Short video traffic dynamic transaction and multi-dimensional value evaluation intelligent system

Abstract

The invention discloses a short video traffic dynamic transaction and multi-dimensional value evaluation intelligent system, which relates to the field of electronic commerce, and is characterized in that in a delivery window optimization module, a fixed delivery window length twinwindow is determined; according to the method, the delivery period can be adaptively adjusted according to the fractal robustness of the short video under multiple time scales such as minute scale, hour scale and day scale, and when the multi-time-dimension fractal continuous robust index FCSms is relatively high, the fixed delivery window length twinwindow is prolonged to utilize the long tail effect of the short video; and when the multi-time-dimension fractal continuous robust index FCSms is relatively low, the fixed delivery window length twinwindow is shortened to avoid excessive delivery. Compared with an existing fixed-time-length delivery mode, the system enables a related short-time delivery window of a short video platform to be closer to the life cycle of video traffic, and improves the accuracy and timeliness of advertisement delivery.

Description

Technical Field

[0001] The present invention relates to the field of e-commerce technology, and in particular to an intelligent system for dynamic short video traffic transactions and multi-dimensional value assessment. Background Art

[0002] With the development of e-commerce, the dynamic management and value assessment of internet short video content, in particular, is being explored. More specifically, this involves an intelligent system for dynamic short video traffic transactions and multi-dimensional value assessment based on multi-timescale fractal feature analysis. This system treats the changes in short video playback volume at multiple timescales, such as minutes, hours, and days, as multi-scale fractal curves. By combining multi-dimensional characteristic indicators such as fractal dimension, porosity, and slope variance, it establishes a quantitative model for traffic continuity and stability. The analysis results are directly applied to optimizing advertising resource delivery windows and prioritizing transactions, enabling the refined exploration and utilization of the commercial value of short videos.

[0003] Currently, in existing short video traffic evaluation and advertising delivery systems, most methods use fixed time window statistics, such as 24-hour total playback volume and 7-day incremental playback volume, as the basis for video value judgment. This single-time-scale statistical method ignores the self-similarity characteristics of short video playback volume at different time scales and cannot reflect the scalability and multi-stage growth model of traffic. At the same time, existing methods generally use static priority or single-indicator ranking when allocating advertising resources, lacking cross-time-scale traffic structure stability analysis, resulting in some videos with long-term growth potential being prematurely eliminated, or the resource allocation rhythm being seriously misaligned with the video popularity curve, thereby reducing the overall return on advertising and user reach.

[0004] The main cause of the above defects is that the existing traffic evaluation system fails to capture the fractal characteristics and gap structure characteristics of video playback volume at multiple time scales, and lacks the ability to model dynamic factors such as popularity outbreak delay and cross-scale stability. When this defect occurs, it will lead to the following abnormal effects: First, videos that explode in popularity in the later period cannot obtain sufficient delivery support in the early stage, missing the optimal detonation period; second, videos that have a short-term surge but quickly decay occupy a large amount of high-value resources in the early stage of delivery, resulting in a decrease in advertisers' return on investment; third, the advertising delivery window is seriously mismatched with the video traffic life cycle, resulting in serious waste of exposure in some delivery cycles and insufficient resource supply in other cycles. The above problems not only reduce the efficiency of advertising resource utilization, but may also affect the overall commercial revenue of the platform and the user's viewing experience. Summary of the Invention

[0005] In response to the shortcomings of the existing technology, the present invention provides an intelligent system for dynamic trading and multi-dimensional value assessment of short video traffic, which solves the problems mentioned in the background technology.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: an intelligent system for dynamic short video traffic trading and multi-dimensional value assessment, comprising: Traffic multi-dimensional collection module: This module collects video playback data from short video platforms in real time through collection tools, and transmits the video playback data to the cross-scale fractal analysis module; Cross-scale fractal analysis module: This module pre-processes video playback data to obtain a standardized data set, and then extracts features based on the standardized data set to obtain a fractal data set. Fractal robustness evaluation module: It calculates and outputs the multi-time-dimensional fractal persistence robustness index FCSms based on the fractal data set, and presets the multi-time-dimensional fractal persistence robustness index threshold Fth for preliminary comparative evaluation; Prioritization module: triggers transaction priority strategies through preliminary comparative evaluation, calculates and outputs the Fractal Survival Delivery Window Score (FSWms) based on the multi-time-dimensional Fractal Continuous Robustness Index (FCSms), and prioritizes video transactions based on the Fractal Survival Delivery Window Score (FSWms); Delivery window optimization module: Determine the delivery window length twindow based on the multi-time dimension fractal persistence robustness index FCSms, and allocate advertising resources and delivery time based on the ranking results of the fractal survival delivery window score FSWms.

[0007] Preferably, the traffic multi-dimensional acquisition module includes an acquisition unit and a transmission unit; The collection unit collects video playback data from the short video platform at the minute, hour, and day time scales by setting a collection tool on the short video platform; The acquisition tool includes a data acquisition terminal and an acquisition control program; The data acquisition terminal is deployed at the back-end data interface node of the short video platform, located between the platform's data exchange layer and the business processing layer, to obtain the original playback data that has not been aggregated and compressed. The data acquisition terminal has data buffering, real-time data packaging and network communication functions; The acquisition control program runs in the data acquisition terminal, calls the platform open interface API, collects the log stream of the short video platform, and extracts the video playback data and additional parameters from the log stream; The video playback volume data includes a unique video identifier, a corresponding timestamp, a scale consistency slope variance Xslope, a heat burst delay time TB, a reference delay time TR, future exposure availability EA, a burst skewness SB, a minute-level playback volume sequence Pmin, an hour-level playback volume sequence Phour, and a day-level playback volume sequence Pday; The sampling step of the minute-level playback volume sequence Pmin is 1 minute, and the box counting scale r∈{1,2,4,8,16} minutes; The sampling step of the hour-level playback volume sequence Phour is 1 hour, and the box counting scale r∈{1,2,4,8} hours; The sampling step of the daily playback volume sequence Pday is 1 day, and the box counting scale r∈{1,2,4} days.

[0008] Preferably, the transmission unit packages the video playback data into data packets with a unique video identifier and a corresponding timestamp according to the time scale, and sends them to the cross-scale fractal analysis module in an encrypted manner through the high-speed message queue transmission protocol MQTT. After receiving the data packets, the cross-scale fractal analysis module reconstructs the order of the data packets and generates a receipt confirmation receipt.

[0009] Preferably, the cross-scale fractal analysis module includes a data preprocessing unit for preprocessing the video playback volume data to generate a standardized data set; The preprocessing includes time synchronization processing, dimension unification processing and data structuring and storage; The time synchronization process aligns the time stamps of the minute-level playback volume sequence Pmin, the hour-level playback volume sequence Phour, and the day-level playback volume sequence Pday in the video playback volume data according to the global time reference, and fills in the missing time data points through the interpolation algorithm, wherein the interpolation method is linear interpolation; The dimension unification process uses the Z-Score standardization method to perform numerical standardization on all parameters in the video playback data, so that the parameters of different physical dimensions are converted into dimensionless values ​​with a mean of 0 and a standard deviation of 1; The data structuring and storage is achieved by grouping the video playback volume data after dimension unification according to time scale and parameter type and storing them as a standardized data set, and adding a video unique identifier and a corresponding timestamp to each record.

