DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 3/11/2026 has been entered.
Claim Status
Claims 1-20 are pending in this Office Action.
Claims 1, 4, 11, 14, and 20 are amended.
Response to Arguments
Applicant’s arguments with respect to claims 1, 11, and 20 have been fully considered, but are not persuasive.
Applicant argues a careful review of the other reference cited by the Examiner fails to teach or suggest the first set of user interactions comprises interactions of a current user with a plurality of items via the first GUI during the current streaming session.
The Examiner respectfully disagrees. Jia teaches the session context comprises a plurality of data elements recorded by the session analysis module 106a. The recorded data elements typically comprise content interaction data relating to the digital content objects with which the user is interacting during the session, and interaction rating data associated with the content interactions where the user provided a rating to one or more of the digital content objects with which the user interacted (col. 7, line 56 through col. 8, line 26 and col. 8, lines 52-55, Fig. 2).
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-20 are rejected under 35 U.S.C. 103 as being unpatentable over Swaminathan et al. (US 2017/0251258) in view of Jia (US 10,244,286).
Regarding claims 1, 11, and 20, Swaminathan teaches: A computer-implemented method for generating recommendations for users during streaming sessions [(abstract)], the method comprising:
generating a first plurality of feature values for a first plurality of features based on a first set of user interactions that have occurred via a first graphical user interface (GUI) during a current streaming session [determining features including session consumption data and ratings, such as session progress values and ratings via a GUI (par. 18-20, 27, and 30, Fig. 3)]
aggregating the first plurality of feature values and a first feature value associated with metadata for a first item to generate a first feature vector [combining the consumption data and ratings with other features, such as video item specific attribute data into a feature vector (par. 23, 29, 34, 37, 45-46, Fig. 5 and 6)]
executing a model on at least the first feature vector to generate a first score for the first item [recommendation engine uses a feature vector and a model, such as a factorization machine, to predict user interest in the video, such as a rating for the video (par. 20, 28, 37, 41, and 44-46)]
aggregating the first plurality of feature values and a second feature value associated with metadata for a second item to generate a second feature vector [combining the consumption data and ratings with other features, such as video item specific attribute data for another video option, into a feature vector (par. 23, 29, 34, 37, 45-46, Fig. 5 and 6)]
executing the trained machine learning model on at least the second feature vector to generate a second score for the second item [recommendation engine uses a model, such as a factorization machine, to predict user interest in second video option, such as a rating for the video option (par. 20, 28, 37, 41, and 44-46)]
generating a first recommendation based on the first score and the second score [recommendation engine predicts the interest that the user will have in multiple video options in that particular session context using the model and provides a recommendation based on those predictions (par. 44, Fig. 7)] and
displaying the first recommendation within the first GUI [providing the recommendation to the computing device of the user in the particular session context (par. 17, 44, and 57, Fig. 3)].
Swaminathan does not explicitly disclose: the first set of user interactions comprises interactions of a current user with a plurality of items via the first GUI during the current streaming session; and the model is a trained machine learning model.
Jia teaches: the first set of user interactions comprises interactions of a current user with a plurality of items via the first GUI during the current streaming session [The session context comprises a plurality of data elements recorded by the session analysis module 106a. The recorded data elements typically comprise content interaction data relating to the digital content objects with which the user is interacting during the session, and interaction rating data associated with the content interactions where the user provided a rating to one or more of the digital content objects with which the user interacted (col. 7, line 56 through col. 8, line 26 and col. 8, lines 52-55, Fig. 2)]
the model is a trained machine learning model [a trained machine learning model (col. 12, lines 3-5, Fig. 2, 6, and 7)].
It would have been obvious to one of ordinary skill in the art, having the teachings of Swaminathan and Jia before the effective filing date of the claimed invention to modify the computer-implemented method of Swaminathan by incorporating the teaching of Jia such that the first set of user interactions comprises interactions of a current user with a plurality of items via the first GUI during the current streaming session; and the model is a trained machine learning model. The motivation for doing so would have been to enable real-time recommendations of content that a user will predictably interact with (Jia – col. 1, line 66 through col. 2, line 4) and to generate recommendations in a data-efficient manner, with less computational resources (Jia – col. 2, lines 4-10). Therefore, it would have been obvious to combine the teachings of Swaminathan and Jia to obtain the invention as specified in the instant claim.
Regarding claims 2 and 12, Swaminathan and Jia teach the computer-implemented method of claim 1, Jia further teaches: generating the first recommendation comprises computing a ranked item list based on the first score and at least the second score [ranking the content according to the highest rating in a list (col. 12, lines 14-21)].
Regarding claim 3, Swaminathan and Jia teach the computer-implemented method of claim 1, Swaminathan further teaches: computing the first feature value based on first metadata associated with the first item [determining item feature information from genre or other video-specific or item-specific context (par. 29 and 34-35)].
