FORM 2
THE PATENTS ACT, 1970
(39 OF 1970)
AND
THE PATENT RULES, 2003
COMPLETE SPECIFICATION
(See section 10 and rule 13)
“SYSTEM AND METHOD FOR GENERATING A TARGET MEDIA FEED AND
MEDIA RECOMMENDATIONS THEREOF”
We, Novi Digital Entertainment Private Limited, an Indian National, of Star House, Urmi
Estate, 95, Ganapatrao Kadam Marg, Lower Parel West, Mumbai 400013, Maharashtra, India.
The following specification particularly describes the invention and the manner in which it is to be
performed.
2
SYSTEM AND METHOD 5 FOR GENERATING A TARGET MEDIA FEED AND MEDIA
RECOMMENDATIONS THEREOF
TECHNICAL FIELD:
10 The present disclosure generally relates to the field of delivery of digital audio‐video content.
More particularly, the present disclosure relates to a method and a system for generating a
target short media feed based on a media feed provided on a set of user devices and
generating media recommendations thereof.
15 BACKGROUND OF THE DISCLOSURE:
The following description of the related art is intended to provide background information
pertaining to the field of the disclosure. This section may include certain aspects of the art
that may be related to various features of the present disclosure. However, it should be
appreciated that this section is used only to enhance the understanding of the reader with
20 respect to the present disclosure, and not as admission of the prior art.
Media streaming is increasingly popular for delivering television, movies, and other content
to viewers. It involves point‐to‐point or multicast media streaming of digitized content over
the Internet or a similar network. This method is commonly used for video on demand (VOD)
25 services, digital video recorder (DVR) services, Internet Protocol Television (IPTV), and other
convenient services. Typically, the media stream is played back in real‐time as it is delivered
to the player.
In some cases, the video can be sideloaded or cached at the player to enable faster delivery
30 than in real‐time. However, there are instances when viewers might only be interested in
specific programs or genres, like sports games, without watching them in their entirety. Due
to the abundance of content, many viewers prefer to watch highlights or the most interesting
parts of various programs. Identifying such highlights can be challenging, especially when
multiple broadcasts are happening simultaneously, requiring continuous switching between
35 different broadcasts. Further, the advent of social networking has transformed how people
interact, and social media often includes short media clips that users conveniently view using
3
portable devices. These clips may contain highlights of news, 5 sports, TV shows, or other
media, offering quick references to interesting or amusing content. Currently, users may have
to create a compilation of the one or more media frames such as reels or short clips,
comprising most watched media frames themselves in accordance with the interest of the
user, which can be time‐consuming and may lead to them missing out on desired segments.
10 Alternatively, users may search for highlights provided by media content producers, but
without knowing what users want, these highlights may not be optimal. Media content
producers may find it difficult to identify the most desirable portions of content for their
audience.
15 In summary, the increasing number of media streaming platforms available today has
revolutionized the way users consume entertainment content. With the rise of various
streaming services, viewers now have access to a vast abundance of movies, TV shows,
documentaries, and other media content at their fingertips. However, this wealth of options
has also brought about a new challenge for users: the difficulty of locating specific content of
20 interest. With so much content available on these platforms, it has become like finding a
needle in a haystack to discover the exact shows or movies one desires. Browsing through
endless libraries and categories can be overwhelming and time‐consuming, making it
frustrating for users to find what they want to watch. Additionally, the existing
content recommendation systems and methods are based on metadata similarity comparison
25 which does not accurately identify the content of users interest explicitly the interest of the
user. As a result, many users feel the need for better content curation and recommendation
systems to help them navigate through the sea of options and locate their preferred content
with ease. Therefore, the importance of finding new ways for media providers to enable users
to locate content of their interest cannot be overstated. In today's highly competitive media
30 streaming landscape, user experience plays a pivotal role in retaining customers and
increasing viewership. With an overwhelming amount of content available, viewers expect
seamless and personalized content discovery. By implementing advanced content
recommendation systems and intuitive search features, media providers can enhance user
satisfaction and keep users engaged on their platforms. Tailoring content suggestions based
35 on a user's viewing history, preferences, and behavior can create a more enjoyable and
relevant viewing experience, increasing the likelihood of users staying loyal to the platform.
4
5
Therefore, in light of these limitations, there is a need for a solution that allows for generating
recommendations with context for a user based on his past preference to engage and enable
the user to identify media content of interest. Hence, a system and a method for generating
a short media feed of a user interest based on a media feed and generating media
10 recommendations based on the user interest or the generated short media feed.is required.
OBJECTS OF THE DISCLOSURE
Some of the objects of the present disclosure, which at least one embodiment disclosed
herein satisfies are listed herein below.
15
It is an object of the present disclosure to provide a system and a method for generating a
short media feed based on a media feed provided to a set of user devices and for identifying
the most watched scenes from the media feed to generate one or more media
recommendations based on the user interest.
20
It is another object of the present disclosure to provide a solution that may enhance the
utilization of media frame data in a short media feed generation process by extracting
relevant information from the media frame data to create a more contextually relevant short
media feed.
25
It is another object of the present disclosure to provide a solution for identifying one or more
top frames from a set of media frames of a media feed based on a comparison of start time,
end time, and total runtime associated with the media feed to efficiently select the most
appropriate frames. Furthermore, the object is to provide the solution that intend to integrate
30 user media traffic data related to the media feed, which reflects media playout actions
initiated at user devices, into the top frame identification process, thus enhancing the
accuracy of selecting top frames that align with user preferences.
Yet another object of the present disclosure is to provide a solution to incorporate a media
35 frame data associated with one or more top frames of a media feed into a short media feed
data generation process, such as additional metadata or context information; thus, the object
5
of the present disclosure is to provide a solution that aims to 5 create more informative and
engaging short media feed data and generating one or more media recommendations for the
user based for instance on watch history related to the short media feed data.
SUMMARY OF THE DISCLOSURE
10 This section is provided to introduce certain aspects of the present disclosure in a simplified
form that are further described below in the detailed description. This summary is not
intended to identify the key features or the scope of the claimed subject matter.
In order to achieve the aforementioned objectives, one aspect of the disclosure relates to a
method for generating a target short media feed. The method comprises identifying, by a
15 processing unit via an explicit signal (user feedback) collecting module, a media playout action
on a set of user devices, wherein the media playout action is associated with a media feed
and the media feed comprises at least a set of media frames and a set of media frame data.
Further, the method comprises receiving, by the processing unit via a normalization module,
a set of metadata from the set of user devices based on the media playout action, wherein
20 the set of metadata comprises at least a start time associated with the media playout action,
an end time associated with the media playout action, and a total runtime associated with
the media feed. The method further comprises identifying, by the processing unit via a top
frame identification module, one or more top frames from the set of media frames based on
the set of metadata. The method further encompasses generating, by the processing unit via
25 a ML/AI‐based short‐form feed curation module, a short media feed associated with the
media feed based on the one or more top frames. Further, the method comprises generating,
by the processing unit via a title and description curation module, a short media feed data
associated with the short media feed based on the set of media frame data, wherein the short
media feed data comprises at least one of a title associated with the short media feed and a
30 description associated with the short media feed. Thereafter, the method encompasses
generating, by the processing unit via a recommendation module, the target short media feed
based on the short media feed and the short media feed data.
Another aspect of the present disclosure relates to a system generating a target short media
feed. The system comprises a processing unit, configured to identify, via an explicit signal
6
(user feedback) collecting module, a media playout action on 5 a set of user devices, wherein
the media playout action is associated with a media feed and the media feed comprises at
least a set of media frames and a set of media frame data. The processing unit is further
configured to receive, via a normalization module, a set of metadata from the set of user
devices based on the media playout action, wherein the set of metadata comprises at least a
10 start time associated with the media playout action, an end time associated with the media
playout action, and a total runtime associated with the media feed. The processing unit is
further configured to identify, via a top frame identification module, one or more top frames
from the set of media frames based on the set of metadata. Further, the processing unit is
configured to generate, via a ML/AI‐based short‐form feed curation module, a short media
15 feed associated with the media feed based on the one or more top frames. The processing
unit is further configured to generate, via a title and description curation module, a short
media feed data associated with the short media feed based on the set of media frame data,
wherein the short media feed data comprises at least one of a title associated with the short
media feed and a description associated with the short media feed. Thereafter, the processing
20 unit is configured to generate, via a recommendation module, the target short media feed
based on the short media feed and the short media feed data.
BRIEF DESCRIPTION OF DRAWINGS
The accompanying drawings, which are incorporated herein, and constitute a part of this
25 disclosure, illustrate exemplary embodiments of the disclosed methods and systems in which
like reference numerals refer to the same parts throughout the different drawings.
