DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Amendment
In response to the office action from 12/17/2025, the applicant has submitted an amendment, filed 3/13/2026, amending claims 1, 3, 5, 6, 10, 12, 15, 17, 18, 20, cancelling claims 2, 16, 19, adding claims 21-23, while arguing to traverse the prior art and 101 rejections. Applicant’s arguments have been fully considered but are moot with respect to new grounds of rejections further in view of KWON et al. (KR20230075322) and for the reasons explained in the response to arguments.
Response to Arguments
Following a broad overview of the last office action on page 10, and a diagram associated with the “2024 Guidance Update on Patent Subject Matter Eligibility” on page 11, on page 12 the 2nd ¶, following a copy of the claims 1, 15 and 18 as were examined (i.e., not the latest amended version), it is concluded that: “Such recitations or particular processing operations are clearly not mental processes as alleged, as human mind is not equipped to perform these claim limitations”.
Please visit the previous (as well as the current office action that follows) for how it was shown that each one of the steps could be carried out by a human without any use of any particular machine (i.e. last office action page 3 last ¶).
On page 12 the last ¶, it is recited: “Applicant” “submits that previously-presented independent claims 1, 15 and 18 clearly recite arrangements providing an improvement in computer technology” “For example” “for enhancing machine learning models such as large language models (LLMs) to enable the LLMs to provide relevant answers to queries based on table data utilizing retrieval augmented generation (RAG) processing” “This enhance RAG processing with table comprehension achieves improved results for queries related to tabular data”.
Respectfully there are two flaws here: 1) there is no discussion on how each of the quoted claim limitations have impacted the additional elements (including the “machine learning” as well as the “one or more processors”) in reverse or what was explained in the office action as setting any limitations on “practicing the abstract idea” (page 5 ¶ 1); 2) improvement is assessed based on the impact of the quoted claim limitations on the said additional elements recited in the claims; please see Enfish and McRo quoted here: “teachings that the claimed invention achieves other benefits over conventional databases, such as increased flexibility, faster search times, and smaller memory requirements” Enfish memorandum of 5/19/2016); and/or “An “improvement in computer-related technology” is not limited to improvements in the operation of a computer or a computer network per se, but may also be claimed as a set of “rules” (basically mathematical relationships) that improve computer-related technology by allowing computer performance of a function not previously performable by a computer” (MCRO memorandum of 11/2/2016). In conclusion respectfully the applicant should find teachings in which the claim’s “processors” and/or “machine learning” can meet such criteria in order to fulfill the required “improvement” criteria.
In conclusion the examiner must respectfully disagree with the conclusion: “It is therefore respectfully submitted that the that the § 101 rejection is improper and should be withdrawn” (page 13 ¶ 2).
The remainder of the arguments (page 13 last ¶ through the end of the 2nd ¶ on page 14) are directed as to why prior art of record fails to teach the latest amendments and the new claims.
Please visit the new office action for further details.
Finally on page 14 lines 7-5 above the bottom it is recited: “The dependent claims are believed to be patentable by virtue of their respective dependencies from the independent claims.
Since applicants have not argued the merits of these dependent claims, but assert patentability solely through their dependence on the allegedly patentable parent claims, they stand or fall with said parent claims and hence no further response to applicant’s arguments is necessary.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1, 3-15, 17-18, 20-23 stand rejected:
Claims 1, 15 and 18 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claims recite a “query” directed at “a portion of content” that is believed to be present in a “tabular data” (basically “a table” (Sp. ¶ 0027 L 19)) in a “document” among a plurality of “documents”. The “query” even “identif[ies]” those “documents” required to be searched as a “context” accompanying a “search text”, where the “search text” and the “context” together form the “query”.
So following reception of the “query”, the “apparatus” and/or “model” “generate” “plurality of document chunks” where each “chunk” possesses “representations that maintain tabular formatting” (basically tables which may possess “images included in the given tabular data structure” (Sp. ¶ 0046 S1) also called “placeholders” and/or “representations” “compris[ing]” “plain-text” (Sp. ¶ 0044 S1)). Then among these “document chunks” “a plurality” in which the “search text” of the “query” indicate a “similarity” above a “threshold” (Sp. ¶ 0045 line 7 above the bottom) with those “document chunks” are “select[ed]” and used as “prompts” to a “machine learning system” “to generate an output” which comprises “at least a portion of content from at least one or the one or more tabular data structures”. Prior to submission of the “document chunks” to the “machine learning”, a “determinat[ion]” of an “alignment of the given tabular data structure” “relative to additional data structure” associated with the basically “document chunks” is made and used in “generating the document chunks”. These limitations, as drafted, are processes that, under their broadest reasonable interpretation, cover performance in the mind but for the recitation of generic computer components. That is other than reciting using “a processor” (claim 1), “processing device” (claim 18), nothing in these limitations precludes their steps from practically being performed in the mind.
For example, but for the “processor” (claim 1), and “processing device” (claim 18), the context of these claims encompass, e.g. in a hypothetical situation a professor could ask (prompt) a graduate student researcher to look for a “specific item” believed to be in a table and/or tables within a class and/or of books, and/or scientific journal articles and even give the title of those books and/or articles to the student to search in a library. The “specific item” may adhere to a certain “alignment” format in the table which could make the search easier if that format is given to the student in which case when he is searching he could skip pages not containing tables not adhering to such format or even easier not possessing any tables. The student could spend perhaps days and/or even months but he can certainly identify tables within each book and/or journal and compare their entries with the professor’s “specific item” and determine potential matches. But in a realistic scenario if there are hundreds if not thousands of tables in the said documents, is it really possible for the student to search all of them, but here the claim recites “one or more documents”, which brings it to a domain manageable by a human. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claims recite an abstract idea.
