Smart standby data analysis system and method based on television

Abstract

The invention discloses a smart standby data analysis system and method based on a television, and relates to the technical field of smart televisions, and the method comprises the steps: collecting user operation behavior data and content state data in real time; based on user operation behavior data collected in a historical time period, analyzing the distribution of use behaviors on multidimensional features in a time unit, and constructing a user active cycle model; when the non-operation duration of the television reaches a preset threshold value, collecting and analyzing current real-time context information; performing a multi-factor logic weighting decision by combining the active period classification corresponding to the current time point in the user active period model and the real-time context information, and generating a standby control instruction; and executing standby operation according to the standby control instruction. According to the method and the device, recoverable marking is performed on the playing session, so that the interruption node can be directly recovered when the user wakes up the television subsequently, the content does not need to be searched again, and the use convenience is improved.

Description

Technical Field

[0001] This invention relates to the field of smart TV technology, specifically a smart standby data analysis system and method based on a television set. Background Technology

[0002] Televisions generally have an automatic standby function, which typically uses a simple timer mechanism. This means that after a preset period of inactivity, the television automatically enters standby mode to save energy. However, this standby strategy is too mechanical and fails to fully consider the user's actual usage scenarios and intentions. Its low level of intelligence leads to a significant conflict between user experience and energy-saving performance.

[0003] Traditional standby methods cannot distinguish between a user's natural departure after prolonged viewing and a brief interruption (such as answering a phone call or doing housework). Closing the program when the user only briefly leaves for the moment forcibly interrupts viewing continuity, causing user resentment. Secondly, it fails to perceive the status of the content being played. For example, entering standby at the end of a movie versus during the middle of the movie has completely different meanings for the user, and current strategies cannot differentiate between these scenarios. A fixed standby threshold cannot accommodate the varying usage habits of different users or the same user at different times. For instance, during high-activity periods such as weekend evenings, the likelihood of a user returning after a brief absence is much higher than on weekday mornings; a fixed threshold cannot meet personalized needs. Summary of the Invention

[0004] The purpose of this invention is to provide a smart standby data analysis system and method based on a television set, so as to solve the problems raised in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a smart standby data analysis method based on a television set, the method comprising: S100: During the operation of the television, user operation behavior data and content status data are collected in real time; S200. Based on user operation behavior data collected within historical time periods, a user activity cycle model is constructed by analyzing the distribution of usage behavior on multidimensional features within time units. The user activity cycle model is used to describe the typical TV usage time patterns of users. S300: When the duration of no operation on the TV reaches a preset threshold, collect and analyze the current real-time context information; the real-time context information includes content status information used to indicate the playback status and operation sequence information used to reflect the user's intent. S400: Combining the current time point with the active period classification in the user's active period model and real-time context information, perform multi-factor logical weighted decision-making to generate standby control instructions; standby control instructions include standby trigger timing and standby prompt methods; S500: Execute the corresponding standby operation according to the standby control command; the standby operation includes: when the decision result is that standby needs to be delayed, resetting the no-operation timer and outputting a prompt message that matches the decision result; when the decision result is to enter standby immediately, marking the playback session as recoverable according to the current content status, and controlling the TV to enter standby state.

[0006] According to the above scheme, user operation behavior data includes power-on time, power-off or standby time, channel switching operation, application launch operation, volume adjustment operation, pause operation, and return to the main interface operation; content status data includes the type of currently playing content and the playback progress status.

[0007] According to the above scheme, step S200 includes: S210. Divide the historical timeline into predefined periodic time units and conduct behavioral pattern analysis; S220. Based on the power-on time point, power-off time point, or standby time point data within the time unit, calculate the user activity intensity of the time unit; the user activity intensity is quantified by analyzing the multi-dimensional characteristics of user behavior in terms of time distribution, usage frequency, and usage duration within the time unit. User activity intensity is calculated using the following formula: I t =α×F t +β×D t +γ×L t ; Among them, I t F represents the user activity intensity in the t-th time unit; t D represents the percentage of user usage frequency in the t-th time unit; t L represents the percentage of user usage days in the t-th time unit; t This represents the percentage of average user usage time in the t-th time unit; α, β, and γ represent weighting coefficients. S230. Based on the user activity intensity of each time unit, classify all time units into different active period types; the active period types include high-activity periods and low-activity periods. S240. Map the active period classification results to time units to generate a user active period model; the user active period model is used to predict the expected active period type for any given time point.