Components in the drawings are not necessarily to scale, emphasis instead being placed upon
clearly illustrating the principles of the present disclosure. Some drawings may indicate the
components using block diagrams and may not represent the internal circuitry of each
30 component. It will be appreciated by those skilled in the art that disclosure of such drawings
includes disclosure of electrical components, electronic components or circuitry commonly
used to implement such components.
FIG.1 illustrates an exemplary block diagram depicting an exemplary network architecture
diagram [100], in accordance with exemplary embodiments of the present disclosure.
7
FIG.2 illustrates an exemplary block diagram of a system 5 [200], for generating a target short
media feed, in accordance with exemplary embodiments of the present disclosure.
FIG.3 illustrates an exemplary method flow diagram [300], for generating a target short media
feed, in accordance with exemplary embodiments of the present disclosure.
The foregoing shall be more apparent from the following more detailed description of the
10 disclosure.
DETAILED DESCRIPTION
In the following description, for the purposes of explanation, various specific details are set
forth in order to provide a thorough understanding of embodiments of the present disclosure.
It will be apparent, however, that embodiments of the present disclosure may be practiced
15 without these specific details. Several features described hereafter can each be used
independently of one another or with any combination of other features. An individual
feature may not address any of the problems discussed above or might address only some of
the problems discussed above.
20 The ensuing description provides exemplary embodiments only, and is not intended to limit
the scope, applicability, or configuration of the disclosure. Rather, the ensuing description of
the exemplary embodiments will provide those skilled in the art with an enabling description
for implementing an exemplary embodiment. It should be understood that various changes
may be made in the function and arrangement of elements without departing from the spirit
25 and scope of the disclosure as set forth.
Specific details are given in the following description to provide a thorough understanding of
the embodiments. However, it will be understood by one of ordinary skill in the art that the
embodiments may be practiced without these specific details. For example, circuits, systems,
30 processes, and other components may be shown as components in block diagram form in
order not to obscure the embodiments in unnecessary detail.
8
Also, it is noted that individual embodiments may be described as a process 5 which is depicted
as a flowchart, a flow diagram, a data flow diagram, a structure diagram, or a block diagram.
Although a flowchart may describe the operations as a sequential process, many of the
operations can be performed in parallel or concurrently. In addition, the order of the
operations may be re‐arranged. A process is terminated when its operations are completed
10 but could have additional steps not included in a figure.
The word “exemplary” and/or “demonstrative” is used herein to mean serving as an example,
instance, or illustration. For the avoidance of doubt, the subject matter disclosed herein is not
limited by such examples. In addition, any aspect or design described herein as “exemplary”
15 and/or “demonstrative” is not necessarily to be construed as preferred or advantageous over
other aspects or designs, nor is it meant to preclude equivalent exemplary structures and
techniques known to those of ordinary skill in the art. Furthermore, to the extent that the
terms “includes,” “has,” “contains,” and other similar words are used in either the detailed
description or the claims, such terms are intended to be inclusive—in a manner similar to the
20 term “comprising” as an open transition word—without precluding any additional or other
elements.
As used herein, a “processing unit” or “processor” or “operating processor” includes one or
more processors, wherein processor refers to any logic circuitry for processing instructions. A
25 processor may be a general‐purpose processor, a special purpose processor, a conventional
processor, a digital signal processor, a plurality of microprocessors, one or more
microprocessors in association with a DSP core, a controller, a microcontroller, Application
Specific Integrated Circuits, Field Programmable Gate Array circuits, any other type of
integrated circuits, etc. The processor may perform signal coding data processing,
30 input/output processing, and/or any other functionality that enables the working of the
system according to the present disclosure. More specifically, the processor or processing
unit is a hardware processor.
As used herein, “a client device”, “a user equipment”, “a user device”, “a smart‐user‐device”,
35 “a smart‐device”, “an electronic device”, “a mobile device”, “a handheld device”, “a wireless
communication device”, “a mobile communication device”, “a communication device” may
9
be any electrical, electronic and/or computing device 5 or equipment, capable of implementing
the features of the present disclosure. The user equipment/device may include, but is not
limited to, a mobile phone, smart phone, laptop, a general‐purpose computer, desktop,
personal digital assistant, tablet computer, television, smart TVs, streaming sticks, gaming
consoles, wearable device or any other computing device which is capable of implementing
10 the features of the present disclosure. Also, the user device may contain at least one input
means configured to receive an input from at least one of a transceiver unit, a processing unit,
a storage unit, an input unit and any other such unit(s) which are required to implement the
features of the present disclosure.
15 As used herein, “storage unit” or “memory unit” refers to a machine or computer‐readable
medium including any mechanism for storing information in a form readable by a computer
or similar machine. For example, a computer‐readable medium includes read‐only memory
(“ROM”), random access memory (“RAM”), magnetic disk storage media, optical storage
media, flash memory devices or other types of machine‐accessible storage media. The storage
20 unit stores at least the data that may be required by one or more units of the system to
perform their respective functions. The storage unit is configured to store a data required to
implement the features of the present solution as disclosed below in the present disclosure.
As used herein the term "media feed" refers to a digital collection of media content, such as
25 images, videos, audio, and/or a combination thereof, presented in a sequential or nonsequential
format. A media feed may be curated based on specific themes, events, topics, or
user preferences. It can be accessed and consumed by users through various electronic
devices, applications, or platforms. Media feed may include but is not limited to social media
feeds, news feeds, entertainment feeds, or any other form of digital content feeds accessible
30 over computer networks or communication systems. The media feed may be sourced from
various content creators, users, or content providers and may be updated regularly or
periodically to reflect current or relevant content. The term "media feed" as used in the
present disclosure encompasses any digital content stream or channel designed for user
consumption, regardless of the specific format or delivery mechanism.
35
10
As used herein, “similar” and “same” may 5 be used interchangeably in this patent specification
and may convey the same meaning. The use of these terms may not be interpreted as
implying any difference in meaning or scope.
The present disclosure relates to a system and a method for generating a target short media
10 feed. The present disclosure provides a solution that effectively overcomes the shortcomings
of the existing solutions. The existing solutions suffer from several limitations, such as these
solutions lack efficient and accurate frame identification techniques. Further, the
conventional approaches often rely on manual curation or rudimentary techniques, leading
to time‐consuming and error‐prone processes that may result in the inclusion of irrelevant or
15 outdated content in the short media feed. Moreover, the absence of user media traffic data
integration in the existing solutions restrict the ability to tailor the content to user
preferences, resulting in generic and less engaging media feeds. Furthermore, the existing
solutions do not provide sufficient flexibility in the selection process of top frames. In these
solutions, the lack of customizable threshold levels for user media traffic data hinders the
20 adaptability of short media feed generation to diverse user needs and preferences. This
limitation results in the delivery of suboptimal media feeds that may not align with users'
specific interests or real‐time content consumption patterns. Additionally, the existing
methods often overlook the potential of leveraging target media frame data to enhance the
short media feed data. Neglecting to incorporate additional metadata and context
25 information related to selected frames may lead to incomplete or less informative short
media feed descriptions, limiting the overall user experience. Overall, the shortcomings in the
existing solutions underscore the necessity for the presented novel solution, which addresses
these limitations by introducing advanced frame identification techniques, incorporating user
media traffic data with customizable threshold levels, and maximizing the utilization of target
30 media frame data. The proposed solution significantly improves the efficiency,
personalization, and relevance of short media feed generation, overcoming the deficiencies
found in the conventional methods and offering a superior user experience.
Hereinafter, exemplary embodiments of the present disclosure will be described in detail with
reference to the accompanying drawings so that those skilled in the art can easily carry out
35 the solution provided by the present disclosure.
11
Referring to Figure 1, 5 Figure 1 illustrates an exemplary block diagram depicting an exemplary
network architecture diagram [100], in accordance with exemplary embodiments of the
present disclosure. As shown in Figure 1, the exemplary network architecture diagram [100]
comprises a set of user devices (UD) [102(1)], [102(2)], ….[102(n)] (hereinafter collectively
referred to as user device [102] or the set of user device [102] for clarity purpose) in
10 communication with at least one server device [106], and a content delivery network [108]
wherein in an implementation the server device [106] further comprises a system [200]
configured to implement the feature of the present disclosure. In an implementation system
[200] may be in connection with the server device [106] and the user device [102], in a manner
as obvious to a person skilled in the art to implement the features of the present disclosure.
15 Also, in Figure 1 only a single the server device [106] is shown, however, there may be multiple
such server devices [106] or there may be any such numbers of said server device [106] as
obvious to a person skilled in the art or as required to implement the features of the present
disclosure.