The judicial exception is not integrated into a practical application. In particular, the claims only recited additional elements, namely the hardware component “processor” and the software “machine learning” are recited at a high-level of generality; i.e., the single “processor” is responsible for the steps of “obtain” “a query”…, “generate” “document chunks” …, “select” “document chunks”, “generate” “prompt” …, “apply the prompt …”, “provide an answer …”, without any specific details on how each the “processor” is involved in each step. Likewise the no details are provided on how the “machine learning” is utilized in “generat[ing]” “an output”, and also no specific “machine learning” technique is even spelled out throughout the disclosure, and the claims do not assign a task to the said “machine learning” which really cannot be done without it. Accordingly, these additional elements do not impose any meaningful limits on practicing the abstract idea. The claims are thus directed to an abstract idea.
The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to integration of the abstract idea into a practical application, the additional elements of using a processor to perform all the claimed steps amount to no more than mere instructions to apply the exception using a generic computer component. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. The claims are not patent eligible.
Regarding claims 3, 4, determining coordinates (including horizontal alignment) of a table in a document, book and/or research paper will only require a ruler and/or measuring tape; identification of textual content within the said table will require an ability to read and write.
Regarding claims 5, 17 and 20, the student could decipher any “placeholder” he encounters in any of the tables of the document or book or article that he finds. Deciphering a “placeholder” would not require any particular machine.
Regarding claims 6, and 21, determining coordinates of a table in an article will require a ruler and/or measuring tape. Deciphering a “placeholder” would not require any particular machine.
Regarding claims 7, and 22, the student could present an answer to his professor’s question with (augmenting it with) an image copy of the table the answer was taken.
Regarding claims 8, and 23, the student would intuitively choose entries of a table that are most similar (both in terms of key words and conceptually) to the queries subject to search.
Regarding claim 9, recognition of table caption will only require the ability to read and write.
Regarding claim 10, to summarize contents of table would also require grammar school level literacy.
Regarding claim 11, the professor instructing his student, could provide him (prompt him) with an organized list of items (a template) to search for. This template may be done in writing on a sheet of pager.
Regarding claim 12, this could arguably be performed through mental steps with the aid of pen and paper, e.g. a human (e.g., the graduate student) mentally deciding to add an image, where to add it, and drawing it using a pen on the paper when he is presenting the result of his searches in a table containing that image and add it (augment it) with other content (likely textual content) of relevant portions of the table.
Regarding claims 13 and 14, to direct the methods to information technology (and even additionally to associated trouble shooting) amounts to an intended use, and the method limitations do not indicate they can enhance performance of a query system to an information technology application.
Claim Rejections - 35 USC § 102
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.
Claim(s) 1, 15, 18 is/are rejected under 35 U.S.C. 102(a)(1) as being anticipated by KWON HYUKCHUL et al. (KR20230075322A) (Kwon et al.).
Regarding claim 1, Kwon et al. do teach an apparatus comprising: at least one processing device comprising a processor coupled to a memory (Page 1 line 4: “System” (an apparatus or processing device comprising processor and memory) “and Method for Realtime Table Question Answering for Long Sequence”);
the at least one processing device being configured:
to obtain a query, the query comprising search text and a context, the context identifying one or more documents to be searched using the search text, at least one of the one or more documents comprising one or more tabular data structures (¶ 0041: “the user inputs” (to input a query) “table data” (comprising tabular data structure as context) “for fining correct answer” “and the question” (also part of the query) “for which the correct answer is to be found”, ¶ 0024 page 10 lines 2-4: “input to the question-answering” (e.g., the “question” (query)) “model is input by simplifying it into” “entity text” (is converted into search text); ¶ 0042: “table data” (the tabular data comprising of “row and column” (tabular data structure (¶ 0045))) “that can be entered includes table data existing in web documents or office documents”(is included into one or more identified documents));
to generate a plurality of document chunks by parsing the one or more documents, each of the plurality of document chunks comprising a portion of content of one of the one or more documents, wherein the one or more tabular data structures are replaced in the plurality of document chunks by one or more tabular data structure representations that maintain a tabular formatting of the one or more tabular data structures (¶ 0045: “In the table” (the table included in the document or part of the document) “flattening and refining unit” “the two dimensional table information is converted into a one dimensional token sequence”(is parsed) “form, and the index of the row and column” “of each cell” (into “cells” (document chunks)) “had before being converted are attached to provide two-dimensional information” (i.e., with a tabular data structure representation; each “cell” maintains a portion of the “table” content, and each “table” is included in the “document”, so the “cell” (document chunk) has a portion of the content of the document) ;
