Intelligent overseas media content distribution system and method based on artificial intelligence

Abstract

The invention relates to the technical field of content recommendation, and particularly discloses an intelligent overseas media content distribution system and method based on artificial intelligence, the system comprises a division unit and a group division unit, the region division unit divides a digital map into a plurality of user fields and marks buildings in the user fields; the group division module divides males and females in a user field into a first group and a second group, divides the first group and the second group into a plurality of third groups according to the news field to which the highest interest score of each user belongs, and divides the first group and the second group into a plurality of third groups; the news interest of each person in the third group with the activity range and the news interest tendency highly overlapped is considered in the recommendation process, and the same topic is provided for the users in the same user field in the extracurricular and chatting process in the recommendation mode, so that the users are indirectly helped to review the news content in life, and the learning result is consolidated.

Description

Technical Field

[0001] The present invention relates to the technical field of content recommendation, and more specifically, to an artificial intelligence-based intelligent distribution system and method for overseas media content. Background Art

[0002] Currently, there are various development directions for recommending overseas media content in China, with the application of this technology to exams being a particularly important area. Some applications offer bilingual introductions to foreign political and economic events relevant to the exam process, helping users understand these events and improve their English reading skills. Generally speaking, overseas media content recommendations for exams focus on hot news and exam-related content based on user data. However, this recommendation model is largely based on individual preferences, making it difficult to promote subsequent review and study.

[0003] In view of this, the present invention proposes an overseas media content intelligent distribution system and method based on artificial intelligence. Summary of the Invention

[0004] The purpose of the present invention is to provide an AI-based intelligent distribution system and method for overseas media content, which solves the following technical problems:

[0005] How to solve the problem that the traditional recommendation model only considers personal preferences and makes recommendations, which cannot promote subsequent review and study.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] An AI-based intelligent distribution system for overseas media content, comprising:

[0008] An information acquisition module, a classification module, and a recommendation module are used within the application software to acquire user information, including user gender, user location, and the user's interest score for each news category. The number of news categories is preset, including at least politics, economy, entertainment, and military. The interest score is calculated based on the user's browsing history and search history. The interest score for a specific news category can be calculated using a preset calculation formula based on the frequency and number of occurrences of a keyword in a news category in the browsing history, combined with the number of searches for that news category in the active search history.

[0009] The division module includes a region division unit and a group division unit. The region division unit divides a digital map into several user areas and marks buildings in the user areas. The user area is the area where the user resides or conducts daily activities, and can generally be a campus including surrounding accommodation areas. The group division module divides males and females in a user area into a first group and a second group, and divides the first group and the second group into several third groups based on the news area to which each user has the highest interest score. For example, the user with the highest interest score in the political field in the first group is divided into a third group, and the user with the highest score in the political field in the second group is also divided into a third group.

[0010] The recommendation module includes a first recommendation position and a second recommendation position. The recommendation module matches the news content with each user in a third group to obtain a matching score, and calculates a recommendation coefficient based on the matching score. The first recommendation list and the second recommendation list are generated based on the recommendation coefficient. The first recommendation list includes news content in the same interest field as the current third group, and the second recommendation list includes news content in a different interest field from the current third group.

[0011] Through the above technical solution: a process of group division and obtaining a first recommendation list and a second recommendation list according to the group division results is provided. The present invention recommends news to users through the first recommendation list and the second recommendation list. Compared with the traditional recommendation method that only considers personal interests, the present invention considers the news interests of each person in the third group whose activity scope and news interest tendencies are highly overlapping during the recommendation process, thereby obtaining a "common denominator" of news content for recommendation. Under this recommendation mode, the same topics are provided to users in the same user field during extracurricular activities and chats, thereby indirectly helping users review news content in life and consolidate learning outcomes.

[0012] As a further technical solution of the present invention, the process of calculating and obtaining the recommendation coefficient according to the matching score includes:

[0013] Based on the current news content, through the formula Get the recommended coefficient In, where δ i is the weight coefficient of the i-th person in the current third group, P i is the matching score between the i-th person in the current third group and the current news content, i is a non-zero positive integer not greater than N, and N is the total number of people in the current third group.

[0014] Through the above technical solution: a process for obtaining a recommendation coefficient is given. The recommendation coefficient of the present invention is obtained based on the interests of each person in the third group after grouping. Its purpose is to provide the same topics and chat content for users in the same user field, thereby helping users consolidate and remember overseas news knowledge in the process of daily life.

