System and method for arranging convergence media program list

Abstract

The invention discloses a converged media program list arrangement system and a generation method, and belongs to the technical field of media content management and intelligent scheduling. The system comprises a content acquisition module, a feature extraction module, a user portrait and platform model module, a program optimization and arrangement module, an intelligent conflict detection module, a self-adaptive generation module and a feedback optimization module. Unified management of program resources is realized through multi-source content collection and semantic tagging analysis; a multi-objective optimization model is constructed based on the user portrait and the platform strategy, and program lists adapted to different terminals are automatically generated; and dynamically adjusting a programming result by utilizing a conflict detection and self-learning feedback mechanism. According to the method, the intelligent level of convergence media program management can be remarkably improved, the cross-platform content distribution efficiency is optimized, and automatic generation and accurate pushing of the program list are realized.

Description

Technical Field

[0001] This invention relates to the field of media content management and intelligent scheduling technology, specifically to a converged media program schedule arrangement system and method. Background Technology

[0002] With the increasingly prominent trend of media convergence, traditional radio and television program schedules can no longer meet the needs of contemporary multi-terminal and multi-channel content dissemination. Existing program scheduling systems are mostly oriented towards a single platform such as a television station or a web platform, lacking the function of coordinated scheduling and intelligent matching of multimedia content such as video, audio, short videos, and text and image push.

[0003] Existing program schedule generation typically relies on manual editing or fixed templates, resulting in low scheduling efficiency, uneven resource utilization, and an inability to dynamically adjust based on audience profiles, platform characteristics, or real-time dissemination effects. Furthermore, in cross-platform content delivery, the program playback logic of different platforms varies, such as duration, scheduling, and tagging rules. The lack of unified standards and automated generation mechanisms easily leads to content duplication, time slot conflicts, or poor dissemination effects.

[0004] Therefore, there is an urgent need for a system and method that can intelligently analyze program content characteristics, audience needs and platform strategies, and automatically generate the optimal integrated media program schedule to improve scheduling efficiency and dissemination accuracy, and achieve true multi-platform integrated program scheduling and intelligent generation. Summary of the Invention

[0005] (a) Technical problems to be solved

[0006] To address the shortcomings of existing technologies, this invention provides a multimedia program schedule arrangement system and method, solving the following problems: Existing program schedule generation typically relies on manual editing or fixed templates, resulting in low arrangement efficiency, uneven resource utilization, and an inability to dynamically adjust based on audience profiles, platform characteristics, or real-time dissemination effects. Furthermore, in cross-platform content delivery, the program playback logic of different platforms differs, such as duration, scheduling, and tagging rules, lacking unified standards and automated generation mechanisms, which easily leads to content duplication, time slot conflicts, or poor dissemination effects.

[0007] (II) Technical Solution

[0008] The purpose of this invention is to provide a converged media program schedule arrangement system and generation method. Through technologies such as artificial intelligence, knowledge graphs, and big data analysis, a unified management, intelligent scheduling, and adaptive generation system for multi-source content is constructed. This system enables automatic generation, real-time updating, and intelligent optimization of program schedules across different platforms and terminals, improving program scheduling efficiency and audience matching. This invention is achieved through the following technical solution: A converged media program schedule arrangement system, comprising: a content acquisition module: used to acquire multi-source content information from television station databases, online video platforms, audio podcasts, news clients, and self-media systems, including program metadata and historical playback records;

[0009] Feature extraction module: Performs feature analysis on the collected program content and extracts structured tag information, including topic classification, audience attributes, sentiment, media type, and broadcasting platform characteristics.

[0010] User profiling and platform model module: Construct user profiles through data mining, including information on age, region, interests and preferences, and active time periods; establish platform broadcasting models, analyze the program characteristics of different platforms, and form platform broadcasting strategy templates;

[0011] Program selection and scheduling module: Based on program characteristics, user profiles and platform models, it uses multi-objective optimization algorithms to select and automatically sort programs, and generate a draft program list for multi-platform collaboration.