to select a subset of the plurality of document chunks based at least in part on determining a similarity between content of the plurality of document chunks and the search text, the selected subset of the plurality of document chunks comprising at least one document chunk comprising at least one of the one or more tabular data structure representations (¶ 0033 last 4 lines: “selecting rows and columns containing correct answers and selecting” (to select) “correct cells” (a subset of document chunks) “in a correct answer output unit”; where the “correct cells” are obtained using a “similarity” calculation; i.e., Abstract lines 17+: “a sentence vector output unit outputting a sentence vector for the similarity comparison” (determining a similarity) “between questions” (between the query search text and) “and cells” (and a subset of document chunks that obey the “similarity” (similarity)) “of the table” (corresponding to tabular data); ¶ 0066 last sentence: “the location of the cell” (document chunk) “indicating the correct answer” (selected (i.e., with the correct answer)) “can be predicted through the row and column with the highest similarity” (obtained according to the similarity calculation wherein the “cells” (at least one document chunk) are defined by the “two-dimensional information” (the data structure representation (¶ 0045)));
to generate, based at least in part on the query, a prompt for input to a machine learning system, the prompt comprising the selected subset of the plurality of document chunks (¶ 0044: “The table data” (i.e., “cells” (plurality of document chunks comprising their selected subset)) “can be input” (are used to generate a prompt) “into a neural network-based language model” (to a machine learning system); ¶ 0061 last sentence: “embedding vector of each cell of the table data” (i.e., “table data” comprises of “cell” (document chunk) data which itself also comprises of the “correct” “cells” (the subset of document chunks)));
to apply the prompt to the machine learning system to generate an output (¶ 0060: “to search for the correct answer” (to generate an output of “correct” “cells” (the selected subset of the document chunks)) “input” (prompt) “to a neural network” (to the machine learning system) “language model”); and
to provide an answer to the query based at least in part on the output of the machine learning system, the answer comprising at least a portion of content from at least one of the one or more tabular data structures (¶ 0016 last sentence: “a question-answering learning unit for learning a task in which a neural network” (the machine learning) “based language model finds” (provides) “a correct answer” (an answer to the “question” (query)) “for rows and columns” (corresponding to one or more tabular data structure or document chunks which are associated with those rows and columns which correspond to “cells” (which obey the “similarity” condition)));
wherein generating a given one of the plurality of document chunks comprising a given portion of content of a given one of the one or more documents including a given one of the one or more tabular data structures, the given tabular data structure comprising first textual content comprises:
determining an alignment of the given tabular data structure in the given portion of the content of the given document relative to additional textural content of the given portion of the content of the given document (¶ 0045: “In the table” (the table included in the document or part of the document used) “flattening and refining unit” “the two dimensional table information” (determining an alignment information) “is converted” “into a one dimensional token sequence” “form, and the index” (between the tabular data structure) “of the row and column” (and additional textual content or first textual content (i.e., ¶ 0021 line 3: “texts” (i.e. additional or first textual content) “existing in the selected rows and columns”) “of each cell” (relative to “cells” (portion of the document content)) “had before being converted are attached to provide two-dimensional information”);
extracting the given tabular data structure from the given portion of the content of the given document (¶ 0045 “the two dimensional table information” is used to extract “the index” (the tabular data structure) “of the row and column” of “each cell” (the given portion of the given document));
generating a plain-text representation of the first textual content and the tabular formatting of the given tabular data structure (for each “row and column” “of each cell” (first textual content) according to ¶ 0045 an “index” (tabular data structure in plain text) is determined and also according to ¶ 0021 “texts” (plain text) “existing in the rows and columns” (in the first content associated with the tabular data));
extracting the additional textual content from the given portion of the content of the given document (¶ 0045: “In the table” (the table included in the document or part of the document used) “flattening and refining unit” “the two dimensional table information” (with the determined alignment) “is converted” (to generate) “into a one dimensional token sequence” “form, and the index of the row and column” (using additional textual content (i.e., ¶ 0021 “texts” (i.e. additional textual content) “existing” (extracted) “in the selected rows and columns”)) “of each cell” (from the given portion of the document));
and generating the given document chunk by merging the plain-text representation of the first textual content and the tabular formatting of the given tabular data structure with the additional textual content extracted from the given portion of the content of the given document while maintaining the determined alignment (¶ 0045: “In the table” (the table included in the document or part of the document used) “flattening and refining unit” “the two dimensional table information” (with the determined alignment) “is converted” “into a one dimensional token sequence” “form, and the index of the row and column” (by merging with the additional textual content or content’s plain text representation (i.e., ¶ 0021 “texts” (i.e. textual content) “existing in the selected rows and columns”)) “of each cell” (to generate “cells” (document chunks)) “had before being converted are attached to provide two-dimensional information”).
