news platform is ≥50 people, the system determines that the corresponding news category released by the news platform is a news category with high fluctuation, and reduces the push frequency of the news category in the future period;

[0015] If the daily fluctuation in the number of people who browse a certain news category published by the news platform is ≥30, the system will determine that the corresponding news category published by the news platform is a news category with medium fluctuation, and will not change the push frequency of the news category in the future period;

[0016] If the daily fluctuation in the number of people viewing a certain news category published by a news platform is ≤20, the system will determine that the corresponding news category published by the news platform is a news category with low fluctuations, and increase the push frequency of the news category in the future period;

[0017] The fluctuation data verification module is used to preset the corresponding news category as the reference news category based on the high fluctuation change on each natural day, and collect the browsing data of the corresponding number of viewers for news categories other than the reference news category; if the browsing data is not collected, the system determines that there is still browsing demand for the reference news category among the corresponding number of viewers, otherwise, no determination is made.

[0018] As a preferred solution of the present invention, in the target population news data acquisition module, the news browsing population in the preset area is differentiated, including being differentiated into hard news browsing population and soft news browsing population; wherein the hard news includes political, economic and social news events;

[0019] Soft news includes cultural, entertainment, sports and educational news events.

[0020] As a preferred solution of the present invention, in the fluctuation data verification module, if the browsing data of the corresponding number of viewers for news categories other than the reference news category is collected, the corresponding news category is sorted according to the browsing data according to θ1, θ2, ..., θ n The news categories are distinguished, n represents the nth news category, and the number of viewers of each news category is counted among the number of viewers, the top three news categories are obtained, and the news categories are marked. At the same time, in the future time period, the push of the marked news categories is increased among the viewing population corresponding to the high fluctuation change.

[0021] As a preferred solution of the present invention, the confidence calculation is performed on the marked news category, and the calculation method includes taking every ten days as a calculation period, and within the calculation period, taking every two days as a calculation node to perform the confidence calculation on the news category, and calculating according to the following formula:

[0022] in, Indicates that the tth computing node The number of people who viewed the marked news category during the period;

[0023] In the formula, w is a positive integer, indicating the increase in the number of people browsing the marked news category; β o Indicates o The βth browsing time of the marked news category among the number of people calculated this time;

[0024] The browsing time of the top three news categories is collected in the t-th computing node, wherein the news category with the longest browsing time is determined as a news category with high confidence, and the news category with high confidence includes news categories with a browsing time of more than one hour; those with a browsing time less than the browsing time are determined as news categories with low confidence, and in the future period, the news category with high confidence and the news category with low confidence are respectively given as the main pushed news category and the secondary pushed news category.

[0025] As a preferred solution of the present invention, in which: in the news category with high confidence, the browsing time period of the corresponding browsing population is obtained, and the obtaining method includes obtaining 2 to 3 most concentrated browsing time periods in one of the computing nodes, and in the browsing time period, the news category with high confidence is given as the news category of the first push series, and outside the browsing time period, the reference news category is given as the news category of the second push series.

[0026] As a preferred solution of the present invention, in which: in the news category of the first push series, the word count characteristics of the corresponding news category are collected to obtain the word count range; the acquisition method includes obtaining the total number of browsed words of the corresponding browsing population and the number of people in the corresponding browsing population in a browsing period, and according to the total number of browsed words, the number of news items corresponding to the average number of words in the total number of browsed words is pushed to the corresponding viewers in the future period.

[0027] As a preferred solution of the present invention, among the news items, the browsing method of the corresponding browsing group is obtained, and the browsing method includes voice broadcast, and the time period occupied by the voice broadcast is obtained in the 2 to 3 most concentrated browsing time periods obtained, and the corresponding news is pushed for voice broadcast in the future time period.

[0028] As a preferred solution of the present invention, in which: in the 2 to 3 most concentrated browsing time periods, the time periods are divided into a front browsing time period, a middle browsing time period and a rear browsing time period, and based on the front browsing time period, the random gradient descent method is used to segment and word the last voice-announced news in the time period, and a number of word segments are divided, and keywords involved in the voice-announced content are extracted from each of the divided word segments, and the corresponding news events are searched based on the keywords to obtain the news category corresponding to the news event, and in the corresponding browsing time period, if the viewer does not switch the voice-announced news, the news category will be used as the first news category to be pushed in the subsequent browsing time period; otherwise, it will not be pushed.