In general, the content delivery network (CDN) [108] is a geographically distributed network
20 of servers strategically placed in various locations around the world. CDNs facilitate efficient
delivery of content, such as personalized content layout/ or catalogue (whether dynamic or
static), web pages, images, videos, and other digital assets, to end‐users, reducing latency and
improving load times.
25 The CDN [108] is further connected to various other CDNs (not shown in figures) present at
different geographical locations, creating a hierarchy of interconnected networks. Each CDN
in the hierarchy can have its own set of servers (e.g., the server device [106]) located in
different regions, forming a distributed network infrastructure. It is to be noted that the CDN
may be public CDN, private CDN, Telco CDN or the ISP CDN.
30 When the user device [102] communicates with the content delivery server system after
initial authentication or validation of subscription of the user via the user device, the
12
communication request is routed 5 through the CDN network. The CDN [108] then uses
intelligent routing algorithms to direct the communication request to the nearest server or
the server with the lowest load (e.g., the server device [106]), minimizing the distance and
optimizing response times. After the direction of the communication request, the content
catalogue layout is then routed through the CDN [108].
10 Further, in the implementation where the system [200] is present in the server device [106],
based on the implementation of the features of the present disclosure, a target short media
feed is generated at the server device [106] by the system [200], wherein the target short
media feed is based on the short media feed and the short media feed data. A set of metadata
from the set of user devices [102] is received via the normalization module [202(b)] at the
15 server device [106] based on the media playout action, wherein the set of metadata
comprises at least a start time associated with the media playout action, an end time
associated with the media playout action, and a total runtime associated with the media feed.
Further, one or more top frames from a set of media frames are identified at the server device
[106] based on the set of metadata. Then a short media feed associated with the media feed
20 is generated at the server device [106] based on the one or more top frames and a short
media feed data associated with the short media feed is also generated at the server device
[106] based on a set of media frame data, wherein the short media feed data comprises at
least one of a title associated with the short media feed and a description associated with the
short media feed.
25 Further, the processing unit [202] as referred herein may include various modules aimed at
performing specific functions as disclosed by the present disclosure, such as a Top frame
identification module [202(a)], a Normalization Module [202(b)], a ML/AI‐based Short‐Form
Feed Curation Module [202(c)], a Title and Description Curation Module [202(d)], an Explicit
Signal (User Feedback) Collecting Module [202(e)], a Content Identification module [202(f)],
30 a Recommendation Module [202(g)] and a Media playout action collecting module [202(h)].
It's important to note that the modules outlined above are not exhaustive, and those skilled
in the relevant field would understand that these modules can operate in diverse
combinations to effectively implement the method as disclosed herein. The flexibility in
13
combining these modules is 5 recognized as inherent to the implementation of the present
disclosure.
Further, referring to Figure 2, an exemplary block diagram of a system [200], for generating a
target short media feed at the server device [106] is shown. The system [200] comprises at
10 least one processing unit [202] and at least one storage unit [204]. Also, all of the
components/ units of the system [200] are assumed to be interacting with each other in a
manner that may be obvious to a person skilled in the art in light of the present disclosure, to
implement the solution as disclosed herein unless otherwise indicated below. Also, in Fig. 2
only a few units are shown, however, the system [200] may comprise multiple such units or
15 the system [200] may comprise any such numbers of said units, as required to implement the
features of the present disclosure. Further, in an implementation, the system [200] may be
present in the server device [102] to implement the features of the present disclosure. The
system [200] may be a part of the server device [106]/ or may be independent but in
communication with the server device [106].
20
The system [200] is configured for generating the target short media feed at the server device
[106], with the help of the interconnection between the components/units of the system
[200], wherein the generated target short media feed is then provided at the set of user
devices [102] by the server device [106].
25
Further, in an implementation of the present solution, the system [200] is configured for
generating one or more media recommendations via a recommendation module [202(g)]
based on the one or more top frames and the short media feed data associated with the short
media feed generated at the server device [106] based on the set of media frame data,
30 wherein the one or more media recommendations is at least one of an existing media feed
based on at least the one or more top frames, the short media feed data, and a long media
feed generated based on the one or more top frames and the short media feed data. Further,
the long media feed may comprise the existing media feed based on at least one of the one
or more top frames and the short media feed data.
35
14
Further, the long media feed may be retrieved from the storage unit [5 204] based on at least
one of the one or more top frames and the short media feed data. In an exemplary
implementation of the present solution as disclosed herein, the long media feed may
comprise a predefined number the one or more top frames, wherein the one or more top
frames are detected by the processing unit [202] based on a set of user preference
10 parameters such as the one or more top frames similarity with their past media preferences
in the long form media feed, a user action such as media like action associated with long
media feed of similar genre etc. Further, in another implementation of the present solution,
as disclosed herein, the long media feed may comprise a plurality of frames including but not
limited to the predefined number of the one or more top frames, wherein the one or more
15 top frames are detected by the processing unit [202] based on a user action associated with
the short form feed such as a media like action, a media love action, a media dislike action, a
media reject action, a comment action etc.
Further, in another implementation of the present disclosure, to recommend the target
20 media feed to the set of user devices [102], the solution firstly comprises recommending, by
the processing unit [202] via the recommendation module [202(g)] to the set of user devices
[102], the target short media feed. Then, the solution leads to receiving, by the processing
unit [202] via an explicit signal (user feedback) collecting module [202(e)] from the set of user
devices [102], a user feedback on the target short media feed. Thereafter, said solution
25 encompasses identifying, by the processing unit [202], a set of target short media feeds based
on the user feedback. In an implementation, in an event of receipt of the user feedback on
the target short media feed, the processing unit [202] via the ML/AI‐based short‐form feed
curation module [202(c)] identifies one or more keywords and/or title related to the target
short media feed. Thereafter, the processing unit [202] extracts one or more target keywords
30 and one or more target titles from a database of keywords and title, wherein such extraction
is based on the identified one or more keywords and/or the identified title related to the
target short media feed. Further, the processing unit [202] identifies the set of target short
media feeds based on the extracted one or more target keywords and the extracted one or
more target titles.
35
15
Further, after identification of the set of target short media feeds, the solution 5 also comprises
identifying, by the processing unit [202], a set of user preference parameters based on at least
one of one or more user actions on the set of target short media feeds, the media playout
actions and the one or more top frames. The set of user preference parameters indicates an
area of interest of the user in the media feed comprising the set of media frames and the set
10 of media frame data based on the media playout action for e.g., the area of interest of the
user is a character 'A’ present in the set of media frames or a genre associated with the media
feed A. Then the solution of the present disclosure comprises detecting, by the processing
unit [202] via a content identification module [202(f)], at least one of the one or more existing
media feeds and the one or more long media feeds based on the set of user preference
15 parameters for e.g., a media feed B is detected where the character ‘A’ is featured in the
media feed or a portion of the media feed and wherein the media feed is of a comedy genre.
In an exemplary implementation in order to detect at least one of the one or more existing
media feeds and the one or more long media feeds, the processing unit [202] is firstly
configured to identify via the content identification module [202(f)], a title and/or one or
20 more keywords related to one or more short media feeds from the set of target short media
feeds, wherein said identification is based on the set of user preference parameters. For
example, if one or more user preference parameters indicate a user’s interest in action
movies, the processing unit [202] via the content identification module [202(f)], identifies a
title and/or one or more keywords related to one or more short media feeds from the set of
25 target short media feeds that are related to action movies. It is pertinent to note that such an
example is non‐limiting and may include other exemplary implementations that may be
obvious to a person skilled in the art in light of the present disclosure.
Thereafter, the processing unit [202] via the content identification module [202(f)], is
30 configured to detect from a database at least one of the one or more existing media feeds
and the one or more long media feeds based on the identified title and/or the identified one
or more keywords. Particularly, in an exemplary implementation the processing unit [202]
respectively matches the identified title and/or the identified one or more keywords with one
or more titles and/or one or more keywords that are stored in the database and are related
35 to at least one of one or more pre‐identified existing media feeds and one or more preidentified
long media feeds. The processing unit [202] is then configured to detect from at
16
least one of the one or more pre‐identified 5 existing media feeds and the one or more preidentified
long media feeds, at least one of the one or more existing media feeds and the one
or more long media feeds, based on a successful match.
The processing unit [202] after detection of at least one of the one or more existing media
10 feeds and the one or more long media feeds, identifies one or more target short media feeds
from the set of target short media feeds in at least one of the one or more existing media
feeds and the one or more long media feeds. Further the processing unit [202] ranks at least
one of the one or more existing media feeds and the one or more long media feeds based on
the identified one or more target short media feeds in at least one of the one or more existing
15 media feeds and the one or more long media feeds.