Regarding claim 15, Kwon et al. do teach a computer program product comprising a non-transitory processor-readable storage medium having stored therein program code of one or more software programs (Page 1 line 4: “System” (an apparatus or processing device associated with a medium) “and Method for Realtime Table Question Answering for Long Sequence”),
wherein the program code when executed by the at least one processing device causes the at least one processing device:
to obtain a query, the query comprising search text and a context, the context identifying one or more documents to be searched using the search text, at least one of the one or more documents comprising one or more tabular data structures (¶ 0041: “the user inputs” (to input a query) “table data” (comprising tabular data structure as context) “for fining correct answer” “and the question” (also part of the query) “for which the correct answer is to be found”, ¶ 0024 page 10 lines 2-4: “input to the question-answering” (e.g., the “question” (query)) “model is input by simplifying it into” “entity text” (is converted into search text); ¶ 0042: “table data” (the tabular data comprising of “row and column” (tabular data structure (¶ 0045))) “that can be entered includes table data existing in web documents or office documents”(is included into one or more identified documents));
to generate a plurality of document chunks by parsing the one or more documents, each of the plurality of document chunks comprising a portion of content of one of the one or more documents, wherein the one or more tabular data structures are replaced in the plurality of document chunks by one or more tabular data structure representations that maintain a tabular formatting of the one or more tabular data structures (¶ 0045: “In the table” (the table included in the document or part of the document) “flattening and refining unit” “the two dimensional table information is converted into a one dimensional token sequence”(is parsed) “form, and the index of the row and column” “of each cell” (into “cells” (document chunks)) “had before being converted are attached to provide two-dimensional information” (i.e., with a tabular data structure representation; each “cell” maintains a portion of the “table” content, and each “table” is included in the “document”, so the “cell” (document chunk) has a portion of the content of the document) ;
to select a subset of the plurality of document chunks based at least in part on determining a similarity between content of the plurality of document chunks and the search text, the selected subset of the plurality of document chunks comprising at least one document chunk comprising at least one of the one or more tabular data structure representations (¶ 0033 last 4 lines: “selecting rows and columns containing correct answers and selecting” (to select) “correct cells” (a subset of document chunks) “in a correct answer output unit”; where the “correct cells” are obtained using a “similarity” calculation; i.e., Abstract lines 17+: “a sentence vector output unit outputting a sentence vector for the similarity comparison” (determining a similarity) “between questions” (between the query search text and) “and cells” (and a subset of document chunks that obey the “similarity” (similarity)) “of the table” (corresponding to tabular data); ¶ 0066 last sentence: “the location of the cell” (document chunk) “indicating the correct answer” (selected (i.e., with the correct answer)) “can be predicted through the row and column with the highest similarity” (obtained according to the similarity calculation wherein the “cells” (at least one document chunk) are defined by the “two-dimensional information” (the data structure representation (¶ 0045)));
to generate, based at least in part on the query, a prompt for input to a machine learning system, the prompt comprising the selected subset of the plurality of document chunks (¶ 0044: “The table data” (i.e., “cells” (plurality of document chunks comprising their selected subset)) “can be input” (are used to generate a prompt) “into a neural network-based language model” (to a machine learning system); ¶ 0061 last sentence: “embedding vector of each cell of the table data” (i.e., “table data” comprises of “cell” (document chunk) data which itself also comprises of the “correct” “cells” (the subset of document chunks)));
to apply the prompt to the machine learning system to generate an output (¶ 0060: “to search for the correct answer” (to generate an output of “correct” “cells” (the selected subset of the document chunks)) “input” (prompt) “to a neural network” (to the machine learning system) “language model”); and
to provide an answer to the query based at least in part on the output of the machine learning system, the answer comprising at least a portion of content from at least one of the one or more tabular data structures (¶ 0016 last sentence: “a question-answering learning unit for learning a task in which a neural network” (the machine learning) “based language model finds” (provides) “a correct answer” (an answer to the “question” (query)) “for rows and columns” (corresponding to one or more tabular data structure or document chunks which are associated with those rows and columns which correspond to “cells” (which obey the “similarity” condition)));
wherein generating a given one of the plurality of document chunks comprising a given portion of content of a given one of the one or more documents including a given one of the one or more tabular data structures, the given tabular data structure comprising first textual content comprises:
determining an alignment of the given tabular data structure in the given portion of the content of the given document relative to additional textural content of the given portion of the content of the given document (¶ 0045: “In the table” (the table included in the document or part of the document used) “flattening and refining unit” “the two dimensional table information” (determining an alignment information) “is converted” “into a one dimensional token sequence” “form, and the index” (between the tabular data structure) “of the row and column” (and additional textual content or first textual content (i.e., ¶ 0021 line 3: “texts” (i.e. additional or first textual content) “existing in the selected rows and columns”) “of each cell” (relative to “cells” (portion of the document content)) “had before being converted are attached to provide two-dimensional information”);
extracting the given tabular data structure from the given portion of the content of the given document (¶ 0045 “the two dimensional table information” is used to extract “the index” (the tabular data structure) “of the row and column” of “each cell” (the given portion of the given document));
generating a plain-text representation of the first textual content and the tabular formatting of the given tabular data structure (for each “row and column” “of each cell” (first textual content) according to ¶ 0045 an “index” (tabular data structure in plain text) is determined and also according to ¶ 0021 “texts” (plain text) “existing in the rows and columns” (in the first content associated with the tabular data));
extracting the additional textual content from the given portion of the content of the given document (¶ 0045: “In the table” (the table included in the document or part of the document used) “flattening and refining unit” “the two dimensional table information” (with the determined alignment) “is converted” (to generate) “into a one dimensional token sequence” “form, and the index of the row and column” (using additional textual content (i.e., ¶ 0021 “texts” (i.e. additional textual content) “existing” (extracted) “in the selected rows and columns”)) “of each cell” (from the given portion of the document));
and generating the given document chunk by merging the plain-text representation of the first textual content and the tabular formatting of the given tabular data structure with the additional textual content extracted from the given portion of the content of the given document while maintaining the determined alignment (¶ 0045: “In the table” (the table included in the document or part of the document used) “flattening and refining unit” “the two dimensional table information” (with the determined alignment) “is converted” “into a one dimensional token sequence” “form, and the index of the row and column” (by merging with the additional textual content or content’s plain text representation (i.e., ¶ 0021 “texts” (i.e. textual content) “existing in the selected rows and columns”)) “of each cell” (to generate “cells” (document chunks)) “had before being converted are attached to provide two-dimensional information”).