[0029] As a preferred solution of the present invention, the intermediate browsing period is divided into 2 to 5 browsing nodes, each browsing node is given as ten minutes, and the switching data of the first pushed news category by the browser in a certain browsing node is obtained. When it is collected at the corresponding browsing node that the first pushed news category is switched during the broadcast process, the news category is cancelled as the first pushed news category in the future period; otherwise, it is not cancelled.

[0030] Beneficial effects:

[0031] The present invention calculates the average number of views per capita for news released by a news publishing platform in a preset area with each natural day as the collection period, thereby obtaining the browsing behavior of the browsing population in the preset area, and then pushing the corresponding news category to the browsing population in the preset area in a targeted manner. On the basis of the targeted pushing of news categories, the fluctuation of the number of viewers of the news category can be further collected to more accurately collect the browsing preferences of viewers. When a certain news category is obtained to have a high fluctuation, news that viewers outside the news category are willing to browse can be obtained, and then relevant news can be pushed to viewers in a more comprehensive manner. In addition, the browsing time period with the most concentrated browsing population in the preset area can be collected, and the browsing method can be obtained during this time period, so as to improve the browsing experience of viewers from a more humane perspective, avoid the shallowness and fragmentation of news push, and ensure the dissemination effect of news. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. Among them:

[0033] Figure 1A schematic diagram of a modular structure of a media convergence management system according to an embodiment of the present invention;

[0034] Figure 2 It is a schematic diagram of the process structure of an embodiment of the present invention;

[0035] Numbers in the figure: 110 - target population news data acquisition module; 120 - target news data browsing fusion calculation unit; 1201 - news data feature acquisition module; 1202 - browsing volume fluctuation analysis module; 1203 - determination module; 130 - fluctuation data verification module. DETAILED DESCRIPTION

[0036] In order to make the purpose, technical solution and advantages of the embodiment of the present invention clearer, the technical solution of the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings of the embodiment of the present invention. Obviously, the described embodiment is a part of the embodiment of the present invention, not all of the embodiments. Based on the described embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field belong to the scope of protection of the present invention.

[0037] As the media content in today's society is becoming more and more entertainment-oriented, it is difficult for today's integrated media to conduct basic data analysis and public opinion analysis based on functions and applications, and cannot provide scientific data support for operational decisions, resulting in the shallowness and fragmentation of news, which is not conducive to ensuring the effectiveness of news dissemination.

[0038] Based on this, the present invention proposes a media convergence management system, which can obtain the browsing behavior of the browsing population in the preset area by calculating the average number of views per capita for news released on the news release platform in a preset area with each natural day as the collection period, and then push the corresponding news categories to the browsing population in the preset area in a targeted manner, and can also collect the browsing time periods with the most concentrated browsing population in the preset area, and obtain the browsing methods during these time periods, so as to improve the browsing experience of viewers from a more humane perspective, avoid the shallowness and fragmentation of news push, and ensure the dissemination effect of news.

[0039] The present invention is further described in detail below through embodiments and in conjunction with the accompanying drawings.

[0040] Reference Figure 1 to Figure 2 , is an embodiment of the present invention, which provides a media convergence management system, including:

[0041] The target group news data acquisition module 110 is used to distinguish the news browsing groups in a preset area, obtain their browsing content, and mark the corresponding groups according to the browsing content;

[0042] It should be noted that in this embodiment, the news browsing population in the preset area is differentiated into hard news browsing population and soft news browsing population; wherein hard news includes political, economic and social news events;

[0043] Soft news includes cultural, entertainment, sports and educational news events;

[0044] The target news data browsing fusion calculation unit 120 is used to calculate the average number of views per capita for the news publishing platform in the preset area with each natural day as the collection period, and to sort the news publishing platforms in the preset area in a sequential order based on the calculated average number of views per capita with each month as a statistical period. The target news data browsing fusion calculation unit 120 includes a news data feature collection module 1201, a pageview fluctuation analysis module 1202 and a determination module 1203;

