Intelligent television integrated service quick starting and channel switching system

Abstract

The invention discloses a smart television integrated service quick start and channel switching system, and particularly relates to the technical field of big data, which comprises a data acquisition module, a user behavior prediction module, a resource intelligent preloading module and an intelligent response and switching acceleration module, the data acquisition module is used for recording multi-dimensional behavior data of a smart television user, and the user behavior prediction module is connected with the data acquisition module, constructs and trains a user behavior prediction model based on the multi-dimensional behavior data, and predicts operation performed by the user in a future time period. The resource intelligent preloading module is connected with the user behavior prediction module and executes personalized preloading operation of resources according to a prediction result, the intelligent response and switching acceleration module is connected with the resource intelligent preloading module and is connected with an operating system and a user interaction interface of the smart television, and when a user carries out channel switching operation, the resource intelligent preloading module carries out the personalized preloading operation of the resources according to the prediction result. An acceleration response is provided.

Description

Technical Field

[0001] This invention relates to the field of big data technology, and more specifically, to a smart TV integrated service quick start and channel switching system. Background Technology

[0002] With the rapid development of multimedia technology and the continuous improvement of network bandwidth, smart TVs have evolved from simple broadcast receiving devices into comprehensive information terminals that integrate multiple functions such as film and television entertainment, online education, interactive games, and smart home control.

[0003] However, in terms of user experience, current smart TV systems still face some common challenges, especially in the two key areas of "quick startup" and "channel switching." Existing smart TVs often require a long waiting time during startup. This is mainly due to the operating system of the smart TV, which requires various applications to initialize, load resources, and establish network connections, among other complex operations. After pressing the power button, users typically have to wait tens of seconds or even longer to access the main interface or begin watching content. This long wait severely impacts the user's immediate experience. Regarding channel switching, although technologies such as digital TV and IPTV are widespread, smart TVs often experience noticeable delays when switching channels, especially when switching from one application to another or from one live stream source to another. This includes channel buffering time, decoding time, and the time required for context switching between different applications. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, embodiments of the present invention provide a smart TV integrated service quick start and channel switching system to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a smart TV integrated service quick start and channel switching system, specifically including a data acquisition module, a user behavior prediction module, a resource intelligent preloading module, and an intelligent response and switching acceleration module;

[0006] The data acquisition module is used to record multi-dimensional behavioral data of smart TV users, including channel switching sequences, application usage records, time and date data, device status and environmental data;

[0007] The user behavior prediction module is connected to the data acquisition module and is used to receive the multi-dimensional behavior data, and to build and train a user behavior prediction model based on the multi-dimensional behavior data to predict the user's operations in the future time period.

[0008] The intelligent resource preloading module is connected to the user behavior prediction module and is used to receive the prediction results output by the user behavior prediction module and perform personalized preloading operations of resources based on the prediction results.

[0009] The intelligent response and switching acceleration module is connected to the intelligent resource preloading module, and is also connected to the operating system and user interface of the smart TV, providing accelerated response when the user performs channel switching operations.

[0010] Preferably, as a preferred embodiment of the integrated smart TV service quick start and channel switching system of the present invention, it includes the data acquisition module for recording multi-dimensional behavioral data of smart TV users, including channel switching sequences, application usage records, time and date data, device status and environmental data, specifically including the following:

[0011] The channel switching sequence records the starting channel ID and target channel ID for each user switching operation, the specific time point of each channel switching, and the duration the user stays on each channel; the application usage record records the application IDs and names, startup times, usage durations, usage frequencies, and in-application operations that the user has launched and used; the time and date data are precise timestamps associated with all behavioral data; the device status and environment data record the current network status of the smart TV, device load, user login information, and external input source status;

[0012] The collected data is preprocessed, including data cleaning, session segmentation, feature engineering and vectorization, to transform it into a standardized dataset;

[0013] The session segmentation is performed at the dimension of a single user ID, distinguishing different users through user login information, and further includes:

[0014] S1. User identification and isolation: Using the "user login information" recorded in the device status and environment data, the user identity of the current device is identified, and all collected behavioral data is associated with the corresponding user ID;

[0015] S2. Session segmentation based on time intervals: For all behavioral events of a single user, they are sorted according to their timestamps and the following rules are used for session segmentation:

[0016] a. Set an inactivity time threshold The first action event of the current session is taken as the first event of the current session, starting from the user's first action event.