Regarding claim 18, Kwon et al. do teach a method comprising:
obtaining a query, the query comprising search text and a context, the context identifying one or more documents to be searched using the search text, at least one of the one or more documents comprising one or more tabular data structures (¶ 0041: “the user inputs” (to input a query) “table data” (comprising tabular data structure as context) “for fining correct answer” “and the question” (also part of the query) “for which the correct answer is to be found”, ¶ 0024 page 10 lines 2-4: “input to the question-answering” (e.g., the “question” (query)) “model is input by simplifying it into” “entity text” (is converted into search text); ¶ 0042: “table data” (the tabular data comprising of “row and column” (tabular data structure (¶ 0045))) “that can be entered includes table data existing in web documents or office documents”(is included into one or more identified documents));
generating a plurality of document chunks by parsing the one or more documents, each of the plurality of document chunks comprising a portion of content of one of the one or more documents, wherein the one or more tabular data structures are replaced in the plurality of document chunks by one or more tabular data structure representations that maintain a tabular formatting of the one or more tabular data structures (¶ 0045: “In the table” (the table included in the document or part of the document) “flattening and refining unit” “the two dimensional table information is converted into a one dimensional token sequence”(is parsed) “form, and the index of the row and column” “of each cell” (into “cells” (document chunks)) “had before being converted are attached to provide two-dimensional information” (i.e., with a tabular data structure representation; each “cell” maintains a portion of the “table” content, and each “table” is included in the “document”, so the “cell” (document chunk) has a portion of the content of the document) ;
selecting a subset of the plurality of document chunks based at least in part on determining a similarity between content of the plurality of document chunks and the search text, the selected subset of the plurality of document chunks comprising at least one document chunk comprising at least one of the one or more tabular data structure representations (¶ 0033 last 4 lines: “selecting rows and columns containing correct answers and selecting” (to select) “correct cells” (a subset of document chunks) “in a correct answer output unit”; where the “correct cells” are obtained using a “similarity” calculation; i.e., Abstract lines 17+: “a sentence vector output unit outputting a sentence vector for the similarity comparison” (determining a similarity) “between questions” (between the query search text and) “and cells” (and a subset of document chunks that obey the “similarity” (similarity)) “of the table” (corresponding to tabular data); ¶ 0066 last sentence: “the location of the cell” (document chunk) “indicating the correct answer” (selected (i.e., with the correct answer)) “can be predicted through the row and column with the highest similarity” (obtained according to the similarity calculation wherein the “cells” (at least one document chunk) are defined by the “two-dimensional information” (the data structure representation (¶ 0045)));
generating, based at least in part on the query, a prompt for input to a machine learning system, the prompt comprising the selected subset of the plurality of document chunks (¶ 0044: “The table data” (i.e., “cells” (plurality of document chunks comprising their selected subset)) “can be input” (are used to generate a prompt) “into a neural network-based language model” (to a machine learning system); ¶ 0061 last sentence: “embedding vector of each cell of the table data” (i.e., “table data” comprises of “cell” (document chunk) data which itself also comprises of the “correct” “cells” (the subset of document chunks)));
applying the prompt to the machine learning system to generate an output (¶ 0060: “to search for the correct answer” (to generate an output of “correct” “cells” (the selected subset of the document chunks)) “input” (prompt) “to a neural network” (to the machine learning system) “language model”); and
providing an answer to the query based at least in part on the output of the machine learning system, the answer comprising at least a portion of content from at least one of the one or more tabular data structures (¶ 0016 last sentence: “a question-answering learning unit for learning a task in which a neural network” (the machine learning) “based language model finds” (provides) “a correct answer” (an answer to the “question” (query)) “for rows and columns” (corresponding to one or more tabular data structure or document chunks which are associated with those rows and columns which correspond to “cells” (which obey the “similarity” condition)));
wherein generating a given one of the plurality of document chunks comprising a given portion of content of a given one of the one or more documents including a given one of the one or more tabular data structures, the given tabular data structure comprising first textual content comprises:
determining an alignment of the given tabular data structure in the given portion of the content of the given document relative to additional textural content of the given portion of the content of the given document (¶ 0045: “In the table” (the table included in the document or part of the document used) “flattening and refining unit” “the two dimensional table information” (determining an alignment information) “is converted” “into a one dimensional token sequence” “form, and the index” (between the tabular data structure) “of the row and column” (and additional textual content or first textual content (i.e., ¶ 0021 line 3: “texts” (i.e. additional or first textual content) “existing in the selected rows and columns”) “of each cell” (relative to “cells” (portion of the document content)) “had before being converted are attached to provide two-dimensional information”);
extracting the given tabular data structure from the given portion of the content of the given document (¶ 0045 “the two dimensional table information” is used to extract “the index” (the tabular data structure) “of the row and column” of “each cell” (the given portion of the given document));
generating a plain-text representation of the first textual content and the tabular formatting of the given tabular data structure (for each “row and column” “of each cell” (first textual content) according to ¶ 0045 an “index” (tabular data structure in plain text) is determined and also according to ¶ 0021 “texts” (plain text) “existing in the rows and columns” (in the first content associated with the tabular data));
extracting the additional textual content from the given portion of the content of the given document (¶ 0045: “In the table” (the table included in the document or part of the document used) “flattening and refining unit” “the two dimensional table information” (with the determined alignment) “is converted” (to generate) “into a one dimensional token sequence” “form, and the index of the row and column” (using additional textual content (i.e., ¶ 0021 “texts” (i.e. additional textual content) “existing” (extracted) “in the selected rows and columns”)) “of each cell” (from the given portion of the document));
and generating the given document chunk by merging the plain-text representation of the first textual content and the tabular formatting of the given tabular data structure with the additional textual content extracted from the given portion of the content of the given document while maintaining the determined alignment (¶ 0045: “In the table” (the table included in the document or part of the document used) “flattening and refining unit” “the two dimensional table information” (with the determined alignment) “is converted” “into a one dimensional token sequence” “form, and the index of the row and column” (by merging with the additional textual content or content’s plain text representation (i.e., ¶ 0021 “texts” (i.e. textual content) “existing in the selected rows and columns”)) “of each cell” (to generate “cells” (document chunks)) “had before being converted are attached to provide two-dimensional information”);
and wherein the method is performed by at least one processing device comprising a processor coupled to a memory (Page 1 line 4: “System” (an apparatus or processing device comprising processor and memory) “and Method for Realtime Table Question Answering for Long Sequence”).
