# US 12,225,268 B2 · Method and system for facilitating content recommendation to content viewers > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,225,268 B2 - Document ID: US12225268B2 - Status: Granted - Kind: B2 - Application number: 18/185,322 - Filed: 2023-03-16 - Published: 2023-10-12 - Granted: 2025-02-11 - Earliest claimed: 2022-03-22 - Assignees: Star India Private Limited - Applicants: Star India Private Limited - Inventors: Rachit Jain; Shenglong Lyu - Web page: https://patentgrep.com/patent/US12225268B2 ## CPC - H04N 21/4668: {for recommending content, e.g. movies} - H04N 21/251: {Learning process for intelligent management, e.g. learning user preferences for recommending movies (details of learning user preferences for the retrieval of video data in a video database G06F16/739; computer systems using learning methods G06N3/08)} - H04N 21/44204: {Monitoring of content usage, e.g. the number of times a movie has been viewed, copied or the amount which has been watched (monitoring of user activities for profile generation for accessing a video database G06F16/739; protecting generic digital content where the protection is independent of the precise nature of the content G06F21/10; arrangements for monitoring the use made of the broadcast services in broadcast systems H04H60/31)} - H04N 21/4826: {using recommendation lists, e.g. of programmes or channels sorted out according to their score} - H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION ## Abstract Various embodiments provide a method and a system for facilitating content recommendation to content viewers. The method performed by the system includes accessing content viewer behavior data associated with a content viewer. The method includes classifying the content viewer in a viewer cohort based on the content viewer behavior data. The viewer cohort includes a plurality of related content viewers. The method includes accessing and aggregating content viewer interaction data from the plurality of related content viewers. The content viewer interaction data includes a popularity index. The method includes computing via a machine learning model, a performance metric for each recommended content item based on the content viewer interaction data. The method includes determining an optimal content sequence for displaying the plurality of recommended content items based on the performance metrics for the plurality of recommended content items. ## Claims ### Claim 1 A computer-implemented method, comprising: accessing, by a system, content viewer behavior data associated with a content viewer from a database associated with the system; classifying, by the system, the content viewer in a viewer cohort from a plurality of viewer cohorts based, at least in part, on the content viewer behavior data, the viewer cohort comprising a plurality of related content viewers; accessing and aggregating, by the system, content viewer interaction data from the plurality of related content viewers in the viewer cohort, the content viewer interaction data comprising a popularity index of each recommended content item from a plurality of recommended content items that is being served on a plurality of electronic devices associated with the plurality of related content viewers by a digital platform server; computing, by the system via a machine learning model, a performance metric for each recommended content item of the plurality of recommended content items based, at least in part, on the content viewer interaction data; and determining, by the system, an optimal content sequence for displaying the plurality of recommended content items based, at least in part, on the performance metrics for the plurality of recommended content items, the optimal content sequence indicating a sequence for displaying each recommended content item of the plurality of recommended content items on an electronic device of the content viewer; wherein computing the performance metric comprises performing a plurality of operations iteratively for a predefined threshold number of attempts, the plurality of operations comprising: computing, by the system, a score for each recommended content item of the plurality of recommended content items based, at least in part, on the content viewer interaction data and a scoring policy; ranking, by the system, each recommended content item of the plurality of recommended content items based, at least in part, on the score related to the each recommended content item and a ranking policy; rearranging, by the system, the plurality of ranked recommended content items at a predefined time interval based, at least in part, on a shuffling policy; facilitating, by the system, the display of a plurality of rearranged recommended content items on User Interfaces (UIs) of the plurality of electronic devices associated with the plurality of related content viewers; and computing and updating, by the system, the content viewer interaction data from the plurality of related content viewers based, at least in part, on monitoring interactions of the plurality of related content viewers with the plurality of rearranged recommended content items. ### Claim 2 (depends on 1) The computer-implemented method as claimed in claim 1, wherein the predefined threshold is determined based, at least in part, on a total number of the plurality of recommended content items and a number of recommended content items that can be displayed on a display screen of the electronic device at a given instant. ### Claim 3 (depends on 1) The computer-implemented method as claimed in claim 1, further comprising: accessing, by the system, phonebook data of the content viewer from the electronic device of the content viewer, the phonebook data comprising a plurality of contacts of the content viewer; and upon determining that at least one contact from the plurality of contacts is one of a plurality of related content viewers of the viewer cohort adjusting, by the system, the score for each recommended content item of the plurality of recommended content items based, at least in part, on contact behavior data of the at least one contact, wherein the contact behavior data is accessed from the database. ### Claim 4 (depends on 1) The computer-implemented method as claimed in claim 1, further comprising: accessing, by the system, historical content viewing data of the content viewer from the database; and determining, by the system, the content viewer behavior data associated with the content viewer based, at least in part, on the historical content viewing data. ### Claim 5 (depends on 4) The computer-implemented method as claimed in claim 4, further comprising: accessing, by the system, real-time content viewing data of the content viewer associated with a