[0008] According to the above scheme, the content status information includes whether the TV is currently in a live broadcast state, a video-on-demand state, a paused state, or an idle state of the main interface. The recent operation sequence information is a sequence of user operations within a preset time period arranged in chronological order. The operation sequence information includes an operation type sequence and an operation intent feature determined based on the continuous distribution characteristics of the operation type. The operation types include channel switching, application launch, volume adjustment, pause, and return to the main interface. The operation intent features include content selection, parameter adjustment, or playback interruption intent.

[0009] According to the above scheme, step S400 includes: S410, predefine decision weights for the active period classification corresponding to the current time point, the content status information in the real-time context information, and the operation intention characteristics, respectively; Predefine decision weights, the specific steps are as follows: S411. Assign a set of preset initial decision weights to the active period classification quantification value, content status information quantification value, and operation intention feature quantification value, and the sum of the weights is 1; during system operation, monitor the user's manual correction operation on the automatic standby decision in real time; when the user's correction operation occurs, record the quantification value of each factor corresponding to this decision and the real intention label represented by the user's correction behavior, form a feedback data sample and store it in the sample set. S412. Using a predetermined period or a predetermined number of samples as trigger conditions, read the feedback sample set; using the user's real intent label in the sample set as the optimization objective and the decision weight as the optimization variable, construct a least squares loss function with respect to the decision weight; using the gradient descent algorithm, calculate the partial derivative of the loss function with respect to each decision weight, and iteratively update each decision weight according to the direction of the partial derivative and the preset learning rate, so that the loss function value tends to be minimized. S413. During the weight iteration and update process, constraints are imposed on the numerical range of each decision weight to ensure that it is within a preset reasonable range, so as to avoid a single factor from excessively dominating the decision result; at the same time, a scenario-based configuration interface is provided, allowing different initial weight preferences or locking of specific weights according to external scenario instructions, so as to achieve adaptation to different application scenarios. S420. Classify the active period corresponding to the current time point, couple it with the content status information and operation intention characteristics indicated by the real-time context information, and perform a weighted evaluation based on the predefined decision weights to output a comprehensive decision value. The comprehensive decision value is calculated using the following formula: S = ω1 × A + ω2 × B + ω3 × C; Where S represents the comprehensive decision value; A represents the quantified value of the active period classification; B represents the quantified value of the content status information; C represents the quantified value of the operational intent characteristics; ω1, ω2 and ω3 represent the decision weights; S430. Compare the comprehensive decision value with the preset decision threshold to determine the standby triggering time; wherein, when the comprehensive decision value is greater than the first decision threshold, it is determined that standby needs to be delayed; when the comprehensive decision value is less than the second decision threshold, it is determined that standby needs to be started immediately; when the comprehensive decision value is between the second decision threshold and the first decision threshold, it is determined that standby needs to be canceled; the first decision threshold is greater than the second decision threshold. S440. Generate corresponding standby control instructions based on the determined standby trigger timing and real-time context information; the standby control instructions include the standby trigger timing and the standby prompt method corresponding to the trigger timing and context information.

[0010] The system includes a preset threshold for inactivity duration, a first decision threshold, and a second decision threshold. It employs a reinforcement learning algorithm to jointly optimize the system by reducing the number of times users manually cancel standby and lowering overall standby power consumption. The thresholds are dynamically adjusted based on historical user feedback data to adapt to user habits. Additionally, a settings interface is provided, allowing users to manually adjust the thresholds according to their preferences.

[0011] According to the above scheme, step S500 includes: S510. Parse the standby control command and obtain the standby trigger timing and standby prompt method contained in the standby control command; S520. If the parsed standby trigger time is delayed standby, then reset the no-operation timer and control the user interface to output delayed standby prompt information in the standby prompt mode; S530. If the parsed standby trigger time is immediate standby, then mark the playback session as recoverable according to the current content status, and control the user interface to output immediate standby prompt information in the standby prompt mode, and then control the TV to enter standby mode. S540. If the parsed standby trigger time is to cancel standby, then reset the no-operation timer and control the user interface to output the standby cancellation notification information in the standby prompt mode.

