DETAILED ACTION
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 statements (IDS) submitted on 21 November 2024 and 02 March 2026, respectively, are in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Claim Objections
Claims 1 and 9 are objected to because of the following informalities: The phrase “a ranking of slots, for filling in conversations” should read “. Appropriate correction is required.
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.
Claim(s) 1-17 and 20 is/are rejected under 35 U.S.C. 103 as being unpatentable over 12531056, hereinafter referred to as Liu et al., in view of CN 111694932, hereinafter referred to as Chen et al.
Regarding claim 1, Liu et al. discloses a system, comprising:
a processor (Liu et al., fig. 8(804).) that executes computer executable components stored in memory (Liu et al., fig. 8(806).), wherein the computer executable components comprise:
a prompt generation component that builds, based on usage logs (Liu et al., fig. 2(140) – prompt generation 140 is built using data 130 which comprises user interaction history data 210, user preference data 215, and dialog data 220 (i.e., usage logs).),
a hybrid grounding component that grounds information pertaining to the subset of slots for the virtual assistant, wherein the hybrid grounding component utilizes intent determination to fill one or more of the slots (“Some embodiments involve prompt generation/engineering using the prompt generation component 140. The prompt generation component 140, along with the data selector 142, is used to generate prompts 230 to ground the most relevant knowledge from the LM 160 under a given context,” Liu et al., para [0064]. Also, “In some embodiments, the prompt generation component 140 may be optimized to generate the prompt 230 so that the words 232 correspond to an optimized coverage rate. Coverage rate may refer to the overlap of the context generated from LM 160 with a ground truth (e.g., words that are included in the spoken input). A higher overlap between the LM-generated context and the ground truth, the better the quality of the context data. The coverage rate may be optimized by variations in the prompt 230,” Liu et al., para [0086]. Here, the slots are filled during the prompt generation using intent (i.e., contextual) determination data. And, “The user device 110 may include a wakeword detection component 620 configured to compare the audio data 611 to stored models used to detect a wakeword (e.g., “Alexa”) that indicates to the device 110 that the audio data 611 is to be processed for determining NLU output data (e.g., slot data that corresponds to a named entity, label data, and/or intent data, etc.),” Liu et al., para [0154].).
Although Liu et al. teaches ranking of words to be included in a prompt (Liu et al., para [0057].), Chen et al. is cited to explicitly disclose a ranking of slots, for filling in conversations (“Step 201, analyze the content of the conversation sent by the user using the terminal device to obtain slot information corresponding to the slots in the slot set,” Chen et al., p. 9, highlighted section. And, “Here, the information entropy of the slot is used to represent the uncertainty of the value of the slot information of the slot. In practice, it can be assumed that each retrieval result obtained by the slot S in step 202 has n values: S 1 , ..., S i , ..., S n , and the corresponding probabilities of each value are: p 1 , ..., p i , ..., p n …,” Chen et al., p. 10, highlighted section. The entropy of the slot (i.e., uncertainty) provides a ranking of the slot.), and dynamically and incrementally prompts a subset of slots based on the ranking for a virtual assistant (“Step 204: Determine the target slot according to the information entropy of each slot in the slot set,” Chen et al., p. 10, last highlight. And, “Then, in this embodiment, the above-mentioned execution body may adopt various implementation manners to determine the target slot according to the information entropy of each slot in the slot set calculated in step 203. Here, the target slot is a slot that requires user supplementary information, and the target slot may be one slot or multiple slots,” Chen et al., p. 11, highlighted section. Thus, a subset of slots may be selected. The examiner also notes that a subset may be the set itself.). Chen et al. benefits Liu et al. by proposing a method which limits the amount of information that needs to be collected from a user (Chen et al., Background technique), thereby providing answers more efficiently to the user. Therefore, it would be obvious for one skilled in the art to combine the teachings of Liu et al. with those of Chen et al. to improve the automatic speech recognition of Liu et al.
As to claim 9, method claim 9 and system claim 1 are related as system and method of using same, with each claimed element’s function corresponding to the system step. Accordingly claim 9 is similarly rejected under the same rationale as applied above with respect to system claim.
Regarding claim 2, Liu et al., as modified by Chen et al., discloses the system of claim 1, further comprising a hybrid authoring component that generates steps for the virtual assistant to follow (Chen et al., p. 9-11, steps 201-205, are dialog authoring steps for the virtual assistant to follow.).
As to claim 10, method claim 10 and system claim 2 are related as system and method of using same, with each claimed element’s function corresponding to the system step. Accordingly claim 10 is similarly rejected under the same rationale as applied above with respect to system claim.
Regarding claim 3, Liu et al., as modified by Chen et al., discloses the system of claim 1, further comprising an artificial intelligence component that trains a large language model to detect the grounded information (“In order to apply the machine learning techniques, the machine learning processes themselves need to be trained. Training a machine learning component such as, in this case, one of the trained models, requires establishing a “ground truth” for the training examples. In machine learning, the term “ground truth” refers to the accuracy of a training set's classification for supervised learning techniques,” Liu et al., para [0169].).
