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 .
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 14-15 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
With respect to claim 14, it is unclear what comprises “a recommendation model…and a user-item matrix....” regarding to the limitation “transmitting, by the electronic device, a recommendation request to the server comprising a recommendation model pre-trained based on past item histories of multiple users and a user-item matrix predicted by the recommendation model” as recited in lines 3-5 of claim 14.
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
The following is a quotation of the first paragraph of pre-AIA 35 U.S.C. 112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
Claims 14-15 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the enablement requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to enable one skilled in the art to which it pertains, or with which it is most nearly connected, to make and/or use the invention. According to the specification at paragraph 0036 of the present application, the server 101 includes a recommendation model 120 pre-trained based on past item histories of multiple users and a user-item matrix 107 predicted by the recommendation model 120. Accordingly, the claimed subject matter “transmitting, by the electronic device, a recommendation request to the server comprising a recommendation model pre-trained based on past item histories of multiple users and a user-item matrix predicted by the recommendation model” of claim 14 was not described in the specification in such a way as to enable one skilled in the art to which it pertains, or with which it is most nearly connected, to make and/or use the invention. (Emphasis added).
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 1-13 are rejected under 35 U.S.C. 103 as being unpatentable over Dektyarev et al. (US 20240012861 A1) in view of Giggs et al. (US 7,685,232 B2).
Regarding claim 1, Dektyarev teaches a recommendation system (100 – see FIG. 1) comprising:
a server (112 within system 100) comprising:
a recommendation model pre-trained based on past item histories of multiple users (a model within recommendation system 180 trained based on data indicative of previous interactions between a plurality of users and a plurality of digital items – see 0122-0124, 0133, 0171, 0183); and
a user-item matrix predicted by the recommendation model (user-item matrix calculated by the model within recommendation system 180 – see 0133, 0170, 0171, 0185, 0187);
wherein the server is configured to, based on receiving a recommendation request (the server 112 receives a request 150 for content recommendation from electronic device 104 – see 0093, 0119, 0198); and
an electronic device (104 within system 100) configured to:
transmit the recommendation request (e.g., request 150 - see 0093, 0119, 0198); and
provide a target user with a recommended item using a dedicated recommendation model with the first submatrix based on an item history of the target user at a time of transmitting the recommendation request (provide to user 102 content recommendation using a given recommendation model according to user-item interaction data associated with the user 102 – see 0091, 0125, 0132, 0135, 0179-0181).
Dektyarev lacks to teach extracting a first submatrix for approximating the recommendation model to the dedicated recommendation model from the predicted user-item matrix and transmit the first submatrix; and the dedicated recommendation model trained
based on a second submatrix generated by reprocessing the first submatrix based on user’s item. Gibbs teaches systems and methods for making recommendations of items personally tailored to individual anonymous users using Collaborative Filtering model that employs matrix factorization. Particularly, the server derives from the rating matrix a user feature matrix and an item feature matrix using matrix factorization by approximation. The server then sends the item feature matrix to a client device. The client device, after receiving the item feature matrix from the server, calculates client’s user feature vector using the item feature matrix and client’s rating vector, and makes personalized content recommendations based on own user feature vector and item feature matrix. See abstract; col. 8, line 47 to col. 9, line 30; col. 11, lines 6-21; col. 11, line 63 to col. 12, line 18; col. 16, lines 62-67. It would have been obvious to one of ordinary skill in the art at the time invention was made to modify Dektyarev by extracting a first submatrix for approximating the recommendation model to a given recommendation model from the user-item matrix and transmit the first submatrix; and the dedicated recommendation model trained based on a second submatrix generated by reprocessing the first submatrix based on user’s item as taught or suggested by Gibbs to locally make personalized recommendations to user while preserving user privacy by the client device.
Regarding claim 2, Dektyarev in combination with Gibbs further teaches a memory configured to store the recommendation model and the predicted user-item matrix representing a degree of interactions between the multiple users and multiple items predicted by the recommendation model (a memory stores the model and data representing interactions between the users and items - see Dektyarev: 0118, 0122-0129, 0133, 0137, 0138, 0160, 0161, 0170); a processor configured to extract the first submatrix from the predicted user-item matrix; and a communication interface configured to receive the recommendation request and transmit the first submatrix to the electronic device (the server, via the processor, aggregates the rating vectors received from the client devices into a rating matrix, factorizes the rating matrix into a user feature matrix and an item feature matrix, and sends, via a communication interface, the item feature matrix to the client device – see Gibbs: FIGs 4, 5B; col. 19, line 20 to col. 20, line 20).
