Patent Application
20240129559
Referring to
The content provider 100 typically provides the video programming to the cable TV distributor 110 via a satellite link 110 and/or a high-speed broadband network system, such as, for example, the Internet. In other cases, the content provider provides the video programming to the cable TV distributor on a recording media (e.g., digital video disc, hard drive). The cable TV distributor 110 provides the customers 120 with video programming via the cable network 130. The cable network 130 may be composed of, for example, coaxial and/or fiber-optic cabling. Video programming is distributed to customers in either an analog or digital (MPEG) format via the cable network 130. The video programming is received by a consumer premise device associated with the customer 120 that is capable of receiving the analog or the digital signal and displaying video programming represented by the signal(s) on an associated display device. The customer premise equipment may be, for example, a television, a set top box, or otherwise. The typical cable TV distributor 110 includes media storage server for storing programming content received from the content provider 100 for a predetermined time, typically until it is distributed to customers via the cable network 130. Referring also to
Referring to
Many customer devices associated with a cable network are configured to generate and/or display a programming guide from which to select video content. Some programming guides are generally referred to an electronic programming guide in the form of a grid pattern of times and channels, such as what is traditionally provided with cable television.
In other embodiments, customers with an Internet connection may access video content that is provided by the cable television distributor 110 or other video content delivery services directly through an Internet based content delivery video service, such as for example, www.hulu.com, www.abc.com, www.netflix.com, and www.historychannel.com. In this case, typically the video content delivery service organizes the content in some manner, such as using a graphical interface. The video content is selectable by the customer to be viewed, normally on-demand.
The foregoing and other objectives, features, and advantages of the invention may be more readily understood upon consideration of the following detailed description of the invention, taken in conjunction with the accompanying drawings.
Referring to
The content providers, or other rights holder, grant rights to a particular storefront, such as a video on demand service, to provide the video content to particular customers for a limited period of time or availability window. The availability window may have a start date and/or start time (inherent from the time the content is provided if no start date or start time is provided) and the availability window may have an end date and/or end time (perpetual if no end date or end time is provided). The content providers typically provide the video content in a master file format (e.g., ingest file format provided by the content provider), which is then processed by the video content management system 450 into a format suitable for the storefront which may include a plurality of different video files, each of which have different characteristics. The content providers make their content available to the video content management system 450 with different lead times before the start date and time of the availability window. For example, some content may be provided to the video content management system 450 several days or several weeks prior to the respective start time of the availability window. In cases with sufficient lead time, there is normally available time for the video content management system 450 to process the video content and make it available to the storefront in a timely manner. For example, some content may be provided to the video content management system 450 a few hours or a few minutes prior to the respective start date and time of the availability window. In cases with a limited amount of lead time, there is unlikely available time for the video content management system 450 to process the video content and make it available to the storefront before the start time of the availability window. As a result, the storefront may miss out on revenue, and the customers may get frustrated with that storefront and seek an alternative storefront from which to obtain the desired video content. Also, the date and time at which the video content is provided to the video content management system 450 is generally unpredictable and the date and time at which different content providers provide video content to the video content management system 450 is likewise unpredictable. Further, the amount of content provided by any particular content provider at any particular date and time is likewise unpredictable, from a single video file and/or metadata file to thousands of video files and/or metadata files. To further complicate matters, the content providers often provide updated metadata which identifies the manner of distribution of the video content and the manner of presentation of the video content in the storefront. To also further complicate matters, the content providers often provide updated video content without updated metadata content which requires reprocessing of the video content. To even further complicate matters, the content providers often provide new video content with new metadata which identifies the manner of processing the video content, the distribution of the video content, and the manner of presentation of the video content in the storefront. Having video content being made available on a storefront on a timely basis, coinciding with the start date and time of the availability window, is important to the content providers and/or the storefront providers, to maximize the duration of the content being available which is especially desirable when the content is being paid for on an individual basis by customers.
Referring to
After the discovery process 504, the package is ingested by an ingest process 506. The ingest process separates each of the video files and/or metadata files from the respective package files for subsequent processing in an orderly manner. In this manner, each of the package files from the storage device 502 may be ingested in a manner consistent with the processing capabilities of the system to process the contents of the package files.