[0012] Intelligent conflict detection module: Automatically detects whether there are time conflicts, copyright conflicts or duplicate scheduling issues in the program schedule, and provides optimization suggestions;

[0013] Adaptive generation module: Automatically generates program guides in the corresponding format according to the display rules of different media terminals, supporting multiple output formats such as JSON, XML, and EPG;

[0014] Feedback and Optimization Module: Collects program playback effect data in real time, provides feedback and optimization to the generated model, and realizes dynamic adjustment and self-learning update of the program list.

[0015] As a further preferred embodiment of the present invention, the content acquisition module includes: a multi-source access unit, which enables interface connection with TV station databases, OTT platforms, short video platforms, podcast platforms, and self-media systems; adopts multiple protocol access methods such as API calls, RSS subscriptions, and FTP synchronization; sets the update frequency and capture range according to the access source; a data parsing unit, which parses the acquired program data, identifies metadata fields, including program name, type, duration, broadcast time, host, and copyright information; supports multiple data structure parsing formats such as JSON, XML, and CSV; a content cleaning unit, which cleans, deduplicates, and completes redundant or incomplete data; performs fuzzy matching and merging of similar program entries to ensure data consistency; and a unified storage unit, which establishes a unified program content index library by combining a relational database and a distributed file system; and provides a fast retrieval interface based on keywords, themes, and broadcast times.

[0016] As a further preferred embodiment of the present invention, the feature extraction module includes a text semantic analysis unit that performs natural language processing (NLP) on the program synopsis, script, and subtitle text to extract topic keywords and semantic tags; and supports sentiment analysis and hot topic identification.

[0017] The audio and video feature extraction unit performs image recognition and audio recognition on the program video content; it uses a convolutional neural network (CNN) to extract image features and a Transformer model to analyze the audio structure.

[0018] Tag mapping and standardization units map automatically extracted feature tags to a unified tag system; establish a multi-dimensional tag system to realize the semantic computation of programs.

[0019] As a further preferred embodiment of the present invention, the user profile and platform model module includes: a user behavior collection unit that collects user viewing behavior; supporting multi-terminal collection, including mobile apps, TV set-top boxes, and web pages; a user profile modeling unit that generates user profiles based on big data analysis technology; profile dimensions including age, gender, interests, viewing time, and commonly used devices; a platform characteristic modeling unit that establishes broadcast strategy models for different platforms; defining program broadcast duration, format, and tag weight rule templates for each platform; and an audience matching prediction unit that uses a deep learning model to predict the audience matching degree of a specific program; outputting a program-user matching score to provide input parameters for subsequent scheduling.

[0020] As a further preferred embodiment of the present invention, the program selection and scheduling module includes a weight allocation unit that sets adaptive weights for program features, user profiles, and platform weights; dynamically adjusts the weight coefficients based on real-time data; and a relevance calculation unit that utilizes a relevance calculation model between content features and audience features.

[0021] ,in For content feature matching degree, To improve the matching degree of user profiles Platform preference coefficient Content feature weighting coefficient User profile weighting coefficient Platform strategy weighting coefficient;

[0022] The system calculates the overall program matching degree; the sorting optimization unit uses genetic algorithms or reinforcement learning algorithms to sort programs; it supports multi-objective optimization; the dynamic scheduling unit automatically generates a preliminary program list based on the platform's playback timeline; and it performs intelligent time slot allocation based on time slot priority.

[0023] As a further preferred embodiment of the present invention, the intelligent conflict detection module includes: a time conflict detection unit that checks whether program playback times overlap, have insufficient intervals, or exceed the broadcast schedule; automatically calculates and adjusts playback time periods; a copyright conflict detection unit that compares with the program copyright database to prevent illegal repeated broadcasts or over-authorized playback; a duplicate broadcast identification unit that counts the number of times the same program is broadcast on different platforms; triggers an adjustment mechanism when the duplication rate exceeds a set threshold; and an intelligent correction unit that combines optimization algorithms to automatically replace or adjust the order of conflicting programs.