Regarding claim 3, Dektyarev in combination with Gibbs further teaches that wherein the processor is configured to: identify first similar users similar to the target user corresponding to the electronic device among the multiple users (determining a similarity between users included a target user and other users. For example, users 1 and 3 have the similar user-item interaction with a value of 1 indicative of “like” on track 3 – see Dektyarev: FIG. 2, 0185, 0186; see Gibbs: Table 3; col. 15, lines 35-38); identify candidate items to be provided to the target user among the multiple items (determines a set of media contents for content recommendation – see Dektyarev: 0135, 0198; Gibbs: col. 12, lines 9-11); and extract the first submatrix corresponding to the first similar users and the candidate items from the predicted user-item matrix (for example, the training set 250 includes a portion of matrix 260 obtained from interaction matrix 200 corresponding to the similarity of users and the set of media contents for content – see Dektayarev: FIG. 2; 0190; the server derives from the rating matrix a user feature matrix and an item feature matrix using matrix factorization by approximation and sending the media feature matrix to the client device via proxy - see Gibbs: see col. 11, lines 6-14).
Regarding claim 4, Dektyarev in combination with Gibbs further teaches the features of identifying the first similar users based on at least one of additional information of the target user or a past item history of the target user, wherein the additional information of the target user comprises one or more of a gender, an age, an occupation, a residence, or an interest of a user (identifying the similarity between users based on past item history and additional information about the users included the target user – see Dektyarev: 0185, 0186; Gibbs: col. 15, lines 35-38; the additional information about the users includes user-profile data associated with respective users of the recommendation system such as, but not limited to: name, age, gender, user-selected types of digital content that he/she desires, and the like. See Dektyarev: 0159).
Regarding claim 5, Dektyarev in combination with Gibbs further teaches the features of identifying the first similar users based on a result of comparing similarities between items included in the predicted user-item matrix and items included in the past item history of the target user (identifying the similarity between users according to analyzing similarities among items included in the item history and the predicted user-item matrix – see Gibbs: col. 12, lines 30-64; Dektyarev: 0022, 0051, 0186).
Regarding claim 6, Dektyarev further teaches the features of identifying the first similar users based on a result of comparing similarities between users included in the predicted user-item matrix and the additional information of the target user (identifying the similarity between users according to analyzing similarities between users included in the predicted user-item matrix and the data associated with the target user, e.g., different types of user events such as interaction time, purchased, ordered or downloaded a given item, etc. – see 0159-0169, 0185, 0186).
Regarding claim 7, Dektyarev in combination with Gibbs teaches that wherein the processor is configured to identify the candidate items based on past item histories of the first similar users (determines a subset of digital content for recommendation associated with user-item interactions of the similar users – see Dektyarev: 0135, 0198; Gibbs: col. 12, lines 9-11), recommended items predicted for the first similar users by the recommendation model (the subset of digital items are used as training features for generating the given ISDT sub-model such that previous user-item interactions between the plurality of users and the subset of digital items are used as values for respective training features - see Dektyarev: 0193), and items selected from among the multiple items other than the recommended item (other training set may be generated for other digital item where user-interaction data for that other digital item is used as other training target set, and other subset of digital items are used as other training features -see Dektyarev: 0197).
Regarding claim 8, Dektyarev in combination with Gibbs further teaches that a communication interface (within 104 – see FIG. 1) configured to transmit a recommendation request to a server (device 104 sends a request 150 to a server 112 – see Dektyarev: FIG.1, 0093, 0119, 0198) in response to the recommendation request; a processor configured to extract second similar users from the first submatrix based on an item history of the target user at the time of transmitting the recommendation request, extract the second submatrix corresponding to the second similar users from the first submatrix, and provide the target user with a recommended item using the dedicated recommendation model trained based on the second submatrix; and a memory configured to store the dedicated recommendation model (provide to user 102 content recommendation using a given recommendation model according to user-item interaction data associated with the user 102 – see Dektyarev: 0091, 0125, 0132, 0135, 0179-0181; the client device, after receiving the item feature matrix from the server, calculates client’s user feature vector using the item feature matrix and client’s rating vector, makes personalized content recommendations based on own user feature vector and item feature matrix, and stores the trained recommendation model. See Gibbs: abstract; col. 8, line 47 to col. 9, line 30; col. 11, lines 6-21; col. 11, line 63 to col. 12, line 18; col. 16, lines 62-67; col. 19, lines 29-39).