[0010] Preferably, the cross-scale fractal analysis module further includes a fractal feature extraction unit for performing feature extraction on the standardized data set to obtain a fractal data set; The fractal data set includes minute-level fractal dimension Dmin, hour-level fractal dimension Dhour, day-level fractal dimension Dday, geometric mean porosity LG, minute-level porosity Amin, hour-level porosity Ahour, day-level porosity Aday, scale consistency slope variance Xslope, heat outbreak delay time TB, reference delay time TR, future exposure availability EA and outbreak skewness SB.

[0011] Preferably, the fractal robustness evaluation module includes a fractal robustness analysis unit; The fractal robustness analysis unit extracts the heat outbreak delay time TB and the reference delay time TR from the fractal data set to form a slow-heating factor, constructs the scale consistency slope variance Xslope into a stability factor, and constructs the minute-level fractal dimension Dmin, the hour-level fractal dimension Dhour, and the day-level fractal dimension Dday in combination with the geometric mean porosity LG into a complexity and spatial normalization factor. After multiplying and summarizing the slow-heating factor, the stability factor and the complexity and spatial normalization factor, the multi-time dimension fractal persistence robustness index FCSms is calculated and output; it is used to measure the complexity, stability, gap structure characteristics and heat growth pattern of the flow change curve of a short video at different time scales, and reflect its long-term sustainability potential through a single numerical value.

[0012] Preferably, the fractal robustness evaluation module includes an evaluation unit; The evaluation unit calculates the multi-time dimension fractal persistence robustness index FCSms in batches on the historical sample set of the same category to obtain the empirical distribution, takes the 80th percentile as the initial multi-time dimension fractal persistence robustness index threshold Fth, and re-evaluates it on a monthly rolling basis; Then, a preliminary comparative evaluation is conducted between the real-time multi-time fractal persistence robustness index FCSms and the multi-time fractal persistence robustness index threshold Fth to determine the persistence robustness of the short video traffic structure at multiple time scales. The trigger is based on the preliminary comparative evaluation results. The specific evaluation contents are as follows: When the multi-time dimension fractal persistence robustness index FCSms ≥ the multi-time dimension fractal persistence robustness index threshold Fth, the current short video traffic structure is determined to be robust, and the transaction priority strategy is triggered at this time; When the multi-time dimension fractal persistence robustness index FCSms is less than the multi-time dimension fractal persistence robustness index threshold Fth, it is determined that the current short video traffic structure is not robust and is processed according to the conventional strategy.

[0013] Preferably, the prioritization module includes a fractal survival delivery window analysis unit; The fractal survival delivery window analysis unit obtains the effective fractal dimension Deff by performing mean calculation based on the minute-level fractal dimension Dmin, the hour-level fractal dimension Dhour and the day-level fractal dimension Dday after the transaction priority strategy is triggered; At the same time, the multi-time dimension fractal continuous robustness index FCSms output by the fractal robustness evaluation module is received and combined with the effective fractal dimension Deff, as well as the scale consistency slope variance Xslope, future exposure availability EA and outbreak skewness SB to calculate and output the fractal survival delivery window score FSWms.

[0014] Preferably, the prioritization module further includes a sorting unit; The sorting unit generates a priority queue for all videos that trigger the transaction priority strategy in descending order of the fractal survival delivery window score FSWms to sort the transactions. The specific sorting steps are as follows: The 85th percentile P85 and the 95th percentile P95 of the distribution of the Fractal Survival Delivery Window Score (FSWms) based on historical short videos are used as step thresholds. The real-time Fractal Survival Delivery Window Score (FSWms) is then compared with the step thresholds for a secondary evaluation, and a priority queue is generated based on the secondary comparative evaluation. When the fractal survival delivery window score FSWms ≥ 95th percentile P95, it is classified as secondary price increase and loaded first; When the 85th percentile P85≤Fractal Survival Window Score FSWms<95th percentile P95, it is classified as Level 1 surcharge and conventional loading; When the fractal survival delivery window score FSWms is less than the 85th percentile P85, it is divided into three levels of price increase and is divided into the basic layer loading.

[0015] Preferably, the delivery window optimization module includes a delivery optimization unit; The delivery optimization unit determines a fixed delivery window length twindow based only on the multi-time fractal persistence robustness index FCSms when the target short video passes the preliminary comparative evaluation. The specific determination form is: twindow = t0 × (1 + FCSms), where t0 represents the basic delivery duration set by the platform; And register to form a fixed delivery window [ts,te], ts represents the delivery window start time, te represents the delivery window end time; The delivery optimization unit does not accept any increase or decrease in the fixed delivery window length twindow by the fractal survival delivery window score FSWms; when the fixed delivery windows [ts, te] of different videos overlap and platform resources are limited, the loading intensity and starting order of advertising resources are allocated only within each fixed delivery window [ts, te] according to the ranking results of the fractal survival delivery window score FSWms, keeping the start and end boundaries of each fixed delivery window [ts, te] unchanged.

[0016] The present invention provides an intelligent system for dynamic short video traffic trading and multi-dimensional value assessment. It has the following beneficial effects: (1) The system determines a fixed delivery window length twindow in the delivery window optimization module, so that the delivery cycle can be adaptively adjusted according to the fractal robustness of short videos at multiple time scales such as minutes, hours, and days. When the multi-time dimension fractal persistence robustness index FCSms is high, the fixed delivery window length twindow is extended to take advantage of the long tail effect of the video; when the multi-time dimension fractal persistence robustness index FCSms is low, the fixed delivery window length twindow is shortened to avoid over-delivery. Compared with the existing fixed-duration delivery method, this system makes the delivery window more closely fit the life cycle of video traffic, improving the accuracy and timeliness of advertising delivery.

[0017] (2) In the delivery window optimization module, the start and end boundaries of the fixed delivery window [ts,te] are not affected by the adjustment of the fractal survival delivery window score FSWms. Only within the fixed window, the advertising resource loading intensity and the starting order are prioritized according to the fractal survival delivery window score FSWms. When platform resources are limited and delivery windows of different videos overlap, this strategy ensures fairness in window length while allocating high-value resources to videos with higher scores through priority queues, avoiding resources being occupied by low-potential videos, thereby improving resource allocation efficiency.

[0018] (3) The system uses the multi-time fractal persistence robustness index FCSms to determine the delivery time boundary and the fractal survival delivery window score FSWms to optimize the resource allocation order, thus achieving a division of labor and collaboration between the time dimension and the resource dimension. This collaborative mechanism effectively prevents low-robustness videos from occupying high-value exposure resources during the window period, while ensuring that high-potential videos receive sufficient delivery support during the critical time period, thereby improving the conversion rate of advertising delivery and overall commercial benefits and reducing resource waste. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is a structural diagram of an intelligent system for dynamic short video traffic trading and multi-dimensional value assessment according to the present invention.