[0024] As a further preferred embodiment of the present invention, the adaptive generation module includes: a format adaptation unit that selects the output format according to the target platform; supporting multiple formats such as EPG, XML, JSON, RSS, and Excel; an interface template generation unit that automatically generates program list page templates according to the platform's UI rules; providing responsive layout and multi-language support; and a multi-terminal compatibility unit that automatically adjusts the resolution, time format, and program cover size to ensure consistent display across multiple terminals.

[0025] As a further preferred embodiment of the present invention, the feedback and optimization module includes a playback data acquisition unit that monitors program playback data in real time, including clicks, completion rate, interaction rate, and comment sentiment index; an effect evaluation unit that quantitatively evaluates the overall effect of the program list and outputs a comprehensive dissemination index; a model self-learning unit that updates the parameters of the algorithm model based on the evaluation results; a reinforcement learning method that optimizes the program recommendation logic to achieve continuous evolution; a strategy adjustment unit that sends the optimization results back to the program selection module to form a closed-loop feedback; and support for manual intervention and strategy template updates.

[0026] A method for generating a multimedia program schedule includes the following steps: Step S1: Multi-source content acquisition and tagging processing.

[0027] We collect content resources from multiple platforms and perform semantic analysis on the program content using natural language processing (NLP) and image recognition technologies to generate standardized metadata tags.

[0028] Step S2: User Profile and Platform Strategy Modeling

[0029] A multi-dimensional user profile model is built based on user behavior data and interaction data, while the playback strategies and broadcast restrictions of each platform are defined.

[0030] Step S3: Program Matching and Optimal Ranking

[0031] By using a matching degree calculation model between content features and user profiles, the relevance of program resources is scored, and a preliminary program list is generated based on a weighted ranking algorithm or reinforcement learning algorithm.

[0032] Step S4: Intelligent Conflict Detection and Correction

[0033] The program schedule is detected for time conflicts, duplicate broadcasts, and mismatches in duration, and the scheduling results are readjusted through the automatic correction module.

[0034] Step S5: Generation and Publishing of Multi-Format Program Guides

[0035] Based on the target platform type, a program schedule file in an adapted format is automatically generated and pushed to the corresponding platform's broadcast control system.

[0036] Step S6: Playback Feedback and Self-Learning Optimization

[0037] Collect program playback effect data, adjust recommendation weights through model training, and achieve intelligent self-optimization of program schedule generation.

[0038] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of a method for generating a multimedia program schedule.

[0039] (III) Beneficial Effects

[0040] This invention provides a converged media program schedule arrangement system and method, which has the following beneficial effects:

[0041] Unified management of multi-source content: Connecting multiple content channels such as radio, television, online video, and social media to achieve unified integration and tagging of content resources.

[0042] Intelligent program scheduling: Through artificial intelligence algorithms, program selection and dynamic sorting are achieved, greatly reducing the workload of manual scheduling.

[0043] Cross-platform adaptive generation: Supports automatic adaptation of program playlist formats to different playback platforms and terminals, reducing manual transcoding and publishing costs.

[0044] Real-time feedback optimization mechanism: The program scheduling strategy can be continuously optimized based on audience feedback data to achieve precise dissemination.

[0045] Highly scalable: The system can be flexibly expanded according to the business scale of the media center, and is suitable for regional converged media centers and large broadcasting organizations. Attached Figure Description

[0046] Figure 1 This is a schematic diagram of the system principle framework of the present invention;

[0047] Figure 2 This is a schematic diagram of the method flow of the present invention. Detailed Implementation

[0048] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0049] Please see Figure 1-2 The present invention provides a technical solution: a converged media program schedule arrangement system, including: a content acquisition module: used to acquire multi-source content information from TV station databases, online video platforms, audio podcasts, news clients, and self-media systems, including program metadata and historical playback records;

[0050] Feature extraction module: Performs feature analysis on the collected program content and extracts structured tag information, including topic classification, audience attributes, sentiment, media type, and broadcasting platform characteristics.