[0045] The news data feature collection module 1201 is used to collect the news categories released by the corresponding news platform on each natural day, and the news categories include news news, communication news, commentary news and feature news;

[0046] It should be noted that the news in this embodiment is news that briefly and quickly reports on recent events;

[0047] Newsletter news is long-form news that provides detailed and in-depth coverage of news events;

[0048] Commentary news refers to news that expresses opinions and views on news events, including editorials, commentator articles and letters to readers;

[0049] Feature news is news that reports on a certain news event or person;

[0050] The pageview fluctuation analysis module 1202 responds to the news categories collected by the news data feature collection module and published on the corresponding news platform, and is used to analyze the fluctuation changes in the number of daily pageviews of different news categories, and generate a data set based on a statistical cycle;

[0051] At the same time, the fluctuation of the number of daily viewers of different news categories is divided into high fluctuation, medium fluctuation and low fluctuation.

[0052] The determination module 1203 is used to limit the frequency of news categories released by the news platform in the future period by using the daily browsing data of every 100 people as reference data. In the reference data, if the daily browsing fluctuation number of a certain news category released by the news platform is ≥50 people, the system determines that the corresponding news category released by the news platform is a news category with high fluctuation, and reduces the push frequency of the news category in the future period;

[0053] If the daily fluctuation in the number of viewers of a certain news category published by a news platform is ≥30, the system will determine that the corresponding news category published by the news platform is a news category with medium fluctuation, and will not change the push frequency of the news category in the future period;

[0054] If the daily fluctuation in the number of people who browse a certain news category published by a news platform is ≤20, the system will determine that the corresponding news category published by the news platform is a news category with low fluctuations, and increase the push frequency of the news category in the future period;

[0055] The fluctuation data verification module 130 is used to preset the corresponding news category as the reference news category based on the high fluctuation change on each natural day, and collect the browsing data of the corresponding number of viewers for the news category other than the reference news category; if the browsing data is not collected, the system determines that there is still browsing demand for the reference news category among the corresponding number of viewers, otherwise, no determination is made;

[0056] It should be emphasized in this embodiment that if browsing data of the corresponding number of viewers for news categories other than the reference news category is collected, the corresponding news category is sorted according to the browsing data in the order of θ1, θ2, ..., θ n The number of viewers of each news category is counted, and the top three news categories are obtained. The news categories are marked, and the marked news categories are pushed to the viewers corresponding to the high fluctuation changes in the future time period.

[0057] Furthermore, this embodiment performs confidence calculation on the marked news category, and the calculation method includes taking every ten days as a calculation period, and within the calculation period, taking every two days as a calculation node to perform confidence calculation on the news category, and calculating according to the following formula:

[0058] in, Indicates that the tth computing node The number of people who viewed the marked news category during the period;

[0059] In the formula, w is a positive integer, indicating the increase in the number of people browsing the marked news category; β o Indicates o The βth browsing time of the marked news category among the number of people calculated this time;

[0060] The browsing time of the top three news categories is collected in the t-th computing node, among which the news category with the most browsing time is determined as a news category with high confidence, and the news category with high confidence includes the news category with browsing time of more than one hour; the news category with less than the browsing time is determined as a news category with low confidence, and in the future period, the news category with high confidence and the news category with low confidence are respectively given as the main pushed news category and the secondary pushed news category;

[0061] At the same time, in this embodiment, in the news categories with high confidence, the browsing time period of the corresponding browsing population is obtained, and the obtaining method includes obtaining 2 to 3 most concentrated browsing time periods in one computing node, and in the browsing time period, the news categories with high confidence are given as the news categories of the first push series, and outside the browsing time period, the reference news categories are given as the news categories of the second push series;

[0062] On the basis of the above, this embodiment collects the word count features of the corresponding news category in the news category of the first push series to obtain the word count range; the acquisition method includes obtaining the total number of browsed words and the number of people in the corresponding browsing group in a browsing period, and according to the total number of browsed words, pushing the number of news items corresponding to the average number of words in the total number of browsed words to the corresponding viewers in the future period.