[0020] Figure 2 Schematic diagram of the time scale playback sequence. DETAILED DESCRIPTION

[0021] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0022] Example 1: The present invention provides an intelligent system for dynamic short video traffic trading and multi-dimensional value assessment. Figure 1 and Figure 2 ,include: Traffic multi-dimensional collection module: This module collects video playback data from short video platforms in real time through collection tools, and transmits the video playback data to the cross-scale fractal analysis module; Cross-scale fractal analysis module: This module pre-processes video playback data to obtain a standardized data set, and then extracts features based on the standardized data set to obtain a fractal data set. Fractal robustness evaluation module: It calculates and outputs the multi-time-dimensional fractal persistence robustness index FCSms based on the fractal data set, and presets the multi-time-dimensional fractal persistence robustness index threshold Fth for preliminary comparative evaluation; Prioritization module: triggers transaction priority strategies through preliminary comparative evaluation, calculates and outputs the Fractal Survival Delivery Window Score (FSWms) based on the multi-time-dimensional Fractal Continuous Robustness Index (FCSms), and prioritizes video transactions based on the Fractal Survival Delivery Window Score (FSWms); Delivery window optimization module: Determine the delivery window length twindow based on the multi-time dimension fractal persistence robustness index FCSms, and allocate advertising resources and delivery time based on the ranking results of the fractal survival delivery window score FSWms.

[0023] In this embodiment, the system deploys a data acquisition terminal and acquisition control program in the short video platform through the traffic multi-dimensional acquisition module, and collects video playback data in real time at the minute, hour, and day time scales, including the video unique identifier, timestamp, and scale consistency slope variance Xslope, heat burst delay time TB, reference delay time TR, future exposure available EA, burst skewness SB and other parameters, and sends the data to the cross-scale fractal analysis module via an encrypted transmission protocol. Secondly, the cross-scale fractal analysis module performs time synchronization, dimension unification, and data structuring preprocessing on the received video playback data to generate a standardized data set; based on the standardized data set, fractal feature parameters are extracted, including minute-level fractal dimension Dmin, hour-level fractal dimension Dhour, day-level fractal dimension Dday, geometric mean porosity LG, etc., to form a fractal data set, providing multi-dimensional input for subsequent robustness analysis. Next, the fractal robustness assessment module uses the fractal dataset to calculate and output the multi-time fractal persistence robustness index (FCSms) through a structural combination of slow-heating factors, stability factors, and complexity and spatial normalization factors. It also determines the multi-time fractal persistence robustness index threshold (Fth) based on the historical distribution of videos in the same category. It then conducts a preliminary comparative assessment of the real-time multi-time fractal persistence robustness index (FCSms) against the threshold value to determine whether the video's traffic structure at multiple time scales exhibits persistence robustness. When the judgment result triggers a transaction priority strategy, the prioritization module combines the effective fractal dimension (Deff), the multi-time fractal persistence robustness index (FCSms), the scale consistency slope variance (Xslope), the future exposure availability (EA), and the burst skewness (SB) to calculate and output the fractal survival delivery window score (FSWms). It then ranks all videos that trigger the transaction priority strategy by score, generating a priority queue that is divided into priority loading, regular loading, and base-layer loading levels, thereby optimizing the order and intensity of advertising resource loading. Finally, the delivery window optimization module determines a fixed delivery window length, twindow, based on the multi-time fractal persistence robustness index FCSms, forming a fixed delivery time interval [ts, te]. Within this window, advertising resources are allocated based on the ranking results of the fractal survival delivery window score FSWms. This ensures that high-potential videos receive priority exposure resources when different video windows overlap and resources are limited. Through this implementation, the system achieves multi-scale feature quantification of video traffic, precise robustness assessment, and hierarchical priority allocation of advertising resources. This not only improves the timeliness and accuracy of advertising delivery, but also reduces resource waste, maximizes the commercial conversion potential of high-potential videos, and achieves the goal of optimal traffic value allocation in a dynamic competitive environment.

[0024] Example 2: Please refer to Figure 1 and Figure 2 ,Specifically: the traffic multi-dimensional collection module includes a collection unit and a transmission unit; The collection unit sets up collection tools on the short video platform to collect video playback data from the short video platform at the minute, hour, and day time scales; The acquisition tools include data acquisition terminals and acquisition control programs; The data collection terminal is deployed at the back-end data interface node of the short video platform, located between the platform's data exchange layer and business processing layer, to obtain the original playback data that has not been aggregated and compressed. The data collection terminal has data buffering, real-time data packaging and network communication functions. The acquisition control program runs in the data acquisition terminal, calls the platform's open interface API, collects the log stream of the short video platform, and extracts video playback data and additional parameters from the log stream; Video play data includes the video's unique identifier, corresponding timestamp, scale consistency slope variance Xslope, popularity outbreak delay time TB, reference delay time TR, future exposure availability EA, outbreak skewness SB, minute-level play sequence Pmin, hour-level play sequence Phour, and day-level play sequence Pday. The sampling step of the minute-level playback volume sequence Pmin is 1 minute, and the box counting scale r∈{1,2,4,8,16} minutes; The sampling step of the hour-level playback volume sequence Phour is 1 hour, and the box counting scale r∈{1,2,4,8} hours; The sampling step of the daily playback volume sequence Pday is 1 day, and the box counting scale r∈{1,2,4} days; in: The scale consistency slope variance Xslope represents the stability of the video playback volume change rate, i.e., the slope, at different time scales (minutes, hours, and days). First, the playback volume change slope sequence is calculated at each time scale. We then compare the consistency of slope changes at different scales and take their variance as the scale consistency slope variance Xslope. Smaller variances indicate more consistent playback trends across scales, while larger variances indicate greater differences in changes at different scales. This is used to measure the multi-scale stability of video traffic and is an important input for determining fractal persistence. The burst latency (TB) refers to the time interval from the time a video is released to the first significant increase in views, reflecting the lag in the explosion of video popularity. It is useful for determining whether a video is suitable for short-term or long-term delivery windows during advertising. The reference delay time TR is the relative difference in the target video's delay time based on the average hot burst delay time of similar videos. It is used as a reference in fractal analysis to correct the delivery window prediction model. The Future Exposure Availability EA predicts the total amount of exposure a video may receive in the platform's recommendation algorithm over a period of time. This prediction is based on the ARIMA model, combined with historical playback trends, current platform recommendation weights, and user activity models. This directly influences the bidding strategy and placement sequence in advertising transactions. The burst skewness SB reflects the symmetry of the video popularity outbreak curve, describes the difference in playback volume distribution before and after the outbreak, determines the morphological characteristics of the video popularity life cycle, and assists in selecting the duration of the delivery.

[0025] The transmission unit packages the video playback data into data packets with unique video identifiers and corresponding timestamps according to the time scale, and sends them to the cross-scale fractal analysis module in an encrypted manner through the high-speed message queue transmission protocol MQTT. After receiving the data packets, the cross-scale fractal analysis module reconstructs the order of the data packets and generates a receipt confirmation receipt.