Then the solution of the present disclosure comprises recommending, by the processing unit
[202] via the recommendation module [202(g)] to the set of user devices [102], a target media
feed comprising at least one of the one or more existing media feeds and the one or more
20 long media feeds based on the ranking of at least one of the one or more existing media feeds
and the one or more long media feeds. For e.g., a media feed C of comedy genre is
recommended to the user based on the area of interest of the user i.e., character ‘A’ or a
media feed comprising at least a portion of media feed B and/or media feed ‘A’ is
recommended to the user based on identifying character ‘A’ as the area of interest. Further,
25 in an implementation of the present solution as disclosed herein, each of the existing media
feed and the long media feed includes one or more short media feeds from the set of target
short media feeds.
More specifically, in order to implement the features of the present disclosure, the processing
30 unit [202] is configured to identify, via a media playout action collecting module [202(h)], a
media playout action on a set of user devices, wherein the media playout action is associated
with a media feed and the media feed comprises at least a set of media frames and a set of
media frame data, wherein one or more media frame data from the set of media frame data
comprises at least a metadata associated with one or more media frames from the set of
35 media frames. As used herein “metadata associated with one or more media frames” refers
to a collection of structured data elements intricately linked with the one or more media
17
frames of the media feed. The metadata 5 associated with one or more media frames
encompasses essential information that enhances the identification and processing of the
one or more media frames during the generation of one or more short media feeds. Further,
the metadata associated with one or more media frames may include a media frame identifier
data, which uniquely identifies each media frame within the media feed to facilitate efficient
10 referencing and tracking. Additionally, the metadata associated with one or more media
frames may include one or more data related to a subtitle of the one or more media frames,
encompassing textual information pertaining to the subtitles or a caption associated with
each frame from the one or more media frames. These subtitle data may be associated with
one or more spoken languages such as Hindi, English, Gujrati etc., and can be represented in
15 machine‐readable formats. Further, in an implementation of the present solution, the
metadata associated with one or more media frames may include an audio output associated
with the with each frame from the one or more media frames, wherein the audio output may
be converted in a predefined format such as vector format. Furthermore, in an
implementation of the present solution the metadata associated with one or more media
20 frames may include the audio output associated with the with each frame from the one or
more media frames mapped with the caption associated with said each frame from the one
or more media frames. Furthermore, the metadata may comprise time stamp data for each
media frame from the one or more media frames, such as data related to a start time
associated with each media frame and an end time associated with each media frame.
25 Furthermore, the scope of these terms may extend to include any obvious variations or
developments that would be apparent to a person skilled in the art in light of technological
advancements or industry developments. Therefore, the interpretation of these terms should
be broad and flexible to accommodate future discoveries or innovations in the field.
30 Further, the term "playout action on a set of user devices" as used in the present disclosure
refers to actions initiated by users on their respective user devices for starting or stopping the
playback of a media feed. These actions may include a "start playing media feed" action,
whereby users initiate the playback of a media feed on their user devices, and a "stop playing
media feed" action, whereby users halt the ongoing playback of a media feed on their user
35 devices. The term is intended to cover actions related to the initiation and termination of
media feed playback on the set of user devices [102]. It is important to note that the definition
18
provided above is for illustrative purposes only and shall not be limited to 5 the specific actions
mentioned. The term "playout action on a set of user devices" may encompass any variations
or related actions that would be obvious to a person skilled in the art. Such variations may
include additional media control actions, such as pause, rewind, or fast‐forward, initiated by
users during media feed playback on their user devices. Furthermore, the term may also
10 extend to cover actions initiated by user interfaces, applications, or automated systems that
manage media feed playback on user devices.
Further, the processing unit [202] of the system [200] is configured to receive, via a
normalization module [202(b)], a set of metadata from the set of user devices based on the
15 media playout action, wherein the set of metadata comprises at least a start time associated
with the media playout action, an end time associated with the media playout action, and a
total runtime associated with the media feed. For ease of understanding, in an example,
wherein the total runtime associated with the media feed is 5 mins and the media playout
action to initiate consuming the media feed is identified by a processing unit [202], on a set
20 of user devices [102] e.g., Device A, Device B and Device C is 1 min, 2 min and 3 min
respectively. Further, the media playout action to terminate consuming the media feed
identified by the processing unit [202], on a set of user devices i.e., Device A, Device B and
Device C is 3 min, 4 min and 5 min respectively as shown in table 1 below.
S. NO USER DEVICE START TIME END TIME
1 Device A 1 min 3 min
2 Device B 2 min 4 min
3 Device C 3min 5 min
25
TABLE 1
For the purpose of the present disclosure, "set of metadata" refers to a structured collection
of data that is received by the processing unit [202] from the set of user devices [102] based
30 on the media playout action. The set of metadata includes specific information essential for
short media feed generation. It comprises, but is not limited to, the following data elements:
19
5
● Start Time: The term "start time" denotes the temporal point at which the media
playout action is initiated on the user devices. It signifies the exact moment when the
playback of the media feed begins.
● End Time: The term "end time" indicates the temporal point at which the media
10 playout action is terminated on the user devices. It marks the precise moment when
the playback of the media feed concludes.
● Total Runtime: The term "total runtime" represents the total duration of the media
feed's playback available on the user devices. It provides the overall length of time the
media feed may be played.
15 It is to be noted that the start time and the end time of the media feed may be different from
the total runtime of the media feed such as the total length of the media feed i.e., total
runtime of the media feed is 10 min but the user may have performed the media playout
action on the user device to initiate watching the media feed at 2 min mark of the media feed
and again performed the playout action to terminate watching the media feed at 5 min mark
20 of the media feed i.e., the start time associated with the media playout action is 2 min and
the end time associated with the media playout action is 5 min mark whereas the total
runtime associated with the media feed is 10 min. Further, it is also to be noted that the
definitions do not limit the disclosure to these specific data elements and may include any
other relevant data that would be obvious to a person skilled in the art and necessary for
25 generating the short media feed.
Further, the processing unit [202] of the system [200] is configured to identify, via a top frame
identification module [202(a)], one or more top frames from the set of media frames based
on the set of metadata, wherein the one or more top frames from the set frames are
30 identified via the ML/AI‐based short‐form feed curation module [202(c)]based on a
comparison of the start time associated with the media playout action, the end time
associated with the media playout action, and the total runtime associated with the media
feed. Further, in an implementation of the present solution the one or more top frames from
the set frames are identified via the ML/AI‐based short‐form feed curation module [202(c)]
20
based on a user media traffic data, wherein the 5 user media traffic data is based on the media
playout action initiated at the set of user devices.
Further, in another implementation of the present solution, the one or more top frames from
the set frames are identified via the ML/AI‐based short‐form feed curation module [202(c)]
10 based on a predefined threshold level associated with the user media traffic data. It is to be
noted that the predefined threshold level associated with the user media traffic data, may be
a system‐defined threshold level associated with the user media traffic data or a dynamically
defined threshold level associated with the user media traffic data based on certain
conditions. Further, the choice of the threshold level, as well as the identification of one or
15 more top frames from the set frames based on this threshold level, is subject to different
factors, including technological advancements, user preferences, and system requirements.
Therefore, any variations or alternative methods of determining the threshold level or
identifying top frames that may become evident to skilled individuals in the relevant field are
also intended to be covered within the scope of this disclosure. It is also important to
20 recognize that advancements in technology or the introduction of new methodologies may
lead to innovative ways of setting and applying the threshold level in the context of user
media traffic data.
In an implementation of the present solution, the one or more top frames from the set of
25 media frames may be identified by the processing unit [202] via the top frame identification
module [202(a)], based on a set of historic metadata associated with past media playout
action of one or more users retrieved from the storage unit [204], wherein the historic
metadata may be associated with the target media generated based on the one or more top
frames and the short media feed data or a recommended media based at least one of the one
30 or more top frames, the short media feed data, and the long media feed. In another
implementation of the present solution, the one or more top frames from the set of media
frames may be identified by the processing unit [202] based on generating a graphical
representation of the set of historic metadata associated with the past media playout action
of the one or more users retrieved from the storage unit [204] via a predefined data analysis
35 method such as a heatmap, a bar graph etc. In another implementation of the present
solution, the one or more top frames from the set of media frames may be identified by the
21
processing unit [202] 5 via the top frame identification module [202(a)] based on generating a
graphical representation based on at least one of the set of historic metadata associated with
the past media playout action of the one or more users retrieved from the storage unit [204]
and the metadata associated with the media playout action of one or more users received by
the processing unit [202] in real time via a predefined data analysis method such as a
10 heatmap, a bar graph etc.