[0017] b. Iterate through the user's subsequent action events and calculate the time interval between the current event and the previous event. ,when If the user is determined to be in a state of continuous activity, it means that the two events belong to the same session, and the current event is added to that session;

[0018] c. When the calculated time interval Once it is determined that a user's continuous viewing has ended, it indicates that the previous session has ended. The previous event is taken as the end point of the current session, and the current event is immediately taken as the start point of a new session to begin building a new sequence of behaviors.

[0019] S3. Constructing a sequence of behaviors within a session: Based on the rules above, the sessions containing behaviors are organized into an ordered sequence of behaviors according to their chronological order of occurrence.

[0020] Preferably, as a preferred embodiment of the intelligent TV integrated service quick start and channel switching system of the present invention, it includes a user behavior prediction module connected to a data acquisition module, used to receive the multi-dimensional behavior data, and to construct and train a user behavior prediction model based on the multi-dimensional behavior data to predict the user's operations in the future time period, specifically including the following:

[0021] The user behavior prediction model is a deep learning-based sequence prediction model that captures the dynamics, contextual dependencies, and long short-term memory characteristics of smart TV user behavior to achieve accurate prediction of future user behavior. It further includes:

[0022] Receive a standardized dataset from the data acquisition module after data cleaning, session segmentation, feature engineering, and vectorization. For each user behavior session, represent it as an ordered sequence. Where N is the number of events in the session, and each action event... After feature engineering and vectorization, it is represented as a multidimensional feature vector. This includes behavioral type characteristics, behavioral target characteristics, temporal context characteristics, user session characteristics, and device and environment characteristics;

[0023] LSTM is used as the sequence prediction model. An LSTM network is set to contain L layers, and the multidimensional feature vector of the behavioral event is received through the input layer of the LSTM. The output of each layer is a hidden state. At each time t, the LSTM state update is represented as: ,in, It is the feature vector of the t-th behavioral event. and These are the hidden state and the cell state of the previous moment, respectively;

[0024] The output of the LSTM layer The input is fed into a fully connected layer to generate prediction results, yielding the category probability distribution of the user's next action. Where W is the weight matrix, b is the bias term, and y is the predicted behavior category probability vector, including channel ID and application ID. It is the hidden state of the last time step;

[0025] Each user's behavior sequence is input into the LSTM model, and the multidimensional feature vector of each event in the sequence is extracted as input data. The LSTM model is trained using historical behavior data. During training, the model weights are adjusted to reduce prediction error. After training, for new user behavior data, the trained LSTM model is input to obtain the prediction result of the next behavior. The prediction result is output in the form of the TOP-K highest probability behavior categories and their corresponding prediction probabilities, and then transmitted to the resource intelligent preloading module.

[0026] Preferably, as a preferred embodiment of the intelligent TV integrated service quick start and channel switching system of the present invention, it includes a resource intelligent preloading module connected to a user behavior prediction module, used to receive the prediction results output by the user behavior prediction module, and perform personalized resource preloading operations according to the prediction results, specifically including the following:

[0027] The received prediction results are analyzed and strategy planning is performed. A preloading instruction is generated based on a multi-threshold hierarchical model. This model sets different preloading levels and corresponding resource loading depths according to the prediction probability, including the following hierarchical strategies:

[0028] Full preloading instruction: When the probability value of the predicted target is higher than the first threshold. Upon determining that the user's intent is clear, the "full preload" instruction is executed, pre-starting the application and loading core data. This involves performing the most in-depth resource preparation and the most complete application warm-up for both the live streaming channel and the application, further including:

[0029] For live streaming channels: Perform the most thorough resource preparation, with the following sequence of operations:

[0030] 1) Establish a complete network socket connection and complete session authentication;

[0031] 2) The buffer is sufficient to cover the initial delay of the TS stream data;

[0032] 3) Initialize the decoder hardware context in advance and complete all parameter configurations before decoding;

[0033] For applications: Perform the most complete application warm-up, with the following sequence of operations:

[0034] 1) Start the application process completely in the background;

[0035] 2) Load and render its main user interface;

[0036] 3) Pre-request and load its initial core data to put the application in a "click-to-use" ready state;

[0037] Partial preloading instruction: When the probability value of the predicted target is within the second threshold With the first threshold In between, the "partial preloading" instruction is executed to perform lightweight preparation of the critical path and caching of basic resources for both the live streaming channel and the application, further including:

[0038] For live streaming channels: Perform lightweight preparation for the critical path, with the following sequence of operations:

[0039] 1) Only perform DNS pre-resolution and TCP three-way handshake pre-connection;

[0040] 2) Pre-allocate but do not initialize decoder resources;

[0041] For applications: perform caching of underlying resources, with the following operation sequence:

[0042] 1) Preload the application's installation package, core code library, and critical resource files into memory;

[0043] 2) Lock it using the system interface to prevent it from being quickly reclaimed;

[0044] Lightweight prefetch instruction: When the probability value of the predicted target is below the second threshold It does not perform any substantial resource loading operations; for live channels, it only prefetches the channel's electronic program guide information; for applications, it only prefetches their icons and names.

[0045] Based on the generated preloading instructions, perform the corresponding preloading operations, including channel live stream preloading, application pre-start and data preloading, interface element and image material preloading, and system service preloading. After completing the above preloading operations, record and update the target resource ID, current preloading level, specific status of each preloading item and timestamp key information to the internally maintained preloading resource pool.

[0046] The preloaded resource pool is a dynamic data structure maintained in memory, used to store and manage all resource instances in the preloaded state. It intelligently manages the resources preloaded locally and receives resource hit signals sent by the intelligent response and switching acceleration module. Based on this signal, the resource is marked as "used" and its timestamp is updated. When user behavior does not match the prediction, unnecessary preloaded tasks are stopped, and the status records of loaded but unused resources are cleared from the preloaded resource pool, releasing system resources.

[0047] Preferably, as a preferred embodiment of the intelligent TV integrated service rapid start-up and channel switching system of the present invention, it includes the intelligent response and switching acceleration module connected to the resource intelligent preloading module, and connected to the operating system and user interaction interface of the intelligent TV, providing accelerated response when the user performs channel switching operations, specifically including the following:

[0048] Listen for and intercept channel switching and application startup commands from the user interaction interface. Within milliseconds, extract the target resource identifier from the command and quickly match and query it with the preloaded resource pool maintained in the intelligent resource preloading module to obtain the current preload level and specific preload status of the resource.

[0049] Based on the matching query results, the corresponding acceleration response strategy is activated, including instant path switching, fast activation path, and normal startup path, and further includes:

[0050] The instantaneous switching path is to skip the startup process and immediately submit the content to the display system for rendering when the target resource is in a "fully preloaded" state in the preloaded resource pool;

[0051] The rapid activation path is to quickly activate the target resource when it is in a "partially preloaded" state in the preloaded resource pool, requiring only the completion of the remaining key steps.

[0052] The conventional startup path is to start the target resource in the preloaded resource pool when it is only in a "lightweight prefetch" state, following the conventional process.

[0053] It also sends a resource hit signal to the resource intelligent preloading module, notifying that the resource has been actually accessed by the user.

[0054] On the other hand, the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, wherein when the computer program is executed by the processor, it implements the functional modules of a smart TV integrated service quick start and channel switching system as described above.

[0055] On the other hand, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements a smart TV integrated service quick start and channel switching system as described above.

[0056] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0057] This invention utilizes intelligent preloading and accelerated response mechanisms to provide users with a smoother and more immediate experience when operating smart TVs, thereby improving user satisfaction. The preloaded resource pool can intelligently manage preloaded resources, dynamically adjusting and eliminating them based on predicted probabilities and cache validity periods to avoid ineffective resource occupation. Through user behavior prediction and personalized preloading technology, the startup time of smart TV services and channel switching latency are significantly shortened, greatly enhancing the user experience. Attached Figure Description

[0058] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0059] Figure 1 This is a flowchart of a method for a smart TV integrated service quick start and channel switching system according to the present invention. Detailed Implementation

[0060] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0061] In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0062] In the description of this application, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in this application.

[0063] Example 1

[0064] This embodiment provides, for example Figure 1 The system shown is a smart TV integrated service quick start and channel switching system, which specifically includes a data acquisition module, a user behavior prediction module, a resource intelligent preloading module, and an intelligent response and switching acceleration module.

[0065] The data acquisition module is used to record multi-dimensional behavioral data of smart TV users, including channel switching sequences, application usage records, time and date data, device status and environmental data;

[0066] The user behavior prediction module is connected to the data acquisition module and is used to receive the multi-dimensional behavior data, and to build and train a user behavior prediction model based on the multi-dimensional behavior data to predict the user's operations in the future time period.