[0026] In this embodiment, the system's multi-dimensional traffic acquisition module works in coordination through the acquisition unit and the transmission unit. The acquisition unit collects video playback data and additional parameters in real time at the minute, hour, and day time scales of the short video platform. The minute level captures short-term fluctuations and immediate hot spots, the hourly level reflects mid-term recommendations and active rhythm, and the day level depicts the long-tail effect and sustained attractiveness, ensuring the comprehensiveness and stability of the traffic structure at multiple time scales; the data acquisition terminal is deployed between the platform's data exchange layer and the business processing layer, directly obtaining unaggregated and uncompressed raw data, avoiding data accuracy loss caused by internal aggregation, sampling, or caching of the platform, thereby obtaining a high-fidelity traffic curve in subsequent fractal analysis and reducing the risk of underestimation of the fractal dimension; the scale consistency slope variance Xslope is used to measure the consistency of the changing trend of video playback volume at multiple time scales. The smaller the variance, the more stable the trend, which can effectively identify videos suitable for continuous delivery and eliminate short-lived traffic; the dual-parameter structure of the hot burst delay time TB and the reference delay time TR is used to distinguish between instant burst types For slow-burning videos, window matching errors can be avoided. For example, slow-burning videos can be allocated a longer delivery cycle to release their value; the future exposure supply EA predicts the future redeemable traffic scale based on the ARIMA model, ensuring that advertising resources match potential exposure capabilities and preventing over-delivery or supply shortage; the burst skewness SB reflects the symmetry of the heat outbreak curve, distinguishes between concentrated outbreaks and steady growth patterns, and guides the timing of delivery to avoid a drop in conversion rate; the transmission unit transmits data through MQTT encryption and reconstructs the time sequence to ensure the integrity and consistency of the time series and prevent fractal calculation deviations caused by data disorder or tampering; the above implementation process, through precise collection, robust feature extraction and secure transmission throughout the entire link, not only improves the authenticity and stability of the fractal analysis input data, but also physically ensures that the obtained multi-time scale traffic characteristics can accurately reflect the dissemination rhythm and life cycle characteristics of short videos, thereby providing a high-quality data foundation for subsequent robustness evaluation and priority sorting, and achieving precise advertising delivery timing and resource allocation and maximizing benefits.

[0027] Example 3: Please refer to Figure 1 ,Specifically: the cross-scale fractal analysis module includes a data ,preprocessing unit for preprocessing the video playback data and ,generating a standardized data set; Preprocessing includes time synchronization, dimension unification, and data structuring and storage; Time synchronization involves aligning the timestamps of the minute-level playback sequence Pmin, hour-level playback sequence Phour, and day-level playback sequence Pday in the video playback data according to the global time base. Missing data points are then filled in using a linear interpolation algorithm to ensure that data at different time scales are comparable at the same reference time. Dimensional unification processing uses the Z-Score standardization method to perform numerical standardization on all parameters in the video playback data, so that parameters with different physical dimensions are converted into dimensionless values ​​with a mean of 0 and a standard deviation of 1; Data structuring and storage: The video playback data after dimension unification is grouped and stored as a standardized data set according to time scale and parameter type, and a unique video identifier and corresponding timestamp are added to each record to ensure that subsequent fractal calculations can be directly called.

[0028] The cross-scale fractal analysis module also includes a fractal feature extraction unit for performing feature extraction on the standardized data set to obtain a fractal data set; The fractal data set includes minute-level fractal dimension Dmin, hour-level fractal dimension Dhour, day-level fractal dimension Dday, geometric mean porosity LG, minute-level porosity Amin, hour-level porosity Ahour, day-level porosity Aday, scale consistency slope variance Xslope, heat outbreak delay time TB, reference delay time TR, future exposure availability EA and outbreak skewness SB; Among them, the minute-level fractal dimension Dmin, hour-level fractal dimension Dhour and day-level fractal dimension Dday are respectively extracted from the standardized data set by extracting the minute-level playback volume sequence Pmin, hour-level playback volume sequence Phour and day-level playback volume sequence Pday, arranging the playback volume sequence of each time scale in ascending order by timestamp and constructing a two-dimensional curve of time and playback volume. The fractal dimension of the minute-level, hour-level and day-level curves is calculated based on the box counting method, and the minute-level calculation result is recorded as the minute-level fractal dimension Dmin, the hour-level calculation result is recorded as the hour-level fractal dimension Dhour, and the day-level calculation result is recorded as the day-level fractal dimension Dday. The steps are as follows: Step 1: Set multiple box counting scales ri, where the unit is the minimum time interval of the corresponding time scale; Step 2: Divide the time axis into intervals using multiple box counting scales ri, and count the minimum number of boxes N(ri) required to cover the curve; Step 3: Calculate lnN(ri) and ln(1 / ri) and perform linear regression fitting to obtain the fractal dimension ; Wherein, d represents the integral function and ln represents the natural logarithm; The minute-level porosity Amin, hour-level porosity Ahour, and day-level porosity Aday are obtained by the sliding box porosity calculation method, the steps of which are as follows: select a preset box counting scale set R for the minute-level playback volume sequence Pmin, hour-level playback volume sequence Phour, and day-level playback volume sequence Pday on the corresponding time scale, and slide the playback volume time series according to the box counting scale ri to obtain the mass Mr, where the mass Mr is defined as the binary occupied mass; calculate the porosity curve, and take the arithmetic mean of each box counting scale ri as the index to obtain the scale porosity; Minute-level porosity Amin: This indicates the sparseness of the video playback curve at the minute-level time scale. This refers to the proportion of time in the entire minute-level time series where the playback volume is below zero, as well as the distribution density of these low-value intervals on the time axis. Hourly porosity (Ahour): measures the distribution and proportion of low-play volume intervals at an hourly granularity, also based on the play volume curve, at an hourly time scale. Aday: This indicates the proportion of "empty windows" of video playback over a long period of time at the daily time scale. For example, it is the ratio of the video's low-interest period to the entire day. The geometric mean porosity LG is obtained by averaging the minute-level porosity Amin, hour-level porosity Ahour, and day-level porosity Aday, and is used to penalize the hollow complexity.

[0029] In this embodiment, the cross-scale fractal analysis module of the system first completes time synchronization processing, dimensional unification processing and data structuring and storage by the data preprocessing unit: the time synchronization process aligns the minute-level playback volume sequence Pmin, the hour-level playback volume sequence Phour and the day-level playback volume sequence Pday according to a unified benchmark and fills the gaps with linear interpolation, in order to eliminate the false fluctuations caused by cross-scale time dislocation; the dimensional unification process performs Z-Score standardization on all parameters to convert different physical dimensions into comparable dimensionless values ​​to avoid a large-scale quantity "drowning" the fine-grained signal; the data structuring and storage groups the standardized multi-scale data according to "time scale × parameter type" and provides a unique video identifier. and timestamp to ensure that the fractal calculation can be directly called without loss and traceability; then, the fractal feature extraction unit extracts the minute-level fractal dimension Dmin, hour-level fractal dimension Dhour, and day-level fractal dimension Dday from the time-playback volume curve based on the box counting method to characterize the self-similar complexity; and uses the sliding box porosity method to obtain the minute-level porosity Amin, hour-level porosity Ahour, day-level porosity Aday and geometric mean porosity LG to quantify the "window" sparsity of the flow; at the same time, the scale consistency slope variance Xslope, heat outbreak delay time TB, reference delay time TR, future exposure availability EA and outbreak skewness SB are retained as robustness and rhythm judgment factors. This is done because in real platform environments, common situations such as "holiday breakpoints," "time drift caused by batch new product releases," and "scale imbalance caused by promotional peaks" can directly distort fractal dimension and porosity assessments. Data misalignment and dimensional bias can be effectively avoided through alignment, standardization, and structuring. Using a combination of fractal dimension and porosity to simultaneously characterize "complexity and sparsity" can avoid misjudgments caused by looking only at peak values ​​or only at mean values, such as short spikes being mistaken for long tails or slow-burning videos being eliminated early. The resulting beneficial effects are: more robust feature input, significantly improved cross-scale comparability, and more reliable differentiation between long tails and short peaks. This directly improves the accuracy of the subsequent multi-time-dimensional fractal persistence robustness index FCSms and the fractal survival delivery window score FSWms, thereby reducing mismatches and resource waste on the delivery side and improving the yield rate of priority sorting and window configuration.

