DETAILED ACTION
This action is in response to the application filed on November 27th, 2024. Claims 1-20 are pending and have been examined.
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 .
Specification
The disclosure is objected to because of the following informalities:
Line 2 of [0055] of the specification reads “to case the topic generation model”, but should read ”to cause the topic generation model”.
The sentence ending on line 6 of [0060] of the specification ends abruptly and is incomplete.
Appropriate correction is required.
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.
Claim(s) 1, 2, 4, 8, 10, 16, 17, and 20 is/are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Bogdan et al. (US Pat. Pub. No. 2021/0232764 A1 hereinafter Bogdan).
Regarding claim 1, Bogdan discloses a method comprising: generating, utilizing a topic generation model, a topic data structure by: providing a plurality of verbatims to the topic generation model, each verbatim comprising natural language user input text (Bogdan, Fig. 15; [0048]: "FIG. 15 depicts an illustrative embodiment of a method 1500 used by system 100 to process unstructured text data in the documents 130 to produce presentable content 150."; [0025]: " Documents 130 may comprise fielded data in the form of text strings, such as comment data, status data, periodic satisfaction surveys, project management reports, trouble ticket logs, etc. Co-collected structured data 117 may comprise fielded and structured data that is collected with documents 130. System 100 converts the unstructured data contained in documents 130 into structured data according to topics contained within the unstructured data."); and receiving, as output from the topic generation model, the topic data structure comprising topics and keywords, wherein the topic data structure is disconnected from the plurality of verbatims (Bogdan, Fig. 15, 1502; [0048]: "The process starts with step 1502, where the system 100 determines a plurality of topics from the documents 130."; [0031]: "The semantic analysis may include keyword assessment in the text from the document associated with the topic."); generating, utilizing a reverb correlation model, a correlation between a verbatim and a topic of the topic data structure; and assigning the topic to the verbatim based on the correlation between the verbatim and the topic to connect the verbatim to the topic (Bogdan, Fig. 15, 1506; [0050]: "In step 1506, the system 100 maps each document in the documents 130 to one or more topics in the subset of topics, thereby creating topic-document pairs, as described above.").
Regarding claim 2, the rejection of claim 1 is incorporated. Bogdan discloses all of the elements of the current invention as stated above. Bogdan further discloses wherein the topic generation model is a large language model trained to generate a plurality of topics based on a plurality of verbatims, the plurality of topics conveying underlying themes and a plurality of keywords associated with the plurality of topics (Bogdan, [0026]: "As shown in FIG. 1, computer 110 comprises a topic model tool, TMT 112, a tool for verbatim analysis, TVA 114, and a tool for cross data analysis, TCDA 116. TMT 112 performs a natural language processing statistical analysis of the unstructured data in documents 130 to discern topics within the data and create structured data from the unstructured data."; [0027]: "By observing the words in a document, and their frequency of occurrence, TMT 112 can determine the likely probability that a word is attributed to one particular topic or another.").
Regarding claim 4, the rejection of claim 1 is incorporated. Bogdan discloses all of the elements of the current invention as stated above. Bogdan further discloses receiving additional verbatims comprising additional natural language user input text; and updating the topic data structure by providing the additional verbatims to the topic generation model (Bogdan, [0065]: "In an alternative embodiment, the batch processing of the plurality of text documents as set forth above may be modified to process a stream of text documents, or processing on a rolling basis.").
Regarding claim 8, the rejection of claim 1 is incorporated. Bogdan discloses all of the elements of the current invention as stated above. Bogdan further discloses generating a heatmap representing correlations between the verbatim and the topic; and providing the heatmap for display on a client device (Bogdan, [0035]: "FIG. 2 depicts an exemplary layout 200 of presentable content 150 generated by the system 100 of FIG. 1. As shown in FIG. 2, presentable content 150 comprises a heat map comprising a plurality of clustered topics 210, a title 211 for each cluster, a topic area 215 within the clustered topics 210, and a bias legend 220. Each topic area 215 can be a geometric area that illustrates the bias and frequency of occurrence of the topic or another summary statistic within the documents 130.").
