DETAILED ACTION
This office action is responsive to the above application filed on April 30th, 2024. Claims 1-20 are pending, all examined and rejected.
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 .
Information Disclosure Statement
The information disclosure statement (IDS) were submitted on April 30th, 2024. The submissions are in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Specification
The disclosure is objected to because of the following informalities: “[…] ping time data associated with the web browsing session, and temporal data associated with the web browsing session time.” Should be “[…] ping time data associated with the web browsing session, or temporal data associated with the web browsing session time.” [Page 14 ¶ 3]
Appropriate correction is required.
Claim Objections
Claim 8 is objected to because of the following informalities: “[…] ping time data associated with the web browsing session, and temporal data associated with the web browsing session time.” Should be “[…] ping time data associated with the web browsing session, or temporal data associated with the web browsing session time.”
Appropriate Correction is required.
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 1-3, 5, 6, 8-12, 14-18, and 20 are rejected under 35 U.S.C. 101 because the claimed inventions are directed to abstract ideas without significantly more.
With regard to Claims 1, 10, and 16:
Step 1, MPEP 2106.03
Claim 1. A computer implemented method [Statutory category of invention]
Claim 10. A non-transitory processor-readable storage medium having stored therein program code of one or more software programs [Statutory category of invention]
Claim 16. An apparatus [Statutory category of invention]
Step 2A, Prong 1, MPEP 2106.04
Claim 1, 10, and 16. Generating one or more predictions of one or more web pages, associated with the at least one web application, to be sought in connection with the web browsing session by processing at least a portion of the activity-related data using one or more statistical algorithms [The courts consider a mental process that can be practically performed in the human mind, or by a human using a pen and paper to be an abstract idea. MPEP 2106.04(a)(2), for example a person predicting which page a user will visit based on their browsing history]
Step 2A, Prong 2, MPEP 2106.04
Claim 1, 10, and 16. Obtaining activity-related data associated with at least one user device and at least one web application during a web browsing session [Insignificant extra solution activity of mere data gathering. MPEP 2106.05(g)]
Claim 1, 10, and 16. Automatically preloading, in connection with the at least one web application, at least one of the one or more web pages for use in the web browsing session by the at least one user device [Mere instructions to apply an exception. MPEP 2106.05(f)]
Claim 1. Wherein the method is performed by at least one processing device comprising a processor coupled to a memory. [Field of use and technological environment. MPEP 2106.05(h)]
Claim 10. A non-transitory processor-readable storage medium having stored therein program code of one or more software programs, wherein the program code when executed by at least one processing device causes the at least one processing device: to [...]. [Field of use and technological environment. MPEP 2106.05(h)]
Claim 16. At least one processing device comprising a processor coupled to a memory; the at least one processing device being configured: to [...]. [Field of use and technological environment. MPEP 2106.05(h)]
Step 2B, MPEP 2106.05
Claim 1, 10, and 16. Obtaining activity-related data associated with at least one user device and at least one web application during a web browsing session [Insignificant extra solution activity of mere data gathering. MPEP 2106.05(g)]
Claim 1, 10, and 16. Automatically preloading, in connection with the at least one web application, at least one of the one or more web pages for use in the web browsing session by the at least one user device [Well understood routine and conventional activity. MPEP 2106.05(d)]
Claim 1. wherein the method is performed by at least one processing device comprising a processor coupled to a memory. [Field of use and technological environment. MPEP 2106.05(h)]
Claim 10. A non-transitory processor-readable storage medium having stored therein program code of one or more software programs, wherein the program code when executed by at least one processing device causes the at least one processing device: to [...]. [Field of use and technological environment. MPEP 2106.05(h)]
Claim 16. At least one processing device comprising a processor coupled to a memory; the at least one processing device being configured: to [...]. [Field of use and technological environment. MPEP 2106.05(h)]
Dependent Claims
With regard to Claims 2, 11, and 17:
Step 1, MPEP 2106.03
Same statutory category as its respective parent.