[0030] Example 4: Please refer to Figure 1 ,Specifically: the fractal robustness evaluation module includes a fractal robustness ,analysis unit; The fractal robustness analysis unit extracts the heat outbreak delay time TB and reference delay time TR from the fractal dataset to form a slow-heating factor, constructs the scale consistency slope variance Xslope into a stability factor, and combines the minute-level fractal dimension Dmin, hour-level fractal dimension Dhour, and day-level fractal dimension Dday with the geometric mean porosity LG to construct a complexity and spatial normalization factor. After multiplying the slow-heating factor, stability factor, and complexity and spatial normalization factor, the multi-time-dimensional fractal persistence robustness index FCSms is calculated and output. It is used to measure the complexity, stability, void structure characteristics, and heat growth pattern of a short video's flow change curve at different time scales, and reflects its long-term sustainability potential through a single value. The multi-time dimension fractal persistence robustness index FCSms is calculated and output by the following algorithm formula; ; Where, ln represents the natural logarithm; Basic sources: Box-counting dimension: a basic quantity in fractal geometry, the minimum number of boxes N(ri) for the time-playback curve under box size r. The linear regression is approximated by the slope; Porosity / Lacunarity: A "sparseness / voidity" indicator commonly used in texture statistics and fractal analysis, and a sliding box quality Mr: Scale consistency idea: The variance of the multi-scale regression slope of multiple box counting scales ri measures the cross-scale stability. The smaller the variance, the more stable it is. Logarithmic stabilization of time ratio: Taking ln(1+⋅) we get a dimensionless, robust "slow-heat" enhancement term; Structural changes, deduction based on the above Fractal complexity fusion: using As a geometric mean fractal dimension across the three scales of minute / hour / day, it avoids subjective weighting; Gap Penalization Normalization: Apply a logarithmic penalty to the "hollow complexity" with 1+ln(LG); when the overall sparsity increases, the exponent is naturally suppressed; Stability gain / rejection: Suppress cross-scale unstable samples; Slow heat amplification: Reward the “slow-burn, long-tail” feature; in, Indicates the slow-burn factor. When the heat burst delay time TB is greater than the reference delay time TR, the index increases, favoring slow-burn videos with "long-lasting momentum." When the heat burst delay time TB=0, this item is 0, suppressing samples that immediately peak and then quickly decay. It represents the stability factor. The more stable the cross-scale slope is (Xslope→0), the factor →1. If the multi-scale morphology is unstable (Xslope is large), the overall index is depressed. Represents the complexity and spatial normalization factor, numerator: the higher the fractal dimension, the stronger the fluctuation complexity / self-similarity; Denominator: When the geometric mean porosity LG→1 (uniform and continuous), ln(LG)→0, no penalty; when the geometric mean porosity LG increases and the pores become sparser, ln(⋅) increases, forming a logarithmic penalty to avoid "hollow complexity"; Dimensional consistency and rationality verification: Fractal dimension D∗: dimensional quantity, naturally dimensionless; Geometric mean porosity LG: composed of the expectation / variance ratio, dimensionless; ln(LG) is also dimensionless, 1+ln(⋅) is dimensionless; Scale consistency slope variance Xslope: the variance of the regression slope, the slope is dimensionless and the variance is also dimensionless; dimensionless; Time ratio : time / time is dimensionless; ln(1+⋅) is dimensionless; Multiplication and division structure: dimensionless × dimensionless × dimensionless. The multi-time dimension fractal persistence robustness index FCSms is dimensionless, which is in line with common sense.

[0031] The fractal robustness assessment module also includes an assessment unit; The evaluation unit calculates the multi-time-dimensional fractal persistence robustness index FCSms in batches on historical sample sets of the same category to obtain an empirical distribution. The 80th percentile is taken as the initial multi-time-dimensional fractal persistence robustness index threshold Fth, and is re-evaluated on a monthly rolling basis. Then, a preliminary comparative evaluation is conducted between the real-time multi-time fractal persistence robustness index FCSms and the multi-time fractal persistence robustness index threshold Fth to determine the persistence robustness of the short video traffic structure at multiple time scales. The trigger is based on the preliminary comparative evaluation results. The specific evaluation contents are as follows: When the multi-time dimension fractal persistence robustness index FCSms ≥ the multi-time dimension fractal persistence robustness index threshold Fth, the current short video traffic structure is determined to be robust, and the transaction priority strategy is triggered at this time; When the multi-time dimension fractal persistence robustness index FCSms is less than the multi-time dimension fractal persistence robustness index threshold Fth, it is determined that the current short video traffic structure is not robust and is processed according to the conventional strategy.

[0032] In this embodiment, the fractal robustness analysis unit of the fractal robustness evaluation module of the system first constructs three factors from the fractal data set and performs product aggregation to output the multi-time dimension fractal persistence robustness index FCSms: First, the slow heat factor is composed of the heat outbreak delay time TB and the reference delay time TR, the purpose of which is to distinguish between "immediate outbreak" and "slow heat long tail". The physical meaning is that when TB / TR is larger, it means that the heat release is more lasting and can maintain effective reach within a longer window; second, the stability factor is composed of the scale consistency slope variance Xslope, the purpose of which is to constrain the trend consistency across minutes / hours / days. The physical meaning is that the smaller the Xslope, the better. The more in-phase the scale curves are, the less likely it is to experience a "short surge followed by a sharp drop", thereby reducing the risk of deployment; thirdly, the complexity and spatial normalization factor is composed of the geometric mean of the minute-level fractal dimension Dmin, the hour-level fractal dimension Dhour, and the day-level fractal dimension Dday, and the geometric mean porosity LG. The purpose is to simultaneously characterize the self-similar complexity and the "window" sparsity. The physical meaning is that the higher the fractal dimension, the richer the structure, and the larger 1+ln(LG), the more empty the traffic is and the more it needs to be suppressed; the reason for adopting the above-mentioned structured combination is that in real scenarios, "short spikes", "holiday breakpoints", "recommendation strategy switching" and other situations are common, which will lead to Misjudgment of a single indicator, such as misjudging a short-lived video rushing to the top by only looking at the peak value as having high potential, can be effectively avoided by linking TB / TR, Xslope and fractal dimension - LG; the evaluation unit then batch calculates the multi-time dimension fractal persistence robustness index FCSms on historical samples of the same category and takes the 80th percentile to construct the multi-time dimension fractal persistence robustness index threshold Fth, the purpose of which is to provide an objective admission line that can be rolled over time. The physical meaning is to limit the lower bound of "stability" by the statistical steady state of the group; during online judgment, if the multi-time dimension fractal persistence robustness index FCSms ≥ the multi-time dimension fractal persistence robustness index threshold Fth, the transaction priority strategy is directly triggered, otherwise it is entered. For example, for a "slow-burning, low-gap, cross-scale stable" video, the multi-time dimension fractal persistence robustness index FCSms is usually higher than the multi-time dimension fractal persistence robustness index threshold Fth, and can be stably converted within a longer window; on the contrary, a video with a "single-peak steep drop, high porosity, and large Xslope" will be suppressed to prevent high-value resources from being occupied by instantaneous noise; the resulting beneficial effects are: the admission judgment is more noise-resistant, reducing short-lived content that mistakenly enters the high window, the window and rhythm are more matched, slow-burning videos are not eliminated early, and the risk of delivery and waste of resources are significantly reduced, thereby improving the hit rate and overall ROI of subsequent priority sorting and delivery window optimization.