[0015] As a further technical solution of the present invention: the process of obtaining the weight coefficient of the i-th person in the third group includes:

[0016] By formula Get the weight coefficient δ of the i-th person in the third group i , where n1 is the total number of selected people in the third group.

[0017] Through the above technical solution: a method for obtaining the weight coefficient of the i-th person in the third group is provided. In the present invention, the initial weight coefficient of each person in the third group is 1. With the appearance of the selected group, the weight of the selected group in the recommendation coefficient acquisition process is increased, and dynamic weight adjustment is achieved. Different recommended content is obtained in different situations, which is more in line with the user's usage habits.

[0018] As a further technical solution of the present invention: a method for obtaining the selected population in the third group includes:

[0019] Set the monitoring period and obtain user information for the corresponding date of the previous monitoring period;

[0020] Based on the acquired user information, the number of people who stayed in each marked building of the third group in the user domain during each hour of the activity time is determined. The total number of stays is obtained based on the number of stays. The sum of the number of stays in each hour of the activity time is the total number of stays. The number of stays in the current hour is determined by setting a critical time, such as half an hour. If a user of the current third group stays in the current building for more than half an hour during the current hour of the activity time, it is counted as one stayer.

[0021] If the ratio of the total number of stays in a marked building in the user's area to the total number of people in the third group exceeds 1 / 3, the activity coefficient of the current building is calculated;

[0022] The activity coefficient is used to determine whether the users corresponding to the total number of stays in a marked building belong to the selected group.

[0023] As a further technical solution of the present invention, the process of calculating and obtaining the activity coefficient of the current building includes:

[0024] By formula Get the activity coefficient Ac, where λ1 and λ2 are the preset first and second weight coefficients λ1+λ2=1, and λ1∈[0.2,0.4], λ2∈[0.6,0.8], n a is the total number of users staying in a marked building on that day, T a is the duration of overlapping activities, c is the total duration of activities, c∈[8,16] hours.

[0025] Through the above technical solution: a method for obtaining a selected group of people is provided. The selected group of people of the present invention first obtains an activity coefficient based on the historical data of the activity time of several users who meet specific requirements in the third group. The activity coefficient is proportional to the total number of users who stay in a marked building on that day, and is proportional to the length of overlapping activities. As the total number of users who stay in a marked building on that day and the length of overlapping activities increase, the larger the value of the activity coefficient is, which means that the several users who meet specific requirements in the third group will have more chances to meet in the current building during the activity time, thereby providing more opportunities for these users to chat with the news content in the second recommendation list.

[0026] As a further technical solution of the present invention, the process of obtaining the duration of overlapping activities includes:

[0027] Get the number of people staying in each marked building of the third group during each hour of the activity time;

[0028] The duration of overlapping activities is equal to the number of hours of the activity time multiplied by the number of people staying in each marked building within each hour minus one.

[0029] As a further technical solution of the present invention, the process of determining whether the users corresponding to the total number of stays at a marked building are selected groups according to the activity coefficient includes:

[0030] Set a judgment threshold and compare the activity coefficient with the judgment threshold;

[0031] If the activity coefficient is greater than the judgment threshold, the users corresponding to the total number of stays in the marked building are the selected group, otherwise they are judged as not being the selected group.

[0032] As a further technical solution of the present invention: the first recommendation list and the second recommendation list are both used for recommending content on the homepage after the application software is opened, and the second recommendation list is used for a pop-up window before the application software is opened.

[0033] An artificial intelligence-based method for intelligent distribution of overseas media content includes the following steps:

[0034] Obtain user information, including user gender, user location, and user interest rating for each news area;

[0035] Divide a number of user areas on a digital map and mark the buildings in the user areas;

[0036] Divide males and females in a user domain into the first group and the second group, and divide the first group and the second group into several third groups based on the news domain with the highest interest score of each user;

[0037] Match the news content with each user in a third group to obtain a matching score, and calculate a recommendation coefficient based on the matching score;

[0038] A first recommendation list and a second recommendation list are generated according to the recommendation coefficient.

[0039] Beneficial effects of the present invention:

[0040] (1) The present invention takes into account the news interests of each person in the third group whose activity scope and news interest tendencies are highly overlapping during the recommendation process. Under this recommendation model, the same topics are provided to users in the same user field during extracurricular activities and chats, thereby indirectly helping users review news content in their lives and consolidate their learning outcomes.