Regarding claim 10, Bogdan discloses a system comprising: at least one processor (Bogdan, [0061]: “In accordance with various embodiments of the subject disclosure, the operations or methods described herein are intended for operation as software programs or instructions running on or executed by a computer processor or other computing device, and which may include other forms of instructions manifested as a state machine implemented with logic components in an application specific integrated circuit or field programmable gate array.”); and at least one non-transitory computer readable storage medium comprising instructions that, when executed by the at least one processor (Bogdan, [0062]: “While the tangible computer-readable storage medium 1622 is shown in an example embodiment to be a single medium, the term “tangible computer-readable storage medium” should be taken to include a single medium or multiple media (e.g., a centralized or distributed database, and/or associated caches and servers) that store the one or more sets of instructions. The term “tangible computer-readable storage medium” shall also be taken to include any non-transitory medium that is capable of storing or encoding a set of instructions for execution by the machine and that cause the machine to perform any one or more of the methods of the subject disclosure.”), cause the system to: generate, utilizing a topic generation model, a topic data structure by: providing a plurality of verbatims to the topic generation model, each verbatim comprising natural language user input text (Bogdan, Fig. 15; [0048]: "FIG. 15 depicts an illustrative embodiment of a method 1500 used by system 100 to process unstructured text data in the documents 130 to produce presentable content 150."; [0025]: " Documents 130 may comprise fielded data in the form of text strings, such as comment data, status data, periodic satisfaction surveys, project management reports, trouble ticket logs, etc. Co-collected structured data 117 may comprise fielded and structured data that is collected with documents 130. System 100 converts the unstructured data contained in documents 130 into structured data according to topics contained within the unstructured data."); and receiving, as output from the topic generation model, the topic data structure comprising topics and keywords, wherein the topic data structure is disconnected from the plurality of verbatims (Bogdan, Fig. 15, 1502; [0048]: "The process starts with step 1502, where the system 100 determines a plurality of topics from the documents 130."; [0031]: "The semantic analysis may include keyword assessment in the text from the document associated with the topic."); generate, utilizing a reverb correlation model, a correlation between a verbatim and a topic of the topic data structure; and assign the topic to the verbatim based on the correlation between the verbatim and topic to connect the verbatim to the topic (Bogdan, Fig. 15, 1506; [0050]: "In step 1506, the system 100 maps each document in the documents 130 to one or more topics in the subset of topics, thereby creating topic-document pairs, as described above.").
Regarding claim 16, Bogdan discloses a non-transitory computer-readable medium storing instructions that, when executed by at least one processor (Bogdan, [0062]: “While the tangible computer-readable storage medium 1622 is shown in an example embodiment to be a single medium, the term “tangible computer-readable storage medium” should be taken to include a single medium or multiple media (e.g., a centralized or distributed database, and/or associated caches and servers) that store the one or more sets of instructions. The term “tangible computer-readable storage medium” shall also be taken to include any non-transitory medium that is capable of storing or encoding a set of instructions for execution by the machine and that cause the machine to perform any one or more of the methods of the subject disclosure.”), cause a computing device to: generate, utilizing a topic generation model, a topic data structure by: providing a plurality of verbatims to the topic generation model, each verbatim comprising natural language user input text (Bogdan, Fig. 15; [0048]: "FIG. 15 depicts an illustrative embodiment of a method 1500 used by system 100 to process unstructured text data in the documents 130 to produce presentable content 150."; [0025]: " Documents 130 may comprise fielded data in the form of text strings, such as comment data, status data, periodic satisfaction surveys, project management reports, trouble ticket logs, etc. Co-collected structured data 117 may comprise fielded and structured data that is collected with documents 130. System 100 converts the unstructured data contained in documents 130 into structured data according to topics contained within the unstructured data."); and receiving, as output from the topic generation model, the topic data structure comprising topics and keywords, wherein the topic data structure is disconnected from the plurality of verbatims (Bogdan, Fig. 15, 1502; [0048]: "The process starts with step 1502, where the system 100 determines a plurality of topics from the documents 130."; [0031]: "The semantic analysis may include keyword assessment in the text from the document associated with the topic."); generate, utilizing a reverb correlation model, a correlation between a verbatim and a topic of the topic data structure; and assign the topic to the verbatim based on the correlation between the verbatim and topic to connect the verbatim to the topic (Bogdan, Fig. 15, 1506; [0050]: "In step 1506, the system 100 maps each document in the documents 130 to one or more topics in the subset of topics, thereby creating topic-document pairs, as described above.").