In an implementation of the present solution, where the identification of the one or more top
frames from the set of media frames is based on the set of metadata, wherein the set of
metadata may be a real‐time metadata or a historic metadata, said identification of the one
15 or more top frames may be normalized based on a comparison of the real‐time metadata and
the historic metadata via the processing unit [202]. In an implementation the normalization
indicates a probability of the one or more top media frames being consumed at the set of
user devices [102]. Such as in an example if top media frames are identified by the processing
unit [202] based on a historic metadata at 15 sec, 1min 30 sec and 2min 45 sec time mark
20 from the runtime associated with the media feed, now the identification of the one or more
top frames may be normalized based on a comparison of the real‐time metadata i.e., the start
time associated with the media playout action, the end time associated with the media
playout action, and the total runtime associated with the media feed of the users User A, User
B and User C on the user devices i.e., Device A, Device B and Device C respectively, received
25 in real time as shown in table 1, and the one or more top framed identified based on the
historic metadata via the processing unit [202]. The normalization of identification of the one
or more top frames indicates the probability of the one or more top media frames at 15 sec
1min 30 sec and 2min 45 sec time mark of the media feed being consumed by the users on
the Device A, Device B and Device C, wherein the probability is as shown in table 2 below.
30
S.
NO
USER
DEVICE
START
TIME
END
TIME
PROBABILITY
OF MEDIA
FRAME
CONSUMED
AT 15 sec
PROBABILITY
OF MEDIA
FRAME
CONSUMED
PROBABILITY
OF MEDIA
FRAME
CONSUMED
22
AT 1min 30
sec
AT 2min 45
sec
1 Device A 1 min 3 min Low High High
2 Device B 2 min 4 min Low Low High
3 Device C 3min 5 min Low Low Low
5
Table 2
As indicated above in table 2, the probability of consuming the top media frames at 15 sec is
the lowest as all of the devices (i.e., Device A, Device B and Device C) have a low probability
to consume the top media frames at 15 sec. However, the probability of consuming the top
10 media frames at 1 min 30 sec is the medium probability as only Device A has a high probability
to consume the top media frames at 1 min 30 sec, whereas the probability of consuming the
top media frames at 2 min 45 sec is the higher as the devices i.e., Device A and Device B have
a high probability to consume the top media frames at 2 min 45 sec.
15 A person skilled in the art would appreciate that normalization plays a pivotal role in the
implementation of the described solution for discerning top frames within a set of media
frames. Further, normalization enhances accurate identification of frames, regardless of the
varying conditions and contexts in which they are viewed. In this context, two types of
metadata are crucial, i.e., real‐time metadata and historical metadata. The real‐time
20 metadata reflects current circumstances, while the historical metadata offers insights from
previous instances. These data sets can significantly differ, potentially causing discrepancies
in the identification process. Normalization intervenes by standardizing the comparison
between these two types of metadata. This process harmonizes data, rectifying disparities in
timeframes and other pertinent parameters. This proves particularly critical in instances
25 where the timing of media consumption is paramount. For example, if a specific frame
historically garnered popularity at particular time intervals, normalization ensures that this
historical data aligns with the real‐time circumstances of the current media feed. In essence,
normalization elevates the precision and dependability of the identification process,
facilitating a more accurate forecast of frames likely to be viewed by users on their respective
23
devices in real time. This 5 insight proves invaluable for optimizing content delivery and
enriching user experiences, establishing it as an indispensable facet of the overall solution.
Further, the processing unit [202] of the system [200] is configured to generate, via a ML/AIbased
short‐form feed curation module [202(c)], a short media feed associated with the
10 media feed based on the one or more top frames. For ease of understanding continuing from
the above example, wherein the one or more top media frames are identified by the
processing unit [202] at 15 sec, 1min 30 sec and 2min 45 sec mark of the media feed, now the
processing unit [202] may generated the short media feed associated with the media feed
comprising said one or more top media frames i.e., the top media frames at 15 sec, 1min 30
15 sec and 2min 45 sec mark of the media feed. In another exemplary implementation of the
present solution, where the identification of the one or more top media frames are
normalized the processing unit [202] may generate the short media feed associated with the
media feed comprising the one or more top media frames, wherein the probability of
consuming the one or more top media frame is at least one of the medium probability and
20 the high probability as discussed above i.e., the processing unit [202] may generated the short
media feed associated with the media feed comprising said one or more top media frames
i.e., the top media frames at 1min 30 sec and 2min 45 sec mark of the media feed.
Further, the processing unit [202] of the system [200] is configured to generate, via a title and
25 description curation module [202(d)], a short media feed data associated with the short
media feed based on the set of media frame data, wherein the short media feed data
comprises at least one of a title associated with the short media feed and a description
associated with the short media feed. In an implementation of present solution, the short
media feed data is generated via the recommendation module [202(g)] based on one or more
30 target media frame data associated with the one or more top frames. Continuing from above
example, where the short media feed is generated based on the media frame at 1min 30 sec
and 2min 45 sec mark of the media feed and the title associated with the short media feed
and a description associated with the short media feed may be generated based on the
retrieving the short media feed data from the set of media frame data i.e., the short media
35 feed data associated with the media frame data at 1min 30 sec and 2min 45 sec mark of the
media feed. Further, in another exemplary implementation of the present invention, where
24
5 the short media feed is generated based on the media frame at 1min 30 sec and 2min 45 sec
mark of the media feed and the title associated with the short media feed and a description
associated with the short media feed may be generated based on the retrieving the media
frame data from the set of media frame data associated with the one or more top frames
detected by the processing unit [202] i.e., the media frame data associated with the top
10 frames at 15 sec, 1min 30 sec and 2min 45 sec mark.
Further, in another exemplary implementation of the present solution the title associated
with the short media feed and a description associated with the short media feed may be
generated via a large language model (LLM). It should be noted that the use of a large
15 language model (LLM) for generating the title and description for the short media feed, as
mentioned in the present disclosure, is provided solely as an exemplary implementation. It
should be understood that any other method that may be obvious to a person skilled in the
art or any other method that may be developed in the future for the same purpose can also
be employed. The mention of LLM‐based generation should not be interpreted as restricting
20 or limiting the scope of the present disclosure, which encompasses all methods and
techniques for generating short media feed data associated with the set of media frame data,
including but not limited to those using LLM or any other language model.
25 In an exemplary implementation, the solution encompasses the Large Language Model (LLM),
wherein to generate the title associated with the short media feed and to generate the
description associated with the short media feed the LLM processes contextual information
and learned patterns to create concise and engaging content descriptors aligned with a media
i.e., the short media feed and the long media feed. Subsequently, explicit user signals such as
30 likes, dislikes, comments, and other interactions are collected and analyzed. This user
feedback serves as a vital input for iterative refinement of the LLM‐based generation process,
enhancing the alignment between generated content and user preferences. Additionally, the
system leverages the generated short media feed to recommend the long media feed,
including but not limited to movies, songs, albums, podcasts, etc. These recommendations
35 are curated through a comprehensive analysis of content characteristics, user behavior, and
engagement metrics.
25
5
Thereafter, the processing unit [202] of the system [200] is configured to generate, via the
recommendation module [202(g)], the target short media feed based on the short media feed
and the short media feed data. In an exemplary implementation of the present solution, the
target short media feed generated by the processing unit [202] may be transmitted for
10 consumption to one or more users of one or more user devices based on their past media
preferences, a media similarity with their past media preferences, a user action such as a
media like action to a similar media of similar genre etc. or may be transmitted for
consumption to the one or more users based on any other parameter that may be obvious to
the person skilled in the art.
15
Now, referring to Figure 3, an exemplary method flow diagram [300], for generating a target
short media feed, in accordance with exemplary embodiments of the present disclosure is
shown. In an implementation the method [300] is performed by the system [200]. Further, in
an implementation, the system [200] may be present in the user device [102] to implement
20 the features of the present disclosure. Also, as shown in Figure 3, the method [300] starts at
step [302].
Further, in an implementation of the present solution, the system [200] is configured for
generating one or more media recommendations via the recommendation module [202(g)]
25 based on the one or more top frames associated with the short media feed generated at the
server device [106] based on the set of media frame data, wherein the one or more media
recommendations is at least one of an existing media feed based on at least the one or more
top frames, the short media feed data, and a long media feed generated based on the one or
more top frames and the short media feed data. Further, the long media feed may comprise
30 the existing media feed based on at least the one or more top frames and the short media
feed data. Further, the long media feed may be retrieved from the storage unit [204] based
on at least one of the one or more top frames and the short media feed data.