live-viewing session from the digital platform server; and dynamically updating, by the system, the content viewer behavior data based, at least in part, on the real-time content viewing data. ### Claim 6 (depends on 1) The computer-implemented method as claimed in claim 1, further comprising: accessing, by the system, a set of content viewer behavior data associated with a plurality of content viewers from the database; determining, by the system, a relation metric for the plurality of content viewers based, at least in part, on the set of content viewer behavior data, the relation metric indicating a content preference of each of the plurality of content viewers; generating, by the system, the plurality of viewer cohorts, each of the plurality of viewer cohorts indicating a content category for the plurality of related content viewers with a same content preference; and classifying, by the system, the plurality of content viewers between the plurality of viewer cohorts based, at least in part, on the relation metric, each of the plurality of viewer cohorts comprising the plurality of related content viewers. ### Claim 7 (depends on 1) The computer-implemented method as claimed in claim 1, further comprising: tracking, by the system, interaction information related to each of the plurality of related content viewers, the interaction information comprising click-through rates, average view duration, watch history, search terms, comments, ratings, and number of views related to each recommended content item placed at different positions on UIs of the plurality of electronic devices associated with each of the plurality of related content viewers; and determining, by the system, the popularity index of each recommended content item based, at least in part, on interaction information related to each of the plurality of related content viewers. ### Claim 8 (depends on 1) The computer-implemented method as claimed in claim 1, further comprising: accessing, by the system, a content library of the digital platform server from the database, the content library comprising a plurality of content items; and determining, by the system, the plurality of recommended content items to be served to the content viewer from the content library based, at least in part, on the content viewer behavior data. ### Claim 9 (depends on 1) The computer-implemented method as claimed in claim 1, wherein the machine learning model is a recursive neural network (RNN) model. ### Claim 10 A system, the system comprising: a memory for storing instructions; and a processor configured to execute the instructions and thereby cause the system, at least in part, to: access content viewer behavior data associated with a content viewer from a database associated with the system; classify the content viewer in a viewer cohort from a plurality of viewer cohorts based, at least in part, on the content viewer behavior data, the viewer cohort comprising a plurality of related content viewers; access and aggregate content viewer interaction data from the plurality of related content viewers in the viewer cohort, the content viewer interaction data comprising a popularity index of each recommended content item from a plurality of recommended content items that is being served on a plurality of electronic devices associated with the plurality of related content viewers by a digital platform server; compute via a machine learning model, a performance metric for each recommended content item of the plurality of recommended content items based, at least in part, on the content viewer interaction data; and determine an optimal content sequence for displaying the plurality of recommended content items based, at least in part, on the performance metrics for the plurality of recommended content items, the optimal content sequence indicating a sequence for displaying each recommended content item of the plurality of recommended content items on an electronic device of the content viewer; wherein to compute the performance metric the system is caused to perform a plurality of operations iteratively for a predefined threshold number of attempts, the plurality of operations comprising: compute a score for each recommended content item of the plurality of recommended content items based, at least in part, on the content viewer interaction data and a scoring policy; rank each recommended content item of the plurality of recommended content items based, at least in part, on the score related to the each recommended content item and a ranking policy; rearrange the plurality of ranked recommended content items at a predefined time interval based, at least in part, on a shuffling policy; facilitate the display of a plurality of rearranged recommended content items on User Interfaces (UIs) of the plurality of electronic devices associated with the plurality of related content viewers; and compute and update the content viewer interaction data from the plurality of related content viewers based, at least in part, on monitoring interactions of the plurality of related content viewers with the plurality of rearranged recommended content items. ### Claim 11 (depends on 10) The system as claimed in claim 10, wherein the predefined threshold is determined based, at least in part, on a total number of the plurality of recommended content items and a number of recommended content items that can be displayed on a display screen of the electronic device at a given instant. ### Claim 12 (depends on 10) The system as claimed in claim 10, wherein the system is further caused, at least in part to: access phonebook data of the content viewer from the electronic device of the content viewer, the phonebook data comprising a plurality of contacts of the content viewer; and upon determining that at least one contact from the plurality of contacts is one of a plurality of related content viewers of the viewer cohort adjust the score for each recommended content item of the plurality of recommended content items based, at least in part, on contact behavior data of the at least one contact, wherein the contact behavior data is accessed from the database. ### Claim 13 (depends on 10) The system as claimed in claim 10, wherein the system is further caused, at least in part to: access historical content viewing data of the content viewer from the database; and determine, the content viewer behavior data associated with the content viewer based, at least in part, on the historical content viewing data. ### Claim 14 (depends on 13) The system as claimed in claim 13, wherein the system is further caused, at least in part to: access real-time content viewing data of the content viewer associated with a live-viewing session from the digital platform server; and dynamically