Regarding claim 17, the rejection of claim 16 is incorporated. Bogdan discloses all of the elements of the current invention as stated above. Bogdan further discloses wherein: the topic generation model is a large language model trained to generate a plurality of topics conveying underlying themes and a plurality of keywords associated with the plurality of topics; and the reverb correlation model is a large language model trained to recognize topics within verbatims based on semantic associations between the verbatims and the topics (Bogdan, [0026]: "As shown in FIG. 1, computer 110 comprises a topic model tool, TMT 112, a tool for verbatim analysis, TVA 114, and a tool for cross data analysis, TCDA 116. TMT 112 performs a natural language processing statistical analysis of the unstructured data in documents 130 to discern topics within the data and create structured data from the unstructured data."; [0027]: "By observing the words in a document, and their frequency of occurrence, TMT 112 can determine the likely probability that a word is attributed to one particular topic or another.").
Regarding claim 20, the rejection of claim 16 is incorporated. Bogdan discloses all of the elements of the current invention as stated above. Bogdan further discloses wherein the verbatim is selected from the plurality of verbatims (Bogdan, [0025]: "System 100 is configured to process unstructured data, such as documents 130. Documents 130 may comprise fielded data in the form of text strings, such as comment data, status data, periodic satisfaction surveys, project management reports, trouble ticket logs, etc. Co-collected structured data 117 may comprise fielded and structured data that is collected with documents 130. System 100 converts the unstructured data contained in documents 130 into structured data according to topics contained within the unstructured data.").
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) 3, 6, 7, 9, 11, 15 is/are rejected under 35 U.S.C. 103 as being unpatentable over Bogdan as applied to claims 1, 2, 4, 8, 10, 16, 17, and 20 above, and further in view of Rogynskyy et al. (US Pat. Pub. No. 2025/0045308 A1 hereinafter Rogynskyy).
Regarding claim 3, the rejection of claim 1 is incorporated. Bogdan discloses all of the elements of the current invention as stated above. However, Bogdan fails to expressly recite wherein the reverb correlation model is a large language model trained to recognize topics within verbatims based on semantic associations between the verbatims and the topics.
Rogynskyy teaches wherein the reverb correlation model is a large language model trained to recognize topics within verbatims based on semantic associations between the verbatims and the topics (Rogynskyy, [0400]: "Based on the execution, the large language model can generate (i) a first set of topics, (ii) a first set of references indicating one or more subsets of the first set of text strings, each subset of text strings corresponding to a different topic of the first set of topics, and (iii) an attribute for each of the first set of topics. For each text string in the subset of text strings corresponding to a topic, the large language model can generate a reference. Each reference can function as a pointer that identifies specific text strings within the first set of strings (e.g., chunk) that contributed to generating the topic. Each attribute for a topic can indicate a level of relevance of the topic to the opportunity record object.").
Bogdan and Rogynskyy are analogous arts because they each belong to the same field of data processing. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the data analytics system of Bogdan to incorporate the teachings of Rogynskyy to recognize topics within text based on semantic associations. Recognizing topics in this way helps the system efficiently and accurately analyze large volumes of data (Rogynskyy, [0002]). This ensures that the system provides the best results to the user in a reasonable amount of time.
Regarding claim 6, the rejection of claim 1 is incorporated. Bogdan discloses all of the elements of the current invention as stated above. However, Bogdan fails to expressly recite further comprising generating the topic data structure by: extracting a plurality of topics from the plurality of verbatims, each topic of the plurality of topics associated with one or more keywords; and generating the topics by selecting a subset of the plurality of topics based on a relative semantic similarity of the plurality of topics to the plurality of verbatims.