Next, at step [304], the method [300] comprises identifying, by a processing unit [202] via a
35 media playout action collecting module [202(h)], a media playout action on a set of user
devices, wherein the media playout action is associated with a media feed and the media feed
26
comprises at least a set of media frames and a set 5 of media frame data. In an implementation
of the solution, one or more media frame data from the set of media frame data comprises
at least a metadata associated with one or more media frames from the set of media frames.
As used herein “metadata associated with one or more media frames” refers to a collection
of structured data elements intricately linked with the one or more media frames of the media
10 feed. The metadata associated with one or more media frames encompasses essential
information that enhances the identification and processing of the one or more media frames
during the generation of one or more short media feeds. Further, the metadata associated
with one or more media frames may include a media frame identifier data, which uniquely
identifies each media frame within the media feed to facilitate efficient referencing and
15 tracking. Additionally, the metadata associated with one or more media frames may include
one or more data related to a subtitle of the one or more media frames, encompassing textual
information pertaining to the subtitles or a caption associated with each frame from the one
or more media frames. These subtitle data may be associated with one or more spoken
languages such as Hindi, English, Gujrati etc., and can be represented in machine‐readable
20 formats. Further, in an implementation of the present solution, the metadata associated with
one or more media frames may include an audio output associated with the with each frame
from the one or more media frames, wherein the audio output may be converted in a
predefined format such as vector format. Furthermore, in an implementation of the present
solution the metadata associated with one or more media frames may include the audio
25 output associated with the with each frame from the one or more media frames mapped with
the caption associated with said each frame from the one or more media frames.
Furthermore, the metadata may comprise time stamp data for each media frame from the
one or more media frames, such as data related to a start time associated with each media
frame and an end time associated with each media frame. Furthermore, the scope of these
30 terms may extend to include any obvious variations or developments that would be apparent
to a person skilled in the art in light of technological advancements or industry developments.
Therefore, the interpretation of these terms should be broad and flexible to accommodate
future discoveries or innovations in the field.
35 Further, the term "playout action on a set of user devices" as used in the present disclosure
refers to actions initiated by users on their respective user devices for starting or stopping the
27
playback of 5 a media feed. These actions may include a "start playing media feed" action,
whereby users initiate the playback of a media feed on their user devices, and a "stop playing
media feed" action, whereby users halt the ongoing playback of a media feed on their user
devices. The term is intended to cover actions related to the initiation and termination of
media feed playback on the set of user devices [102]. It is important to note that the definition
10 provided above is for illustrative purposes only and shall not be limited to the specific actions
mentioned. The term "playout action on a set of user devices" may encompass any variations
or related actions that would be obvious to a person skilled in the art. Such variations may
include additional media control actions, such as pause, rewind, or fast‐forward, initiated by
users during media feed playback on their user devices. Furthermore, the term may also
15 extend to cover actions initiated by user interfaces, applications, or automated systems that
manage media feed playback on user devices.
Next, at step [306], the method [300] comprises receiving, by the processing unit [202] via a
normalization module [202(b)], a set of metadata from the set of user devices based on the
20 media playout action, wherein the set of metadata comprises at least a start time associated
with the media playout action, an end time associated with the media playout action, and a
total runtime associated with the media feed. For ease of understanding, in an example,
wherein the total runtime associated with the media feed is 5 mins and the media playout
action to initiate consuming the media feed is identified by a processing unit [202], on a set
25 of user devices [102] e.g., Device A, Device B and Device C is 1 min, 2 min and 3 min
respectively. Further, the media playout action to terminate consuming the media feed
identified by the processing unit [202], on a set of user devices i.e., Device A, Device B and
Device C is 3 min, 4 min and 5 min respectively as shown in table 3 below.
S. NO USER DEVICE START TIME END TIME
1 Device A 1 min 3 min
2 Device B 2 min 4 min
3 Device C 3min 5 min
30
TABLE 3
28
5
For the purpose of the present disclosure, "set of metadata" refers to a structured collection
of data that is received by the processing unit [202] from the set of user devices [102] based
on the media playout action. The set of metadata includes specific information essential for
short media feed generation. It comprises, but is not limited to, the following data elements:
10
● Start Time: The term "start time" denotes the temporal point at which the media
playout action is initiated on the user devices. It signifies the exact moment when the
playback of the media feed begins.
● End Time: The term "end time" indicates the temporal point at which the media
15 playout action is terminated on the user devices. It marks the precise moment when
the playback of the media feed concludes.
● Total Runtime: The term "total runtime" represents the total duration of the media
feed's playback available on the user devices. It provides the overall length of time the
media feed may be played.
20 It is to be noted that the start time and the end time of the media feed may be different from
the total runtime of the media feed such as the total length of the media feed i.e., total
runtime of the media feed is 10 min but the user may have performed the media playout
action on the user device to initiate watching the media feed at 2 min mark of the media feed
and again performed the playout action to terminate watching the media feed at 5 min mark
25 of the media feed i.e., the start time associated with the media playout action is 2 min and
the end time associated with the media playout action is 5 min whereas the total runtime
associated with the media feed is 10 min. Further, it is also to be noted that the definitions
do not limit the disclosure to these specific data elements and may include any other relevant
data that would be obvious to a person skilled in the art and necessary for generating the
30 short media feed.
Next, at step [308], the method [300] comprises identifying, by the processing unit [202] via
the top frame identification module [202(a)], one or more top frames from the set of media
frames based on the set of metadata. In a preferred implementation of the present solution,
35 the identifying via the ML/AI‐based short‐form feed curation module [202(c)] the one or more
29
top frames from the set frames is based on a comparison 5 of the start time associated with
the media playout action, the end time associated with the media playout action, and the
total runtime associated with the media feed. Further, in another implementation of the
present solution the identifying via the ML/AI‐based short‐form feed curation module [202(c)]
the one or more top frames from the set frames is further based on a user media traffic data,
10 wherein the user media traffic data is based on the media playout action initiated at the set
of user devices.
Further, in another implementation of the present solution, the one or more top frames from
the set frames are identified via the ML/AI‐based short‐form feed curation module [202(c)]
15 based on a predefined threshold level associated with the user media traffic data. It is to be
noted that the predefined threshold level associated with the user media traffic data, may be
a system‐defined threshold level associated with the user media traffic data or a dynamically
defined threshold level associated with the user media traffic data based on certain
conditions. Further, the choice of the threshold level, as well as the identification of one or
20 more top frames from the set frames based on this threshold level, is subject to different
factors, including technological advancements, user preferences, and system requirements.
Therefore, any variations or alternative methods of determining the threshold level or
identifying top frames that may become evident to skilled individuals in the relevant field are
also intended to be covered within the scope of this disclosure. It is also important to
25 recognize that advancements in technology or the introduction of new methodologies may
lead to innovative ways of setting and applying the threshold level in the context of user
media traffic data.
In an implementation of the present solution, the one or more top frames from the set of
30 media frames may be identified by the processing unit [202], based on a set of historic
metadata associated with past media playout action of one or more users retrieved from the
storage unit [204], wherein the historic metadata may be associated with the target media
generated based on the one or more top frames and the short media feed data or a
recommended media based at least one of the one or more top frames, the short media feed
35 data, and the long media feed. In another implementation of the present solution, the one or
more top frames from the set of media frames may be identified by the processing unit [202]
30
based on generating a graphical representation 5 of the set of historic metadata associated with
the past media playout action of the one or more users retrieved from the storage unit [204]
via a predefined data analysis method such as a heatmap, a bar graph etc. In another
implementation of the present solution, the one or more top frames from the set of media
frames may be identified by the processing unit [202] based on generating a graphical
10 representation based on at least one of the set of historic metadata associated with the past
media playout action of the one or more users retrieved from the storage unit [204] and the
metadata associated with the media playout action of one or more users received by the
processing unit [202] in real time via a predefined data analysis method such as a heatmap, a
bar graph etc.
15
In an implementation of the present solution, where the identification of the one or more top
frames from the set of media frames is based on the set of metadata, wherein the set of
metadata may be a real‐time metadata or a historic metadata, said identification of the one
or more top frames may be normalized based on a comparison of the real‐time metadata and
20 the historic metadata via the processing unit [202]. In an implementation the normalization
indicates a probability of the one or more top media frames being consumed at the set of
user devices [102]. Such as in an example if top media frames are identified by the processing
unit [202] based on a historic metadata at 15 sec, 1min 30 sec and 2min 45 sec time mark
from the runtime associated with the media feed, now the identification of the one or more
25 top frames may be normalized based on a comparison of the real‐time metadata i.e., the start
time associated with the media playout action, the end time associated with the media
playout action, and the total runtime associated with the media feed of the users User A, User
B and User C on the user devices i.e., Device A, Device B and Device C respectively, received
in real time as shown in table 3, and the one or more top framed identified based on the
30 historic metadata via the processing unit [202]. The normalization of identification of the one
or more top frames indicates the probability of the one or more top media frames at 15 sec
1min 30 sec and 2min 45 sec time mark of the media feed being consumed by the users on
the Device A, Device B and Device C, wherein the probability is as shown in table 4 below.