As to claim 11, method claim 11 and system claim 3 are related as system and method of using same, with each claimed element’s function corresponding to the system step. Accordingly claim 11 is similarly rejected under the same rationale as applied above with respect to system claim.
Regarding claim 4, Liu et al., as modified by Chen et al., discloses the system of claim 3, wherein the artificial intelligence component utilizes the detected information to fill slots (Liu et al., para [0064], [0086], and [0154].).
As to claim 12, method claim 12 and system claim 4 are related as system and method of using same, with each claimed element’s function corresponding to the system step. Accordingly claim 12 is similarly rejected under the same rationale as applied above with respect to system claim.
Regarding claim 5, Liu et al., as modified by Chen et al., discloses the system of claim 1, wherein the prompt generation component performs the ranking by employing a machine learning model with an objective function of predicting the subset of slots to prompt given a current conversation and a frequency of slots filled on previous utterances (“The target slot is determined according to the information entropy of each slot in the slot set calculated in step 203. Here, the target slot is a slot that requires user supplementary information, and the target slot may be one slot or multiple slots,” Chen et al., p. 11, highlighted section. The function H(S) takes into account the probability for a given slot value, which is determined based on past conversation frequency data.).
Regarding claim 6, Liu et al., as modified by Chen et al., discloses the system of claim 1, wherein the hybrid grounding component determines at least one of: detected intent, detected entities, context of a conversation, session history, goal statement, format constraints regarding one or more slots, business policy constraints, application programming interface (API) call or environment information, previous error results, or feedback from a human user (Liu et al., para [0064] and [0086]. Here, the slots are filled during the prompt generation using intent (i.e., contextual) determination data. And, Liu et al., para [0154], teaches slot data that corresponds to a named entity, label data, and/or intent data, etc.).
As to claim 15, method claim 15 and system claim 6 are related as system and method of using same, with each claimed element’s function corresponding to the system step. Accordingly claim 15 is similarly rejected under the same rationale as applied above with respect to system claim.
Regarding claim 7, Liu et al., as modified by Chen et al., discloses the system of claim 1, wherein the hybrid grounding component integrates natural language processing results of a trained virtual assistant into the grounding of information pertaining to respective slots (Liu et al., para [0154] and [0169].).
Regarding claim 8, Liu et al., as modified by Chen et al., discloses the system of claim 1, wherein the prompt generation component builds the ranking of slots based at least in part on usage logs (Liu et al., user interaction history data 210. And, Chen et al., p. 9-11, steps 201-205.).
Regarding claim 13, Liu et al., as modified by Chen et al., discloses the method of claim 9, further comprising predicting which slots to prompt (Chen et al., p. 10, highlighted sections.).
Regarding claim 14, Liu et al., as modified by Chen et al., discloses the method of claim 10, wherein the predicting of slots to prompt is based at least in part on a current conversation and a frequency of slots filled on previous utterances (Chen et al., p. 10 – H(S) is the calculated information entropy of the slot S. This function takes into account the probability for a given slot value, which is determined based on past conversation frequency data.).
Regarding claim 16, Liu et al., as modified by Chen et al., discloses the method of claim 9, wherein the virtual assistant fills the slots (Liu et al., para [0154].).
Regarding claim 17, Liu et al., as modified by Chen et al., discloses the method of claim 10, wherein the slots are filled in an incremental, accumulative manner (Chen et al., p. 10, last highlight and p. 11, highlighted section.).
Regarding claim 20, Liu et al. discloses a computer program product comprising a computer readable storage medium having program instructions embodied therewith (“In operation, each of these systems may include computer-readable and computer-executable instructions that reside on the respective device (120/125),” Liu et al., para [0175].), the program instructions executable by a processor to cause the processor to:
ground information pertaining to the slots (“Some embodiments involve prompt generation/engineering using the prompt generation component 140. The prompt generation component 140, along with the data selector 142, is used to generate prompts 230 to ground the most relevant knowledge from the LM 160 under a given context,” Liu et al., para [0064]. Also, “In some embodiments, the prompt generation component 140 may be optimized to generate the prompt 230 so that the words 232 correspond to an optimized coverage rate. Coverage rate may refer to the overlap of the context generated from LM 160 with a ground truth (e.g., words that are included in the spoken input). A higher overlap between the LM-generated context and the ground truth, the better the quality of the context data. The coverage rate may be optimized by variations in the prompt 230,” Liu et al., para [0086]. Here, the slots are filled during the prompt generation using intent (i.e., contextual) determination data. And, “The user device 110 may include a wakeword detection component 620 configured to compare the audio data 611 to stored models used to detect a wakeword (e.g., “Alexa”) that indicates to the device 110 that the audio data 611 is to be processed for determining NLU output data (e.g., slot data that corresponds to a named entity, label data, and/or intent data, etc.),” Liu et al., para [0154].);