Rogynskyy teaches further comprising generating the topic data structure by: extracting a plurality of topics from the plurality of verbatims, each topic of the plurality of topics associated with one or more keywords; and generating the topics by selecting a subset of the plurality of topics based on a relative semantic similarity of the plurality of topics to the plurality of verbatims (Rogynskyy, [0262]: "The text string generator 910 can store text strings (e.g., insights) that the text string generator 910 generates in the database 911. The database 911 can be a database (e.g., a relational database) with a semantic index. The semantic index can include word embeddings (e.g., vectors) in which words with similar meanings are represented closer to each other than other words."; [0401]: "The computer can repeat the chunking process iteratively for each set of text strings. After processing an initial set of text strings, the inferences from the large language model (e.g., topics and their corresponding subsets) can be used to update and/or fine-tune the analysis for subsequent sets of text strings. For instance, the computer can identify one or more subsets of text strings corresponding to highly relevant topics that the large language model identified based on the initial set of text strings. The computer can identify the one or more subsets based on their attributes indicating that the topics have a level of relevancy above a threshold with the record object and/or that the attributes of the topics satisfy some other criteria. The computer can retrieve the identified subsets of text strings via their respective references output by the large language model. The computer can input the retrieved subsets of text strings, along with a second set of text strings (e.g., a second set of text strings generated from electronic activities generated and/or received sequentially after the first set of electronic activities) that correspond to a different set of text strings from the first set of text strings, into the large language model. Based on the input, the large language model can generate a second set of topics, references to text strings from which the second set of topics were generated, and/or attributes for the second set of topics indicating a relevance of the second set of topics to the record object. The computer can iteratively repeat this process any number of times to until processing each set or chunk of text strings for the record object."; [0431]: "In some embodiments, the data processing system can be configured to select the chunks based on themes or keywords."). The same motivation for claim 3 applies equally to claim 6.
Regarding claim 7, the rejection of claim 1 is incorporated. Bogdan discloses all of the elements of the current invention as stated above. However, Bogdan fails to expressly recite further comprising generating the topic data structure by: determining a first subset of topics from a first subset of the plurality of verbatims; determining a second subset of topics from a second subset of the plurality of verbatims; and generating the topics by combining the first subset of topics and the second subset of topics.
Rogynskyy teaches further comprising generating the topic data structure by: determining a first subset of topics from a first subset of the plurality of verbatims (Rogynskyy, [0400]: "Based on the execution, the large language model can generate (i) a first set of topics, (ii) a first set of references indicating one or more subsets of the first set of text strings, each subset of text strings corresponding to a different topic of the first set of topics, and (iii) an attribute for each of the first set of topics. For each text string in the subset of text strings corresponding to a topic, the large language model can generate a reference. Each reference can function as a pointer that identifies specific text strings within the first set of strings (e.g., chunk) that contributed to generating the topic. Each attribute for a topic can indicate a level of relevance of the topic to the opportunity record object."); determining a second subset of topics from a second subset of the plurality of verbatims (Rogynskyy, [0401]: "The computer can repeat the chunking process iteratively for each set of text strings. After processing an initial set of text strings, the inferences from the large language model (e.g., topics and their corresponding subsets) can be used to update and/or fine-tune the analysis for subsequent sets of text strings. For instance, the computer can identify one or more subsets of text strings corresponding to highly relevant topics that the large language model identified based on the initial set of text strings. The computer can identify the one or more subsets based on their attributes indicating that the topics have a level of relevancy above a threshold with the record object and/or that the attributes of the topics satisfy some other criteria. The computer can retrieve the identified subsets of text strings via their respective references output by the large language model."); and generating the topics by combining the first subset of topics and the second subset of topics (Rogynskyy, [0401]: "The computer can input the retrieved subsets of text strings, along with a second set of text strings (e.g., a second set of text strings generated from electronic activities generated and/or received sequentially after the first set of electronic activities) that correspond to a different set of text strings from the first set of text strings, into the large language model. Based on the input, the large language model can generate a second set of topics, references to text strings from which the second set of topics were generated, and/or attributes for the second set of topics indicating a relevance of the second set of topics to the record object. The computer can iteratively repeat this process any number of times to until processing each set or chunk of text strings for the record object.").
Bogdan and Rogynskyy are analogous arts because they each belong to the same field of data processing. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the data analytics system of Bogdan to incorporate the teachings of Rogynskyy to generate various subsets of topics while determining the relation between topics and input text. This allows the system to identify certain subcategories of topics that may be particularly important or require further processing (Rogynskyy, [0401]). As such, the system can effectively identify different types of data that the user may want handled in different ways.