[0063] Among them, this embodiment obtains the browsing method of the corresponding browsing group in the number of news items, and the browsing method includes voice broadcast. The time period occupied by the voice broadcast is obtained in the 2 to 3 most concentrated browsing time periods, and the corresponding news is pushed for voice broadcast in the future time period.

[0064] This embodiment further divides the time periods into a front browsing period, a middle browsing period and a back browsing period in the 2 to 3 most concentrated browsing time periods, and uses the random gradient descent method based on the front browsing period to perform sentence segmentation and word segmentation processing on the last voice-announced news in the period, and divides a number of word segments, extracts keywords involved in the voice-announced content from each of the divided word segments, searches for corresponding news events based on the keywords, and obtains the news category corresponding to the news event. If the viewer does not switch to the voice-announced news in the corresponding browsing period, the news category will be used as the first news category to be pushed in the subsequent browsing period; otherwise, it will not be pushed.

[0065] On the basis of the above, this embodiment specifically divides the intermediate browsing period into 2 to 5 browsing nodes, each browsing node is given as ten minutes, and the switching data of the browser for the first pushed news category in a certain browsing node is obtained. When it is collected at the corresponding browsing node that the first pushed news category is switched during the broadcast process, the news category is cancelled as the first pushed news category in the future period; otherwise, it is not cancelled.

[0066] In summary, the present invention calculates the average number of views per capita for news released by a news publishing platform in a preset area with each natural day as the collection period, thereby obtaining the browsing behavior of the browsing population in the preset area, and then pushing the corresponding news categories to the browsing population in the preset area in a targeted manner. On the basis of the targeted push of news categories, the fluctuation of the number of viewers of the news category can be further collected to more accurately collect the browsing preferences of viewers. When a certain news category is obtained to have a high fluctuation, news that viewers outside the news category are willing to browse can be obtained, and then relevant news can be pushed to viewers in a more comprehensive manner. In addition, the browsing time period with the most concentrated browsing population in the preset area can be collected, and the browsing method can be obtained during this time period, so as to improve the browsing experience of viewers from a more humane perspective, avoid the shallowness and fragmentation of news push, and ensure the dissemination effect of news.

[0067] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A media integration management system, characterized in that: include: The target population news data acquisition module is used to distinguish the news browsing population in a preset area, obtain their browsing content, and mark the corresponding population according to the browsing content; The target news data browsing fusion calculation unit is used to calculate the average number of views per capita for the news publishing platform in the preset area with each natural day as the collection period, and based on the calculated average number of views per capita, the news publishing platform in the preset area is sorted in order with each month as a statistical period. The target news data browsing fusion calculation unit includes a news data feature collection module, a pageview fluctuation analysis module and a determination module; The news data feature collection module is used to collect news categories released by the corresponding news platform on each natural day, and the news categories include news, communication news, commentary news and feature news; The pageview fluctuation analysis module responds to the news categories collected by the news data feature collection module and published on the corresponding news platform, and is used to analyze the fluctuation changes in the number of daily pageviews of different news categories, and generate a data set based on one of the statistical cycles; At the same time, the fluctuation of the number of daily viewers of different news categories is divided into high fluctuation, medium fluctuation and low fluctuation. The determination module is used to limit the frequency of news categories released by the news platform in the future period by using the daily browsing data of every 100 people as reference data, wherein, in the reference data, if the daily browsing fluctuation number of a certain news category released by the news platform is ≥50 people, the system determines that the corresponding news category released by the news platform is a news category with high fluctuation, and reduces the push frequency of the news category in the future period; If the daily fluctuation in the number of people who browse a certain news category published by the news platform is ≥30, the system will determine that the corresponding news category published by the news platform is a news category with medium fluctuation, and will not change the push frequency of the news category in the future period; If the daily fluctuation in the number of people viewing a certain news category published by a news platform is ≤20, the system will determine that the corresponding news category published by the news platform is a news category with low fluctuations, and increase the push frequency of the news category in the future period; The fluctuation data verification module is used to preset the corresponding news category as the reference news category based on the high fluctuation change on each natural day, and collect the browsing data of the corresponding number of viewers for news categories other than the reference news category; if the browsing data is not collected, the system determines that there is still browsing demand for the reference news category among the corresponding number of viewers, otherwise, no determination is made.

