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 .
Response to Arguments
Applicant's arguments with respect to 35 U.S.C. 103 rejection of claims 1 and 17 have been considered and found persuasive, and the rejection has been withdrawn. See detailed reason for allowance below. The claims would be allowable if the Applicant can overcome the 101 Abstract Idea set forth.
Applicant's arguments with respect to 35 U.S.C. 101 Abstract Idea in regard to claims 1, 17 and 20 have been considered, however are not found to be persuasive due to the following reasons. Claims 1, 17 and 20 still recite collecting website information, evaluating whether an entity falls within a prohibited category and deciding whether that entity may use messaging services. These are information analysis and service-access decisions. Although the actual ML calculations may not be performed mentally, using a computer to automate the underlying evaluation does not by itself remove the abstract idea.
The communication control limitations also do not clearly establish a practical application. Preventing transmission, denying service, and updating a database state the results of the classification but do not explain a technical manner of controlling the communications network. Likewise, claim 1’s multistage training, threshold based relabeling, and model selection are not tied to an identified improvement in model accuracy, efficiency, architecture, or computer operation. Calling the process “real time,” “closed loop” or “system level” does not supply the missing technical improvement.
Applicant's arguments with respect to 35 U.S.C. 102 in regard to claim 20 has been considered but are moot due to new grounds of rejection necessitated by amendments. See detailed rejection below.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-2, 5-17 and 19-23 rejected under 35 U.S.C. 101 Abstract Idea.
Under step one of the Alice/Mayo analysis, claims 1, 17 and 20 are directed to the abstract idea of collecting and evaluating information to make a compliance decision. The claims gather website text, label or classify an entity, determine whether the entity falls within a prohibited category, and decide whether the entity may use messaging services. These are information analysis and rule-based decision-making activities that are ordinarily be performed though human judgement, even though the claims automate them using computers and ML models.
The additional limitations do not integrate the abstract idea into a specific technological improvement. Claim 1 recites multistage training, confidence thresholds, relabeling, and model selection; claim 17 recites generating labeled training data and training a production model; and claim 20 recites converting website information into a vector and computing a category score. However, the claims do not specify a new model architecture, training algorithm, computer structure, or messaging network operation. They primarily use ML as a tool for performing the abstract classification and compliance decision more efficiently.
Under step two of Alice/Mayo, the claims do not contain an inventive concept sufficient to transform the abstract idea into patent-eligible subject matter. The processors, compliance system, websites, ML models, and database are recited according to their ordinary functions. Blocking communications, denying messaging access, and storing classifications merely implement the resulting compliance decision. Therefore, claims 1, 17 and 20 are rejected under 101 Abstract Idea.
The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception because the claims are (i) mere instructions to implement the idea on a computer, and/or (ii) recitation of generic computer structure that serves to perform generic computer functions that are well-understood, routine, and conventional activities previously known to the pertinent industry. Viewed as a whole, these additional claim element(s) do not provide meaningful limitation(s) to transform the abstract idea into a patent eligible application of the abstract idea such that the claim(s) amounts to significantly more than the abstract idea itself. Therefore, the claim(s) are rejected under 35 U.S.C. 101 as being directed to non-statutory subject matter. There is further no improvement to the computing device.
Dependent claims 2, 5-16, 19 and 21-23 further recite an abstract idea performable by a human and do not amount to significantly more than the abstract idea as they do not provide steps other than what is conventionally known.
Claim 2, Saving the second dataset after a threshold check is just storing the result of an abstract scoring decision, which is routine computer activity, not a technical improvement.
Claim 5: Comparing model predictions for a match is just checking whether results agree, which is an abstract mental comparison.
Claim 6: Building a third dataset from matching predictions and retraining just repeats the same abstract relabel-and-retrain idea in another round.
Claim 7: Identifying disputed-category websites is just flagging disagreements, which is another abstract judgment step.
Claim 8: Relabeling disputed websites and training again just adds another round of abstract data review and model retraining.
Claim 9: Finding low-count categories and relabeling to boost them is just statistical data balancing, not a concrete technological improvement.
Claim 10: Defining a low-count category as less than 10% is just adding a numerical rule or threshold to the same abstract process.
Claim 11: Identifying edge data points below 50% confidence and retraining is just thresholding uncertainty and reusing the data, which is still abstract data analysis.
Claim 12: Saying the production model is decision-tree based only picks a conventional ML model type, not a specific technical improvement.
Claim 13: Saying the production model is gradient boosting likewise only picks a conventional ML model type, not a patent-eligible improvement.