[0012] A smart standby data analysis system based on a television set, comprising: a data acquisition module, a user behavior analysis module, a context analysis module, an intelligent decision-making module, and a standby control module; The data acquisition module is used to collect user operation behavior data and content status data in real time during the operation of the TV; The user behavior analysis module, connected to the data acquisition module, is used to construct a user activity cycle model based on user operation behavior data collected within historical time periods by analyzing the distribution of usage behavior on multi-dimensional features within a time unit. The context analysis module, connected to the data acquisition module, is used to collect and analyze the current real-time context information when the TV is inactive for a period of time that reaches a preset threshold. The real-time context information includes content status information to indicate the playback status and operation sequence information to reflect the user's intent. The intelligent decision-making module connects to the user behavior analysis module and the context analysis module respectively. It is used to combine the active period classification corresponding to the user's active period in the current time point in the user's active period model with real-time context information to perform multi-factor logical weighted decision-making and generate standby control instructions. The standby control instructions include standby trigger timing and standby prompt method. The standby control module, connected to the intelligent decision-making module, is used to execute standby operations according to standby control commands.

[0013] According to the above scheme, the user behavior analysis module includes a periodic modeling unit and a model prediction unit. The periodic modeling unit is used to divide the historical timeline into predefined periodic time units, calculate the user activity intensity based on the data within the time units, classify the active period, and generate a user activity period model. The model prediction unit is used to store the user activity period model and predict the active period classification based on the time point.

[0014] According to the above scheme, the intelligent decision-making module includes a weight management unit and a decision analysis unit; the weight management unit is used to predefine decision weights for active period classification, content status information and operation intention characteristics; the decision analysis unit is used to couple and weight the active period classification, content status information and operation intention characteristics, output a comprehensive decision value, and determine the standby triggering time based on the analysis results of the comprehensive decision value and the preset threshold.

[0015] According to the above scheme, the standby control module includes an instruction parsing unit and an action execution unit; the instruction parsing unit is used to parse standby control instructions and obtain the standby trigger timing and standby prompt mode; the action execution unit is used to execute the following actions based on the parsed standby trigger timing: resetting the no-operation timer, marking the playback session as recoverable, controlling the TV to enter standby mode, and controlling the user interface to output prompt information.

[0016] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention constructs a user activity cycle model and combines it with real-time context information for multi-factor weighted decision-making, which transforms standby judgment from a single time dimension to an intelligent analysis that integrates multi-dimensional intent, greatly reducing misjudgments and making the standby triggering time more in line with the user's real needs. 2. In the immediate standby scenario, this invention marks the playback session as recoverable, ensuring that when the user wakes up the TV later, they can directly resume to the interrupted point without having to search for content again; the differentiated prompts for delayed standby and canceled standby scenarios reduce user operation steps and improve ease of use; 3. Based on the user activity cycle model, this invention analyzes low-activity and high-activity periods, effectively balancing the energy-saving needs of smart TVs with user experience. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating the steps of a smart standby data analysis method based on a television set according to the present invention. Figure 2 This is a schematic diagram of the structure of a smart standby data analysis system based on a television set according to the present invention. Detailed Implementation

[0018] 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.

[0019] Example: Figures 1-2 As shown, the present invention provides a technical solution, a smart standby data analysis method based on a television set, the method comprising the following steps: S100: During the operation of the television, user operation behavior data and content status data are collected in real time; Specifically, user operation data includes power-on time, power-off or standby time, channel switching, application launch, volume adjustment, pause, and return to the main screen; content status data includes the type of currently playing content and playback progress status.