[0051] User profiling and platform model module: Construct user profiles through data mining, including information on age, region, interests and preferences, and active time periods; establish platform broadcasting models, analyze the program characteristics of different platforms, and form platform broadcasting strategy templates;

[0052] Program selection and scheduling module: Based on program characteristics, user profiles and platform models, it uses multi-objective optimization algorithms to select and automatically sort programs, and generate a draft program list for multi-platform collaboration.

[0053] Intelligent conflict detection module: Automatically detects whether there are time conflicts, copyright conflicts or duplicate scheduling issues in the program schedule, and provides optimization suggestions;

[0054] Adaptive generation module: Automatically generates program guides in the corresponding format according to the display rules of different media terminals, supporting multiple output formats such as JSON, XML, and EPG;

[0055] Feedback and Optimization Module: Collects program playback effect data in real time, provides feedback optimization to the generated model, and enables dynamic adjustment and self-learning updates of the program schedule; the data flow between modules is as follows:

[0056] The content acquisition module provides the raw program data to the feature extraction module;

[0057] The feature extraction module outputs structured tags to the program selection module and the user profiling module;

[0058] The user profile and platform model module outputs audience matching scores and platform strategy parameters;

[0059] The program selection and scheduling module generates a draft program list based on the input data, which is then corrected by the conflict detection module.

[0060] The revised program schedule is output and published by the adaptive generation module;

[0061] The final feedback data is collected by the feedback and optimization module, which then adjusts the model weights in reverse to form an intelligent closed loop.

[0062] The content acquisition module includes: a multi-source access unit, which interfaces with TV station databases, OTT platforms, short video platforms, podcast platforms, and self-media systems; it adopts multiple access protocols such as API calls, RSS subscriptions, and FTP synchronization; and it sets the update frequency and capture range according to the access source. A data parsing unit parses the acquired program data, identifies metadata fields including program name, type, duration, broadcast time, host, and copyright information; and supports multiple data structure parsing formats such as JSON, XML, and CSV. A content cleaning unit cleans, deduplicates, and completes redundant or incomplete data; it performs fuzzy matching and merging of similar program entries to ensure data consistency. A unified storage unit uses a combination of relational databases and distributed file systems to establish a unified program content index; and it provides a fast search interface based on keywords, themes, and broadcast times. This module supports unified acquisition of heterogeneous data and multi-protocol fusion, forming the technical foundation for the system's "multi-source integration."

[0063] The feature extraction module includes a text semantic analysis unit that performs natural language processing on program descriptions, scripts, and subtitle texts to extract thematic keywords and semantic tags; it also supports sentiment analysis and hot topic identification.

[0064] The audio and video feature extraction unit performs image recognition and audio recognition on the program video content; it uses a convolutional neural network (CNN) to extract image features and a Transformer model to analyze the audio structure.

[0065] Tag mapping and standardized units map automatically extracted feature tags to a unified tag system; a multi-dimensional tag system is established to realize the computationalization of program semantics. The introduction of AI semantic understanding and multimodal feature extraction upgrades program schedule arrangement from "manual classification" to "intelligent tag-driven".

[0066] The user profiling and platform model module includes: a user behavior collection unit that gathers user viewing behavior data; supporting multi-terminal collection, including mobile apps, TV set-top boxes, and web pages; a user profiling modeling unit that generates user profiles based on big data analytics; profile dimensions include age, gender, interests, viewing time, and commonly used devices; a platform characteristic modeling unit that establishes broadcast strategy models for different platforms; defining program broadcast duration, format, and tag weighting rule templates for each platform; and an audience matching prediction unit that uses deep learning models to predict the audience matching degree of specific programs; outputting program-user matching scores to provide input parameters for subsequent scheduling. This module integrates "user behavior characteristics" with "platform broadcast strategies" in a unified model, realizing the basic logic for generating personalized program schedules.