31
S.
NO
USER
DEVICE
START
TIME
END
TIME
PROBABILITY
OF MEDIA
FRAME
CONSUMED
AT 15 sec
PROBABILITY
OF MEDIA
FRAME
CONSUMED
AT 1min 30
sec
PROBABILITY
OF MEDIA
FRAME
CONSUMED
AT 2min 45
sec
1 Device A 1 min 3 min Low High High
2 Device B 2 min 4 min Low Low High
3 Device C 3min 5 min Low Low Low
5
Table 4
As indicated above in table 4, the probability of consuming the top media frames at 15 sec is
the lowest as all of the devices (i.e., Device A, Device B and Device C) have a low probability
to consume the top media frames at 15 sec. However, the probability of consuming the top
10 media frames at 1 min 30 sec is the medium probability as only Device A has a high probability
to consume the top media frames at 1 min 30 sec, whereas the probability of consuming the
top media frames at 2 min 45 sec is the higher as the devices i.e., Device A and Device B have
a high probability to consume the top media frames at 2 min 45 sec.
15 A person skilled in the art would appreciate that normalization plays a pivotal role in the
implementation of the described solution for discerning top frames within a set of media
frames. Further, normalization enhances accurate identification of frames, regardless of the
varying conditions and contexts in which they are viewed. In this context, two types of
metadata are crucial, i.e., real‐time metadata and historical metadata. The real‐time
20 metadata reflects current circumstances, while the historical metadata offers insights from
previous instances. These data sets can significantly differ, potentially causing discrepancies
in the identification process. Normalization intervenes by standardizing the comparison
between these two types of metadata. This process harmonizes data, rectifying disparities in
timeframes and other pertinent parameters. This proves particularly critical in instances
25 where the timing of media consumption is paramount. For example, if a specific frame
historically garnered popularity at particular time intervals, normalization ensures that this
32
historical data aligns w 5 ith the real‐time circumstances of the current media feed. In essence,
normalization elevates the precision and dependability of the identification process,
facilitating a more accurate forecast of frames likely to be viewed by users on their respective
devices in real time. This insight proves invaluable for optimizing content delivery and
enriching user experiences, establishing it as an indispensable facet of the overall solution.
10
Next, at step [310], the method [300] comprises generating, by the processing unit [202] via
a ML/AI‐based short‐form feed curation module [202(c)], a short media feed associated with
the media feed based on the one or more top frames. For ease of understanding continuing
from the above example, wherein the one or more top media frames are identified by the
15 processing unit [202] at 15 sec, 1min 30 sec and 2min 45 sec mark of the media feed, now the
processing unit [202] may generated the short media feed associated with the media feed
comprising said one or more top media frames i.e., the top media frames at 15 sec, 1min 30
sec and 2min 45 sec mark of the media feed. In another exemplary implementation of the
present solution, where the identification of the one or more top media frames are
20 normalized the processing unit [202] may generate the short media feed associated with the
media feed comprising the one or more top media frames, wherein the probability of
consuming the one or more top media frame is at least one of the medium probability and
the high probability as discussed above i.e., the processing unit [202] may generated the short
media feed associated with the media feed comprising said one or more top media frames
25 i.e., the top media frames at 1min 30 sec and 2min 45 sec mark of the media feed.
Next, at step [312], the method [300] comprises generating, by the processing unit [202] via
a title and description curation module [202(d)], a short media feed data associated with the
short media feed based on the set of media frame data, wherein the short media feed data
30 comprises at least one of a title associated with the short media feed and a description
associated with the short media feed. Further, in an implementation of the present solution,
the generating via a recommendation module [202(g)] the short media feed data is further
based on the one or more top frames associated with the short media feed. Continuing from
above example, where the short media feed is generated based on the media frame at 1min
35 30 sec and 2min 45 sec mark of the media feed and the title associated with the short media
feed and a description associated with the short media feed may be generated based on the
33
retrieving the short media feed data from the set of media frame data 5 i.e., the short media
feed data associated with the media frame data at 1min 30 sec and 2min 45 sec mark of the
media feed. Further, in another exemplary implementation of the present solution, where
the short media feed is generated based on the media frame at 1min 30 sec and 2min 45 sec
mark of the media feed and the title associated with the short media feed and a description
10 associated with the short media feed may be generated based on the retrieving the media
frame data from the set of media frame data associated with the one or more top frames
detected by the processing unit [202] i.e., the media frame data associated with the top
frames at 15 sec, 1min 30 sec and 2min 45 sec mark.
15 Further, in another exemplary implementation of the present solution the title associated
with the short media feed and a description associated with the short media feed may be
generated via a large language model (LLM). It should be noted that the use of a large
language model (LLM) for generating the title and description for the short media feed, as
mentioned in the present disclosure, is provided solely as an exemplary implementation. It
20 should be understood that any other method that may be obvious to a person skilled in the
art or any other method that may be developed in the future for the same purpose can also
be employed. The mention of LLM‐based generation should not be interpreted as restricting
or limiting the scope of the present disclosure, which encompasses all methods and
techniques for generating short media feed data associated with the set of media frame data,
25 including but not limited to those using LLM or any other language model.
In an exemplary implementation, the solution encompasses the Large Language Model (LLM),
wherein to generate the title associated with the short media feed and to generate the
description associated with the short media feed the LLM processes contextual information
30 and learned patterns to create concise and engaging content descriptors aligned with a media
i.e., the short media feed and the long media feed. Subsequently, explicit user signals such as
likes, dislikes, comments, and other interactions are collected and analyzed. This user
feedback serves as a vital input for iterative refinement of the LLM‐based generation process,
enhancing the alignment between generated content and user preferences. Additionally, the
35 system leverages the generated short media feed to recommend the long media feed,
including but not limited to movies, songs, albums, podcasts, etc. These recommendations
34
are curated through a comprehensive 5 analysis of content characteristics, user behavior, and
engagement metrics.
Next, at step [314], the method [300] comprises generating, by the processing unit [202] via
the recommendation module [202(g)], the target short media feed based on the short media
10 feed and the short media feed data. In an exemplary implementation of the present solution,
the target short media feed generated by the processing unit [202] may be transmitted for
consumption to one or more users of one or more user devices based on their past media
preferences, a media similarity with their past media preferences, a user action such as a
media like action to a similar media of similar genre etc. or may be transmitted for
15 consumption to the one or more users based on any other parameter that may be obvious to
the person skilled in the art.
Thereafter, the method terminates at step [316].
20 Therefore, the present disclosure introduces a novel solution for generating a target short
media feed with unparalleled efficiency and customization. Unlike conventional methods, the
proposed technique leverages advanced processing capabilities to identify media playout
actions on user devices and obtain essential metadata, including start time, end time, and
total runtime associated with media feeds. This information facilitates the swift and accurate
25 identification of one or more top frames from the media frames, ensuring the selection of the
most captivating and relevant content for the short media feed. A significant advantage of
the disclosed method lies in its ability to incorporate user media traffic data, acquired from
user‐initiated media playout actions, to enhance the frame identification process. By
analyzing user behavior, the method aligns the short media feed with user preferences,
30 delivering a personalized and engaging experience. Furthermore, the customizable threshold
levels associated with user media traffic data offer unprecedented flexibility, empowering
users or administrators to tailor the short media feed generation to specific preferences or
requirements.
35 Moreover, the novel solution as disclosed herein utilizes the target media frame data
associated with the selected top frames. This enriched data, including additional metadata
35
and context information, elevates the quality 5 of the short media feed data, providing users
with comprehensive and informative titles and descriptions. The integration of media frame
data ensures that the target short media feed is not only visually appealing but also
contextually relevant, resulting in a captivating and immersive user experience.
10 In conclusion, the combination of advanced frame identification techniques, user media
traffic data, customizable threshold levels, and target media frame data presents a
revolutionary method for generating short media feeds. This innovation leads to an enhanced
and personalized user experience, making the disclosed method a standout solution in the
field of media content curation and consumption. The technical advantages of this solution,
15 including efficient content selection, real‐time adaptability to user preferences, and high
customization levels, have the potential to transform how short media feeds are created and
consumed. As a result, this approach holds significant promise in delivering a more engaging
and tailored media experience to users, ultimately driving higher customer retention and
revenue generation for the platform.