Claim 14: Fetching and parsing website text, even with Beautiful Soup, is just routine data gathering and preprocessing around the abstract idea.
Claim 15: Vectorizing parsed text is just turning data into a mathematical representation for analysis, which stays abstract.
Claim 16: Creating a vector and computing a category score is the core abstract scoring/classification step itself, not a practical technological application.
Claim 21: Merely telling the method to how fast to complete the process. Does not improve a particular application.
Claim 22: Merely telling the method to how fast to complete the process. Does not improve a particular application.
Claim 23: Merely telling the method to how fast to complete the process. Does not improve a particular application.
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.
Claim(s) 20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Chen et al. (“Automatic detection of pornographic and gambling websites based on visual and textual content using a decision mechanism”; 2020) in view of Chasin et al. (US 7,953,814).
Claim 20,
Chen teaches a computer-implemented method of classifying a prohibited entity comprising (Chen states “our system (PG-VTDM) was designed to automatically detect and filter gambling and pornographic websites” [3.1] [Figs. 1-2]; Chen further states “this module combines multiple ML models for classifying the given website” [3.1.3]):
automatically gathering, by the compliance check system, text from the unlabeled entity's website (Chen obtains “the website source code… via web crawling and scrapping” [3.1] [Fig. 2]; the source code is crawled “by an automatic web crawling procedure” [3.1.1]);
creating a vector from an unlabeled website of the entity (Chen states that “Doc2Vec can map text or sentences to dense real-valued vectors,” and “the preprocessed document text of the website can be mapped into a k-dimensional space vector” [3.3.1]);
computing, using a trained machine learning production model, a score for at least one category for the unlabeled website based on the vector, wherein the at least one category comprises at least one prohibited category (Chen uses website textual feature vectors “to train the SVM,” which predicts subsequent test websites [3.4]; Pttest is “the probability” that a test website is pornographic or gambling related; Pfusion represents the final prohibited-category probability [3.4] [eqs. 4-5]);
The difference between the prior art and the claimed invention is that Chen does not explicitly teach receiving, by a compliance check system, a request from an unlabeled entity to send electronic communications; preventing, by the compliance check system, transmission of the electronic communications and denying access to at least one messaging service when the unlabeled entity is classified as a prohibited entity; and updating a compliance database to store the classification and associated score, and using the stored classification to determine whether to permit the entity to send future electronic communications.
Chasin teaches receiving, by a compliance check system, a request from an unlabeled entity to send electronic communications (Chasin teaches SRG 205 operates by “servicing and responding to SMTP requests… issued by subscribes” [col. 7 lines 23-37] [Fig. 2]; a subscriber is “an individual or company” using services including “text messaging” [col. 5 lines 33-43]);
preventing, by the compliance check system, transmission of the electronic communications and denying access to at least one messaging service when the unlabeled entity is classified as a prohibited entity (Chasin teaches subscriber activity is classified as “indicative of spam, virus, or worm related activity” [claim 1]; reputation determines an action such as “deny” for the current message and whether “subscriber privileges are reduced” [col. 7 lines 1-16]; subsequent messages may be “prohibited from reaching their intended destination [claim 1]; the teaching applies to “mobile messages” [col. 4 lines 16-21]); and
updating a compliance database to store the classification and associated score (Chasin teaches “continuously updates and stores reputation and message sending behavior profiles within a reputation and behavior database 220 for each subscriber” [col. 7 lines 38-49] [Fig. 2]; stored profile data include suspected spam/virus indicators and an “average Bayesian spam filter probability score” [col. 6 lines 13-67]), and
using the stored classification to determine whether to permit the entity to send future electronic communications (Chasin teaches stored reputation data determine “a long-term subscriber action for all subsequent email messages” [col. 3 lines 30-50]; when the score fails a threshold, “any subsequent outbound messages” are prohibited from reaching their destination [claim 1]).
Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the teachings of Chen with teachings of Chasin by modifying the automatic detection of pornographic and gambling websites based on visual and textual content using a decision mechanism as taught by Chen to include receiving, by a compliance check system, a request from an unlabeled entity to send electronic communications; preventing, by the compliance check system, transmission of the electronic communications and denying access to at least one messaging service when the unlabeled entity is classified as a prohibited entity; and updating a compliance database to store the classification and associated score, and using the stored classification to determine whether to permit the entity to send future electronic communications as taught by Chasin for the benefit of allowing service providers filter outbound subscriber email traffic for spam and/or worm-oriented messages using one or more of individual subscriber email sending reputation, real-time analysis of message content and behavior-based anomaly detection (Chasin [col. 1 lines 23-30]).