Step 2A, Prong 1, MPEP 2106.04
Claims 2, 11, and 17. Generating one or more predictions of one or more web pages comprises processing the at least a portion of the activity-related data [The courts consider a mental process that can be practically performed in the human mind, or by a human using a pen and paper to be an abstract idea. MPEP 2106.04(a)(2), for example a person predicting which page a user will visit based on their browsing history]
Step 2A, Prong 2, MPEP 2106.04
Claims 2, 11, and 17. Using at least one bidirectional autoencoder LSTM neural network model. [Mere instructions to apply an exception. MPEP 2106.05(f)]
Step 2B, MPEP 2106.05
Claims 2, 11, and 17. Using at least one bidirectional autoencoder LSTM neural network model. [Mere instructions to apply an exception. MPEP 2106.05(f)]
With regard to Claims 3, 12, and 18:
Step 1, MPEP 2106.03
Same statutory category as its respective parent.
Step 2A, Prong 1, MPEP 2106.04
Claims 3, 12, and 18. Generating one or more predictions of one or more web pages comprises processing the at least a portion of the activity-related data [The courts consider a mental process that can be practically performed in the human mind, or by a human using a pen and paper to be an abstract idea. MPEP 2106.04(a)(2) ), for example a person predicting which page a user will visit based on their browsing history]
Claims 3, 12, and 18. Using at least one Markov chain. [The mathematical concepts abstract idea grouping is defined as mathematical relationships, mathematical formulas or equations, and mathematical calculations. MPEP 2106.04(a)(2), a Markov chain is a mathematical system]
With regard to Claims 5, 14, and 20.
Step 1, MPEP 2106.03
Same statutory category as its respective parent.
Step 2A, Prong 1, MPEP 2106.04
Same judicial exceptions as its respective parent.
Step 2A, Prong 2, MPEP 2106.04
Claims 5, 14, and 20. Automatically preloading at least one of the one or more web pages comprises pre-fetching and caching one or more application programming interfaces (APIs) associated with the at least one of the one or more web pages using one or more single page applications (SPAs). [Mere instructions to apply an exception, MPEP 2106.05(f)]
Step 2B, MPEP 2106.05
Claims 5, 14, and 20. Automatically preloading at least one of the one or more web pages comprises pre-fetching and caching one or more application programming interfaces (APIs) associated with the at least one of the one or more web pages using one or more single page applications (SPAs). [Well understood routine and conventional activity. MPEP 2106.05(d)]
With regard to Claims 6 and 15:
Step 1, MPEP 2106.03
Same statutory category as its respective parent.
Step 2A, Prong 1, MPEP 2106.04
Same judicial exceptions as its respective parent.
Step 2A, Prong 2, MPEP 2106.04
Claims 6 and 15. Wherein automatically preloading at least one of the one or more web pages comprises preloading the at least one of the one or more web pages using one or more server push techniques [Mere instructions to apply an exception, MPEP 2106.05(f)]
Step 2B, MPEP 2106.05
Claims 6 and 15. Wherein automatically preloading at least one of the one or more web pages comprises preloading the at least one of the one or more web pages using one or more server push techniques [Well understood routine and conventional activity. MPEP 2106.05(d)]
With regard to Claim 8:
Step 1, MPEP 2106.03
Same statutory category as its parent.
Step 2A, Prong 1, MPEP 2106.04
Wherein obtaining activity-related data comprises obtaining clickstream data associated with the at least one user device and the at least one web application during the web browsing session, wherein the clickstream data comprises one or more of data identifying a current web page, click location data, time spent on one or more web pages, web application rights associated with the at least one user device, user category associated with the at least one user device, user device location information, one or more user device profile settings [The courts consider a mental process that can be practically performed in the human mind, or by a human using a pen and paper to be an abstract idea. MPEP 2106.04(a)(2), the data that person sees in a user’s browsing behavior can be where they are clicking on the screen]
Step 2A, Prong 2, MPEP 2106.04
Wherein automatically preloading at least one of the one or more web pages comprises preloading the at least one of the one or more web pages using one or more server push techniques [Mere instructions to apply an exception, MPEP 2106.05(f)]
Step 2B, MPEP 2106.05
Wherein automatically preloading at least one of the one or more web pages comprises preloading the at least one of the one or more web pages using one or more server push techniques [Mere instructions to apply an exception, MPEP 2106.05(f)]
With regard to Claim 9:
Step 1, MPEP 2106.03
Same statutory category as its parent.