Regarding claim 9, Dektyarev teaches an electronic device comprising:
a communication interface (within 104 – see FIG. 1) configured to transmit a recommendation request to a server (device 104 sends a request 150 to a server 112 - see FIG.1, 0093, 0119, 0198) in response to the recommendation request; a processor configured to provide the target user with a recommended item using a dedicated recommendation model trained based on an item history of the target user at a time of transmitting the recommendation request (provide to user 102 content recommendation using a given recommendation model according to user-item interaction data associated with the user 102 – see 0091, 0125, 0132, 0135, 0179-0181). Dektyarev lacks to teach that the communication interface of electronic device receives a first submatrix for approximating a recommendation model stored in the server to a dedicated recommendation model for a target user of the electronic device from the server; the processor of the electronic device generates a second submatrix corresponding to second similar users extracted from the first submatrix based on an item history of the target user, using the dedicated recommendation model trained based on the second submatrix; and a memory of the electronic device configured to store the trained dedicated recommendation model. Gibbs teaches systems and methods for making recommendations of items personally tailored to individual anonymous users using Collaborative Filtering model that employs matrix factorization. Particularly, the server derives from the rating matrix a user feature matrix and an item feature matrix using matrix factorization by approximation. The server then sends the item feature matrix to a client device. The client device, after receiving the item feature matrix from the server, calculates client’s user feature vector using the item feature matrix and client’s rating vector, makes personalized content recommendations based on own user feature vector and item feature matrix, and stores the trained recommendation model. See abstract; col. 8, line 47 to col. 9, line 30; col. 11, lines 6-21; col. 11, line 63 to col. 12, line 18; col. 16, lines 62-67; col. 19, lines 29-39. It would have been obvious to one of ordinary skill in the art at the time invention was made to modify Dektyarev by including a client device for receiving a first submatrix for approximating a recommendation model stored in the server to a dedicated recommendation model for a target user of the electronic device from the server; generating a second submatrix corresponding to second similar users extracted from the first submatrix based on an item history of the target user, using the dedicated recommendation model trained based on the second submatrix; and storing the trained dedicated recommendation mode as taught or suggested by Gibbs to locally make personalized recommendations to user while preserving user privacy by the client device.
Regarding claim 10, Dektyarev in combination with Gibbs further teaches that the processor is configured to extract the second similar users from among first similar users based on similarities between the item history of the target user at the time of transmitting the recommendation request and candidate items included in the first submatrix (for example, the training set 250 includes a portion of matrix 260 obtained from interaction matrix 200 corresponding to the similarity of users and the set of media contents for content – see Dektayarev: FIG. 2; 0190; receiving the media feature matrix from the server - see Gibbs: see col. 11, lines 6-14; transmitting a set of media contents for content recommendation – see Dektyarev: 0135, 0198; Gibbs: col. 12, lines 9-11).
Regarding claim 11, Dektyarev in combination with Gibbs further teaches that wherein the processor is configured to extract the second submatrix corresponding to the second similar users from the first submatrix (extracting a portion of matrix 260 obtained from interaction matrix 200 corresponding to the similarity of users and the set of media contents for content – see Dektayarev: FIG. 2; 0190; receiving the media feature matrix from the server - see Gibbs: see col. 11, lines 6-14).
Regarding claim 12, Dektyarev in combination with Gibbs further that the processor is configured to provide the target user with a recommended item corresponding to the time of transmitting the recommendation request and an explanation corresponding to recommended items using the trained dedicated recommendation model (for example, the electronic device generates the request 150 in response to the user 102 providing an explicit indication of the user desire to receive a digital content recommendation – see Dektyarev: 0093; makes personalized content recommendations based on own user feature vector and item feature matrix – see Gibbs: col. 11, lines 63-67).
Regarding claim 13, Dektyarev in view of Gibbs teaches that wherein the dedicated recommendation model is trained by a machine learning technique based on the second submatrix (trained by an algorithm based on the client feature vector and media feature matrix – see Gibbs: col. 11, lines 15-67).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Flanagan (US 20210092203 A1) teaches client machine is configured for an efficient machine learning process that has access to the client data of all clients connected to a server.
Maeda (US 20220215454 A1) teaches generating a user vector that represents an rating state of each of the users based on the ratings for the plurality of objects; generating neighborhood candidate users by excluding a user that has a user vector same as a user vector of a certain user from the plurality of users; selecting a certain number of neighborhood users from the neighborhood candidate users based on similarity of the user vector; and determining a recommended object based on the ratings of each of the neighborhood users.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to NGOC K VU whose telephone number is (571)272-7306. The examiner can normally be reached Monday & Thursday: 10AM-6:30PM EST; Tuesday, Wednesday & Friday: out of office.
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, NATHAN FLYNN can be reached at 571-272-1915. 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.
/NGOC K VU/Primary Examiner, Art Unit 2421
Prosecution Insights