Allowable Subject Matter
Claims 1-2, 5-16, 17 and 19 would be allowable if the Applicant can overcome the 101 Abstract Idea set forth.
The following is a statement of reasons for the indication of allowable subject matter: Chen et al. (“Automatic Detection of Pornographic and Gambling Websites Based on Visual and Textual Content Using a Decision Mechanism”; 2020) in view of Chasin et al. (US 7,953,814):
For claim 1: Chen teaches collecting website source code through web crawling, extracting textual features with Doc2Vec, and representing website text as fixed-length vectors ([3.2-3.3.1] [pgs. 7-8]). Chen then trains different ML models, including SVM, random forest, and logistic regression, to classify websites as pornographic, gambling, or normal and calculates classification probabilities that are compared with a threshold ([3.1.3] [3.4] [pgs. 7, 11-12] [Fig. 4] [eqs. 4-5]). Chen also evaluates different Doc2Vec models and selects the best performing settings based on accuracy and related performance measures ([4.3.1] [pgs. 13-14] [Tables 3-5]). When a website is classified as prohibited, the system blocks access and adds its URL to a blacklist for faster future decisions ([3.1] [pgs. 5-6] [3.4] [pg. 12] [Figs. 1-2]). Thus, Chen teaches much of the claimed website data gathering, model training, classification, scoring, thresholding, blocking and storing for future use.
Chasin teaches monitoring subscriber messages for abuse, storing subscriber profiles and reputation scores, and using that information to control current and future communications. When abuse is detected, the system may deny messages, reduce privileges, or block later messages, including email, SMS and MMS ([col. 3 lines 30-57] [col. 4 lines 16-21] [col. 7 lines 1-56] [Figs. 1-2] [claims 1, 4-6]).
The difference between the prior art and the claimed invention is that Chen nor Chasin explicitly teach selecting a production model from the plurality of candidate label-assist models based on the candidate label-assist model predicting categories having the highest confidence scores; preventing, by the compliance check system, transmission of the electronic communications and denying access to at least one messaging service when the unlabeled entity is classified as a prohibited entity.
Therefore, it would not have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the teachings of Chen and Chasin to include selecting a production model from the plurality of candidate label-assist models based on the candidate label-assist model predicting categories having the highest confidence scores; preventing, by the compliance check system, transmission of the electronic communications and denying access to at least one messaging service when the unlabeled entity is classified as a prohibited entity.
For claim 17: Chen teaches collecting website source code through web crawling, extracting textual features with Doc2Vec, and representing website text as fixed-length vectors ([3.2-3.3.1] [pgs. 7-8]). Chen then trains different ML models, including SVM, random forest, and logistic regression, to classify websites as pornographic, gambling, or normal and calculates classification probabilities that are compared with a threshold ([3.1.3] [3.4] [pgs. 7, 11-12] [Fig. 4] [eqs. 4-5]). Chen also evaluates different Doc2Vec models and selects the best performing settings based on accuracy and related performance measures ([4.3.1] [pgs. 13-14] [Tables 3-5]). When a website is classified as prohibited, the system blocks access and adds its URL to a blacklist for faster future decisions ([3.1] [pgs. 5-6] [3.4] [pg. 12] [Figs. 1-2]). Thus, Chen teaches much of the claimed website data gathering, model training, classification, scoring, thresholding, blocking and storing for future use.
Chasin teaches monitoring subscriber messages for abuse, storing subscriber profiles and reputation scores, and using that information to control current and future communications. When abuse is detected, the system may deny messages, reduce privileges, or block later messages, including email, SMS and MMS ([col. 3 lines 30-57] [col. 4 lines 16-21] [col. 7 lines 1-56] [Figs. 1-2] [claims 1, 4-6]).
The difference between the prior art and the claimed invention is that Chen nor Chasin explicitly teach save the entity name and the associated category to a compliance database and using the saved category for determining whether to prohibit the entity from using toll free numbers (TFN) and short messaging service (SMS) services.
Therefore, it would not have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the teachings of Chen and Chasin to include save the entity name and the associated category to a compliance database and using the saved category for determining whether to prohibit the entity from using toll free numbers (TFN) and short messaging service (SMS) services.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SHREYANS A PATEL whose telephone number is (571)270-0689. The examiner can normally be reached Monday-Friday 8am-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, Pierre Desir can be reached at 571-272-7799. 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.
SHREYANS A. PATEL
Primary Examiner
Art Unit 2653
/SHREYANS A PATEL/Examiner, Art Unit 2659