Regarding claim 11, the rejection of claim 10 is incorporated. Bogdan discloses all of the elements of the current invention as stated above. However, Bogdan fails to expressly recite further comprising instructions that, when executed by the at least one processor, cause the system to iteratively refine the topic data structure by providing the topics to the topic generation model, wherein iteratively refining the topic data structure modifies the topic data structure by expanding at least one topic into one or more subtopics.
Rogynskyy teaches further comprising instructions that, when executed by the at least one processor, cause the system to iteratively refine the topic data structure by providing the topics to the topic generation model, wherein iteratively refining the topic data structure modifies the topic data structure by expanding at least one topic into one or more subtopics (Rogynskyy, [0401]: "The computer can repeat the chunking process iteratively for each set of text strings. After processing an initial set of text strings, the inferences from the large language model (e.g., topics and their corresponding subsets) can be used to update and/or fine-tune the analysis for subsequent sets of text strings. For instance, the computer can identify one or more subsets of text strings corresponding to highly relevant topics that the large language model identified based on the initial set of text strings. The computer can identify the one or more subsets based on their attributes indicating that the topics have a level of relevancy above a threshold with the record object and/or that the attributes of the topics satisfy some other criteria. The computer can retrieve the identified subsets of text strings via their respective references output by the large language model. The computer can input the retrieved subsets of text strings, along with a second set of text strings (e.g., a second set of text strings generated from electronic activities generated and/or received sequentially after the first set of electronic activities) that correspond to a different set of text strings from the first set of text strings, into the large language model. Based on the input, the large language model can generate a second set of topics, references to text strings from which the second set of topics were generated, and/or attributes for the second set of topics indicating a relevance of the second set of topics to the record object. The computer can iteratively repeat this process any number of times to until processing each set or chunk of text strings for the record object.").
Bogdan and Rogynskyy are analogous arts because they each belong to the same field of data processing. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the data analytics system of Bogdan to incorporate the teachings of Rogynskyy to iteratively refine the data structure. Iterative processing helps the system improve accuracy while minimizing hallucinations (Rogynskyy, [0401]). This helps ensure that the system provides error free outputs for the user.
Regarding claim 15, the rejection of claim 10 is incorporated. Bogdan discloses all of the elements of the current invention as stated above. However, Bogdan fails to expressly recite further comprising instructions that, when executed by the at least one processor, cause the system to: determine a first subset of topics from a first subset of the plurality of verbatims; determine a second subset of topics from a second subset of the plurality of verbatims; and generate the topics by combining the first subset of topics and the second subset of topics and deduping duplicate topics.
Rogynskyy teaches further comprising instructions that, when executed by the at least one processor, cause the system to: determine a first subset of topics from a first subset of the plurality of verbatims (Rogynskyy, [0400]: "Based on the execution, the large language model can generate (i) a first set of topics, (ii) a first set of references indicating one or more subsets of the first set of text strings, each subset of text strings corresponding to a different topic of the first set of topics, and (iii) an attribute for each of the first set of topics. For each text string in the subset of text strings corresponding to a topic, the large language model can generate a reference. Each reference can function as a pointer that identifies specific text strings within the first set of strings (e.g., chunk) that contributed to generating the topic. Each attribute for a topic can indicate a level of relevance of the topic to the opportunity record object."); determine a second subset of topics from a second subset of the plurality of verbatims (Rogynskyy, [0401]: "The computer can repeat the chunking process iteratively for each set of text strings. After processing an initial set of text strings, the inferences from the large language model (e.g., topics and their corresponding subsets) can be used to update and/or fine-tune the analysis for subsequent sets of text strings. For instance, the computer can identify one or more subsets of text strings corresponding to highly relevant topics that the large language model identified based on the initial set of text strings. The computer can identify the one or more subsets based on their attributes indicating that the topics have a level of relevancy above a threshold with the record object and/or that the attributes of the topics satisfy some other criteria. The computer can retrieve the identified subsets of text strings via their respective references output by the large language model."); and generate the topics by combining the first subset of topics and the second subset of topics and deduping duplicate topics (Rogynskyy, [0401]: "The computer can input the retrieved subsets of text strings, along with a second set of text strings (e.g., a second set of text strings generated from electronic activities generated and/or received sequentially after the first set of electronic activities) that correspond to a different set of text strings from the first set of text strings, into the large language model. Based on the input, the large language model can generate a second set of topics, references to text strings from which the second set of topics were generated, and/or attributes for the second set of topics indicating a relevance of the second set of topics to the record object. The computer can iteratively repeat this process any number of times to until processing each set or chunk of text strings for the record object.").