Step 2A, Prong 1, MPEP 2106.04
Same judicial exceptions as its respective parent.
Step 2A, Prong 2, MPEP 2106.04
Automatically training at least a portion of the one or more machine learning techniques using feedback related to the one or more predictions [Mere instructions to apply an exception, MPEP 2106.05(f)]
Step 2B, MPEP 2106.05
Automatically training at least a portion of the one or more machine learning techniques using feedback related to the one or more predictions [Well-Understood, Routine, Conventional Activity, MPEP 2106.05(d)]. Automatic training of machine learning techniques based on feedback is a well-known and conventional activity in the art as stated in Lee WO 2025100585 A1 “CONTINUOUS TRAINING METHOD IN LIMITED MOBILE ENVIRONMENT” published May 15, 2025: “[...] the model size must be small and the overall structure must be fixed, whereas conventional continuous learning-related technologies are not easy to apply because certain neural networks are added as the model is learned [...]” [3]. Hebenthal 20200320411 “SYSTEMS AND METHODS FOR ADAPTIVE TRAINING OF A MACHINE LEARNING SYSTEM PROCESSING TEXTUAL DATA” published October 8, 2020 also teaches automatically training machine learning techniques based on feedback from predictions as conventional activity. “Conventional systems and methods typically require continuous training of the ML system to account for such changes” [0016] and “[...] machine learning system that automatically adapts to changes in external factors that can influence the accuracy of results predicted by the ML system” [0018].
Claim Rejections - 35 USC § 102
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 the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claims 1, 4, 7, 10, 13, 16, and 19 are rejected under 35 U.S.C. 102(a)(1) and 35 U.S.C 102(a)(2) as being anticipated by US Patent 11182452B2, Su et. al, patented Nov 21, 2023.
Regarding Claims 1, 10, and 16, Su discloses a computer implemented method, apparatus, or non-transitory processor readable storage medium having stored program code of software programs for obtaining activity-related data associated with at least one user device and at least one web application during a web browsing session (Su Figure 8 and Col 8 ¶3 : “FIG. 8 is a diagram illustrating a method for learning web acceleration based on history, where web statistics that include the history of web page loads can be collected and input into an offline processing unit/server. Offline processing can be performed to obtain resource hints (also referred to herein as “resource data”) from the history.” Web history is activity related data that requires a user with a device to browse applications in a web session [Col 8 ¶3]),
Generating one or more predictions of one or more web pages, associated with the at least one web application, to be sought in connection with the web browsing session by processing at least a portion of the activity-related data using one or more statistical algorithms and one or more machine learning techniques (Su Figure 8, Col 8 ¶3, Col 1 ¶3, Col 11 ¶4, Col 12 ¶2, Col 12 ¶4, and Col 13 ¶3, and : "A level of resource readiness can be determined via a resource readiness processor based on the resource data (1120). Resource readiness for the web page can be prepared by a web acceleration processor based on the level of resource readiness, which can include preconnecting, preloading, or prefetching” [Col 13 ¶3]. Resource data includes confidence level which is the “probability that the web resource will be retrieved in a future visit by a user” [Col 1 ¶3]. Machine learning algorithms were mentioned in Col 12 ¶4 for obtaining web resources. Statistical algorithms for determining the personal confidence level, which is explained to be the probability of a future web page being visited by a user, are mentioned in Col 11 ¶4. The confidence levels can be determined based on a web site list generated from access frequency of the user [Col 12 ¶2]),
Automatically preloading, in connection with the at least one web application, at least one of the one or more web pages for use in the web browsing session by the at least one user device (Su Figure 11: Figure 11 depicts preloading occurring immediately after resource readiness is determined),
{Claim 1} Wherein the method is performed by at least one processing device comprising a processor coupled to a memory (Su Col 15 ¶4 – Col 16 ¶2: Invention can be embodied as a method involving computer hardware).