[0067] The program selection and scheduling module and the weight allocation unit set adaptive weights for program features, user profiles, and platform weights; they dynamically adjust the weight coefficients based on real-time data; the relevance calculation unit utilizes a relevance calculation model between content features and audience features.

[0068] ,in For content feature matching degree, To improve the matching degree of user profiles Platform preference coefficient Content feature weighting coefficient User profile weighting coefficient Platform strategy weighting coefficient;

[0069] The system calculates the overall program matching degree; the sorting optimization unit uses genetic algorithms or reinforcement learning algorithms to sort programs; it supports multi-objective optimization; the dynamic scheduling unit automatically generates a preliminary program list based on the platform's playback timeline; and it performs intelligent time slot allocation based on time slot priority. Machine learning algorithms are introduced for multi-objective program scheduling, enabling the program list to have "adaptive" and "intelligent sorting" capabilities.

[0070] The intelligent conflict detection module includes: a time conflict detection unit that checks whether program playback times overlap, have insufficient intervals, or exceed the broadcast schedule; automatically calculating and adjusting playback periods; a copyright conflict detection unit that compares program copyright databases to prevent illegal duplicate broadcasts or over-authorized playback; a duplicate scheduling identification unit that counts the number of times the same program is scheduled on different platforms; triggering an adjustment mechanism when the duplication rate exceeds a set threshold; and an intelligent correction unit that combines optimization algorithms to automatically replace or adjust the order of conflicting programs. It can automatically identify multiple types of conflicts and has a "self-repair" mechanism, greatly improving the accuracy and completeness of program schedule generation.

[0071] The adaptive generation module includes a format adaptation unit that selects the output format based on the target platform, supporting multiple formats such as EPG, XML, JSON, RSS, and Excel; an interface template generation unit that automatically generates program guide page templates according to the platform's UI rules, providing responsive layout and multi-language support; and a multi-terminal compatibility unit that automatically adjusts resolution, time format, and program cover size to ensure consistent display across multiple terminals. The system enables "generate once, publish to multiple platforms," ​​significantly reducing manual adjustment costs.

[0072] The feedback and optimization module includes a playback data acquisition unit that monitors program playback data in real time, including clicks, completion rate, interaction rate, and comment sentiment indicators; an effect evaluation unit that quantitatively evaluates the overall effect of the program guide and outputs a comprehensive dissemination index; a model self-learning unit that updates the algorithm model parameters based on the evaluation results; and reinforcement learning methods that optimize the program recommendation logic for continuous evolution. The strategy adjustment unit feeds back the optimization results to the program selection module, forming a closed-loop feedback. Manual intervention and strategy template updates are also supported. Through this self-learning closed loop of "real-time data-driven—algorithm update—result optimization," the module achieves dynamic evolution capabilities in program guide generation.

[0073] The method for generating integrated media program schedules includes the following steps: Step S1: Multi-source content acquisition and tagging processing.

[0074] We collect content resources from multiple platforms and perform semantic analysis on the program content using natural language processing (NLP) and image recognition technologies to generate standardized metadata tags.

[0075] Step S2: User Profile and Platform Strategy Modeling

[0076] A multi-dimensional user profile model is built based on user behavior data and interaction data, while the playback strategies and broadcast restrictions of each platform are defined.

[0077] Step S3: Program Matching and Optimal Ranking

[0078] By using a matching degree calculation model between content features and user profiles, the relevance of program resources is scored, and a preliminary program list is generated based on a weighted ranking algorithm or reinforcement learning algorithm.

[0079] Step S4: Intelligent Conflict Detection and Correction

[0080] The program schedule is detected for time conflicts, duplicate broadcasts, and mismatches in duration, and the scheduling results are readjusted through the automatic correction module.