[0033] Example 5: Please refer to Figure 1 ,Specifically: the prioritization module includes a fractal survival delivery window analysis unit; The fractal survival delivery window analysis unit obtains the effective fractal dimension Deff by performing mean calculation based on the minute-level fractal dimension Dmin, hour-level fractal dimension Dhour, and day-level fractal dimension Dday after the transaction priority strategy is triggered. At the same time, the multi-time fractal persistence robustness index FCSms output by the fractal robustness evaluation module is received and combined with the effective fractal dimension Deff, as well as the scale consistency slope variance Xslope, future exposure availability EA and outbreak skewness SB, to calculate and output the fractal survival delivery window score FSWms; The Fractal Survival Window Score (FSWms) is calculated and outputted through the following algorithm formula: ; Where, ln represents the natural logarithm; Numerator: reflects "redeemable ability" and "Sustainable Gains Across Scales" FCSms, ; Denominator: Suppressing "overall sparsity" , “instability across scales” and “peak outbreak risk” ; The origin, modification, and deduction of formulas: fractal dimension (Box-counting), porosity / lacunarity, and long-range correlation / power-law tail concepts (time series, survival analysis); The three-scale fractal dimensions are combined into the effective fractal dimension Deff, and then used Approximately characterize the "long-tail relegation factor". When Deff → 2, the index → ​​0 and the relegation factor → 1, indicating a more stable tail; when Deff decreases, the relegation factor is < 1. The logarithmic return of the saleable inventory is expressed as ln(1+EA) to avoid numerical imbalance caused by extreme scale differences. "Global sparsity", "cross-scale instability" and "spike morphology" are included in the denominator as three independent stability suppression terms to maintain interpretability and monotonicity; Structurally, artificial weights are avoided, and a multiplicative normalized "structural combination" is adopted, which works in conjunction with the multi-time dimension fractal persistence robustness index FCSms: the multi-time dimension fractal persistence robustness index FCSms determines the entry and window on the time side, and the fractal survival delivery window score FSWms determines the resource side ranking and loading intensity; Dimensional consistency and rationality analysis: FCSms, Deff, LG, Xslope, and SB are all dimensionless; ln(1+EA) converts the scale quantity into dimensionless; The numerator, denominator and power term are all dimensionless, and the fractal survival delivery window score FSWms is dimensionless, with consistent dimensions.

[0034] The prioritization module also includes a sorting unit; The sorting unit generates a priority queue for all videos that trigger the transaction priority strategy in descending order of the Fractal Survival Window Score (FSWms) to sort transactions. The specific sorting steps are as follows: The 85th percentile P85 and the 95th percentile P95 of the distribution of the Fractal Survival Delivery Window Score (FSWms) based on historical short videos are used as step thresholds. The real-time Fractal Survival Delivery Window Score (FSWms) is then compared with the step thresholds for a secondary evaluation, and a priority queue is generated based on the secondary comparative evaluation. When the fractal survival delivery window score FSWms ≥ 95th percentile P95, it is classified as secondary price increase and loaded first; When the 85th percentile P85≤Fractal Survival Window Score FSWms<95th percentile P95, it is classified as Level 1 surcharge and conventional loading; When the fractal survival delivery window score FSWms is less than the 85th percentile P85, it is divided into three levels of price increase and is divided into the basic layer loading; Priority loading means that the video will be loaded with advertising resources in the first batch at the beginning of the delivery window, and the loading ratio is high. For example, the allocated exposure quota is ≥ 30% of the total available exposure. It enjoys the highest resource guarantee priority during resource competition. Even if resources are tight, the exposure quota of the video will be guaranteed not to be reduced. In the bidding system, a higher starting bid coefficient can be given or high-value traffic entrances can be allocated first, such as homepage recommendations and hot stream positions, to ensure that the highest-rated videos receive a large amount of exposure early in the delivery window, accelerate the formation of a stable viewing curve and improve conversion efficiency. Regular loading means that ad resources are loaded at regular intervals after the delivery window begins. The loading ratio is moderate, for example, the impression quota accounts for 10%-20% of the total available impressions. It has a medium priority when resources are scarce. The impression quota may be moderately reduced when resources are tight, but it will not be completely eliminated. It is mainly placed in ordinary recommendation positions or regular traffic entrances, and does not occupy core prime exposure positions. This ensures that such videos receive a continuous but relatively moderate delivery rhythm. It is suitable for potential videos or content in the observation period. Basic layer loading means that exposure loading is obtained only when resources are sufficient within the delivery window. The loading ratio is low, for example, the exposure quota accounts for less than 5% of the total available exposure. It is the first to be cut or delayed when there is competition for resources. There may be a long idle period. It is mainly delivered in low-traffic or marginal positions. The exposure value and potential benefits of click-through conversion are relatively low, reducing the waste of resources on low-scoring videos and concentrating resources on high-return videos; but still retaining basic exposure to prevent missing out on individual long-tail videos with "later reversals".