[0020] For example: It was recorded that a user performed a power-on operation at 20:00 on a certain weekday, then performed an on-demand operation for movie A at 20:10, and then performed a pause operation at 20:12; This is just an example and is not a limitation. S200. Based on user operation behavior data collected within historical time periods, a user activity cycle model is constructed by analyzing the distribution of usage behavior on multidimensional features within time units. The user activity cycle model is used to describe the typical TV usage time patterns of users. Specifically, step S200 includes: S210. Divide the historical timeline into predefined periodic time units and conduct behavioral pattern analysis; For example, the historical timeline of the past 4 weeks is divided into time units with 1-hour intervals; S220. Based on the power-on time, power-off time, or standby time data within a time unit, calculate the user activity intensity for that time unit. User activity intensity is quantified by analyzing multi-dimensional characteristics of user behavior within the time unit in terms of time distribution, usage frequency, and usage duration. The formula for user activity intensity is as follows: I t =α×F t +β×D t +γ×L t ; Among them, I t F represents the user activity intensity in the t-th time unit; t D represents the percentage of user usage frequency in the t-th time unit; t L represents the percentage of user usage days in the t-th time unit; t This represents the percentage of average user usage time in the t-th time unit; α, β, and γ represent weighting coefficients. For example: Calculate for the time unit of 20:00-21:00 from Monday to Friday; Within the 4-week statistical period, this time unit was used for 20 days, meaning there was a power-on activity; the total number of power-on times during these 20 days was 30, and the total viewing time was 1500 minutes. To eliminate the difference in magnitude, the daily frequency, percentage of days, and daily duration are normalized to the [0,1] interval using minimum-maximum normalization. F (average daily frequency): X min =0, X max =5; D (percentage of days): X min =0, X max =1; L (average daily duration): X min =0, X max =120; After normalization: F normalized =(1.5-0) / (5-0)=1.5 / 5=0.30; D normalized =(0.714-0) / (1-0)=0.714 / 1=0.714; L normalized =(75-0) / (120-0)=75 / 120=0.625; Activity intensity: I = 0.4 × 0.30 + 0.3 × 0.714 + 0.3 × 0.625 = 0.12 + 0.2142 + 0.1875 = 0.5217; S230. Based on the user activity intensity of each time unit, classify all time units into different active period types; the active period types include high-activity periods and low-activity periods. For example, after calculating the activity intensity of all time units, units with I > 0.4 are classified as high-activity periods, and units with I ≤ 0.4 are classified as low-activity periods; accordingly, 20:00-21:00 from Monday to Friday is classified as a high-activity period. S240. Map the active period classification results to time units to generate a user activity cycle model; the user activity cycle model is used to predict the expected active period type for any given time point; For example, generative models map 20:00-21:00 from Monday to Friday as a period of high activity.

[0021] S300: When the duration of no operation on the TV reaches a preset threshold, collect and analyze the current real-time context information; the real-time context information includes content status information used to indicate the playback status and operation sequence information used to reflect the user's intent. Specifically, the content status information includes whether the TV is currently in live broadcast playback, on-demand movie playback, paused, or idle on the main interface; the recent operation sequence information is a sequence of user operations within a preset time period arranged in chronological order, including an operation type sequence and operation intent features determined based on the continuous distribution characteristics of operation types; operation types include channel switching, application launch, volume adjustment, pause, and return to the main interface; operation intent features include content selection, parameter adjustment, or playback interruption intent.