update the content viewer behavior data based, at least in part, on the real-time content viewing data. ### Claim 15 (depends on 13) The system as claimed in claim 13, wherein the system is further caused, at least in part to: track interaction information associated with each of the plurality of related content viewers, the interaction information comprising click-through rates, average view duration, watch history, search terms, comments, ratings, and number of views related to each recommended content item placed at different positions on UIs of the plurality of electronic devices associated with each of the plurality of related content viewers; and determine the popularity index of each recommended content item based, at least in part, on interaction information related to each of the plurality of related content viewers. ### Claim 16 A non-transitory computer-readable storage medium comprising computer-executable instructions that, when executed by at least a processor of a system, cause the system to perform a method comprising: accessing content viewer behavior data associated with a content viewer from a database associated with the system; classifying the content viewer in a viewer cohort from a plurality of viewer cohorts based, at least in part, on the content viewer behavior data, the viewer cohort comprising a plurality of related content viewers; accessing and aggregating content viewer interaction data from the plurality of related content viewers in the viewer cohort, the content viewer interaction data comprising a popularity index of each recommended content item from a plurality of recommended content items that is being served on a plurality of electronic devices associated with the plurality of related content viewers by a digital platform server; computing via a machine learning model, a performance metric for each recommended content item of the plurality of recommended content items based, at least in part, on the content viewer interaction data; and determining an optimal content sequence for displaying the plurality of recommended content items based, at least in part, on the performance metrics for the plurality of recommended content items, the optimal content sequence indicating a sequence for displaying each recommended content item of the plurality of recommended content items on an electronic device of the content viewer; wherein computing the performance metric comprises performing a plurality of operations iteratively for a predefined threshold number of attempts, the plurality of operations comprising: computing a score for each recommended content item of the plurality of recommended content items based, at least in part, on the content viewer interaction data and a scoring policy; ranking each recommended content item of the plurality of recommended content items based, at least in part, on the score related to the each recommended content item and a ranking policy; rearranging the plurality of ranked recommended content items at a predefined time interval based, at least in part, on a shuffling policy; facilitating the display of a plurality of rearranged recommended content items on User Interfaces (UIs) of the plurality of electronic devices associated with the plurality of related content viewers; and computing and updating the content viewer interaction data from the plurality of related content viewers based, at least in part, on monitoring interactions of the plurality of related content viewers with the plurality of rearranged recommended content items. ### Claim 17 A computer-implemented method, comprising: accessing, by a system, content viewer behavior data associated with a content viewer from a database associated with the system; classifying, by the system, the content viewer in a viewer cohort from a plurality of viewer cohorts based, at least in part, on the content viewer behavior data, the viewer cohort comprising a plurality of related content viewers; accessing, by the system, a content library from a digital platform server, the content library including a plurality of content items; determining and accessing, by the system, a plurality of recommended content items from the plurality of content items of the content library for the content viewer based, at least in part, on the content viewer behavior data; computing, by the system via a machine learning model, a performance metric for each recommended content item of the plurality of recommended content items by performing a plurality of operations iteratively for a predefined threshold number of attempts, the plurality of operations comprising: facilitating, by the system, a display of at least a set of recommended content items from the plurality of recommended content items on UIs of electronic devices for a plurality of related content viewers, accessing and aggregating, by the system, content viewer interaction data from the plurality of related content viewers in the viewer cohort in relation to the plurality of recommended content items, the content viewer interaction data comprising a popularity index of each recommended content item from a plurality of recommended content items that is being served on a plurality of electronic devices associated with the plurality of related content viewers by a digital platform server, computing, by the system, a score for each recommended content item of the plurality of recommended content items based, at least in part, on the content viewer interaction data and a scoring policy, ranking, by the system, each recommended content item of the plurality of recommended content items based, at least in part, on the score related to the each recommended content item and a ranking policy, rearranging, by the system, the plurality of ranked recommended content items at a predefined time interval based, at least in part, on a shuffling policy, facilitating, by the system, the display of a plurality of rearranged recommended content items on User Interfaces (UIs) of the plurality of electronic devices associated with the plurality of related content viewers, and computing and updating, by the system, the content viewer interaction data from the plurality of related content viewers based, at least in part, on monitoring interactions of the plurality of related content viewers with the plurality of rearranged recommended content items; and determining, by the system, an optimal content sequence for displaying the plurality of recommended content items based, at least in part, on the performance metrics for the plurality of recommended content items, the optimal content sequence indicating a sequence for displaying each recommended content item of the plurality of recommended content items on an electronic device of the content viewer.