[0035] In this embodiment, after triggering the transaction priority strategy, the fractal survival delivery window analysis unit of the system first calculates the effective fractal dimension Deff using the minute-level fractal dimension Dmin, the hour-level fractal dimension Dhour, and the day-level fractal dimension Dday to characterize the "roughness / long-tail potential" of the cross-scale curve; then, the multi-time dimension fractal persistence robustness index FCSms and the future exposure availability EA are incorporated into the numerator of the fractal survival delivery window score FSWms to reflect "sustainability × redemption", while the geometric mean porosity LG, scale consistency slope variance Xslope, and burst skewness SB are placed in the denominator to suppress "hollow complexity, cross-scale instability and sharpness". Peak risk" is to: when there is a scenario where "a short spike but the future exposure supply EA seems sufficient", due to the large burst skewness SB and geometric mean porosity LG, the fractal survival delivery window score FSWms is actively lowered to avoid mismatching high-value exposure to content that is about to decay; on the contrary, if the multi-time dimension fractal persistence robustness index FCSms of the two videos is similar, the future exposure supply EA of video A is slightly lower but the scale consistency slope variance Xslope is small and the geometric mean porosity LG is close to 1, then the fractal survival delivery window score FSWms of video A is higher and it obtains resources first. The physical meaning is that "stable and continuous time occupancy" is higher than "Short-term peaks" can be converted into effective reach within the window period; the sorting unit sets a ladder threshold based on the historical 85th percentile P85 and 95th percentile P95 of the fractal survival delivery window score FSWms, and divides the candidate videos into priority loading, ensuring that they obtain ≥30% exposure at the beginning of the delivery window, seize high-value entrances, regular loading of 10%-20% exposure, neutral rhythm and basic layer loading, less than or equal to 5% exposure, and loading only when resources are abundant. Its significance lies in mapping continuous scores into stable and executable operation and maintenance gears to resist frequent repetitions caused by intraday fluctuations; the reason for adopting the "numerator gain × denominator suppression" structure and the "quantile ladder" strategy is because the platform actually Common situations in the service industry, such as "holiday traffic disturbances, temporary adjustments to algorithm weights, and synchronous peaks caused by batch new product releases", will amplify the misjudgment of a single indicator. The multi-factor normalization and step-by-step stratification of the Fractal Survival Delivery Window Score FSWms can effectively avoid misinvestment and over-investment. The resulting beneficial effect is: without changing the fixed delivery window determined by the multi-time-dimensional fractal persistence robustness index FCSms, the Fractal Survival Delivery Window Score FSWms achieves accurate sorting of resource loading intensity and start-up order within the window, significantly reducing exposure waste and mismatch risks, and improving the accessibility and conversion rate of high-potential content in critical periods, thereby improving overall ROI and inventory turnover efficiency.

[0036] Example 6: Please refer to Figure 1 ,Specifically: the delivery window optimization module includes a delivery ,optimization unit; The delivery optimization unit determines a fixed delivery window length twindow based solely on the multi-time fractal persistence robustness index FCSms when the target short video passes the preliminary comparative evaluation. The specific determination format is: twindow = t0 × (1 + FCSms), where t0 represents the basic delivery duration set by the platform and is set according to the short video platform strategy, such as 24 hours or 48 hours. And register to form a fixed delivery window [ts,te], ts represents the delivery window start time, te represents the delivery window end time; The delivery optimization unit does not accept any increase or decrease in the fixed delivery window length twindow by the fractal survival delivery window score FSWms; when the fixed delivery windows [ts, te] of different videos overlap and platform resources are limited, the loading intensity and starting order of advertising resources are allocated only within each fixed delivery window [ts, te] according to the ranking results of the fractal survival delivery window score FSWms, keeping the start and end boundaries of each fixed delivery window [ts, te] unchanged.

[0037] In this embodiment, after the target short video passes the preliminary comparative evaluation, the delivery optimization unit of the system determines the delivery window length twindow only based on the multi-time dimension fractal persistence robustness index FCSms, and registers it to form a fixed delivery window [ts,te]. The reason for "only considering the multi-time dimension fractal persistence robustness index FCSms for timing and not considering the fractal survival delivery window score FSWms for time change" is to decouple "time decision" from "resource decision": on the time side, the multi-time dimension fractal persistence robustness index FCSms reflects the stability and continuity across minutes / hours / days, ensuring the physical meaning of life cycle matching (the higher the multi-time dimension fractal persistence robustness index FCSms, the more stable the curve and the more obvious the long tail, and a longer observable period should be given); on the resource side, the loading intensity and start order are allocated within the established window based on the fractal survival delivery window score FSWms, avoiding scheduling jitter and resource mismatch caused by frequent window changes due to short-term noise, such as temporary peaks during holidays that cause the fractal survival delivery window score FSWm to increase. However, forcibly extending the window will also lock in inefficient periods. When different video windows overlap and resources are limited, only the Fractal Survival Window Score (FSWms) is loaded within each [ts,te], keeping the boundaries unchanged, thus avoiding "crowding" and chain rescheduling. For example, if t0 = 24 hours and a video's multi-temporal fractal persistence robustness index (FCSms) = 1.2, then twindow = 52.8 hours. After the system locks [ts,te], it will not adjust the window due to subsequent small fluctuations in the Fractal Survival Window Score (FSWms), but will only increase its starting order and exposure ratio within the window. The resulting benefits are: window stability and predictable scheduling, reducing rescheduling costs and the risk of mis-delivery. Time matching is more closely aligned with the actual life cycle, reducing the structural waste of "short peak long window" or "long tail short window". When resources conflict, the Fractal Survival Window Score (FSWms) is used for precise allocation within a fixed window, improving the accessibility and conversion rate of high-potential content during critical periods, thereby improving overall ROI and inventory turnover efficiency.

[0038] While embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations may be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An intelligent system for dynamic short video traffic trading and multi-dimensional value assessment, characterized by: include: Traffic multi-dimensional collection module: This module collects video playback data from short video platforms in real time through collection tools, and transmits the video playback data to the cross-scale fractal analysis module; Cross-scale fractal analysis module: This module pre-processes video playback data to obtain a standardized data set, and then extracts features based on the standardized data set to obtain a fractal data set. Fractal robustness evaluation module: It calculates and outputs the multi-time-dimensional fractal persistence robustness index FCSms based on the fractal data set, and presets the multi-time-dimensional fractal persistence robustness index threshold Fth for preliminary comparative evaluation; Prioritization module: triggers transaction priority strategies through preliminary comparative evaluation, calculates and outputs the Fractal Survival Delivery Window Score (FSWms) based on the multi-time-dimensional Fractal Continuous Robustness Index (FCSms), and prioritizes video transactions based on the Fractal Survival Delivery Window Score (FSWms); Delivery window optimization module: Determine the delivery window length twindow based on the multi-time dimension fractal persistence robustness index FCSms, and allocate advertising resources and delivery time based on the ranking results of the fractal survival delivery window score FSWms.

2. The intelligent system for dynamic short video traffic trading and multi-dimensional value assessment according to claim 1, characterized in that: The traffic multi-dimensional acquisition module includes an acquisition unit; The collection unit collects video playback data from the short video platform at the minute, hour, and day time scales by setting a collection tool on the short video platform; The acquisition tool includes a data acquisition terminal and an acquisition control program; The data acquisition terminal is deployed at the back-end data interface node of the short video platform, located between the platform's data exchange layer and the business processing layer, to obtain the original playback data that has not been aggregated and compressed. The data acquisition terminal has data buffering, real-time data packaging and network communication functions; The acquisition control program runs in the data acquisition terminal, calls the platform open interface API, collects the log stream of the short video platform, and extracts the video playback data and additional parameters from the log stream; The video playback volume data includes a unique video identifier, a corresponding timestamp, a scale consistency slope variance Xslope, a heat burst delay time TB, a reference delay time TR, future exposure availability EA, a burst skewness SB, a minute-level playback volume sequence Pmin, an hour-level playback volume sequence Phour, and a day-level playback volume sequence Pday; The sampling step of the minute-level playback volume sequence Pmin is 1 minute, and the box counting scale r∈{1,2,4,8,16} minutes; The sampling step of the hour-level playback volume sequence Phour is 1 hour, and the box counting scale r∈{1,2,4,8} hours; The sampling step of the daily playback volume sequence Pday is 1 day, and the box counting scale r∈{1,2,4} days.