{Claim 10} A non-transitory processor-readable storage medium having stored therein program code of one or more software programs, wherein the program code when executed by at least one processing device causes the at least one processing device: to [...]. (Su Col 2 ¶2: “A computer program product for reducing load time of an encrypted web page in a browser, in accordance with a third aspect of this disclosure, can includes a computer readable storage medium having encoded thereon first program instructions executable by a processor to cause the processor to generate resource data [...]”)
{Claim 16} At least one processing device comprising a processor coupled to a memory; the at least one processing device being configured: to [...]. (Su Col 1 ¶4 – Col 2 ¶1: “A system of reducing load time of a web page in a browser, in accordance with a second aspect of this disclosure, can include a resource data processor that generates [...]”)
Regarding dependent claims 4, 13, and 19, Su discloses generating one or more predictions of one or more web pages comprises calculating a probability value attributed to each of the one or more predictions (Su Col 13 ¶2: Resource data contains the confidence level which in part based on the probability that the user will click on certain links in the webpage),
Wherein automatically preloading at least one of the one or more web pages comprises selecting the at least one of the one or more web pages based at least in part on the probability value attributed to each of the one or more predictions (Su Col 13 ¶2 – Col 14 ¶1, Figure 11: "Preloading the web resource from the domain when the browser is open can be performed when a second level of resource readiness is determined"[Col 13 ¶4] where resource readiness is based on resource data. Resource data being determined in part by a confidence score that indicates “the probability that the web resource will be retrieved in a future visit by a user” [Col 13 ¶2]).
Regarding dependent claim 7, Su discloses automatically preloading at least one of the one or more web pages comprises pre-fetching information associated with the at least one of the one or more web pages during idle time in the web browsing session (Su Col 9 ¶2: "Prefetching can be performed when the web agent is idle and network resources are available").
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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Dependent claims 2, 11, and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Su in view of Chawla “Bidirectional LSTM Autoencoder for Sequence Based Anomaly Detection in Cyber Security” published Oct 2019.
Regarding dependent claim 2, Su teaches all of the limitations of the method, apparatus, or processor readable storage medium of claims 1, 10, or 16,
Su further teaches generating one or more predictions of one or more web pages comprises processing the at least a portion of the activity-related data. See explanation above Col 8 ¶3.
Su fails to disclose processing the data using at least one bidirectional autoencoder long short-term memory (LSTM) neural network model.
Chawla teaches processing data using at least one bidirectional autoencoder long short-term memory (LSTM) neural network model. See e.g. “Our proposed model consists of a Bidirectional Encoder to read the information from both past and future directions with a unidirectional Decoder. Our two significant contributions from the model architecture are as follows. Firstly, it was observed that the CuDNNLSTM language model implementation has substantially reduced the training time when compared to an LSTM model. The time taken for CuDNNLSTM is approximately 10x faster than LSTM due to faster convergence in training” [Page 2 Col 2 ¶4].
Su and Chawla are analogous arts because both pertain to creating predictions based on user actions in web environments. It would have been obvious to one having ordinary skill in the art and the teachings of Su and Chawla before the effective filing date of the art to use a bidirectional autoencoder LSTM model, as taught by Chawla, for the machine learning model that processes the user activity data taught by Su. One would have been motivated to make this combination to reduce the training time of the model as explained above by Chawla.
Regarding dependent claims 11 and 17, claims 11 and 17 are similarly rejected as they recite similar limitations to claim 2.
Dependent claims 3, 12, and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Su in view of Sarukkai “Link prediction and path analysis using Markov chains” published June 2000.
Regarding dependent claim 3, Su teaches all of the limitations of the method, apparatus, or processor readable storage medium of claims 1, 10, or 16,
Su further teaches generating one or more predictions of one or more web pages comprises processing the at least a portion of the activity-related data. See explanation above Col 8 ¶3.
Su fails to teach processing the data using at least one Markov chain.