[0041] (2) The recommendation coefficient of the present invention is obtained based on the interests of each person in the third group after grouping. Its purpose is to provide the same topics and chat content to users in the same user field, thereby helping users consolidate and remember overseas news knowledge in their daily lives.

[0042] (3) In the present invention, the initial weight coefficient of each person in the third group is 1. With the appearance of the selected group, the weight of the selected group in the process of obtaining the recommendation coefficient is increased, and dynamic weight adjustment is achieved. Different recommended content is obtained in different situations, which is more in line with the user's usage habits.

[0043] (4) The activity coefficient of the present invention is proportional to the total number of users who stay in a marked building on that day, and is proportional to the duration of overlapping activities. As the total number of users who stay in a marked building on that day and the duration of overlapping activities increase, the larger the activity coefficient value is, which means that there are more chances for several users in the third group who meet specific requirements to meet in the current building during the activity time, thereby providing more opportunities for these users to chat with each other using the news content in the second recommendation list. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] The present invention will be further described below with reference to the accompanying drawings.

[0045] Figure 1 This is a schematic diagram of the module composition of the intelligent distribution system of the present invention;

[0046] Figure 2It is a flowchart of the steps of the intelligent distribution method of the present invention. DETAILED DESCRIPTION

[0047] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0048] See also Figure 1 As shown, in one embodiment, a system and method for intelligent distribution of overseas media content based on artificial intelligence is provided, including:

[0049] The information acquisition module, the classification module, and the recommendation module are used within the application software to obtain user information, including user gender, user location, and the user's interest score for each news field. The number of news fields is pre-set, including at least politics, economy, entertainment, and military. The interest score is calculated based on the user's browsing history and search history. The interest score of a specific news field can be calculated by combining the frequency and number of occurrences of keywords in a news field in the browsing history with the number of searches for the news field in the active search history using a preset calculation formula. There are many preset calculation formulas for calculating interest scores in related fields, which will not be described in detail. The user's gender is obtained by the user filling out a form, and the user's location is obtained by obtaining positioning permissions;

[0050] It should be noted that the user information obtained in the information acquisition module comes from public databases or data information that the user agrees to disclose.

[0051] The division module includes a region division unit and a group division unit. The region division unit divides a number of user areas on a digital map and marks the buildings in the user areas. The user area is the area where the user resides or conducts daily activities, which can usually be a campus including surrounding accommodation areas. The group division module divides males and females in a user area into a first group and a second group, and divides the first group and the second group into several third groups based on the news area with the highest interest score of each user. For example, the user with the highest interest score in the political field in the first group is divided into a third group, and the user with the highest score in the political field in the second group is also divided into a third group.

[0052] The recommendation module includes a first recommendation position and a second recommendation position. The recommendation module matches the news content with each user in a third group to obtain a matching score, and calculates a recommendation coefficient based on the matching score. The first recommendation list and the second recommendation list are generated based on the recommendation coefficient. The first recommendation list includes news content in the same interest field as the current third group, and the second recommendation list includes news content in a different interest field from the current third group.

[0053] In this embodiment, a process of group division and obtaining a first recommendation list and a second recommendation list based on the group division results is provided. The present invention recommends news to users through the first recommendation list and the second recommendation list. Compared with the traditional recommendation method that only considers personal interests, the present invention considers the news interests of each person in the third group whose activity scope and news interest tendencies are highly overlapping during the recommendation process, thereby obtaining a "common denominator" of news content for recommendation. Under this recommendation mode, the same topics are provided to users in the same user field during extracurricular activities and chats, thereby indirectly helping users review news content in life and consolidate learning outcomes.

[0054] The process of calculating the recommendation coefficient based on the matching score includes:

[0055] Based on the current news content, through the formula Get the recommended coefficient In, where δ i is the weight coefficient of the i-th person in the current third group, P i is the matching score between the i-th person in the current third group and the current news content, i is a non-zero positive integer not greater than N, and N is the total number of people in the current third group.

[0056] In this embodiment, a process for obtaining a recommendation coefficient is provided. The recommendation coefficient of the present invention is obtained based on the interests of each person in the third group after grouping. Its purpose is to provide the same topics and chat content to users in the same user field, thereby helping users consolidate and remember overseas news knowledge in their daily lives.

[0057] The process of obtaining the weight coefficient of the i-th person in the third group includes:

[0058] By formula Get the weight coefficient δ of the i-th person in the third group i , where n1 is the total number of selected people in the third group.

