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 .
Status of the Claims
Claims 1, 3-4, 10, 12, 16 and 18 have been amended, claims 2, 9, 11, 17 have been canceled and 24-26 have been added. Claims 1, 3-8, 10, 12-16, 18-22 and 24-26 are pending.
Claim Rejections - 35 USC § 103
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, 3-8, 10, 12-16, 18-22 is/are rejected under 35 U.S.C. 103 as being unpatentable over Konig et al. (US 20210203784 in view of Shipper et al. (US 20180011678) and in further view of Hardebeck et al. (US 20210029247).
Regarding claim 1, Konig teaches a method of self-learning actions for an automated co-browse session, the method comprising:
initiating an interaction between a user and a chat bot ([0072], [0078], [0108]);
determining a user intent of the user based on the interaction between the user and the chat bot ([0094]);
routing the interaction to a human contact center agent for a co-browse session ([0072], [0108]) of a webpage to enable parallel ([0113] “established in parallel to the primary communication”) interaction with the webpage ([0044] “customers to initiate, manage, and respond to … web-browsing sessions, and other multi-media transactions”, [0053] “media interactions may be … chat, video, text-messaging, web, social media, co-browsing”, [0057]-[0058] “Customers may browse the web pages and get information about the enterprise's products and services”; “contact center via … web chat”, [0060] “configured to record an interaction history for each customer, capturing and storing data,” [0125] “interaction by storing and indexing any messages, documents, files, and other media involved or shared during in the interaction … an image that was shared and annotated by an agent when explaining how to set up a particular piece of equipment”) between the user and the human contact center agent ([0052], [0082], [0108], [0112]);
storing a plurality of actions performed by the human contact center agent ([0060], [0186] “historical record of data reflecting all past interaction between a customer and any contact center,” of a “co-browsing” session [0053]), including a
performing machine learning to determine an optimal solution ([0179], [0195], [0213], [0218], [0264]) for resolving the user intent based on an analysis of the plurality of actions performed by the human contact center agent ([0114], [0216]),
generating an intent configuration file ([0095], [0111] “script may provide multiple templates”, [0127], [0158] “deadline for completing a pending action given an analysis of the interaction script”, [0241]) for the optimal solution ([0039] “optimizes efficiency and promotes repeatability”, [0280] “produce better routing decisions”, [0213] “offering the best chance for successful resolution given the customer”) based on the machine learning ([0192], [0197], [0201]), wherein the intent configuration file defines a sequence of actions to be executed by the chat bot ([0130], [0133], [0135], [0138], [0148], [0160) in an automated co-browse session between the chat bot and another user to resolve the user intent ([0097], [0108], [0111]-[0112]); and storing the intent configuration file in association with the user intent ([0108], [0125]-[0126])(see NOTE).
Konig does not explicitly teach, however Shipper discloses storing a plurality of actions performed by the human contact center agent ([0064] “store and reference historical data about previously highlighted UI elements… .may track historical data based on the same section of the user interface where previously highlighted UI elements have assisted other customers, or based on a customer identifier where previously highlighted UI elements have assisted the specific customer”), including a set of document object model elements of the webpage ([0064] “select all DOM tree elements”, [0065], [0089], [0092], [0095]) and web actions performed by the human contact center agent on one or more of the document object model elements ([0068], [0072], [0083]), during the co-browse session ([0067] “facilitate collaborative application usage such as full collaborative web browsing or co-browsing, or screen sharing with the customer service representative”) to a data store ([0071] “server receives the collection of UI elements in a JSON object”).
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 Konig to include document object model elements tracked during the co-browse session as disclosed by Shipper. Doing so would help improve customer service performance metrics such as lowering average handle time (Shipper [0042]).
Although Konig teaches using machine learning to interactions between users and customer support and Shipper discloses using DOM to collect such interactions, Konig as modified by Shipper does not explicitly teach, however Hardebeck discloses performing machine learning to determine an optimal solution for resolving the user intent based on an analysis of the plurality of actions performed by the human contact center agent, including the set of document object model elements of the webpage and the web actions performed by the human contact center agent on the one or more of the document object model elements, during the co-browse session ([0095] “the co-browsing system receives DOM updates from each of the website visitors on visualization sessions and provides that information to the machine learning algorithm”).
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 Konig to include document object model elements tracked during the co-browse session as disclosed by Hardebeck. Doing so enables the machine learning algorithm to know exactly what the customers are seeing as they interact with the website and what remedy is required to fix the issue (Hardebeck [0095]).