After the ingest process 506 for a package, it is desirable to perform a quality check 508 on the video file to ensure that it is suitable for subsequent processing, such as there are not errors in the video file. It is also desirable to perform the quality check 508 on the metadata file to ensure that its contents do not contain errors. In the event that an error is determined, the package may be flagged as containing an error together with a description of the error, so that the package may be repaired and provided again to the video content management system 450 for subsequent processing. In this manner, package files that include errors are not subsequently processed, thereby consuming computational resources when the resulting processing is going to ultimately result in errors.
After the quality check 508 of the contents of the package, the video file may optionally go through a transcoding process 510. The transcoding process 510 is a digital-to-digital conversion from one encoding to another encoding, such as the video file and/or audio file from an input format to one or more output formats. By way of example, the input format may be an AVI file format and the output may be a HEVC file format. By way of example, the input format may include high-definition progressive file format and the output may be standard-definition interlaced file format. By way of example, the input format may have a first bit rate and the output may have a different second bit rate. By way of example, the input format may have MP3 audio and the output format may have WAV audio. The transcoding process 510 may use a particular transcoder defined by the metadata or the content provider. In other cases, the transcoding process 510 may use any available transcoder, as desired.
After the transcoding process 510, a packaging process 512 may be performed which gathers together the one or more video files, together with the one or more metadata files, into one or more packages for the respective video content.
After the packing process 512, a delivery quality check process 514 is performed on the packages, the one or more video files, and the one or more metadata files, to ensure that it is suitable for subsequent use by the storefront, such that there are not errors. In the event that an error is determined, the package may be flagged as containing an error together with a description of the error, so that the package may be repaired and provided again to the video content management system 450 for subsequent processing. In this manner, package files that include errors are not subsequently processed, thereby consuming computational resources or providing packages unsuitable for use by the storefront.
After the delivery quality check 514, an optional encryption process 516 may be performed on each of the packages, video file(s), and/or metadata file(s).
After the encryption process 516, the video content management system 450 may archive the resulting package 520 of the ingested content, or otherwise deliver the resulting package 522 or an archived package 520 of the ingested content.
The video content management system 450 may store and retrieve the resulting files and processing on the storage device 502 at any point in the process, as desired.
The packages are delivered 522 to a target platform and/or storefront 530. The target platform and/or storefront, may include one or more ways in which the content is provided to the customer, such as on demand. For example, the target platform and/or storefront 530, may include Internet Protocol television 532, satellite 534, over the top service 536, video on demand 538, etc.
In some cases, such as for HTTP Live Streaming a series of different files are created, each for a different bitrate and/or image size, and the different files a chunked. Also, a series of different manifest files are likewise created.
Each of the services of the video content management system 450 may include a queue of one or more files to be processed, a start time and date at which each of the files have started processing, and an end time and date at which each of the files have completed processing. Each of the services of the video content management system 450 may process the files in a serial and/or parallel fashion, depending on processing capabilities of the particular service. For example, there may only be a single hardware-based transcoder for selected files and other processes that need to be processed in a serial fashion. For example, there may be multiple threads and/or processors available for the discovery process and other processes for parallel processing.
Unfortunately, with all the various processes used to process packages for particular target platforms, it is difficult to ensure that the packages are processed in a timely manner, especially in a timely manner relative to an availability window which may be from minutes to hours to days from the time the packages are received. As a result, it is desirable to estimate processing characteristics of each of the processes for a particular type of video file and/or metadata content within the processing pipeline to estimate how long each of the processes will take based upon the current processing characteristics for the particular process. This estimation may likewise take into account whether some processes are serial in nature for each file, or whether some processes are parallel in nature for multiple files. This estimation may further take into account the current processing capabilities of the system, such as the number of processors currently available for such processing for each of the processes. Accordingly, it is desirable to be able to estimate the time for each file to be processed by each process of the processing pipeline based upon characteristics of the current file, characteristics of other files to be processed or being processed, and characteristics of the processing environment.