[0022] For example: On a Wednesday at 20:15, the duration of inactivity on the TV reached a preset threshold. Real-time context data collection yields: Content status information: Analysis shows the TV is currently playing a video-on-demand, and has reached the middle of the video; Recent operation sequence information: The operation sequence within the last 10 minutes is extracted as: Video-on-demand A - Play - Pause; Analysis of this sequence identifies the operation intent as an intention to interrupt playback; S400: Combining the current time point with the active period classification in the user's active period model and real-time context information, perform multi-factor logical weighted decision-making to generate standby control instructions; standby control instructions include standby trigger timing and standby prompt methods; Specifically, step S400 includes: S410, predefine decision weights for the active period classification corresponding to the current time point, the content status information in the real-time context information, and the operation intention characteristics, respectively; For example: Active period category weight 0.5, content status weight 0.4, operation intention weight 0.1; Predefine decision weights, the specific steps are as follows: S411. Assign a set of preset initial decision weights to the active period classification quantification value, content status information quantification value, and operation intention feature quantification value, and the sum of the weights is 1; during system operation, monitor the user's manual correction operation on the automatic standby decision in real time; when the user's correction operation occurs, record the quantification value of each factor corresponding to this decision and the real intention label represented by the user's correction behavior, form a feedback data sample and store it in the sample set. S412. Using a predetermined period or a predetermined number of samples as trigger conditions, read the feedback sample set; using the user's real intent label in the sample set as the optimization objective and the decision weight as the optimization variable, construct a least squares loss function with respect to the decision weight; using the gradient descent algorithm, calculate the partial derivative of the loss function with respect to each decision weight, and iteratively update each decision weight according to the direction of the partial derivative and the preset learning rate, so that the loss function value tends to be minimized. S413. During the weight iteration and update process, constraints are imposed on the numerical range of each decision weight to ensure that it is within a preset reasonable range, so as to avoid a single factor from excessively dominating the decision result; at the same time, a scenario-based configuration interface is provided, allowing different initial weight preferences or locking of specific weights according to external scenario instructions, so as to achieve adaptation to different application scenarios. S420. Classify the active period corresponding to the current time point, couple it with the content status information and operation intention characteristics indicated by the real-time context information, and perform a weighted evaluation based on the predefined decision weights to output a comprehensive decision value. The comprehensive decision value is calculated using the following formula: S = ω1 × A + ω2 × B + ω3 × C; Where S represents the comprehensive decision value; A represents the quantified value of the active period classification; B represents the quantified value of the content status information; C represents the quantified value of the operational intent characteristics; ω1, ω2 and ω3 represent the decision weights; For example: if the current time point is a high-activity period, A is quantized as 1.0; if the content is a video playing, B is quantized as 1.0; if the intent is to interrupt playback, C is quantized as 0.2. Overall decision value: S = 0.5 × 1.0 + 0.4 × 1.0 + 0.1 × 0.2 = 0.5 + 0.4 + 0.02 = 0.92; S430. Compare the comprehensive decision value with the preset decision threshold to determine the standby triggering time; wherein, when the comprehensive decision value is greater than the first decision threshold, it is determined that standby needs to be delayed; when the comprehensive decision value is less than the second decision threshold, it is determined that standby needs to be started immediately; when the comprehensive decision value is between the second decision threshold and the first decision threshold, it is determined that standby needs to be canceled; the first decision threshold is greater than the second decision threshold. For example: the first decision threshold is 0.7, and the second decision threshold is 0.3; since S=0.92>0.7, it is determined that a standby delay is required; S440. Generate corresponding standby control instructions based on the determined standby trigger timing and real-time context information; the standby control instructions include the standby trigger timing and the standby prompt method corresponding to the trigger timing and context information; For example: Generate standby control command: the standby trigger time is delayed standby, and the prompt method is to display a delay countdown prompt.

[0023] The system includes a preset threshold for inactivity duration, a first decision threshold, and a second decision threshold. It employs a reinforcement learning algorithm to jointly optimize the system by reducing the number of times users manually cancel standby and lowering overall standby power consumption. The thresholds are dynamically adjusted based on historical user feedback data to adapt to user habits. Additionally, a settings interface is provided, allowing users to manually adjust the thresholds according to their preferences.

[0024] S500: According to the standby control command, execute the corresponding standby operation; the standby operation includes: when the decision result is that standby needs to be delayed, resetting the no-operation timer and outputting a prompt message that matches the decision result; when the decision result is to enter standby immediately, marking the playback session as recoverable according to the current content status, and controlling the TV to enter standby state. Specifically, step S500 includes: S510. Parse the standby control command and obtain the standby trigger timing and standby prompt method contained in the standby control command; S520. If the parsed standby trigger time is delayed standby, then reset the no-operation timer and control the user interface to output delayed standby prompt information in the standby prompt mode; S530. If the parsed standby trigger time is immediate standby, then mark the playback session as recoverable according to the current content status, and control the user interface to output immediate standby prompt information in the standby prompt mode, and then control the TV to enter standby mode. S540. If the parsed standby trigger time is to cancel standby, then reset the no-operation timer and control the user interface to output the standby cancellation notification information in the standby prompt mode.