Claims 10 and 16 recite substantially the same limitations as claim 1 and is rejected for substantially the same reasons.
Regarding claims 3, 12 and 18, Konig as modified teaches the method, the system and the media, further comprising determining a confidence score indicative of a confidence of the system in the optimal solution based on the machine learning (Konig [0082], [0203[-[0204]); and
wherein generating the intent configuration file for the optimal solution comprises generating the intent configuration file for the optimal solution in response to determining that the confidence score exceeds a threshold confidence level (Konig [0082], [0189], [0212], [0228]).
Regarding claim 4, Konig as modified teaches the method of claim 2, wherein the sequence of actions comprises one or more actions of the plurality of actions performed by the human contact center agent during the co- browse session (Konig [0108], [0235], [0290], Hardebeck [0095], Shipper [0042).
Regarding claims 5, 13 and 19, Konig as modified teaches the method, the system and the media, wherein performing the machine learning comprises analyzing a plurality of sequences of actions performed by human contact center agents during respective co-browse sessions to resolve the user intent (Konig [0096], [0121], [0108], [0235], [0290]).
Regarding claims 7, 15 and 21, Konig as modified teaches the method, the system and the media, wherein the optimal solution is selected from the plurality of sequences of actions performed by the human contact center agents during the corresponding co-browse sessions to resolve the user intent (Konig [0096], [0121], [0108], [0235], [0290], Hardebeck [0042]).
Regarding claim 8, Konig as modified teaches the method of claim 1, wherein the plurality of actions comprises at least one of a mouse movement, mouse interaction, screen pointer, screen change, audio instruction, video instruction, or text entry (Konig [0086], Shipper [0010]).
Regarding claim 9, Konig as modified teaches the method of claim 1, wherein the plurality of actions comprises a plurality of web actions involving interactions with one or more web pages (Konig [0086], Shipper [0068], [0072], [0083]).
Regarding claim 22, Konig as modified teaches the one or more non-transitory machine readable storage media of claim 16, wherein to route the interaction to the human contact center agent for a co-browse session between the user and the human contact center agent comprises to route the interaction to the human contact center agent in response to a determination that the chat bot is unable to resolve the user intent of the user ([0060], Konig [0039] “target the use of human agents for the more difficult customer interactions”, [0052], [0082], [0206], [0208]-[0209]).
NOTE as previously cited Shlomov likewise discloses claim 22 in [0041] (“include an escalated portion that occurs because an automated agent was unable to satisfy and/or understand the human so that the conversation was escalated to a human agent”) and further obviates the teaching of Konig.
Claims 6, 14 and 20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Konig as modified and in further view of Munavalli (US 20210203784).
Regarding claims 6, 14 and 20, Konig as modified does not explicitly teach, however Munavalli discloses the method, the system and the media, wherein analyzing the plurality of sequences of actions performed by the human contact center agents during the corresponding co-browse sessions to resolve the user intent ([0019]) comprises applying a Q-learning reinforcement algorithm ([0022], [0039]) to the plurality of sequences of actions performed by the human contact center agents during the corresponding co-browse sessions to resolve the user intent ([0023]-[0024], [0026]-[0027]).
It would have been obvious to one of ordinary skill in the art at the time of invention to modify the teachings of Konig as modified to include Q-learning reinforcement algorithm as disclosed by Munavalli. Doing so would improve the efficiency and efficacy of the chatbot system through the reinforcement learning framework enabling the machine learning model to adapt to changes, such as changes in the manner in which users interact with the chatbot system (Munavalli [0028]).
Claims 24-25 is/are rejected under 35 U.S.C. 103 as being unpatentable over Konig as modified and in further view of BAER et al. (US 20180219921) and Francis (US 20140214466) and alternatively or additionally in further view of Lewis et al. (US 20220215181).
Regarding claim 24, Konig as modified does not explicitly teach, however BAER discloses the method of claim 1, further comprising:
determining whether data indicative of an incomplete
providing, via the chat bot, the user with an option (see NOTE) to resume the incomplete
It would have been obvious to one of ordinary skill in the art at the time of invention to modify the teachings of Konig as modified to include incomplete session as disclosed by BAER. Doing so would allow the user to restart the conversation again where it was left and not need to start over again (BAER [0004]-[0005]).