Bogdan and Rogynskyy are analogous arts because they each belong to the same field of data processing. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the data analytics system of Bogdan to incorporate the teachings of Rogynskyy to generate various subsets of topics while determining the relation between topics and input text. This allows the system to identify certain subcategories of topics that may be particularly important or require further processing (Rogynskyy, [0401]). As such, the system can effectively identify different types of data that the user may want handled in different ways.
Claim(s) 5 and 14 is/are rejected under 35 U.S.C. 103 as being unpatentable over Bogdan as applied to claims 1, 2, 4, 8, 10, 16, 17, and 20 above, and further in view of Wu at al. (US Pat. Pub. No. 2012/0166438 A1 hereinafter Wu).
Regarding claim 5, the rejection of claim 4 is incorporated. Bogdan discloses all of the elements of the current invention as stated above. However, Bogdan fails to expressly recite further comprising: determining, based on receiving the additional verbatims, a change in a volume of verbatims satisfies a change threshold by comparing a quantity of the additional verbatims to a quantity of the plurality of verbatims; and updating the topic data structure based on satisfying the change threshold.
Wu teaches further comprising: determining, based on receiving the additional verbatims, a change in a volume of verbatims satisfies a change threshold by comparing a quantity of the additional verbatims to a quantity of the plurality of verbatims (Wu, [0040]: "Volume calculation module 204 processes the words and named entities produced by content segmenting module 202 for the current time interval to calculate a volume for each of a predefined number of topics for the current interval."; [0043]: "Once volumes have been calculated for each of the predefined number of topics for the current time interval by volume calculation module 204, topic list generation module 206 compares the current interval volume to a mean volume calculated over a historical time period for each of the predefined number of topics to calculate a deviation for each of the predefined number of topics. The calculated deviation for each of the predefined number of topics comprises the trending score for the topic."); and updating the topic data structure based on satisfying the change threshold (Wu, [0045]: "Topic list generation module 206 then compares the trending score obtained for each of the predefined number of topics (i.e., the deviation measure obtained for each of the predefined number of topics) to a threshold value."; [0046]: "If topic list generation module 206 determines that a topic in the predefined number of topics has a trending score that exceeds the threshold value, then topic list generation module 206 will add the topic to a list of trending topics included in data structure 116.").
Bogdan and Wu are analogous arts because they each belong to the same field of data processing. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the data analytics system of Bogdan to incorporate the teachings of Wu to update a data structure in response to a threshold amount of new data. This helps the system stay up-to-date with time-sensitive information relating to important topics (Wu, [0006]). As such, the system is able to effectively incorporate new data into previously existing data structures.
Regarding claim 14, the rejection of claim 10 is incorporated. Bogdan discloses all of the elements of the current invention as stated above. However, Bogdan fails to expressly recite further comprising instructions that, when executed by the at least one processor, cause the system to: determine, based on receiving additional verbatims, a quantity of additional verbatims satisfies a change threshold based on the quantity of the additional verbatims; and update the topic data structure based on satisfying the change threshold.
Wu teaches further comprising instructions that, when executed by the at least one processor, cause the system to: determine, based on receiving additional verbatims, a quantity of additional verbatims satisfies a change threshold based on the quantity of the additional verbatims (Wu, [0040]: "Volume calculation module 204 processes the words and named entities produced by content segmenting module 202 for the current time interval to calculate a volume for each of a predefined number of topics for the current interval."; [0043]: "Once volumes have been calculated for each of the predefined number of topics for the current time interval by volume calculation module 204, topic list generation module 206 compares the current interval volume to a mean volume calculated over a historical time period for each of the predefined number of topics to calculate a deviation for each of the predefined number of topics. The calculated deviation for each of the predefined number of topics comprises the trending score for the topic."); and update the topic data structure based on satisfying the change threshold (Wu, [0045]: "Topic list generation module 206 then compares the trending score obtained for each of the predefined number of topics (i.e., the deviation measure obtained for each of the predefined number of topics) to a threshold value."; [0046]: "If topic list generation module 206 determines that a topic in the predefined number of topics has a trending score that exceeds the threshold value, then topic list generation module 206 will add the topic to a list of trending topics included in data structure 116.").