[0081] Step S5: Generation and Publishing of Multi-Format Program Guides

[0082] Based on the target platform type, a program schedule file in an adapted format is automatically generated and pushed to the corresponding platform's broadcast control system.

[0083] Step S6: Playback Feedback and Self-Learning Optimization

[0084] Collect program playback effect data, adjust recommendation weights through model training, and achieve intelligent self-optimization of program schedule generation.

[0085] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of a method for generating a multimedia program schedule.

[0086] Example 1: Standardized Processing of Multi-Source Program Content Acquisition and Cleaning

[0087] Implementation Scenarios

[0088] Taking a municipal-level integrated media center as an example, it needs to regularly obtain program content from television stations, local news websites, video accounts, and podcast platforms to support multi-terminal broadcasting tasks. The system is deployed on the content management server in the central computer room and connects to the content libraries of each platform through API interfaces.

[0089] Work steps

[0090] Data source access:

[0091] The multi-source access unit (101) of the content acquisition module defines interface parameters through a configuration file, such as:

[0092] The TV station interface uses FTP data synchronization, and the program EPG table is retrieved once per hour;

[0093] The news client uses a JSON-API interface to access the latest news short videos in real time;

[0094] Short video platforms retrieve video content tagged with "people's livelihood", "education" and "culture and tourism" after authentication via OAuth 2.0.

[0095] Data analysis and standardization:

[0096] The parsing unit (102) parses data from different platforms into a standard structure.

[0097] For example, the EPG fields returned by the TV station are (Title, Time, Duration), while the news client returns (headline, publishTime, videoLength). The system unifies them through field mapping rules as follows:

[0098] Program name, duration, release date, source platform, program type, tags, and copyright status.

[0099] Content cleaning and deduplication:

[0100] A hash-plus-text similarity algorithm is used to identify duplicate programs. If the program information has a title or summary text similarity exceeding 85%, the version with the higher quality score is retained.

[0101] Technically, through this module, the system has achieved automated aggregation, standardized management, and quality screening of program data from multiple platforms. The data update cycle has been shortened from 3 hours to 15 minutes, significantly improving the efficiency and accuracy of content processing.

[0102] Example 2: Program Feature Extraction and Semantic Tag Modeling

[0103] Implementation Scenarios

[0104] To improve the accuracy of program classification and intelligent recommendation, the system extracts semantic features and models multi-dimensional tags for the collected programs.

[0105] Implementation steps

[0106] Text semantic analysis:

[0107] The system uses a BERT-based Chinese semantic model to perform word vector analysis on program descriptions and script content. For example:

[0108] Input text:

[0109] "This episode focuses on rural education reform and explores the real lives of teachers and students in mountainous areas."

[0110] Output tags:

[0111] Themes: Education reform, rural revitalization; Emotional orientation: Positive; Media type: Documentary; Target audience: Education-conscious groups.

[0112] Audio and video feature extraction:

[0113] The video frames are processed by a CNN model to extract key visual features, such as the presenter and scene recognition; the audio features are extracted by a Transformer model to extract speech rate, tone and emotion indicators.

[0114] Tag fusion and standardization:

[0115] Semantic and visual tags are unified into a four-dimensional feature matrix of "theme-type-emotion-media attribute" through tag mapping units. The system automatically generates program feature vectors for subsequent optimization calculations.

[0116] In terms of technical effectiveness, this module achieves multimodal feature fusion, with program tag accuracy exceeding 93%, providing computationally achievable semantic input for subsequent program scheduling.

[0117] Example 3: User Profiling and Platform Strategy Modeling

[0118] Implementation Scenarios

[0119] The system connects to the user data platform to collect viewing behavior data from different terminal users, which is used to build user profiles and platform preference models.

[0120] Implementation steps

[0121] User behavior collection:

[0122] It collects users' viewing records over the past 30 days, including the number of clicks, playback duration, collection rate, and number of interactions.