[0059] In this embodiment, a method for obtaining the weight coefficient of the i-th person in the third group is provided. In the present invention, the initial weight coefficient of each person in the third group is 1. As the selected group appears, the weight of the selected group in the recommendation coefficient acquisition process is increased to achieve dynamic weight adjustment, obtain different recommended content in different situations, and better suit the user's usage habits.

[0060] The methods of obtaining selected people in the third group include:

[0061] Set the monitoring period and obtain the user information of the corresponding date of the previous monitoring period. The monitoring period is usually set to seven days. For example, if the current date is Wednesday, obtain the user information of the corresponding Wednesday of the previous monitoring period.

[0062] Based on the acquired user information, the number of people who stayed in each marked building of the third group in the user domain during each hour of the activity time is determined. The total number of stays is obtained based on the number of stays. The sum of the number of stays in each hour of the activity time is the total number of stays. The number of stays in the current hour is determined by setting a critical time, such as half an hour. If a user of the current third group stays in the current building for more than half an hour during the current hour of the activity time, it is counted as one stayer.

[0063] If the ratio of the total number of stays in a marked building in the user's area to the total number of people in the third group exceeds 1 / 3, the activity coefficient of the current building is calculated;

[0064] The activity coefficient is used to determine whether the users corresponding to the total number of stays in a marked building belong to the selected group.

[0065] The process of calculating the activity coefficient of the current building includes:

[0066] By formula Get the activity coefficient Ac, where λ1 and λ2 are the preset first and second weight coefficients λ1+λ2=1, and λ1∈[0.2,0.4], λ2∈[0.6,0.8], n a is the total number of users staying in a marked building on that day, T a is the duration of overlapping activities, c is the total duration of activities, c∈[8,16] hours.

[0067] In this embodiment, a method for obtaining a selected group of people is provided. The selected group of people of the present invention first obtains an activity coefficient based on the historical data of several users in the third group who meet specific requirements during activity time. The activity coefficient is proportional to the total number of users who stay in a marked building on that day, and is proportional to the length of overlapping activities. As the total number of users who stay in a marked building on that day and the length of overlapping activities increase, the larger the value of the activity coefficient, which means that several users in the third group who meet specific requirements will have more chances to meet in the current building during activity time, thereby providing more opportunities for these users to chat with the news content in the second recommendation list.

[0068] The process of obtaining the overlapping activity duration includes:

[0069] Get the number of people staying in each marked building of the third group during each hour of the activity time;

[0070] The duration of overlapping activities is equal to the product of the number of hours of activity time and the number of people staying in each marked building per hour minus one, that is, through the formula T a =c×(n a -1) Get the duration of overlapping activities T a .

[0071] The process of determining whether the users corresponding to the total number of stays at a marked building belong to the selected group based on the activity coefficient includes:

[0072] Setting a judgment threshold. In this embodiment, the judgment threshold is in the range of 1 / 2 to 2 / 3 of the maximum value of the activity coefficient, and comparing the activity coefficient with the judgment threshold.

[0073] If the activity coefficient is greater than the judgment threshold, the users corresponding to the total number of stays in the marked building are the selected population, otherwise it is judged that they are not the selected population. It should be noted that the process of obtaining the number of selected population also includes: if the selected population does not exist, the number of selected population is counted as zero. If there is no marked building in the user area whose total number of stays exceeds 1 / 3 of the total number of people in the third group, the number of selected population is also output as zero.

[0074] The first recommendation list and the second recommendation list are both used for recommending content on the homepage after the application software is opened, and the second recommendation list is used for the pop-up window before the application software is opened.

[0075] refer to Figure 2 This embodiment also provides an artificial intelligence-based method for intelligent distribution of overseas media content, comprising the following steps:

[0076] S100: Obtain user information, including user gender, user location, and user interest rating for each news area;

[0077] S200, dividing a number of user areas on a digital map and marking buildings in the user areas;

[0078] S300, dividing males and females in a user domain into a first group and a second group, and dividing the first group and the second group into a plurality of third groups based on the news domain to which each user has the highest interest score;

[0079] S400, matching the news content with each user in a third group to obtain a matching score, and calculating a recommendation coefficient based on the matching score;

[0080] S500. Generate a first recommendation list and a second recommendation list based on the recommendation coefficient, wherein the first recommendation list includes news content in the same interest field as the current third group, and the second recommendation list includes news content in a different interest field from the current third group.