Bogdan and Wu are analogous arts because they each belong to the same field of data processing. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the data analytics system of Bogdan to incorporate the teachings of Wu to update a data structure in response to a threshold amount of new data. This helps the system stay up-to-date with time-sensitive information relating to important topics (Wu, [0006]). As such, the system is able to effectively incorporate new data into previously existing data structures.
Claim(s) 12 and 19 is/are rejected under 35 U.S.C. 103 as being unpatentable over Bogdan as applied to claims 1, 2, 4, 8, 10, 16, 17, and 20 above, and further in view of Jeong et al. (US Pat. Pub. No. 2023/0103834 A1 hereinafter Jeong).
Regarding claim 12, the rejection of claim 10 is incorporated. Bogdan discloses all of the elements of the current invention as stated above. However, Bogdan fails to expressly recite further comprising instructions that, when executed by the at least one processor, cause the system to assign the topic to the verbatim by determining the correlation between the verbatim and the topic satisfies a threshold semantic similarity metric associated with a semantic relevance of the topic to the verbatim.
Jeong teaches further comprising instructions that, when executed by the at least one processor, cause the system to assign the topic to the verbatim by determining the correlation between the verbatim and the topic satisfies a threshold semantic similarity metric associated with a semantic relevance of the topic to the verbatim (Jeong, [0042]: "The Semantic Similarity Score Generator 116 is also responsible for generating the overall similarity score for a given topic/sentence combination. Such a calculation may be performed by, for example, averaging the top 5 or 10 similarity scores for a given topic. The resulting calculated score may then be compared to a threshold value (e.g., as specified in the input data for the given topic) in order to determine whether or not the corresponding topic (or keyword) will be marked as relevant to the given sentence of the subject text.").
Bogdan and Jeong are analogous arts because they each belong to the same field of data processing. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the data analytics system of Bogdan to incorporate the teachings of Jeong to determine correlation between a topic and the input data based on a semantic similarity threshold. This ensures that input text is accurately assigned as relevant to a topic or not (Jeong, [0043]). As such, the system will more effectively present only relevant information to the user.
Regarding claim 19, the rejection of claim 16 is incorporated. Bogdan discloses all of the elements of the current invention as stated above. However, Bogdan fails to expressly recite further comprising instructions that, when executed by the at least one processor, cause the computing device to assign the topic to the verbatim based on determining the correlation between the verbatim and the topic satisfies a threshold semantic similarity metric based on a semantic relevance of the topic to the verbatim.
Jeong teaches further comprising instructions that, when executed by the at least one processor, cause the computing device to assign the topic to the verbatim based on determining the correlation between the verbatim and the topic satisfies a threshold semantic similarity metric based on a semantic relevance of the topic to the verbatim (Jeong, [0042]: "The Semantic Similarity Score Generator 116 is also responsible for generating the overall similarity score for a given topic/sentence combination. Such a calculation may be performed by, for example, averaging the top 5 or 10 similarity scores for a given topic. The resulting calculated score may then be compared to a threshold value (e.g., as specified in the input data for the given topic) in order to determine whether or not the corresponding topic (or keyword) will be marked as relevant to the given sentence of the subject text.").
Bogdan and Jeong are analogous arts because they each belong to the same field of data processing. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the data analytics system of Bogdan to incorporate the teachings of Jeong to determine correlation between a topic and the input data based on a semantic similarity threshold. This ensures that input text is accurately assigned as relevant to a topic or not (Jeong, [0043]). As such, the system will more effectively present only relevant information to the user.