Sarukkai teaches processing data using at least one Markov chain. See e.g. “When used in conjunction with a Web server, the same model can be used to predict the probability of seeing a link in the future given a history of accessed links” [Section 1.2 ¶2]. A Markov model is shown as a step in the probabilistic link predictor for storing state transition probabilities based in part on the client data it receives [Figure 1 and Section 3 – System overview]
Su and Sarukkai are analogous arts because both include making predictions as to what link a user will select in a web session based on the user’s data. It would have been obvious to one having ordinary skill in the art and the teachings of Su and Sarukkai before the effective filing date of the art to use a Markov chain, as taught by Sarukkai, for the statistical algorithm used in processing activity related data taught by Su. One would have been motivated to make this combination since “Markov chains have many attractive properties. They can be easily estimated statistically. Since the Markov chain model is also generative, navigation tours can be automatically derived. The Markov chain model can also be adapted on-the-fly with additional user navigation information” as explained by Sarukkai [Section 1.2 ¶2].
Regarding dependent claims 12 and 18, claims 12 and 18 are similarly rejected as they recite similar limitations to claim 3.
Dependent claims 5, 14, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Su in view of Jam “Speed up your SPA by preloading data requests” published December 14, 2023.
Regarding dependent claims 5, Su teaches all of the limitations of the method, apparatus, or processor readable storage medium of claims 1, 10, or 16,
Su further teaches automatically preloading at least one of the one or more web pages. See explanation above Col 8 ¶3.
Su does not teach the preload comprising pre-fetching and caching one or more application programming interfaces (APIs) associated with the at least one of the one or more web pages using one or more single page applications (SPAs).
Jam teaches the preload comprising pre-fetching and caching one or more application programming interfaces (APIs) associated with the at least one of the one or more web pages using one or more single page applications (SPAs). See e.g. “One of the major downsides to a single-page app is that data fetching doesn't begin until your app's JavaScript runs” [The Problem]. A problem with single page applications is listed. “Web browsers have a declarative API for preloading resources. You may have used this before for loading fonts or something to make sure your CSS doesn't block rendering. Well it turns out you can use the same API for preloading arbitrary JS fetch calls” [The Solution]. APIs are fetched before loading the rest of the page.
Su and Jam are analogous arts because both include preloading and prefetching information to reduce the loading times of web pages. It would have been obvious to one having ordinary skill in the art and the teachings of Su and Jam before the effective filing date of the art to combine the API (Application Programming Interface) prefetching in a single page application from Jam with the web caching of Su. One would have been motivated to make this combination because “A typical "naive" client-side React app with a bunch of dependencies chosen by popularity (or at random) can easily end up loading over 1MB of JavaScript” and 1MB of JS to download on a 3G mobile signal will take a long time resulting in an undesirable loading time for the user [Jam, The Problem].
Regarding dependent claims 14 and 20, claims 14 and 20 are similarly rejected as they recite similar limitations to claim 5.
Dependent claims 6 and 15 are rejected under 35 U.S.C. 103 as being unpatentable over Su in view of Beletski WO 2022199806 A1 “Accelerating loading of web pages and web applications” published September 29, 2022.
Regarding dependent claim 6, Su teaches all of the limitations of the method, apparatus, or processor readable storage medium of claims 1, 10, or 16,
Su further teaches automatically preloading at least one of the one or more web pages See explanation above Col 8 ¶3.
Su does not teach preloading comprising preloading the at least one of the one or more web pages using one or more server push techniques.
Beletski teaches preloading comprising preloading the at least one of the one or more web pages using one or more server push techniques. See e.g. HTTP/2 server push method is used to push resources, and the browser caches them until it is ready to load and render the page [Page 20 Paragraph 5 – Page 21 Paragraph 1].
Su and Beletski are analogous arts because both involve accelerating the time it takes to load a web page. It would have been obvious to one having ordinary skill in the art and the teachings of Su and Beletski before the effective filing date of the art to combine the HTTP/2 server push technique from Beletski with the web page loading of Su to further reduce the loading time of web pages as server push techniques allow the server to send expected assets to the browser before the HTML is generated. The increased speed of asset loading results in decreased web page loading times.
Regarding dependent claim 15, claim 15 is similarly rejected as it recites similar limitations to claim 6.