[0025] This invention provides another technical solution: a smart standby data analysis method based on a television set, and a decision weight acquisition method; Based on the common characteristics of television usage scenarios, a set of preset initial decision weights are assigned to the quantitative values ​​of active period classification, content status information, and operation intention characteristics, and the sum of the weights is 1. The active period classification weight ω1=0.3, reflecting the user's potential usage needs at a specific time; the content status information weight ω2=0.4, reflecting the need for continuity of the currently playing content; the operation intention feature weight ω3=0.3, reflecting the user's potential purpose of recent operations; the initial weights satisfy ω1+ω2+ω3=1; During system operation, the system monitors user manual corrections to the automatic standby decision in real time; when a user correction occurs, it is recorded as a sample (S). i ,R i ), where S i R is the comprehensive decision value for this decision. i The actual user needs are tagged, and the delay in standby should be recorded as 1, and the immediate standby should be recorded as -1. These constitute feedback data samples and are stored in the sample set. The system reads a set of feedback samples based on a predetermined period or number of samples. It constructs a least-squares loss function with respect to the decision weights, using the user's real intent labels in the sample set as the optimization target and the decision weights as the optimization variables. It then uses the gradient descent algorithm to calculate the partial derivatives of the loss function with respect to each decision weight and iteratively updates each decision weight based on the direction of the partial derivatives and the preset learning rate, so that the loss function value tends to be minimized. Loss function: L(ω1,ω2,ω3)=Σ(R) i -S i ) 2 S i =ω1A i +ω2B i +ω3C i A i B i and C i This is the quantified value of the sample. The smaller the loss function value, the more closely the weights match user habits. Weight update formula: (k) ω1(k+1)=ω1(k)-η×∂L / ∂ω1(k); ω2(k+1)=ω2(k)-η×∂L / ∂ω2(k); ω3(k+1)=ω3(k)-η×∂L / ∂ω3(k); Where k represents the number of iterations, η represents the learning rate, and ∂L / ∂ω(k) represents the partial derivative of the loss function with respect to the weights ω(k); During the weight iteration and update process, constraints are imposed on the numerical range of each decision weight to ensure that it is within a preset reasonable range, so as to avoid a single factor from excessively dominating the decision result. At the same time, a scenario-based configuration interface is provided, allowing different initial weight preferences or locking of specific weights according to external scenario instructions, so as to achieve adaptation to different application scenarios. For example: Contextualized weight adaptation is tailored to different usage scenarios, allowing adjustment of weight bias through system interfaces: Home scenario: If users are more concerned about content continuity, ω2 can be increased to 0.5, and ω1 can be decreased to 0.2; Hotel scenario: If users are more concerned about energy consumption optimization, ω1 can be increased to 0.4, and ω2 can be decreased to 0.3. Constraints: The weight values ​​are in the range of 0.1≤ω1,ω2,ω3≤0.6 to avoid extreme results caused by a single factor dominating the decision; Update cycle: The system automatically performs weight iteration optimization every 7 days. If the number of user feedback samples exceeds 30, an immediate update is triggered.

[0026] This invention provides another technical solution: a smart standby data analysis system based on a television set, which includes: a data acquisition module, a user behavior analysis module, a context analysis module, an intelligent decision-making module, and a standby control module. The data acquisition module is used to collect user operation behavior data and content status data in real time during the operation of the TV; The user behavior analysis module, connected to the data acquisition module, is used to construct a user activity cycle model based on user operation behavior data collected within historical time periods by analyzing the distribution of usage behavior on multi-dimensional features within a time unit. Specifically, the user behavior analysis module includes a periodic modeling unit and a model prediction unit. The periodic modeling unit is used to divide the historical timeline into predefined periodic time units, calculate user activity intensity based on the data within the time units, classify active periods, and generate a user activity period model. The model prediction unit is used to store the user activity period model and predict the active period classification based on the time point. The context analysis module, connected to the data acquisition module, is used to collect and analyze the current real-time context information when the TV is inactive for a period of time that reaches a preset threshold. The real-time context information includes content status information to indicate the playback status and operation sequence information to reflect the user's intent. The intelligent decision-making module connects to the user behavior analysis module and the context analysis module respectively. It is used to combine the active period classification corresponding to the user's active period in the current time point in the user's active period model with real-time context information to perform multi-factor logical weighted decision-making and generate standby control instructions. The standby control instructions include standby trigger timing and standby prompt method. Specifically, the intelligent decision-making module includes a weight management unit and a decision analysis unit. The weight management unit is used to predefine decision weights for active period classification, content status information, and operation intention characteristics. The decision analysis unit is used to couple and weight the active period classification, content status information, and operation intention characteristics, output a comprehensive decision value, and determine the standby triggering time based on the analysis results of the comprehensive decision value and the preset threshold. The standby control module, connected to the intelligent decision module, is used to execute standby operations according to standby control commands; Specifically, the standby control module includes an instruction parsing unit and an action execution unit; the instruction parsing unit is used to parse standby control instructions and obtain the standby trigger timing and standby prompt mode; the action execution unit is used to execute actions such as resetting the no-operation timer, marking the playback session as recoverable, controlling the TV to enter standby mode, and controlling the user interface to output prompt information, based on the parsed standby trigger timing.