train a large language model to detect the grounded information (“In order to apply the machine learning techniques, the machine learning processes themselves need to be trained. Training a machine learning component such as, in this case, one of the trained models, requires establishing a “ground truth” for the training examples. In machine learning, the term “ground truth” refers to the accuracy of a training set's classification for supervised learning techniques,” Liu et al., para [0169].);
Although Liu et al. teaches ranking of words to be included in a prompt (Liu et al., para [0057].), Chen et al. is cited to disclose building a ranking of slots and prompt slots basted on the ranking (“Step 201, analyze the content of the conversation sent by the user using the terminal device to obtain slot information corresponding to the slots in the slot set,” Chen et al., p. 9, highlighted section. And, “Here, the information entropy of the slot is used to represent the uncertainty of the value of the slot information of the slot. In practice, it can be assumed that each retrieval result obtained by the slot S in step 202 has n values: S 1 , ..., S i , ..., S n , and the corresponding probabilities of each value are: p 1 , ..., p i , ..., p n …,” Chen et al., p. 10, highlighted section. The entropy of the slot (i.e., uncertainty) provides a ranking of the slot.), and dynamically and incrementally prompts a subset of slots based on the ranking for a virtual assistant (“Step 204: Determine the target slot according to the information entropy of each slot in the slot set,” Chen et al., p. 10, last highlight. And, “Then, in this embodiment, the above-mentioned execution body may adopt various implementation manners to determine the target slot according to the information entropy of each slot in the slot set calculated in step 203. Here, the target slot is a slot that requires user supplementary information, and the target slot may be one slot or multiple slots,” Chen et al., p. 11, highlighted section. Thus, a subset of slots may be selected. The examiner also notes that a subset may be the set itself.); and
using the detected information to generate steps for a virtual assistant to follow to fill the slots (Chen et al., p. 9-11, steps 201-205, are dialog authoring steps for the virtual assistant to follow.). Chen et al. benefits Liu et al. by proposing a method which limits the amount of information that needs to be collected from a user (Chen et al., Background technique), thereby providing answers more efficiently to the user. Therefore, it would be obvious for one skilled in the art to combine the teachings of Liu et al. with those of Chen et al. to improve the automatic speech recognition of Liu et al.
Claim(s) 18-19 is/are rejected under 35 U.S.C. 103 as being unpatentable over 12531056, hereinafter referred to as Liu et al., in view of CN 111694932, hereinafter referred to as Chen et al., and further in view of US 11211058, hereinafter referred to as Eakin et al.
Regarding claim 18, Liu et al., as modified by Chen et al., discloses the method of claim 10, but not wherein a user confirms that the slots were filling in correctly. Eakin et al. is cited to dislcose wherein a user confirms that the slots were filling in correctly (“The disambiguation trigger component 630 may store data relating to the dialog session in the feedback data storage 645. The dialog session data may include audio data corresponding the initial user request, two or more ASR hypotheses (along with ASR scores) corresponding to the initial user request, the pair of identified NLU interpretations used by the disambiguation trigger component 630 for explicit disambiguation, the user input selecting or confirming the slot value, and the corrected/final interpretation that is used to generate an output in response to the user's initial request,” Eakin et al., col. 40, lines 14-24.). Eakin et al. benefits Liu et al. by realizing that the speech recognition system has some ambiguity in determining what a user said and thus determines when to request the user for clarification, as well as learns from the clarification feedback received from the user (Eakin et al., col. 2, lines 52-67). Therefore, it would be obvious for one skilled in the art to combine the teachings of Liu et al. with those of Eakin et al. to improve the speech recognition capabilities for Liu et al.
Regarding claim 19, Liu et al., as modified by Chen et al., discloses the method of claim 10, but not wherein a user manually corrects an error made during the slot filling. Eakin et al. is cited to disclose wherein a user manually corrects an error made during the slot filling (Eakin et al., col. 40, lines 14-24.). Eakin et al. benefits Liu et al. by realizing that the speech recognition system has some ambiguity in determining what a user said and thus determines when to request the user for clarification, as well as learns from the clarification feedback received from the user (Eakin et al., col. 2, lines 52-67). Therefore, it would be obvious for one skilled in the art to combine the teachings of Liu et al. with those of Eakin et al. to improve the speech recognition capabilities for Liu et al.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. See attached PTO-892. In particular, the examiner notes Becker et al. as also teaching a plurality of translation scores for the respective facets, a translation score for a respective facet representing a degree to which the translated text corresponds with one or more guidelines assigned to the respective facet. Becker et al. is cited to disclose further comprising generating, using the one or more machine learning models, a plurality of translation scores for the respective facets, a translation score for a respective facet representing a degree to which the translated text corresponds with one or more guidelines assigned to the respective facet (Becker et al., para [0018] and [0021].).
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ANNE L THOMAS-HOMESCU whose telephone number is (571)272-0899. The examiner can normally be reached Mon-Fri 8-6.
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, Bhavesh M Mehta can be reached on 5712727453. 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.
/ANNE L THOMAS-HOMESCU/Primary Examiner, Art Unit 2656
Prosecution Insights