[0123] User clustering modeling:

[0124] The K-Means algorithm is used to divide users into five categories:

[0125] News-focused; Variety and entertainment; Educational and learning; Local lifestyle; High-frequency short videos.

[0126] Platform feature modeling:

[0127] The system creates strategy templates based on the platform's playback characteristics, for example:

[0128] TV: Prioritize longer programs, with a recommended duration of ≥ 20 minutes;

[0129] Mobile devices: Short videos are preferred, with a recommended length of ≤ 3 minutes;

[0130] Website: A mixed layout of text and images with short videos.

[0131] Match prediction:

[0132] The matching degree between programs and user groups is calculated using a deep learning model (DNN). Output a score of 0-1.

[0133] For example, the program "Youth Vision" had a match rate of 0.92 among users aged 18-30 and 0.48 among the elderly.

[0134] In terms of technical effectiveness, this module implements a dual modeling mechanism for audience characteristics and platform features, which can accurately guide program sorting and distribution, thereby improving user stickiness and viewing conversion rates.

[0135] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or basic characteristics. Therefore, the embodiments should be considered illustrative and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of the same elements of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0136] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A converged media program schedule arrangement system, characterized in that, include: Content Acquisition Module: Used to collect content information from multiple sources, including TV station databases, online video platforms, audio podcasts, news clients, and self-media systems, including program metadata and historical playback records; Feature extraction module: Performs feature analysis on the collected program content and extracts structured tag information; User profiling and platform model module: Building user profiles through data mining; Establish a platform broadcasting model, analyze the program characteristics of different platforms, and form a platform broadcasting strategy template; Program selection and scheduling module: Based on program characteristics, user profiles and platform models, it uses multi-objective optimization algorithms to select and automatically sort programs, and generate a draft program list for multi-platform collaboration. Intelligent conflict detection module: Automatically detects whether there are time conflicts, copyright conflicts or duplicate scheduling issues in the program schedule, and provides optimization suggestions; Adaptive generation module: Automatically generates program guides in the corresponding format based on the display rules of different media terminals; Feedback and Optimization Module: Collects program playback effect data in real time, provides feedback and optimization to the generated model, and realizes dynamic adjustment and self-learning update of the program list.

2. The converged media program schedule arrangement system according to claim 1, characterized in that: The content acquisition module includes a multi-source access unit, which enables interface connection with TV station databases, OTT platforms, short video platforms, podcast platforms, and self-media systems. It adopts multiple access methods including API calls, RSS subscriptions, and FTP synchronization; the update frequency and crawling scope are set according to the access source. The data parsing unit analyzes the collected program data, identifies metadata fields including program name, type, duration, broadcast time, host, and copyright information; it supports parsing multiple data structure formats such as JSON, XML, and CSV; the content cleaning unit cleans, deduplicates, and completes redundant or incomplete data; it performs fuzzy matching and merging of similar program entries to ensure data consistency; the unified storage unit uses a combination of relational database and distributed file system to establish a unified program content index library; and it provides a fast search interface based on keywords, themes, and broadcast times.

3. The converged media program schedule arrangement system according to claim 2, characterized in that: The feature extraction module includes a text semantic analysis unit that performs natural language processing on program synopsis, script, and subtitle text to extract thematic keywords and semantic tags; it also supports sentiment analysis and hot topic identification. The audio and video feature extraction unit performs image recognition and audio recognition on the program video content; it uses a convolutional neural network to extract image features and a Transformer model to analyze the audio structure. Tag mapping and standardization units map automatically extracted feature tags to a unified tag system; establish a multi-dimensional tag system to realize the semantic computation of programs.