[0027] 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 essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, 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 equivalents 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.

Claims

1. A television-based intelligent standby data analysis method, characterized in that: The method comprises: S100, collecting user operation behavior data and content state data in real time during the running of the television set; S200, based on the user operation behavior data collected in the historical time period, constructing a user active period model by analyzing the distribution of the use behavior in the time unit in multiple dimensions; the user active period model is used to describe the typical television set use time law of the user; S300, when the television set operation duration reaches a preset threshold, collecting and analyzing the current real-time context information; the real-time context information includes content state information for indicating the playing state and operation sequence information for reflecting the user's intention; S400, combining the active period classification corresponding to the current time point in the user active period model with the real-time context information, performing multi-factor logical weighted decision to generate standby control instructions; the standby control instructions include standby trigger time and standby prompt mode; S500, according to the standby control instructions, performing corresponding standby operation; the standby operation includes, when the decision result is to delay standby, resetting the operation timer and outputting prompt information matched with the decision result; when the decision result is to enter standby immediately, marking the playing session according to the current content state, and controlling the television set to enter standby state.

2. The intelligent standby data analysis method based on a television set according to claim 1, characterized in that: The user operation behavior data includes the power-on time point, the power-off or standby time point, the channel switching operation, the application start operation, the volume adjustment operation, the pause operation and the return to the main interface operation; the content state data includes the type of the current playing content and the playing progress state.

3. The method of claim 1, wherein the method further comprises: Step S200 comprises: S210, dividing the historical time axis into predefined periodic time units and performing behavior law analysis; S220, based on the power-on time point, the power-off or standby time point data in the time unit, calculating the user activity intensity of the time unit; the user activity intensity is quantified by analyzing the multi-dimensional features of the user use behavior in the time unit in time distribution, use frequency and use duration; S230, according to the user activity intensity of each time unit, classifying all time units into different active period types; the active period types include high active period and low active period; S240, mapping the active period classification result with the time unit to generate the user active period model; the user active period model is used to predict the expected active period type of any given time point.

4. The intelligent standby data analysis method based on a television set according to claim 1, characterized in that: The content state information includes that the television set is currently in a live broadcast playing state, a video-on-demand playing state, a pause state or a main interface idle state; The recent operation sequence information is an operation sequence composed of user operations in time sequence within a preset time period, and the operation sequence information includes an operation type sequence and an operation intention feature determined based on a continuous distribution feature of the operation type; the operation type includes a channel switching operation, an application starting operation, a volume adjusting operation, a pause operation, and a return to main interface operation; and the operation intention feature includes a content selection, a parameter adjusting, or a playing interruption intention.

5. The method of claim 1, wherein the method is based on a television set. Step S400 includes: S410, defining decision weights for the active period classification corresponding to the current time point, content state information in the real-time context information, and operation intention features, respectively; S420, coupling the active period classification corresponding to the current time point with the content state information and the operation intention feature indicated by the real-time context information, and performing weighted evaluation according to the predefined decision weights to output a comprehensive decision value; S430, comparing the comprehensive decision value with a preset decision threshold to determine a standby triggering opportunity; when the comprehensive decision value is greater than a first decision threshold, it is determined that standby needs to be delayed; when the comprehensive decision value is less than a second decision threshold, it is determined that standby needs to be immediately; when the comprehensive decision value is between the second decision threshold and the first decision threshold, it is determined that standby needs to be cancelled; and the first decision threshold is greater than the second decision threshold; S440, generating a corresponding standby control instruction according to the determined standby triggering opportunity and the real-time context information; the standby control instruction includes the standby triggering opportunity and a standby prompt mode corresponding to the triggering opportunity and the context information.