Although Konig, Shipper and Hardebeck are all teach collaborative browsing, or co-browsing sessions and thus, it would have been obvious to one of ordinary skill in the art to combine the teachings of BAER’s incomplete session to be incomplete co-browse session already disclosed by Konig, Shipper and Hardebeck. Still, to further obviate the limitation Francis discloses incomplete co-browse session ([0045] “transactions may be directed to … co-browsing sessions”). Francis discloses determining whether data indicative of an incomplete co-browse session is stored (F6:605).
It would have been obvious to one of ordinary skill in the art at the time of invention to modify the teachings of Konig as modified to include incomplete co-browse session as disclosed by Francis. Doing so would interaction through any known communication system (Francis [0002]).
NOTE BAER discloses “user … over chat … can retrieve a dialogue that was started on the previous chat” [0066], which construed to be analogous to the limitation of “providing, via the chat bot, the user with an option to resume the incomplete co-browse session.” However, to further obviate providing … an option to resume the incomplete session, Lewis teaches the same in [0014], [0017]-[0018], [0097]. It would have been obvious to one of ordinary skill in the art at the time of invention to modify the teachings of Konig as modified to provide an option to resume the incomplete session as disclosed by Lewis. Doing so allows for seamlessly resumed conversation (Lewis [0005]).
Regarding claim 25, Konig as modified teaches the method of claim 24, further comprising:
retrieving an intent configuration file (BAER [0143] “the flow together with input data from the user are stored in a database or persistent memory”, “the user may have several flows” i.e. several intents, Konig [0095], [0111], [0127], [0158], [0241]) associated with the incomplete co-browse session (BAER [0048] “session state, identification information but also user id, sequence id of messages related to the ongoing dialogue, timestamp(s), sequence id(s)”);
determining a point in the intent configuration file at which the incomplete co-browse session was terminated (BAER [0038] “information received earlier for continuing the dialogue where it stopped at the chat”, [0047], [0121]-[0142], Konig [0053], Shipper [0067]); and
performing, via the chat bot and during a resumed co-browse session with the user, one or more actions defined by the intent configuration file starting at the point at which the incomplete co-browse session was terminated (BAER [0143], [0158], [0163]).
Claim 26 is/are rejected under 35 U.S.C. 103 as being unpatentable over Konig as modified and in further view of Wang et al. (US 20180033042) and in further view of Mendez et al. (US 20150149557).
Regarding claim 26, Konig as modified does not explicitly teach, however Wang discloses the method of claim 1, further comprising: detecting that the user has remained at a particular webpage for at least a threshold period of time ([0018]); and proactively offering the user assistance via a
It would have been obvious to one of ordinary skill in the art at the time of invention to modify the teachings of Konig as modified to proactively offering the user assistance via a co-browse session as disclosed by Wang. Doing so allows to initiate a personalized, proactive, context-based interaction with a website visitor user, in a timely manner, and without exposing the visitor to queuing time in the contact center (Wang [0019]).
Although Konig, Shipper and Hardebeck are all teach collaborative browsing, or co-browsing sessions and thus, it would have been obvious to one of ordinary skill in the art to combine the teachings of Wang of providing a user assistance via a session to be user assistance via a co-browse session already disclosed by Konig, Shipper and Hardebeck. Still, to further obviate the limitation Mendez discloses incomplete co-browse session ([0107] “agent may initiate a co-browsing session with a visitor”, when “a visitor is having trouble on a web page”; [0172] “Co-browse session may be initiated when the visitor needs help (or the agent or automation rules perceive the visitor may need assistance)”). It would have been obvious to one of ordinary skill in the art at the time of invention to modify the teachings of Konig as modified to include incomplete co-browse session as disclosed by Mendez. Doing so improves customer experience.
Claims 1,3-8,10,12-16,18-22 is/are rejected under 35 U.S.C. 103 as being unpatentable over Konig et al. (US 20210203784 in view of Shipper et al. (US 20180011678) and in further view of Hardebeck et al. (US 20210029247) and alternatively or additionally in further view of SINGH et al. (US 20210158146).