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) 3-4 is/are rejected under 35 U.S.C. 103 as being unpatentable over Kwon et al., and further in view of Poff et al. (US 2022/0318545).
Regarding claim 3, Kwon et al. do not specifically disclose the apparatus of claim 1 wherein determining the alignment of the given tabular data structure in the given portion of the content of the given document relative to the additional textual content of the given portion of the content of the given document comprises:
determining coordinates of the given tabular data structure relative to the additional textual content in the given document chunk; and wherein generating the given document chunk by merging the plain-text representation of the first textual content and the tabular formatting of the given tabular data structure with the additional textual content extracted from the given portion of the content of the given document while maintaining the determined alignment comprises inserting the textual content and the tabular formatting of the given tabular data structure into the given document chunk at the determined coordinates.
Poff et al. do teach:
wherein determining the alignment of the given tabular data structure in the given portion of the content of the given document relative to the additional textual content of the given portion of the content of the given document comprises:
determining coordinates of the given tabular data structure relative to the additional textual content in the given document chunk (Abstract lines 7+: “During first stage, DPMC can identify candidate cells” (generating the document chunk comprises) “of the table based on analysis of the document, including identifying border” (determining coordinates) “lines that can represent cell borders” (of tabular data) “and identifying characters” (relative to additional textual content) “of the candidate cells” (in the given document chunk));
and
wherein generating the given document chunk by merging the plain-text representation of the first textual content and the tabular formatting of the given tabular data structure with the additional textual content extracted from the given portion of the content of the given document while maintaining the determined alignment comprises inserting the textual content and the tabular formatting of the given tabular data structure into the given document chunk at the determined coordinates (Abstract S 2+: “A document processing management component (DPMC) can perform a multi-stage process to extract a table from a document and recreate” “the table” (at the determined coordinates) “including the table structure” (comprising tabular formatting) “and information” “identify candidate cells” (for a given document chunk)“including identifying border lines that” “represent cell borders” “and identifying characters” (and insert e.g. “characters” (textual content)) “of the candidate cells” ).
It would have therefore been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the “document processing management component” for “process[ing]” “table[s]” of Poff et al. into the “table” “column” “row” determination and management procedures of Kwon et al. would enable the combined systems and their associated methods to perform in combination as they do separately and to further enable Kwon et al. “to generate an editable and searchable electronic textual document comprising the recreated table” as disclosed in Poff et al. ¶ 0026 S1.
Regarding claim 4, Kwon et al. do not specifically disclose the apparatus of claim 3 wherein the determined coordinates comprise a horizontal alignment of the given tabular data structure in the given document chunk.
Poff et al. do teach:
the apparatus of claim 3 wherein the determined coordinates comprise a horizontal alignment of the given tabular data structure in the given document chunk (¶ 0140 S 5: “ For instance, the DPMC can determine pairs of candidate cells that can be in line with each other in the horizontal” (doing a horizontal alignment) “direction” “e.g., candidate cell” (of the given document chunk tabular data) “and another candidate cell to the left or right of the candidate cell in the table” “or vertical direction” “e.g., candidate cell and another candidate cell above or below the candidate cell in the table”).
For obviousness to combine Kwon et al. and Poff et al. see claim 3.
Claim(s) 5-8, 17, 20-23 is/are rejected under 35 U.S.C. 103 as being unpatentable over Kwon et al., and further in view of Mathur (US 2019/0384846).
Regarding claim 5, Kwon et al. do not specifically disclose the apparatus of claim 1 wherein the given tabular data structure representation for the given tabular data structure comprises a placeholder for the given tabular data structure.
Mathur does teach the apparatus of claim 1 wherein the given tabular data structure representation for the given tabular data structure comprises a placeholder for the given tabular data structure (¶ 0084: “The set of specific (e.g., Boolean, etc.) values for the set of named variables/parameters v1, v2, v3, . . . vn may be generated at runtime based on what specific combination of columns” (a given tabular data structure representation or document chunks) “in the table” (for the given tabular data structure) “(124) to which a user query is directed, and then used to populate or fill in the” (comprises) “placeholders” (a placeholder)” “:v1”, “:v2”, “:v3”, . . . “:vn” in the union all query statement in expression (1-1) above at runtime. As a result, query results based on column values retrieved for the specific combination of columns in the table (124) to which the user query is directed can be returned as a response to the user query”).
It would have therefore been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate implementation of “placeholder” for “table” search subject to “user query” of Mathur into the corresponding one of Kwon et al. would enable the combined systems and their associated methods to perform in combination as they do separately and to further enable Kwon et al. to also benefits for being able to “dynamically bound to runtime variables to determine what columns are actually to be retrieved at runtime” as disclosed in Mathur ¶ 0026 last sentence.
Regarding claim 6, Kwon et al. do teach the apparatus of claim 5 wherein generating the given document chunk comprises:
determining a tabular data structure identifier for the given tabular data structure (¶ 0045: “In the table” “flattening and refining unit” “the two dimensional table information” (the tabular data structure) “is converted into a one dimensional token sequence” “form, and the index” (comprises an identifier) “of the row and column” “of each cell” “had before being converted are attached to provide two-dimensional information”);
determining coordinates of the given tabular data structure relative to the additional textual content in the given document chunk (¶ 0045: “In the table” “flattening and refining unit” “the two dimensional table information” (the tabular data structure) “is converted into a one dimensional token sequence” “form, and the index” “of the row and column” (relative to the additional textual content (¶ 0021 line 3: “texts” “existing in the selected rows and columns”)) “of each cell”) “of each cell” (for a given document chunk) “had before being converted are attached to provide two-dimensional information” (determining coordinates which also correspond to the tabular data structure relative to surrounding text)).