Regarding claims 4 and 14, Swaminathan and Jia teach the computer-implemented method of claim 1, Swaminathan further teaches: generating the first plurality of feature values comprises computing a third feature value based on at least one of the first set of user interactions, preference-related data associated with the current user, or a snapshot of user interactions associated with both the current user and a previous streaming session [computing completion values or ratings of the user, preferences (par. 29-30), historical session progress data (par. 18)].
Regarding claim 5, Swaminathan and Jia teach the computer-implemented method of claim 1, Swaminathan further teaches: the first set of user interactions indicates at least one of a first amount of time spent hovering over a third item, a selection of the third item, or a second amount of time spent streaming the third item [a percentage of time of the video that the user spent watching the video (par. 18 and 29)].
Regarding claim 6, Swaminathan and Jia teach the computer-implemented method of claim 1, Swaminathan further teaches: monitoring the first GUI during the current streaming session to determine the first set of user interactions [collecting session context information including ratings input by the user and that they are using an application (par. 19, 27, and 30)].
Regarding claims 7 and 17, Swaminathan and Jia teach the computer-implemented method of claim 1, Swaminathan further teaches: generating a snapshot of user interactions based on the first set of user interactions, and storing the snapshot in a first memory [collecting and compiling historical session context information, including the user rating, and storing a record of the information, such as in a dataset (par. 19 and 25)].
Regarding claim 8, Swaminathan and Jia teach the computer-implemented method of claim 7, Swaminathan and Jia further teach: during a different streaming session that is subsequent to the current streaming session, generating a second plurality of feature values for the first plurality of features based on the snapshot; executing the trained machine learning model on at least the second plurality of feature values and the first feature value to generate a third score; generating a second recommendation based on the first score and at least a fourth score that is associated with the second item; and displaying the second recommendation within a second GUI [Swaminathan - a user may watch a video three times during different sessions and each of the sessions can be accounted for in the feature vector (par. 38, Fig. 6). The model may be executed on the feature vector and a recommendation is generated and displayed (par. 44-46). Jia – executing the trained machine learning model (col. 12, lines 3-5, Fig. 2, 6, and 7)].
Regarding claims 9 and 18, Swaminathan and Jia teach the computer-implemented method of claim 1, Jia further teaches: performing one or more machine learning operations on a machine learning model based on preference-related data associated with a plurality of users, a plurality of snapshots of user interactions associated with a plurality of historical streaming sessions, and a plurality of items streamed during the plurality of historical streaming sessions to generate the trained machine learning model [using machine learning techniques to train a machine learning model based on ratings provided by users, prior content interactions in which prior users of the system interacted with one or more digital content objects (col. 6, lines 34-52, col. 8, lines 26-40 and 52-66, and col. 11, lines 22-41)].
Regarding claim 10, Swaminathan and Jia teach the computer-implemented method of claim 9, Jia further teaches: the preference-related data includes at least one of a streaming history or one or more user ratings of one or more items [interaction rating data comprise numeric values corresponding to a preference for the associated digital content objects (col. 6, lines 34-52)].
Regarding claim 13, Swaminathan and Jia teach the one or more non-transitory computer readable media of claim 11, Swaminathan further teaches: generating the first plurality of feature values comprises computing a third feature value based on at least one of the first set of user interactions, a country associated with the current streaming session, a user identifier associated with the current streaming session, or a language associated with the current streaming session [computing completion values or ratings of the user, preferences (par. 29-30) or language (par. 31)].
Regarding claim 15, Swaminathan and Jia teach the one or more non-transitory computer readable media of claim 14, Swaminathan further teaches: the preference-related data includes at least one of a streaming history or one or more user ratings of one or more items [The recommendation module 30 uses the historical content consumption data, ratings data, and/or other information about the interest of many users in many videos to predict individual user's preferences (par. 29-30)].
Regarding claim 16, Swaminathan and Jia teach the one or more non-transitory computer readable media of claim 11, Swaminathan and Jia further teach: computing a second plurality of feature values for the first plurality of features based on the first set of user interactions and at least one other user interaction that occurs via the GUI subsequent to the displaying of the first recommendation within the GUI; executing the trained machine learning model on at least the second plurality of feature values and the first feature value to generate a third score; generating a second recommendation based on the third score and at least a fourth score that is associated with both the second item and the at least one other user interaction; and displaying the second recommendation instead of the first recommendation within the GUI [Swaminathan - a user may watch a video three times during different sessions and each of the sessions can be accounted for in the feature vector (par. 38, Fig. 6). The model may be executed on the feature vector and a recommendation is generated and displayed (par. 44-46). Jia – executing the trained machine learning model (col. 12, lines 3-5, Fig. 2, 6, and 7)].
Regarding claim 19, Swaminathan and Jia teach the one or more non-transitory computer readable media of claim 11, Swaminathan further teaches: the first item comprises a movie, an episode of a television show, or a podcast [movie (Fig. 6)].
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Alexander Boyd whose telephone number is (571)270-0676. The examiner can normally be reached Monday - Friday 9am-5pm PST.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Benjamin Bruckart can be reached at 571-272-3982. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/ALEXANDER BOYD/Examiner, Art Unit 2424
Prosecution Insights