Dependent claim 8 is rejected under 35 U.S.C. 103 as being unpatentable over Su in view of Park KR 102575680 B1 “USER ACTION PATTERN ANALYSIS DEVICE BASED ON ARTIFICIAL INTELLIGENCE AND METHOD FOR CONTROLLING THE SAME” published September 9, 2023.
Regarding dependent claim 8, Su teaches all of the limitations of the method of claim 1.
Su does not teach wherein obtaining activity-related data comprises obtaining clickstream data associated with the at least one user device and the at least one web application during the web browsing session, wherein the clickstream data comprises one or more of data identifying a current web page, click location data, time spent on one or more web pages, web application rights associated with the at least one user device, user category associated with the at least one user device, user device location information, one or more user device profile settings, ping time data associated with the web browsing session, and temporal data associated with the web browsing session time.
Park teaches wherein obtaining activity-related data comprises obtaining clickstream data associated with the at least one user device and the at least one web application during the web browsing session, wherein the clickstream data comprises one or more of data identifying a current web page, click location data, time spent on one or more web pages, web application rights associated with the at least one user device, user category associated with the at least one user device, user device location information, one or more user device profile settings, ping time data associated with the web browsing session, and temporal data associated with the web browsing session time. See e.g. “The user's web browsing behavior data includes click data, scroll data, mouse movement data, page view data, search query data, form input data, session data, user behavior sequence data, and shopping cart behavior data” [0057]. Where “The click data is data about which element the user clicked on and where the user clicked it” [0058].
Su and Park are analogous arts because both involve collecting information from a user during a web browsing session for later use in behavior pattern analysis. It would have been obvious to one having ordinary skill in the art and the teachings of Su and Park before the effective filing date of the art to combine the clickstream analysis of Park with the prediction system of Su. This would be done to fix the problem stated in Park [0004]: “However, conventional technologies are mainly focused on reducing the loading time, and are insufficient in analyzing or improving the user experience, so there is a problem in not knowing the reason why the user stays in the shopping mall and for what reason leaves.”
Dependent claim 9 is rejected under 35 U.S.C. 103 as being unpatentable over Su in view of Aganovic US 20150095756 A1 “Method And Apparatus For Multi-loop, Real-time Website Optimization” published October 1, 2013.
Regarding dependent claim 9, Su teaches all of the limitations of the method of claim 1.
Su further teaches training at least a portion of the one or more machine learning techniques using feedback related to the one or more predictions. See e.g. Col 4 ¶ 5 and Col 12 ¶4: Machine learning can be used for predictions [Col 12 ¶4]. "The system can learn from the history of metadata and performance data (not the content of the web resources) and the acceleration may improve as the data grows" [Col 4 ¶ 5]).
Su does not teach automatically training at least a portion of the one or more machine learning techniques using feedback related to the one or more predictions.
Aganovic teaches automatically training at least a portion of the one or more machine learning techniques using feedback related to the one or more predictions. See e.g. “A key departure of this method from the other prior art approaches is the use of a feedback loop and machine learning algorithm to intelligently adapt to website visitor behavior in real time” [0024]. This adaptation to the web visitor’s behavior is stated to occur in real time [0022].
Su and Aganovic are analogous arts because both involve utilizing automatic machine learning systems to adapt to user web browsing behavior. It would have been obvious to one having ordinary skill in the art and the teachings of Su and Aganovic before the effective filing date of the art to combine the feedback training system from Su with the real time automatic training of Aganovic to more accurately determine the future web page that a user will visit. Aganovic states that predictions for determining user web browsing behavior often become obsolete with the passage of time [0018] due to various complicated factors present in user browsing behavior [Aganovic 0017]. A real time learning system is able to continuously improve its predictions to adapt to the complicated user browsing behavior [0024].
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to RAYAN MIRMOTAHARI whose telephone number is (571)272-0822. The examiner can normally be reached Monday Friday, 8 a.m. 5 p.m. ET..
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, Matthew Ell can be reached at 571-270-3264. 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.
/R.M./ Examiner, Art Unit 2141
/MATTHEW ELL/ Supervisory Patent Examiner, Art Unit 2141