Kwon et al. do not specifically disclose:
and inserting, into the given document chunk, the placeholder for the given tabular data structure at the determined coordinates, the placeholder including a reference to the determined tabular data structure identifier.
Mathur do teach:
and inserting, into the given document chunk, the placeholder for the given tabular data structure at the determined coordinates, the placeholder including a reference to the determined tabular data structure identifier (¶ 0086 S 2: “array elements “arr[0]”, “arr[1]”, “arr[2]”, . . . “arr[n−1]” in the named variables/parameter “arr” may be generated at runtime based on what specific combination of columns” (with reference to determined “specific” (identifier) tabular data structures) “in the table (124)” (a reference to it) “to which a user query is directed, and then used to populate or fill” (inserting) “in the placeholders” (placeholders) “:arr[0]”, “:arr[1]”, “:arr[2]”, . . . “:an[n−1]” in the union all query statement in expression (1-2) above at runtime”).
For obviousness to combine Kwon et al. and Mathur see claim 5.
Regarding claim 7, Kwon et al. do teach the apparatus of claim 5 wherein generating the answer to the query comprises augmenting the output of the machine learning system with an original version of the given tabular data structure extracted from the given document responsive to determining that textual content of the output of the machine learning system is sourced from the given tabular data structure (¶ 0066: “In the question-and-answer for table data, the similarity between the embedding of rows and columns obtained from the sentence vector output unit (104) and the question vector may be compared, and the location of a cell indicating the correct answer” (the output of the machine learning system) “can be predicted through the row and column” (is augmented with “row and column” (with the tabular data structure)) “with the highest similarity”).
Regarding claim 8, Kwon et al. do teach the apparatus of claim 7 wherein generating the given document chunk further comprises determining a textual description for the given tabular data structure (¶ 0047 S 1: “the sentence vector output unit 104 generates expression vectors for row and column” (for each tabular data structure) “expression vectors for text” (generating textual description) “information existing in each cell” (also for each document chunk)),
and wherein augmenting the output of the machine learning system comprises selecting the given tabular data structure from among multiple tabular data structures in the given document chunk responsive to determining that a similarity between the search text of the query and the textual description for the given tabular data structure exceeds a designated similarity threshold (¶ 0066: “In the question-and-answer for table data, the similarity between the embedding of rows and columns” (from multiple tabular data structures) “obtained from the sentence vector output unit (104) and the question” (search text) “vector may be compared, and the location of a cell indicating the correct answer” (a single output is selected) “may be predicted through the row and column” (to be augmented with “row and column” (with the tabular data structure information)) “having the highest similarity” (based on similarity exceeding anything below the “highest similarity” (a designated similarity threshold)).
Regarding claim 17, Kwon et al. do not specifically disclose the computer program product of claim 15 wherein the given tabular data structure representation for the given tabular data structure comprises a placeholder for the given tabular data structure.
Mathur does teach the computer program product of claim 15 wherein the given tabular data structure representation for the given tabular data structure comprises a placeholder for the given tabular data structure (¶ 0084: “The set of specific (e.g., Boolean, etc.) values for the set of named variables/parameters v1, v2, v3, . . . vn may be generated at runtime based on what specific combination of columns” (a given tabular data structure representation) “in the table” (for the given tabular data structure) “(124) to which a user query is directed, and then used to populate or fill in the” (comprises) “placeholders” (a placeholder)” “:v1”, “:v2”, “:v3”, . . . “:vn” in the union all query statement in expression (1-1) above at runtime. As a result, query results based on column values retrieved for the specific combination of columns in the table (124) to which the user query is directed can be returned as a response to the user query”).
It would have therefore been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate implementation of “placeholder” for “table” search subject to “user query” of Mathur into the corresponding one of Kwon et al. would enable the combined systems and their associated methods to perform in combination as they do separately and to further enable Kwon et al. to also benefits for being able to “dynamically bound to runtime variables to determine what columns are actually to be retrieved at runtime” as disclosed in Mathur ¶ 0026 last sentence.
Regarding claim 20, Kwon et al. do not specifically disclose the method of claim 18 wherein the given tabular data structure representation for the given tabular data structure comprises a placeholder for the given tabular data structure.
Mathur does teach the method of claim 18 wherein the given tabular data structure representation for the given tabular data structure comprises a placeholder for the given tabular data structure (¶ 0084: “The set of specific (e.g., Boolean, etc.) values for the set of named variables/parameters v1, v2, v3, . . . vn may be generated at runtime based on what specific combination of columns” (a given tabular data structure representation) “in the table” (for the given tabular data structure) “(124) to which a user query is directed, and then used to populate or fill in the” (comprises) “placeholders” (a placeholder)” “:v1”, “:v2”, “:v3”, . . . “:vn” in the union all query statement in expression (1-1) above at runtime. As a result, query results based on column values retrieved for the specific combination of columns in the table (124) to which the user query is directed can be returned as a response to the user query”).
It would have therefore been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate implementation of “placeholder” for “table” search subject to “user query” of Mathur into the corresponding one of Kwon et al. would enable the combined systems and their associated methods to perform in combination as they do separately and to further enable Kwon et al. to also benefits for being able to “dynamically bound to runtime variables to determine what columns are actually to be retrieved at runtime” as disclosed in Mathur ¶ 0026 last sentence.