A predictive analysis process 550, referred to herein as a Predictive Content Processing Estimator (PCPE), may be used that combines historical processing related data 552 for one or more of the processes of the video content management system 450, characteristics of the processing environment 554, processing characteristics of other files 556 currently in or scheduled to be in the processing pipeline, together with characteristics of a particular package 558 (including video file(s) and/or metadata file(s)) in light of the output package(s) to be generated 560 (including video file(s) and/or metadata file(s)) to provide an estimation of the date and/or time that a particular package will complete its processing 562. The predictive analysis process 550 may output an estimated date/time 570 for each package to complete processing for each process of the processing pipeline and for its completion of the processing pipeline.
The historical processing related data may be based upon training data and/or previously processed packages, as desired. The input from the processing pipeline may include event data, such as the start time/date that each file starts processing by a particular process and the end time/date that each file ends processing by a particular process, so that the predictive analysis process 550 may update its estimate that a particular file will complete its processing for the processing pipeline.
The prediction may use a machine learning process, if desired, such as a Random Forest Algorithm. The Random Forest Algorithm is based upon an ensemble technique, where it analyzes and learns data by constructing multiple decision trees of input data, by row sampling (e.g., sampling rows from the training data), and feature sampling (e.g., sampling on a subset of key parameters). Based on the decision trees that are used for the input, the Random Forest Algorithm may be used to estimate a prediction that fits a majority of the training data. Other machine learning techniques may likewise be used, as desired.
Referring to
As described earlier, a Video Content Management System (VCMS) receives Content Assets (consisting of metadata and associated content) from Content Providers, encodes it and publishes it to one or more Video-on-Demand (VOD) storefront(s). A VOD Service Provider's goal is to keep content refreshed continuously on the Storefront, for an engaging viewing experience.
As also described earlier. VOD storefronts are only able to offer content based on an availability window (e.g., Jan. 1, 2020 at midnight to Dec. 31, 2020 at midnight. Content providers make their content available to storefronts with different lead times relative to the availability window. For example, some content may be provided to the VOD Service Provider days or weeks prior to its availability window, while in other instances it may be provided with only hours of lead time. Content providers therefore make their content available to a VOD service provider on individualized and sometimes unpredictable schedules. Sometimes a content provider provides hundreds or thousands of new content titles, updated content, or updated metadata in a short period of time. Thus, on-time availability of content on the storefront is a performance indicator for VOD Service Providers.
A VOD service provider therefore strives to ensure that all its content makes it to the storefront in time for the content's availability window i.e., the start time of that availability window. However, with content coming into the system at unpredictable times and competing for finite resources, a static priority is not sufficient to ensure the most efficient use of processing and other resources, and by using a static priority, a VCMS may miss the start date needlessly by processing content that could have been deferred. Content that does not make it to the storefront by the start time of the availability window results in lost revenue for the service provider, as customers may become discouraged and rent/buy that content from another provider—or worse, be discouraged and leave that service Provider altogether.
Disclosed is a novel Dynamic Content Delivery Prioritization (DCDP) system that addresses this problem. DCDP uses the remaining time required to deliver content to the storefront to make optimized decisions on prioritization, accomplishing this by leveraging the PCPE, described earlier in this specification, which uses historical data and current system conditions to estimate a content's remaining delivery time to the storefront. As noted previously in this specification, the PCPE in some embodiments may use a machine learning algorithm along with current system conditions to predict the remaining time to delivery.
Specifically, the DCDP disclosed herein considers current system conditions and determines if content will arrive late to the storefront by comparing the estimated remaining delivery time against the availability window start time. If content will arrive late to the storefront, the DCDP will reprioritize the content being processed by the system ensuring content with spare lead time has its processing deferred while content that is in danger of not making it to the storefront by its premiere date is expedited through the system.
The benefits of this system can be seen in reference to
It should be understood that in each of these panels, other content besides the four titles being depicted is being processed based on an earlier priority. Thus, as can be seen by inspecting these panels, from time t0 to t2, resources are available to process only one of the four shown titles, from time t2 to t6 two such titles may be processed, and only after time t6 are there sufficient resources to process three of the four titles depicted.