4. A converged media program schedule arrangement system according to claim 1 or 2, characterized in that: The user profiling and platform model module includes: a user behavior collection unit that collects user viewing behavior; supporting multi-terminal collection, including mobile apps, TV set-top boxes, and web pages; a user profile modeling unit that generates user profiles based on big data analysis technology; profile dimensions including age, gender, interests, viewing time, and commonly used devices; a platform characteristic modeling unit that establishes broadcast strategy models for different platforms; defining program broadcast duration, format, and tag weight rule templates for each platform; and an audience matching prediction unit that uses a deep learning model to predict the audience matching degree of a specific program; outputting a program-user matching score to provide input parameters for subsequent scheduling.

5. A converged media program schedule arrangement system according to claim 1 or 2, characterized in that: The program selection and scheduling module includes a weight allocation unit that sets adaptive weights for program features, user profiles, and platform weights; dynamically adjusting weight coefficients based on real-time data; and a relevance calculation unit that utilizes a relevance calculation model between content features and audience features. ,in For content feature matching degree, To improve the matching degree of user profiles Platform preference coefficient Content feature weighting coefficient User profile weighting coefficient Platform strategy weighting coefficient; The system calculates the overall program matching degree; the sorting optimization unit uses genetic algorithms or reinforcement learning algorithms to sort programs; it supports multi-objective optimization; the dynamic scheduling unit automatically generates a preliminary program list based on the platform's playback timeline; and it performs intelligent time slot allocation based on time slot priority.

6. A converged media program schedule arrangement system according to claim 1 or 2, characterized in that: The intelligent conflict detection module includes: a time conflict detection unit that checks whether program playback times overlap, have insufficient intervals, or exceed the broadcast schedule; automatically calculates and adjusts playback time slots; a copyright conflict detection unit that compares with the program copyright database to prevent illegal repeated broadcasts or over-authorized playback; a duplicate broadcast identification unit that counts the number of times the same program is broadcast on different platforms; and an adjustment mechanism that is triggered when the duplication rate exceeds a set threshold; and an intelligent correction unit that combines optimization algorithms to automatically replace or adjust the order of conflicting programs.

7. A converged media program schedule arrangement system according to claim 1 or 2, characterized in that: The adaptive generation module includes a format adaptation unit that selects the output format according to the target platform; supports multiple formats such as EPG, XML, JSON, RSS, and Excel; an interface template generation unit that automatically generates program list page templates according to the platform's UI rules; provides responsive layout and multi-language support; and a multi-terminal compatibility unit that automatically adjusts the resolution, time format, and program cover size to ensure consistent display across multiple terminals.

8. A converged media program schedule arrangement system according to claim 1 or 2, characterized in that: The feedback and optimization module includes a playback data acquisition unit that monitors program playback data in real time, including clicks, completion rate, interaction rate, and comment sentiment indicators; an effect evaluation unit that quantitatively evaluates the overall effect of the program list and outputs a comprehensive dissemination index; a model self-learning unit that updates the algorithm model parameters based on the evaluation results; a reinforcement learning method that optimizes the program recommendation logic to achieve continuous evolution; a strategy adjustment unit that sends the optimization results back to the program selection module to form a closed-loop feedback; and support for manual intervention and strategy template updates.

9. A method for generating a multimedia program schedule, characterized in that: The method includes the following steps: Step S1: Collect content resources from multiple platforms, and perform semantic analysis on the program content through natural language processing and image recognition technology to generate standardized metadata tags; Step S2: Establish a multi-dimensional user profile model based on user behavior data and interaction data, and define the playback strategy and broadcast restrictions for each platform. Step S3: Utilize the matching degree calculation model between content features and user profiles to score the relevance of program resources, and generate a preliminary program list based on a weighted ranking algorithm or reinforcement learning algorithm. Step S4: Detect time conflicts, duplicate broadcasts, and duration mismatches in the program schedule, and readjust the scheduling results through the automatic correction module; Step S5: Based on the target platform type, automatically generate a program guide file in an adapted format and push it to the corresponding platform's broadcast control system; Step S6: Collect program playback effect data, adjust recommendation weights through model training, and realize intelligent self-optimization of program list generation.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the method of claim 9.

Images