NOTE Konig teaches storing intent configuration files to implement a customer solution. Konig further teaches providing “enhanced or optimized” performance ([0067]), wherein “automated processes can be structured in a way that optimizes efficiency and promotes repeatability” [0039], and “gain insights on what works best … during interactions” [0195], “offering the best chance for successful resolution given the customer” [0213]; “The scripts may be produced via mining data, actions, and dialogue from previous customer interactions. … the sequences of statements made during a request for resolution of a particular issue may be automatically mined from a collection of historical interactions between customers and customer service providers” [0096], “develop a follow-up workflow in which one or more follow-up actions are scheduled” [0160].
Thus, it is reasonable to conclude that scripts with sequences of statements, workflow of actions, which optimizes performance and resolution of customer intent is analogous to the claimed intent configuration files for the optimal solution.
However, alternatively or additionally to further obviate such statement SINGH further discloses generating an intent configuration file for the optimal solution based on the machine learning, wherein the intent configuration file defines a sequence of actions ([0009] “identify the optimal path of actions”, [0012], [0018]-[0019]) to be executed by the chat bot in an automated co-browse session between the chat bot and another user to resolve the user intent; and storing the intent configuration file in association with the user intent ([0011], [0013]-[0014]).
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 Konig to include intent configuration file for the optimal solution as disclosed by SINGH. Doing so would help troubleshoot a technical issue in a consistent, efficient, and effective manner (SINGH [0008]).
Regarding claims 3, 12 and 18, Konig as modified teaches determining a confidence score indicative of a confidence of the system in the optimal solution based on the machine learning (SINGH [0017]-[0018]); and wherein generating the intent configuration file for the optimal solution comprises generating the intent configuration file for the optimal solution in response to determining that the confidence score exceeds a threshold confidence level (SINGH [0071], [0033], [0021]).
The rest of the dependent claims are omitted here for the sake of brevity.
Response to Arguments
Applicant's arguments, filed 07/13/2026, have been fully considered, but they are not deemed persuasive.
◊ With respect to the independent claims now incorporated dependent claims 2, 11 and 17, and the Konig and Singh reference the applicant argues –
“The specification of the subject application describes that each intent configuration file includes one or more sequences of the various actions (e.g., mouse movements/interactions, screen pointers, screen changes, audio/video instructions, text entry, and/or other actions) to be executed by the chat bot in order to automatically resolve the user intent in a manner similar to how it would be resolved by a human agent. This is fundamentally different from Konig's scripts (which automate contact center API calls) and Singh's optimal action sequences (which are troubleshooting steps for network devices)” (see page 5 of the Remarks).
The arguments are not persuasive. With respect to the configuration file, the applicant’s specification discloses in [0039] (PUB version)-
“the intent configuration file defines a sequence of actions to be executed by the chat bot in an automated co-browse session between the chat bot and another user to resolve the user intent.”
Konig analogously teaches –
“The scripts may be produced via mining data, actions, and dialogue from previous customer interactions. Specifically, the sequences of statements made during a request for resolution of a particular issue may be automatically mined from a collection of historical interactions between customers and customer service providers” [0096], “develop a follow-up workflow in which one or more follow-up actions are scheduled” [0160].
It is reasonable and surely obvious to conclude that scripts with sequences of statements, workflow of actions, which optimizes performance and resolution of customer intent is analogous to the claimed intent configuration files for the optimal solution. Singh merely obviates such statement by explicitly showing a record (file) that identifies an optimal path of actions, sequencing of next optimal actions to be performed ([0009]). Which is basically nearly identical to the applicant’s own definition for the configuration file. Wherein Konig shown that such interaction can be based on “co-browsing” [0053]. Thus, Konig on his own, or alternatively in combination with Singh fully and clearly disclose the limitation of –
“generating an intent configuration file for the optimal solution based on the machine learning, wherein the intent configuration file defines a sequence of actions to be executed by the chat bot in an automated co-browse session between the chat bot and another user to resolve the user intent; and storing the intent configuration file in association with the user intent,” as required by the independent claims.
◊ With respect to the combination of references and the limitation of “document object model elements of the webpage,” the applicant argues –
“the prior art of record fails to teach or suggest the integration reflected by the features of independent claim 1, whether considered alone or in combination. … . The cited art simply does not teach or suggest using DOM-level web action data as training input for machine learning to determine optimal solutions” (see page 5 of the Remarks).