[0067] The intelligent resource preloading module is connected to the user behavior prediction module and is used to receive the prediction results output by the user behavior prediction module and perform personalized preloading operations of resources based on the prediction results.

[0068] The intelligent response and switching acceleration module is connected to the intelligent resource preloading module, and is also connected to the operating system and user interface of the smart TV, providing accelerated response when the user performs channel switching operations.

[0069] In this embodiment, the data acquisition module needs to be specifically described. This module records multi-dimensional behavioral data of smart TV users, including channel switching sequences, application usage records, time and date data, device status and environmental data, specifically including the following:

[0070] The channel switching sequence records the starting channel ID and target channel ID for each user switching operation, the specific time point of each channel switching, and the duration the user stays on each channel; the application usage record records the application IDs and names, startup times, usage durations, usage frequencies, and in-application operations that the user has launched and used; the time and date data are precise timestamps associated with all behavioral data; the device status and environment data record the current network status of the smart TV, device load, user login information, and external input source status;

[0071] The collected data is preprocessed, including data cleaning, session segmentation, feature engineering and vectorization, to transform it into a standardized dataset;

[0072] The session segmentation is performed at the dimension of a single user ID, distinguishing different users through user login information, and further includes:

[0073] S1. User identification and isolation: Using the "user login information" recorded in the device status and environment data, the user identity of the current device is identified, and all collected behavioral data is associated with the corresponding user ID;

[0074] S2. Session segmentation based on time intervals: For all behavioral events of a single user, they are sorted according to their timestamps and the following rules are used for session segmentation:

[0075] a. Set an inactivity time threshold The first action event of the current session is taken as the first event of the current session, starting from the user's first action event.

[0076] b. Iterate through the user's subsequent action events and calculate the time interval between the current event and the previous event. ,when If the user is determined to be in a state of continuous activity, it means that the two events belong to the same session, and the current event is added to that session;

[0077] c. When the calculated time interval Once it is determined that a user's continuous viewing has ended, it indicates that the previous session has ended. The previous event is taken as the end point of the current session, and the current event is immediately taken as the start point of a new session to begin building a new sequence of behaviors.

[0078] S3. Constructing a sequence of behaviors within a session: Based on the rules above, the sessions containing behaviors are organized into an ordered sequence of behaviors according to their chronological order of occurrence.

[0079] In this embodiment, the user behavior prediction module is specifically described. This module is connected to the data acquisition module and is used to receive the multi-dimensional behavior data. Based on the multi-dimensional behavior data, it constructs and trains a user behavior prediction model to predict user actions within a future time period. Specifically, this includes the following:

[0080] The user behavior prediction model is a deep learning-based sequence prediction model that captures the dynamics, contextual dependencies, and long short-term memory characteristics of smart TV user behavior to achieve accurate prediction of future user behavior. It further includes:

[0081] Receive a standardized dataset from the data acquisition module after data cleaning, session segmentation, feature engineering, and vectorization. For each user behavior session, represent it as an ordered sequence. Where N is the number of events in the session, and each action event... After feature engineering and vectorization, it is represented as a multidimensional feature vector. This includes behavioral type characteristics, behavioral target characteristics, temporal context characteristics, user session characteristics, and device and environment characteristics;

[0082] LSTM is used as the sequence prediction model. An LSTM network is set to contain L layers, and the multidimensional feature vector of the behavioral event is received through the input layer of the LSTM. The output of each layer is a hidden state. At each time t, the LSTM state update is represented as: ,in, It is the feature vector of the t-th behavioral event. and These are the hidden state and the cell state at the previous moment, respectively. It is the output of the LSTM unit at the current time step t. The cell state represents the model considering the long-term memory information of the previous time step t−1 when processing the input of the current time step t.

[0083] The output of the LSTM layer The input is fed into a fully connected layer to generate prediction results, yielding the category probability distribution of the user's next action. Where W is the weight matrix, b is the bias term, and y is the predicted behavior category probability vector, including channel ID and application ID. It is the hidden state of the last time step;

[0084] Each user's behavior sequence is input into the LSTM model, and the multidimensional feature vector of each event in the sequence is extracted as input data. The LSTM model is trained using historical behavior data. During training, the model weights are adjusted to reduce prediction error. After training, for new user behavior data, the trained LSTM model is input to obtain the prediction result of the next behavior. The prediction result is output in the form of the TOP-K highest probability behavior categories and their corresponding prediction probabilities, and then transmitted to the resource intelligent preloading module.