20 While considerable emphasis has been placed herein on the preferred embodiments, it will
be appreciated that many embodiments can be made and that many changes can be made in
the preferred embodiments without departing from the principles of the disclosure. These
and other changes in the preferred embodiments of the disclosure will be apparent to those
skilled in the art from the disclosure herein, whereby it is to be distinctly understood that the
25 foregoing descriptive matter to be implemented merely as illustrative of the disclosure and
not as limitation.We Claim:
1. A method for generating a target short media feed, the method comprising:
‐ identifying, by a processing unit [202] via a media playout action collecting module [202(h)], a media playout action on a set of user devices [102], wherein the media playout action is associated with a media feed and the media feed comprises at least a set of media frames and a set of media frame data;
‐ receiving, by the processing unit [202] via a normalization module [202(b)], a set of metadata from the set of user devices [102] based on the media playout action, wherein the set of metadata comprises at least a start time associated with the media playout action, an end time associated with the media playout action, and a total runtime associated with the media feed;
‐ identifying, by the processing unit [202] via a top frame identification module [202(a)], one or more top frames from the set of media frames based on the set of metadata;
‐ generating, by the processing unit [202] via a ML/AI‐based short‐form feed curation module [202(c)], a short media feed associated with the media feed based on the one or more top frames;
‐ generating, by the processing unit [202] via a title and description curation module [202(d)], a short media feed data associated with the short media feed based on the set of media frame data, wherein the short media feed data comprises at least one of a title associated with the short media feed and a description associated with the short media feed; and
‐ generating, by the processing unit [202] via a recommendation module [202(g)], the target short media feed based on the short media feed and the short media feed data.
2. The method as claimed in claim 1, wherein one or more media frame data from the set of media frame data comprises at least a metadata associated with one or more media frames from the set of media frames.
3. The method as claimed in claim 1, wherein the identifying the one or more top frames from the set frames is based on a comparison of the start time associated with the

media playout action, the end time associated with the media playout action, and the total runtime associated with the media feed.
4. The method as claimed in claim 3, wherein the identifying the one or more top frames from the set frames is further based on a user media traffic data, wherein the user media traffic data is based on the media playout action initiated at the set of user devices.
5. The method as claimed in claim 4, wherein the identifying the one or more top frames from the set frames is further based on a predefined threshold level associated with the user media traffic data.
6. The method as claimed in claim 1, wherein the generating the short media feed data is further based on the one or more top frames associated with the short media feed.
7. The method as claimed in claim 1, the method further comprises generating one or more media recommendations via the recommendation module [202(g)] based at least on one or more target media frame data associated with the one or more top frames, wherein the one or more media recommendations comprises at least one of one or more existing media feeds and one or more long media feeds.
8. The method as claimed in claim 7, the method further comprises:
‐ recommending, by the processing unit [202] via the recommendation module
[202(g)] to the set of user devices [102], the target short media feed, ‐ receiving, by the processing unit [202] via an explicit signal (user feedback)
collecting module [202(e)] from the set of user devices [102], a user feedback on
the target short media feed, ‐ identifying, by the processing unit [202], a set of target short media feeds based
on the user feedback, ‐ identifying, by the processing unit [202], a set of user preference parameters based
on at least one of one or more user actions on the set of target short media feeds,
the media playout actions and the one or more top frames, ‐ detecting, by the processing unit [202] via a content identification module [202(f)],
at least one of the one or more existing media feeds and the one or more long
media feeds based on the set of user preference parameters,

‐ identifying, by the processing unit [202], one or more target short media feeds from the set of target short media feeds in at least one of the one or more existing media feeds and the one or more long media feeds,
‐ ranking , by the processing unit [202], at least one of the one or more existing media feeds and the one or more long media feeds based on the identified one or more target short media feeds in at least one of the one or more existing media feeds and the one or more long media feeds, and
‐ recommending, by the processing unit [202] via the recommendation module [202(g)] to the set of user devices [102], a target media feed comprising at least one of the one or more existing media feeds and the one or more long media feeds based on the ranking of at least one of the one or more existing media feeds and the one or more long media feeds.
9. The method as claimed in claim 8, wherein the detecting, by the processing unit [202]
via the content identification module [202(f)], at least one of the one or more existing
media feeds and the one or more long media feeds further comprises:
‐ identifying, via the content identification module [202(f)], at least one of a title and one or more keywords related to one or more short media feeds from the set of target short media feeds, wherein said identification is based on the set of user preference parameters, and
‐ detecting, via the content identification module [202(f)] from a database, at least one of the one or more existing media feeds and the one or more long media feeds based on at least one of the identified title and the identified one or more keywords.
10. A system for generating a target short media feed, the system comprises:
‐ a processing unit [202], configured to:
• identify, via a media playout action collecting module [202(h)], a media p l ay o ut acti o n on a s et o f u se r d e v i ce s [10 2 ], w he r e i n th e medi a playout action is associated with a media feed and the media feed comprises at least a set of media frames and a set of media frame data;
• receive, via a normalization module [202(b)], a set of metadata from the set of user devices [102] based on the media playout action, wherein the set of

metadata comprises at least a start time associated with the media playout action, an end time associated with the media playout action, and a total runtime associated with the media feed;
• identify, via the top frame identification module [202(a)], one or more top frames from the set of media frames based on the set of metadata;
• generate, via a ML/AI‐based short‐form feed curation module [202(c)], a short media feed associated with the media feed based on the one or more top frames;
• generate, via a title and description curation module [202(d)], a short media feed data associated with the short media feed based on the set of media frame data, wherein the short media feed data comprises at least one of a title associated with the short media feed and a description associated with the short media feed; and
• generate, via a recommendation module [202(g)], the target short media feed based on the short media feed and the short media feed data.

11. The system as claimed in claim 10, wherein one or more media frame data from the set of media frame data comprises at least a metadata associated with one or more media frames from the set of media frames.
12. The system as claimed in claim 10, wherein the one or more top frames from the set frames are identified based on a comparison of the start time associated with the media playout action, the end time associated with the media playout action, and the total runtime associated with the media feed.
13. The system as claimed in claim 12, wherein the one or more top frames from the set frames are identified based on a user media traffic data, wherein the user media traffic data is based on the media playout action initiated at the set of user devices.
14. The system as claimed in claim 13, wherein the one or more top frames from the set frames are identified based on a predefined threshold level associated with the user media traffic data.
15. The system as claimed in claim 10, wherein the short media feed data is generated based on the one or more top frames associated with the short media feed.

16. The system as claimed in claim 10, wherein the processing unit [202] is further configured to generate one or more media recommendations via the recommendation module [202(g)] based at least on one or more target media frame data associated with the one or more top frames, wherein the one or more media recommendations comprises at least one of one or more existing media feeds and one or more long media feeds.
17. The system as claimed in claim 16, the system comprises processing unit [202] further configured to:
‐ recommend, via the recommendation module [202(g)] to the set of user devices
[102], the target short media feed, ‐ receive, via an explicit signal (user feedback) collecting module [202(e)] from the
set of user devices [102], a user feedback on the target short media feed, ‐ identify, a set of target short media feeds based on the user feedback, ‐ identify, a set of user preference parameters based on at least one of one or more
user actions on the set of target short media feeds, the media playout actions and
the one or more top frames, ‐ detect, via a content identification module [202(f)], at least one of the one or more
existing media feeds and the one or more long media feeds based on the set of
user preference parameters, ‐ identify, one or more target short media feeds from the set of target short media
feeds in at least one of the one or more existing media feeds and the one or more
long media feeds, ‐ ranking, at least one of the one or more existing media feeds and the one or more
long media feeds based on the identified one or more target short media feeds in
at least one of the one or more existing media feeds and the one or more long
media feeds, and ‐ recommend, via the recommendation module [202(g)] to the set of user devices
[102], a target media feed comprising at least one of the one or more existing
media feeds and the one or more long media feeds based on the ranking of at least
one of the one or more existing media feeds and the one or more long media feeds.

18. The method as claimed in claim 17, wherein to detect at least one of the one or more existing media feeds and the one or more long media feeds, the processing unit [202] via the content identification module [202(f)] is further configured to: ‐ identify, at least one of a title and one or more keywords related to one or more
short media feeds from the set of target short media feeds, wherein said
identification is based on the set of user preference parameters, and ‐ detect, from a database, at least one of the one or more existing media feeds and
the one or more long media feeds based on at least one of the identified title and
the identified one or more keywords.