6. The method of claim 1, wherein the method is based on a television set. Step S500 includes: S510, parsing the standby control instruction to obtain the standby triggering opportunity and the standby prompt mode included in the standby control instruction; S520, if the parsed standby triggering opportunity is delay standby, resetting the no-operation timer and controlling the user interface to output delay standby prompt information in the standby prompt mode; S530, if the parsed standby triggering opportunity is immediate standby, marking the playing session as recoverable according to the current content state, and controlling the user interface to output immediate standby prompt information in the standby prompt mode, and then controlling the television to enter the standby state; S540, if the parsed standby triggering opportunity is cancel standby, resetting the no-operation timer and controlling the user interface to output cancel standby notification information in the standby prompt mode.

7. A television-based intelligent standby data analysis system, characterized by: The system includes a data acquisition module, a user behavior analysis module, a context analysis module, an intelligent decision module, and a standby control module; The data acquisition module is configured to acquire user operation behavior data and content state data in real time during the operation of the television; The user behavior analysis module is connected to the data acquisition module and is configured to construct a user active period model by analyzing the distribution of usage behavior in multiple dimensions within a time unit based on the user operation behavior data acquired in a historical time period. The context analysis module is connected to the user behavior analysis module and is configured to analyze the content state of the television at the current time point based on the user active period model and the content state data acquired in real time. The intelligent decision module is connected to the user behavior analysis module and the context analysis module, and is configured to classify the active period corresponding to the current time point, the content state information in the real-time context information, and the operation intention feature, and to perform weighted evaluation according to the predefined decision weights to output a comprehensive decision value. The standby control module is connected to the intelligent decision module and the context analysis module, and is configured to compare the comprehensive decision value with a preset decision threshold to determine a standby triggering opportunity, and to generate a corresponding standby control instruction according to the determined standby triggering opportunity and the real-time context information. The context analysis module is connected with the data collection module, and is configured to collect and analyze current real-time context information when the television is in operation for a duration reaching a preset threshold; the real-time context information includes content state information indicating a playing state and operation sequence information reflecting a user intention; The intelligent decision module is connected with the user behavior analysis module and the context analysis module, and is configured to generate standby control instructions by performing multi-factor logic weighted decision in combination with a real-time context information and an active period classification corresponding to a current time point in the user active period model; the standby control instructions include standby trigger timing and standby prompt mode; The standby control module is connected with the intelligent decision module, and is configured to perform standby operation according to the standby control instructions.

8. The intelligent standby data analysis system based on a television according to claim 7, characterized in that: The user behavior analysis module includes a period modeling unit and a model prediction unit; the period modeling unit is configured to divide a historical time axis into predefined periodic time units, calculate user active intensity and perform active period classification based on data in the time units, and generate a user active period model; and the model prediction unit is configured to store the user active period model and predict an active period classification according to a time point.

9. The intelligent standby data analysis system based on a television according to claim 7, characterized in that: The intelligent decision module includes a weight management unit and a decision analysis unit; the weight management unit is configured to predefine decision weights for active period classification, content state information and operation intention features; and the decision analysis unit is configured to couple and weightedly evaluate the active period classification, the content state information and the operation intention features, output a comprehensive decision value, and determine standby trigger timing according to an analysis result of the comprehensive decision value and a preset threshold.

10. The intelligent standby data analysis system based on a television according to claim 1, characterized in that: The standby control module includes an instruction analysis unit and an action execution unit; the instruction analysis unit is configured to analyze the standby control instructions to obtain standby trigger timing and standby prompt mode; and the action execution unit is configured to perform resetting of an operation timer, marking of a playing session for recoverability, control of the television into a standby state and control of a user interface to output prompt information according to the analyzed standby trigger timing. ​

Images