Claim(s) 13 is/are rejected under 35 U.S.C. 103 as being unpatentable over Bogdan as applied to claims 1, 2, 4, 8, 10, 16, 17, and 20 above, and further in view of Shukla et al. (US Pat. Pub. No. 2025/0321986 A1 hereinafter Shukla).
Regarding claim 13, the rejection of claim 10 is incorporated. Bogdan discloses all of the elements of the current invention as stated above. However, Bogdan fails to expressly recite wherein the reverb correlation model is a zero-shot semantic similarity model and further comprising instructions that, when executed by the at least one processor, cause the system to determine, utilizing the reverb correlation model, the correlation between the verbatim and the topic by evaluating a cosine similarity between the verbatim and the topic.
Shukla teaches wherein the reverb correlation model is a zero-shot semantic similarity model and further comprising instructions that, when executed by the at least one processor, cause the system to determine, utilizing the reverb correlation model, the correlation between the verbatim and the topic by evaluating a cosine similarity between the verbatim and the topic (Shukla, [0047]: "In one or many embodiment(s) described herein, and at least in part, the zero shot text model (132) may include functionality to: be invoked by the multimodal context selector controller (126), where any said invocation(s) is/are provided with a text query, encompassing an arbitrary-length of text posing one or more questions, and any number of topic-related text(s), each representative of a text topic of the any number of text topic(s) identified through hidden thematic structure discovery (described above), as inputs to undergo processing via text classification using classes (or labels) unseen during model training; and return/provide, in response to any said invocation(s) and based on said processing, a similarity metric (e.g., cosine similarity, etc.) value (also referred to herein as an query-topic similarity score) measuring a similarity between the text query and each topic-related text.").
Bogdan and Shukla are analogous arts because they each belong to the same field of data processing. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the data analytics system of Bogdan to incorporate the teachings of Shukla to use a zero shot model to calculate cosine similarity to determine the similarity between input text and a topic. This type of system is capable of fast and accurate processing (Shukla, [0020]), which improves the user’s experience by efficiently providing correct outputs.
Claim(s) 18 is/are rejected under 35 U.S.C. 103 as being unpatentable over Bogdan as applied to claims 1, 2, 4, 8, 10, 16, 17, and 20 above, and further in view of Goel et al. (US Pat. Pub. No. 2018/0308487 A1 hereinafter Goel).
Regarding claim 18, the rejection of claim 18 is incorporated. Bogdan discloses all of the elements of the current invention as stated above. However, Bogdan fails to expressly recite further comprising instructions that, when executed by the at least one processor, cause the computing device to select and train the reverb correlation model based on a target language.
Goel teaches further comprising instructions that, when executed by the at least one processor, cause the computing device to select and train the reverb correlation model based on a target language (Goel, [0042]: "After the mapping of each of the sentences, the dialogue engine 100 trains another neural network to learn the output tags for each of these vectors. The word embeddings are trained on very large amounts of text in the target language. However the classifier is subsequently trains on only a smaller number of tags that are available from the grammar. However the natural proximity of semantically similar phrases in the classifier to the grammar phrases, allow the engine to generalize and tag semantically similar phrases with the correct tags.").
Bogdan and Goel are analogous arts because they each belong to the same field of data processing. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the data analytics system of Bogdan to incorporate the teachings of Goel to train the reverb correlation model based on a target language. This allows the system to be focused on processing a single language with its own specific grammar rules and semantic similarities (Goel, [0042]). This ensures that the system works as accurately as possible in the target language by not adding the increased complexity of grammar and semantics in various languages.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Mathur et al. (US Pat. Pub No. 2022/0292125 A1) discloses a system for generating test scenarios based on failure patterns and themes in customer experiences.
Dhingra et al. (US Pat. Pub. No. 2021/0232766 A1) discloses systems for short text identification.
Marvit et al. (US Pat. No. 9,317,593 B2) discloses a system for modeling topics using statistical distributions.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to TYLER J BECKER whose telephone number is (703)756-1271. The examiner can normally be reached M-Th, 7:15am-5:45pm PT.
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, Daniel Washburn can be reached at (571) 272-5551. 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.
/TYLER BECKER/ Examiner, Art Unit 2657
/DANIEL C WASHBURN/ Supervisory Patent Examiner, Art Unit 2657
Prosecution Insights