[0085] In this embodiment, the resource intelligent preloading module is specifically described. This module is connected to the user behavior prediction module and is used to receive the prediction results output by the user behavior prediction module, and to perform personalized resource preloading operations based on the prediction results. Specifically, this includes the following:

[0086] The received prediction results are analyzed and strategy planning is performed. A preloading instruction is generated based on a multi-threshold hierarchical model. This model sets different preloading levels and corresponding resource loading depths according to the prediction probability, including the following hierarchical strategies:

[0087] Full preloading instruction: When the probability value of the predicted target is higher than the first threshold. Upon determining that the user's intent is clear, the "full preload" instruction is executed, pre-starting the application and loading core data. This involves performing the most in-depth resource preparation and the most complete application warm-up for both the live streaming channel and the application, further including:

[0088] For live streaming channels: Perform the most thorough resource preparation, with the following sequence of operations:

[0089] 1) Establish a complete network socket connection and complete session authentication;

[0090] 2) The buffer is sufficient to cover the initial delay of the TS stream data;

[0091] 3) Initialize the decoder hardware context in advance and complete all parameter configurations before decoding;

[0092] For applications: Perform the most complete application warm-up, with the following sequence of operations:

[0093] 1) Start the application process completely in the background;

[0094] 2) Load and render its main user interface;

[0095] 3) Pre-request and load its initial core data to put the application in a "click-to-use" ready state;

[0096] Partial preloading instruction: When the probability value of the predicted target is within the second threshold With the first threshold In between, the "partial preloading" instruction is executed to perform lightweight preparation of the critical path and caching of basic resources for both the live streaming channel and the application, further including:

[0097] For live streaming channels: Perform lightweight preparation for the critical path, with the following sequence of operations:

[0098] 1) Only perform DNS pre-resolution and TCP three-way handshake pre-connection;

[0099] 2) Pre-allocate but do not initialize decoder resources;

[0100] For applications: perform caching of underlying resources, with the following operation sequence:

[0101] 1) Preload the application's installation package, core code library, and critical resource files into memory;

[0102] 2) Lock it using the system interface to prevent it from being quickly reclaimed;

[0103] Lightweight prefetch instruction: When the probability value of the predicted target is below the second threshold It does not perform any substantial resource loading operations; for live channels, it only prefetches the channel's electronic program guide information; for applications, it only prefetches their icons and names.

[0104] Based on the generated preloading instructions, perform the corresponding preloading operations, including channel live stream preloading, application pre-start and data preloading, interface element and image material preloading, and system service preloading. After completing the above preloading operations, record and update the target resource ID, current preloading level, specific status of each preloading item and timestamp key information to the internally maintained preloading resource pool.

[0105] The preloaded resource pool is a dynamic data structure maintained in memory, used to store and manage all resource instances in the preloaded state. It intelligently manages the resources preloaded locally and receives resource hit signals sent by the intelligent response and switching acceleration module. Based on this signal, the resource is marked as "used" and its timestamp is updated. When user behavior does not match the prediction, unnecessary preloaded tasks are stopped, and the status records of loaded but unused resources are cleared from the preloaded resource pool, releasing system resources.

[0106] In this embodiment, the intelligent response and switching acceleration module needs to be specifically described. This module is connected to the intelligent resource preloading module and also to the smart TV's operating system and user interface. It provides accelerated response when the user performs channel switching operations, specifically including the following:

[0107] Listen for and intercept channel switching and application startup commands from the user interaction interface. Within milliseconds, extract the target resource identifier from the command and quickly match and query it with the preloaded resource pool maintained in the intelligent resource preloading module to obtain the current preload level and specific preload status of the resource.

[0108] Based on the matching query results, the corresponding acceleration response strategy is activated, including instant path switching, fast activation path, and normal startup path, and further includes:

[0109] The instantaneous switching path is to skip the startup process and immediately submit the content to the display system for rendering when the target resource is in a "fully preloaded" state in the preloaded resource pool;

[0110] The rapid activation path is to quickly activate the target resource when it is in a "partially preloaded" state in the preloaded resource pool, requiring only the completion of the remaining key steps.