In panel (a), the four titles are processed according to their arrival time at the VCMS. As can be seen in this figure, only Title 1 meets its availability window start time. In effect, because Title 1 was received so far in advance of its availability start time window, processing this title first used processing resources that were needed for the other three titles to meet their windows.
It might be assumed that this adverse result could be avoided by prioritizing the titles according to their availability widow start times i.e., by processing the title in order of the proximity to their respective availability windows. But panel (b) shows that this is not the case; even in this alternate prioritization scheme Title 2 misses its deadline while both Title 3 and Title 4 were completed well in advance of their respective deadlines.
The dynamic delivery prioritization method shown in panel (c), however, is able to process the titles in an optimal order by which all four titles meet their respective deadlines because it prioritizes titles based not only on the proximity to the availability window start time, but also on the amount of time needed to process the title given current system conditions. This, of course, assumes a means of estimating the time needed to process a title. Thus, the disclosed dynamic delivery prioritization method leverages the PCPE disclosed earlier, which is able to reliably estimate the processing time needed to deliver the content title from the VCMS.
In some preferred embodiments, as detailed below, other factors besides the estimated remaining processing time and the proximity to an availability window may be used to assign relative priorities to content. These other factors may be useful in several circumstances. For example, the DCDP may not be able to provide all titles to the storefront on time, hence other factors may be used to determine which titles will miss their deadlines. In another example, some titles may be of sufficient importance that the risk of its missing its deadline needs to be minimized or reduced. In still other embodiments, these other priorities may be used to temporarily spin up resources, if that is possible, to meet availability windows.
In some embodiments, safety margins may be included in the system 600 by, for example, maintaining spare capacity or processing resources that may be activated when needed. Alternately, or additionally, a buffer amount may be included in front of the availability window start time.
By identifying content in danger of not making it to the storefront in time for its availability window and automatically expediting that content, the service provider can better realize potential revenue and improve customer satisfaction. The automatic nature means operational resources are not expended and the chance that a human operator misses a critical alert is eliminated. Moreover, in preferred embodiments, the prioritization methods disclosed herein are fully automated so as to expedite content without requiring manual intervention by an operations team.
Each functional block or various features in each of the aforementioned embodiments may be implemented or executed by a circuitry, which is typically an integrated circuit or a plurality of integrated circuits. The circuitry designed to execute the functions described in the present specification may comprise a general-purpose processor, a digital signal processor (DSP), an application specific or general application integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic devices, discrete gates or transistor logic, or a discrete hardware component, or a combination thereof. The general-purpose processor may be a microprocessor, or alternatively, the processor may be a conventional processor, a controller, a microcontroller or a state machine. The general-purpose processor or each circuit described above may be configured by a digital circuit or may be configured by an analogue circuit. Further, when a technology of making into an integrated circuit superseding integrated circuits at the present time appears due to advancement of a semiconductor technology, the integrated circuit by this technology is also able to be used.
It will be appreciated that the invention is not restricted to the particular embodiment that has been described, and that variations may be made therein without departing from the scope of the invention as defined in the appended claims, as interpreted in accordance with principles of prevailing law, including the doctrine of equivalents or any other principle that enlarges the enforceable scope of a claim beyond its literal scope. Unless the context indicates otherwise, a reference in a claim to the number of instances of an element, be it a reference to one instance or more than one instance, requires at least the stated number of instances of the element but is not intended to exclude from the scope of the claim a structure or method having more instances of that element than stated. The word “comprise” or a derivative thereof, when used in a claim, is used in a nonexclusive sense that is not intended to exclude the presence of other elements or steps in a claimed structure or method.
The terms and expressions which have been employed in the foregoing specification are used therein as terms of description and not of limitation, and there is no intention, in the use of such terms and expressions, of excluding equivalents of the features shown and described or portions thereof, it being recognized that the scope of the invention is defined and limited only by the claims which follow.
The present application claims priority to U.S. Provisional App. No. 63/408,792 filed Sep. 21, 2022, the contents of which are incorporated herein by reference in their entirety.
| Number | Date | Country | |
|---|---|---|---|
| 63408792 | Sep 2022 | US |