[0081] The above is a detailed description of an embodiment of the present invention. However, the content described is only a preferred embodiment of the present invention and should not be considered to limit the scope of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. An AI-based intelligent distribution system for overseas media content, comprising an information acquisition module, a classification module, and a recommendation module, characterized by: The information acquisition module application software is used to obtain user information, including user gender, user location and user interest score for each news field; The division module includes a region division unit and a group division unit. The region division unit divides a plurality of user areas on a digital map and marks buildings in the user areas. The group division module divides males and females in a user area into a first group and a second group, and further divides the first group and the second group into a plurality of third groups based on the news area with the highest interest score for each user. The recommendation module includes a first recommendation position and a second recommendation position. The recommendation module matches the news content with each user in a third group to obtain a matching score, and calculates a recommendation coefficient based on the matching score. The first recommendation list and the second recommendation list are generated based on the recommendation coefficient. The first recommendation list includes news content in the same interest field as the current third group, and the second recommendation list includes news content in a different interest field from the current third group.

2. The overseas media content intelligent distribution system based on artificial intelligence according to claim 1, characterized in that: The process of calculating the recommendation coefficient based on the matching score includes: Based on the current news content, through the formula Get the recommended coefficient In, where δ i is the weight coefficient of the i-th person in the current third group, P i is the matching score between the i-th person in the current third group and the current news content, i is a non-zero positive integer not greater than N, and N is the total number of people in the current third group.

3. The overseas media content intelligent distribution system based on artificial intelligence according to claim 2, characterized in that: The process of obtaining the weight coefficient of the i-th person in the third group includes: By formula Get the weight coefficient δ of the i-th person in the third group i , where n1 is the total number of selected people in the third group.

4. The overseas media content intelligent distribution system based on artificial intelligence according to claim 3 is characterized in that: The methods of obtaining selected people in the third group include: Set the monitoring period and obtain user information for the corresponding date of the previous monitoring period; Determine the number of people who stay in each marked building of the third group in the user area during each hour of the activity time based on the obtained user information, and obtain the total number of stays based on the number of stays; If the ratio of the total number of stays in a marked building in the user's domain to the total number of people in the third group exceeds 13, the activity coefficient of the current building is calculated; The activity coefficient is used to determine whether the users corresponding to the total number of stays in a marked building belong to the selected group.

5. The overseas media content intelligent distribution system based on artificial intelligence according to claim 4 is characterized in that: The process of calculating the activity coefficient of the current building includes: By formula Get the activity coefficient Ac, where λ1 and λ2 are the preset first and second weight coefficients λ1+λ2=1, and λ1∈[0.2,0.4], λ2∈[0.6,0.8], n a is the total number of users staying in a marked building on that day, T a is the duration of overlapping activities, c is the total duration of activities, c∈[8,16] hours.

6. The overseas media content intelligent distribution system based on artificial intelligence according to claim 5, characterized in that: The process of obtaining the overlapping activity duration includes: Get the number of people staying in each marked building of the third group during each hour of the activity time; The duration of overlapping activities is equal to the number of hours of the activity time multiplied by the number of people staying in each marked building within each hour minus one.

7. The overseas media content intelligent distribution system based on artificial intelligence according to claim 4 is characterized in that: The process of determining whether the users corresponding to the total number of stays at a marked building belong to the selected group based on the activity coefficient includes: Set a judgment threshold and compare the activity coefficient with the judgment threshold; If the activity coefficient is greater than the judgment threshold, the users corresponding to the total number of stays in the marked building are the selected group, otherwise they are judged as not being the selected group.

8. The overseas media content intelligent distribution system based on artificial intelligence according to claim 1 is characterized in that: The first recommendation list and the second recommendation list are both used for recommending content on the homepage after the application software is opened, and the second recommendation list is used for the pop-up window before the application software is opened.

9. An artificial intelligence-based intelligent distribution method for overseas media content, characterized in that: The overseas media content intelligent distribution system based on artificial intelligence as described in claims 1-8 comprises the following steps: Obtain user information, including user gender, user location, and user interest rating for each news area; Divide a number of user areas on a digital map and mark the buildings in the user areas; Divide males and females in a user domain into the first group and the second group, and divide the first group and the second group into several third groups based on the news domain with the highest interest score of each user; Match the news content with each user in a third group to obtain a matching score, and calculate a recommendation coefficient based on the matching score; A first recommendation list and a second recommendation list are generated according to the recommendation coefficient.

Images