2. A media integration management system according to claim 1, characterized in that: In the target population news data acquisition module, the news browsing population in the preset area is differentiated, including being differentiated into hard news browsing population and soft news browsing population; wherein the hard news includes political, economic and social news events; Soft news includes cultural, entertainment, sports and educational news events.

3. A media integration management system according to claim 1, characterized in that: In the fluctuation data verification module, if the browsing data of the corresponding number of viewers for news categories other than the reference news category is collected, the corresponding news category is divided into θ1, θ2, ..., θ n The news categories are distinguished, n represents the nth news category, and the number of viewers of each news category is counted among the number of viewers, the top three news categories are obtained, and the news categories are marked. At the same time, in the future time period, the push of the marked news categories is increased among the viewing population corresponding to the high fluctuation change.

4. A media integration management system as claimed in claim 3, characterized in that: The confidence of the marked news category is calculated, and the calculation method includes taking every ten days as a calculation period, and within the calculation period, taking every two days as a calculation node to calculate the confidence of the news category, and calculating according to the following formula: in, Indicates that the tth computing node The number of people who viewed the marked news category during the period; In the formula, w is a positive integer, indicating the increase in the number of people who browse the marked news category; β o Indicates o The βth browsing time of the marked news category among the number of people calculated this time; The browsing time of the top three news categories is collected in the t-th computing node, wherein the news category with the longest browsing time is determined as a news category with high confidence, and the news category with high confidence includes news categories with a browsing time of more than one hour; those with a browsing time less than the browsing time are determined as news categories with low confidence, and in the future period, the news category with high confidence and the news category with low confidence are respectively given as the main pushed news category and the secondary pushed news category.

5. A media integration management system as claimed in claim 4, characterized in that: In the news category with high confidence, the browsing time period of the corresponding browsing population is obtained, and the obtaining method includes obtaining 2 to 3 most concentrated browsing time periods in one of the computing nodes, and in the browsing time period, the news category with high confidence is given as the news category of the first push series, and outside the browsing time period, the reference news category is given as the news category of the second push series.

6. A media integration management system as claimed in claim 5, characterized in that: In the news categories of the first push series, word count features of the corresponding news categories are collected to obtain word count ranges; The acquisition method includes obtaining the total number of browsed words and the number of people in the corresponding browsing group in a browsing period, and according to the total number of browsed words, pushing the number of news items corresponding to the average number of words in the total number of browsed words to the corresponding viewers in the future period.

7. A media integration management system as claimed in claim 6, characterized in that: From the number of news items, the browsing method of the corresponding browsing group is obtained, the browsing method includes voice broadcast, and the time period occupied by the voice broadcast is obtained in the 2 to 3 most concentrated browsing time periods obtained, and the corresponding news is pushed for voice broadcast in the future time period.

8. A media integration management system as claimed in claim 7, characterized in that: In the 2-3 most concentrated browsing time periods, the time periods are divided into a front browsing time period, a middle browsing time period and a rear browsing time period, and based on the front browsing time period, the last voice broadcast news in the time period is segmented and word segmented using the stochastic gradient descent method to divide a number of word segments, and keywords involved in the voice broadcast content are extracted from each of the divided word segments, and a news event corresponding to the keyword is searched based on the keyword to obtain a news category corresponding to the news event, and if the viewer does not switch the voice broadcast news in the corresponding browsing time period, the news category is used as the first news category to be pushed in the subsequent browsing time period; Otherwise, it will not be pushed.

9. A media integration management system as claimed in claim 8, characterized in that: The intermediate browsing period is divided into 2 to 5 browsing nodes, each of which is given a duration of 10 minutes, and the switching data of the first pushed news category by the browser in a certain browsing node is obtained. When it is collected at the corresponding browsing node that the first pushed news category is switched during the broadcasting process, the news category is cancelled as the first pushed news category in the future period; Otherwise, it will not be cancelled.

Images