3. The intelligent system for dynamic short video traffic trading and multi-dimensional value assessment according to claim 2 is characterized by: The traffic multi-dimensional acquisition module also includes a transmission unit; The transmission unit packages the video playback volume data into data packets with unique video identifiers and corresponding timestamps according to the time scale, and sends them to the cross-scale fractal analysis module in an encrypted manner through the high-speed message queue transmission protocol MQTT. After receiving the data packets, the cross-scale fractal analysis module reconstructs the order of the data packets and generates a receipt confirmation receipt.

4. The intelligent system for dynamic short video traffic trading and multi-dimensional value assessment according to claim 3 is characterized by: The cross-scale fractal analysis module includes a data preprocessing unit for preprocessing video playback data to generate a standardized data set; The preprocessing includes time synchronization processing, dimension unification processing and data structuring and storage; The time synchronization process aligns the time stamps of the minute-level playback volume sequence Pmin, the hour-level playback volume sequence Phour, and the day-level playback volume sequence Pday in the video playback volume data according to the global time reference, and fills in the missing time data points through the interpolation algorithm, wherein the interpolation method is linear interpolation; The dimension unification process uses the Z-Score standardization method to perform numerical standardization on all parameters in the video playback data, so that the parameters of different physical dimensions are converted into dimensionless values ​​with a mean of 0 and a standard deviation of 1; The data structuring and storage is achieved by grouping the video playback volume data after dimension unification according to time scale and parameter type and storing them as a standardized data set, and adding a video unique identifier and a corresponding timestamp to each record.

5. The intelligent system for dynamic short video traffic trading and multi-dimensional value assessment according to claim 4 is characterized by: The cross-scale fractal analysis module further includes a fractal feature extraction unit for performing feature extraction on the standardized data set to obtain a fractal data set; The fractal data set includes minute-level fractal dimension Dmin, hour-level fractal dimension Dhour, day-level fractal dimension Dday, geometric mean porosity LG, minute-level porosity Amin, hour-level porosity Ahour, day-level porosity Aday, scale consistency slope variance Xslope, heat outbreak delay time TB, reference delay time TR, future exposure availability EA and outbreak skewness SB.

6. The intelligent system for dynamic short video traffic trading and multi-dimensional value assessment according to claim 5, characterized in that: The fractal robustness evaluation module includes a fractal robustness analysis unit; The fractal robustness analysis unit extracts the heat outbreak delay time TB and the reference delay time TR from the fractal data set to form a slow-heating factor, constructs the scale consistency slope variance Xslope into a stability factor, and constructs the minute-level fractal dimension Dmin, the hour-level fractal dimension Dhour, and the day-level fractal dimension Dday in combination with the geometric mean porosity LG into a complexity and spatial normalization factor. After multiplying and summarizing the slow-heating factor, the stability factor, and the complexity and spatial normalization factor, the multi-time-dimensional fractal persistence robustness index FCSms is calculated and output; It is used to measure the complexity, stability, gap structure characteristics and popularity growth pattern of a short video's traffic change curve at different time scales, and reflect its long-term sustainability potential through a single value.

7. The intelligent system for dynamic short video traffic trading and multi-dimensional value assessment according to claim 6 is characterized by: The fractal robustness evaluation module includes an evaluation unit; The evaluation unit calculates the multi-time dimension fractal persistence robustness index FCSms in batches on the historical sample set of the same category to obtain the empirical distribution, takes the 80th percentile as the initial multi-time dimension fractal persistence robustness index threshold Fth, and re-evaluates it on a monthly rolling basis; Then, a preliminary comparative evaluation is conducted between the real-time multi-time fractal persistence robustness index FCSms and the multi-time fractal persistence robustness index threshold Fth to determine the persistence robustness of the short video traffic structure at multiple time scales. The trigger is based on the preliminary comparative evaluation results. The specific evaluation contents are as follows: When the multi-time dimension fractal persistence robustness index FCSms ≥ the multi-time dimension fractal persistence robustness index threshold Fth, the current short video traffic structure is determined to be robust, and the transaction priority strategy is triggered at this time; When the multi-time dimension fractal persistence robustness index FCSms is less than the multi-time dimension fractal persistence robustness index threshold Fth, it is determined that the current short video traffic structure is not robust and is processed according to the conventional strategy.

8. The intelligent system for dynamic short video traffic trading and multi-dimensional value assessment according to claim 7 is characterized by: The prioritization module includes a fractal survival delivery window analysis unit; The fractal survival delivery window analysis unit obtains the effective fractal dimension Deff by performing mean calculation based on the minute-level fractal dimension Dmin, the hour-level fractal dimension Dhour and the day-level fractal dimension Dday after the transaction priority strategy is triggered; At the same time, the multi-time dimension fractal continuous robustness index FCSms output by the fractal robustness evaluation module is received and combined with the effective fractal dimension Deff, as well as the scale consistency slope variance Xslope, future exposure availability EA and outbreak skewness SB to calculate and output the fractal survival delivery window score FSWms.

9. The intelligent system for dynamic short video traffic trading and multi-dimensional value assessment according to claim 8, characterized in that: The prioritization module further includes a sorting unit; The sorting unit generates a priority queue for all videos that trigger the transaction priority strategy in descending order of the fractal survival delivery window score FSWms to sort the transactions. The specific sorting steps are as follows: The 85th percentile P85 and the 95th percentile P95 of the distribution of the Fractal Survival Delivery Window Score (FSWms) based on historical short videos are used as step thresholds. The real-time Fractal Survival Delivery Window Score (FSWms) is then compared with the step thresholds for a secondary evaluation, and a priority queue is generated based on the secondary comparative evaluation. When the fractal survival delivery window score FSWms ≥ 95th percentile P95, it is classified as secondary price increase and loaded first; When the 85th percentile P85≤Fractal Survival Window Score FSWms<95th percentile P95, it is classified as Level 1 surcharge and conventional loading; When the fractal survival delivery window score FSWms is less than the 85th percentile P85, it is divided into three levels of price increase and is divided into the basic layer loading.

10. The intelligent system for dynamic short video traffic trading and multi-dimensional value assessment according to claim 9, characterized in that: The delivery window optimization module includes a delivery optimization unit; The delivery optimization unit determines a fixed delivery window length twindow based only on the multi-time fractal persistence robustness index FCSms when the target short video passes the preliminary comparative evaluation. The specific determination form is: twindow = t0 × (1 + FCSms), where t0 represents the basic delivery duration set by the platform; And register to form a fixed delivery window [ts,te], ts represents the delivery window start time, te represents the delivery window end time; The delivery optimization unit does not accept any increase or decrease in the fixed delivery window length twindow by the fractal survival delivery window score FSWms; when the fixed delivery windows [ts, te] of different videos overlap and platform resources are limited, the loading intensity and starting order of advertising resources are allocated only within each fixed delivery window [ts, te] according to the ranking results of the fractal survival delivery window score FSWms, keeping the start and end boundaries of each fixed delivery window [ts, te] unchanged.

Images