[0111] The conventional startup path is to start the target resource in the preloaded resource pool when it is only in a "lightweight prefetch" state, following the conventional process.

[0112] It also sends a resource hit signal to the resource intelligent preloading module, notifying that the resource has been actually accessed by the user.

[0113] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0114] This embodiment also provides a storage medium storing a computer program. When executed by a processor, the program implements the functional modules of a smart TV integrated service quick start and channel switching system as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0115] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A smart TV integrated service quick start and channel switching system, characterized in that: Specifically comprising a data collection module, a user behavior prediction module, a resource intelligent preloading module, and an intelligent response and switching acceleration module; The data collection module is configured to record multi-dimensional behavior data of a smart TV user, including channel switching sequences, application usage records, time and date data, device state and environment data. The user behavior prediction module is connected with the data collection module and configured to receive the multi-dimensional behavior data, and construct and train a user behavior prediction model based on the multi-dimensional behavior data to predict operations of the user in a future time period. The resource intelligent preloading module is connected with the user behavior prediction module and configured to receive a prediction result output by the user behavior prediction module, and perform a personalized preloading operation of resources according to the prediction result. The intelligent response and switching acceleration module is connected with the resource intelligent preloading module, and connected with an operating system and a user interaction interface of the smart TV, and provides an accelerated response when the user performs a channel switching operation.

2. The intelligent TV integrated service fast start and channel switching system of claim 1, wherein: The data collection module is configured to record multi-dimensional behavior data of a smart TV user, including channel switching sequences, application usage records, time and date data, device state and environment data, and specifically includes the following contents: The channel switching sequence records a starting channel ID and a target channel ID of each switching operation of the user, and records a specific time point at which each channel switching occurs and a duration for which the user stays on each channel; the application usage record records an application ID and a name, a starting time, a usage duration, a usage frequency, and an operation behavior in the application of an application that is started and used by the user; the time and date data is an accurate time stamp associated with all behavior data; and the device state and environment data records a current network state, a device load, user login information, and an external input source state of the smart TV. The collected data is preprocessed, including data cleaning, session segmentation, feature engineering, and vectorization operation, to convert the data into a standardized data set.

3. The intelligent TV integrated service fast start and channel switching system of claim 2, wherein: The session segmentation is performed in the dimension of a single user ID, different users are distinguished by user login information, and further includes: S1, user identity recognition and isolation: using the "user login information" recorded in the device state and environment data, the identity of the current device user is determined, and all collected behavior data is associated with the corresponding user ID; S2, session division based on time interval: for all behavior events of a single user, the time stamps thereof are sorted, and the following rules are used for session division: a. Set an inactivity time threshold From the first behavior event of the user, take it as the first event of the current session; b. traversing the user's subsequent behavior events, calculating the time interval between the current event and the previous event , when , determining that the user is in a continuous activity state, indicating that the two events belong to the same session, adding the current event to the session; c. When the calculated time interval , it is determined that the user's continuous watching has ended, indicating that the previous session has ended, the previous event is taken as the end point of the current session, and the current event is immediately taken as the starting event of a new session, and the construction of a new behavior sequence is started; S3, construction of behavior sequence in a session: the behavior events contained in a session divided according to the above rules are arranged in a time sequence to form an ordered behavior sequence.

4. The intelligent TV integrated service fast start and channel switching system of claim 1, wherein: The user behavior prediction model is a sequence prediction model based on deep learning, which captures the dynamic nature, context dependence and long-short term memory characteristics of the behavior of the smart TV user, and realizes accurate prediction of future behavior of the user, and further includes: Receiving the standardized dataset from the data collection module after data cleaning, session segmentation, feature engineering and vectorization processing, for each user behavior session, it is represented as an ordered sequence where N is the number of events in the session, each behavior event After feature engineering and vectorization processing, it is represented as a multi-dimensional feature vector including behavior type features, behavior target features, temporal context features, user session features, and device and environment features; An LSTM is used as a sequence prediction model, and an LSTM network containing L layers is set up, and the multi-dimensional feature vector of the behavior event is received by the input layer of the LSTM The output of each layer is a hidden state The state update of the LSTM at each time t is represented as wherein is the feature vector of the tth behavior event, and are the hidden state and cell state of the previous time, respectively; The output of the LSTM layer The input is fed into a fully connected layer to generate prediction results, yielding the category probability distribution of the user's next action. Where W is the weight matrix, b is the bias term, and y is the predicted behavior category probability vector, including channel ID and application ID. It is the hidden state of the last time step; The behavior sequence of each user is input into an LSTM model, a multi-dimensional feature vector of each event in the sequence is extracted as input data, and the LSTM model is trained using historical behavior data. During the training process, the weights of the model are adjusted to reduce the prediction error. After the training is completed, for new user behavior data, the trained LSTM model is input, and the prediction result of the next behavior is obtained. The prediction result is output in the form of TOP-K behavior categories with the highest probability and their corresponding prediction probabilities, and is transmitted to the resource intelligent preloading module.