The arguments are not persuasive. Konig clearly, throughout the disclosure teaches recording “interaction history (e.g., details of each previous interaction with a customer” [0050], “maintaining a history on how well chats for a particular customer were handled”; “configured to record an interaction history for each customer, capturing and storing data” [0060]; “historical record of data reflecting all past interaction between a customer and any contact center” [0186]; “identify patterns therein correlating one or more input factors to one or more outcomes relevant to the first customer given a particular type of interaction” [0218]. Konig further teaches that such interaction can be from – “chats, text messages, web-browsing sessions, and other multi-media transactions” [0044] and “co-browsing” [0053]. The only thing Konig doesn’t teach is a recording of sequences of the various actions such as mouse movements/interactions, screen pointers, often performed during a co-browsing. However, given that the co-browsing is already explicitly disclosed by Konig, capturing such mouse movements/interactions, screen pointers via a document object model elements of the webpage would be an additional, obvious choice of data extra data recording and observations.
Once again, and as stated on the record, Konig does not teach document object model (DOM). Instead, Konig teaches “historical record of data reflecting all past interaction between a customer and any contact center,” in a “co-browsing” session. All past interaction in the co-browsing session can obviously include screen sharing and mouse movements (i.e. screen sharing is inherent to the co-browsing). Determining additional data, such as sequences of mouse movements, etc., captured by the document object model (DOM) is an extra solution step that would been obvious to those skill in the art.
Konig also teaches applying a machine learning model to the collected historic interactions in order to provide an optimal solution of specific agent routing, based on a customer intent. Thus, applying the machine learning model based on the additional DOM data would likewise be an obvious additional consideration by the machine learning model already disclosed by Konig.
Examiner respectfully notes one cannot show nonobviousness by attacking references individually where the rejections are based on combinations of references. See In re Keller, 642 F.2d 413, 208 USPQ 871 (CCPA 1981); In re Merck& Co., 800 F.2d 1091,231 USPQ 375 (Fed. Cir. 1986).
The reference of Shipper clearly and explicitly teaches capturing all interactive elements on a webpage via the DOM. Thus, given that Konig already teaches using machine learning model on all past interactions between customer and an agent, it would have been obvious to apply the ML model of Konig to the DOM elements of Shipper. The refence of Hardebeck merely obviates such combination and shows that the ML model utilizes Dom elements to facilitate predictions. The motivation used to combine the references is from the references themselves and is not improper.
Thus, the combination of references of Konig, Shipper and Hardebeck fully discloses the limitations of –
“a set of document object model elements of the webpage and web actions performed by the human contact center agent on one or more of the document object model elements, during the co- browse session to a data store; performing machine learning to determine an optimal solution for resolving the user intent based on an analysis of the plurality of actions performed by the human contact center agent, including the set of document object model elements of the webpage and the web actions performed by the human contact center agent on the one or more of the document object model elements, during the co-browse session,” as required.
◊ The applicant further argues –
“motivations are insufficient because they do not explain why a person of ordinary skill would have modified the systems to perform machine learning on DOM elements and web actions to determine optimal solutions. … Shipper teaches away from the proposed combination. … As such, the Applicant respectfully submits that the proposed combination of Konig, Shipper, and Hardebeck is improper” (see page 5 of the Remarks).
The arguments are not persuasive. The examiner recognizes that obviousness may be established by combining or modifying the teachings of the prior art to produce the claimed invention where there is some teaching, suggestion, or motivation to do so found either in the references themselves or in the knowledge generally available to one of ordinary skill in the art. See In re Fine, 837 F.2d 1071, 5 USPQ2d 1596 (Fed. Cir. 1988), In re Jones, 958 F.2d 347, 21 USPQ2d 1941 (Fed. Cir. 1992), and KSR International Co. v. Teleflex, Inc., 550 U.S. 398, 82 USPQ2d 1385 (2007).
In this case, Konig already clearly teaches capturing / storing all interactions in a co-browsing or chat sessions. Shipper merely clarifies that such capturing can be perform via DOM elements capturing in a co-browsing session. These are the only feature that Shipper is relied upon to teach. Other features of Shipper do not need to be included when this modification takes place. Shipper clearly teaches capturing DOM elements in a co-browsing session in paragraphs [0064], [0065], [0089], [0092], [0095]. The motivation used to combine the references is from the references themselves and is not improper.
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 POLINA G PEACH whose telephone number is (571)270-7646. The examiner can normally be reached Monday-Friday, 9:30 - 5:30.
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, Aleksandr Kerzhner can be reached at 571-270-1760. 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.
/POLINA G PEACH/ Primary Examiner, Art Unit 2165 August 3, 2026
Prosecution Insights