5. The intelligent TV integrated service fast start and channel change system of claim 1, wherein: The resource intelligent preloading module is connected with the user behavior prediction module, and is used for receiving the prediction result output by the user behavior prediction module, and performing personalized preloading operation of resources according to the prediction result, specifically including the following contents: The received prediction result is analyzed and strategy planning is performed, and a preloading instruction is generated based on a multi-threshold grading model. The multi-threshold grading model sets different preloading levels and corresponding resource loading depths according to the high and low of the prediction probability, including the following grading strategies: Full preload instruction: when the probability value of the prediction target is higher than a first threshold , it is determined that the user's intention is clear, the "full preload" instruction is executed, the application is started in advance and the core data is loaded, the deepest resource preparation and the most complete application warm-up are performed on the live channel and the application respectively; Partial preloading instruction: when the probability value of the prediction target is between the second threshold and the first threshold , execute the "partial preloading" instruction to perform lightweight preparation of the critical path and cache of the basic resources of the live channel and the application, respectively; Lightweight prefetch instruction: when the probability value of the prediction target is lower than a second threshold without performing any substantial resource loading operations, for live channels only pre-fetching the electronic program guide information of the channel; for applications only pre-fetching their icon and name; According to the generated preloading instruction, the corresponding preloading operation is performed, including channel live streaming preloading, application pre-starting and data preloading, interface element and picture material preloading, and system service preloading. After the above preloading operations are completed, the ID of the target resource, the current preloading level, the specific state of preloading, and the timestamp key information are recorded and updated in the preloading resource pool maintained internally; The preloading resource pool is a dynamic data structure maintained in the memory, used for storing and managing all resource instances in the preloading state. The preloaded resources are intelligently managed, and the resource hit signal sent by the intelligent response and switching acceleration module is received. According to the signal, the resource is marked as "used" and the timestamp is updated. When the user behavior does not match the prediction, the preloading task is stopped, and the state record of the loaded but unused resource is cleared from the preloading resource pool, and the system resource is released.

6. The intelligent TV integrated service fast start and channel change system of claim 1, wherein: The intelligent response and switching acceleration module is connected with the resource intelligent preloading module, and is connected with the operating system and user interaction interface of the smart television. When the user performs channel switching operation, the acceleration response is provided, specifically including the following contents: Listen to and intercept the channel switching and application starting instructions from the user interaction interface. In a millisecond level time, the target resource identifier in the instruction is extracted, and a quick matching query is performed between the resource intelligent preloading module and the preloading resource pool maintained in the resource intelligent preloading module to obtain the current preloading level and specific preloading state of the resource. According to the matching query result, the corresponding acceleration response strategy is started, including the instantaneous switching path, the rapid activation path, and the conventional starting path, further including: The instantaneous switching path is when the target resource is in the "complete preloading" state in the preloading resource pool, directly skipping the starting process and immediately submitting the content to the display system for rendering; The rapid activation path is when the target resource is in the "partial preloading" state in the preloading resource pool, only the remaining key steps need to be completed for rapid activation. The normal starting path is to start according to a normal process when the target resource is only in a "light pre-fetch" state in the pre-loading resource pool; And sending a resource hit signal to the resource intelligent pre-loading module to inform that the resource has been actually accessed by the user. 7.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is characterized in that: The processor implements the function modules of the intelligent television integrated service fast starting and channel switching system of any one of claims 1-6 when executing the computer program.

8. A computer readable storage medium having stored thereon a computer program, characterized in that: The computer program is executed by the processor to implement the function modules of the intelligent television integrated service fast starting and channel switching system of any one of claims 1-6.

Images