Regarding claim 21, Kwon et al. do teach the method of claim 20 wherein generating the given document chunk comprises:
determining a tabular data structure identifier for the given tabular data structure (¶ 0045: “In the table” “flattening and refining unit” “the two dimensional table information” (the tabular data structure) “is converted into a one dimensional token sequence” “form, and the index” (comprises an identifier) “of the row and column” “of each cell” “had before being converted are attached to provide two-dimensional information”);
determining coordinates of the given tabular data structure relative to the additional textual content in the given document chunk (¶ 0045: “In the table” “flattening and refining unit” “the two dimensional table information” (the tabular data structure) “is converted into a one dimensional token sequence” “form, and the index” “of the row and column” (relative to the additional textual content (¶ 0021 line 3: “texts” “existing in the selected rows and columns”)) “of each cell”) “of each cell” (for a given document chunk) “had before being converted are attached to provide two-dimensional information” (determining coordinates which also correspond to the tabular data structure relative to surrounding text)).
Kwon et al. do not specifically disclose:
and inserting, into the given document chunk, the placeholder for the given tabular data structure at the determined coordinates, the placeholder including a reference to the determined tabular data structure identifier.
Mathur do teach:
and inserting, into the given document chunk, the placeholder for the given tabular data structure at the determined coordinates, the placeholder including a reference to the determined tabular data structure identifier (¶ 0086 S 2: “array elements “arr[0]”, “arr[1]”, “arr[2]”, . . . “arr[n−1]” in the named variables/parameter “arr” may be generated at runtime based on what specific combination of columns” (with reference to determined “specific” (identifier) tabular data structures) “in the table (124) to which a user query is directed, and then used to populate or fill” (inserting) “in the placeholders” (placeholders) “:arr[0]”, “:arr[1]”, “:arr[2]”, . . . “:an[n−1]” in the union all query statement in expression (1-2) above at runtime”).
For obviousness to combine Kwon et al. and Mathur see claim 20.
Regarding claim 22, Kwon et al. do teach the method of claim 20 wherein generating the answer to the query comprises augmenting the output of the machine learning system with an original version of the given tabular data structure extracted from the given document responsive to determining that textual content of the output of the machine learning system is sourced from the given tabular data structure (¶ 0066: “In the question-and-answer for table data, the similarity between the embedding of rows and columns obtained from the sentence vector output unit (104) and the question vector may be compared, and the location of a cell indicating the correct answer” (the output of the machine learning system) “may be predicted through the row and column” (is augmented with “row and column” (with the tabular data structure)) “having the highest similarity”).
Regarding claim 23, Kwon et al. do teach the method of claim 22 wherein generating the given document chunk further comprises determining a textual description for the given tabular data structure (¶ 0047 S 1: “the sentence vector output unit 104 generates expression vectors for row and column” (for each tabular data structure) “expression vectors and text” (generating textual description) “information existing in each cell” (also for each document chunk)),
and wherein augmenting the output of the machine learning system comprises selecting the given tabular data structure from among multiple tabular data structures in the given document chunk responsive to determining that a similarity between the search text of the query and the textual description for the given tabular data structure exceeds a designated similarity threshold ((¶ 0066: “In the question-and-answer for table data, the similarity between the embedding of rows and columns” (from multiple tabular data structures) “obtained from the sentence vector output unit (104) and the question vector may be compared, and the location of a cell indicating the correct answer” (a single output is selected) “may be predicted through the row and column” (to be augmented with “row and column” (with the tabular data structure information)) “having the highest similarity” (based on similarity exceeding anything below the “highest similarity” (a designated similarity threshold)).
Claim(s) 9 is/are rejected under 35 U.S.C. 103 as being unpatentable over Kwon et al. in view of Mathur, and further in view of Zoryn et al. (US 2016/0012052).
Regarding claim 9, Kwon et al. in view of Mathur do not specifically disclose the apparatus of claim 8 wherein determining the textual description comprises extracting a table caption for the given tabular data structure.
Zoryn et al. do teach the apparatus of claim 8 wherein determining the textual description comprises extracting a table caption for the given tabular data structure (Abstract S. before last: “Aspects of the invention can be used to match keywords against column names, to match keywords against values in subject and non-subject columns, and to match keywords” (determining table textual descriptions) “against table descriptions like page titles, table captions” (by extracting table caption) “cell values, nearest headings and surrounding text” (in textual format)).
It would have therefore been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the “table” “keyword search” techniques of Zoryn et al. into the corresponding one of Kwon et al. in view of Mathur would enable the combined systems and their associated methods to perform in combination as they do separately and to further enable Kwon et al. in view of Mathur to handle table search with table parameters comprising “table descriptions like page titles, table captions, cell values, nearest headings and surrounding text” as disclosed in Zoryn et al. Abstract S. before last.
Claim(s) 10 is/are rejected under 35 U.S.C. 103 as being unpatentable over Kwon et al. in view of Mathur, and further in view of MOON et al. (US 2024/0242024).
Regarding claim 10, Kwon et al. in view of Mathur do not specifically disclose the apparatus of claim 8 wherein determining the textual description comprises applying at least one of natural language processing summarization and natural language processing topic extraction to the first textual content of the given tabular data structure.
MOON et al. do teach the apparatus of claim 8 wherein determining the textual description comprises applying at least one of natural language processing summarization and natural language processing topic extraction to the first textual content of the given tabular data structure (¶ 0028 S 1+: “In embodiments, table embeddings may be used to solve table-to-text tasks such as table summarization” (doing a natural language processing summarization for a table or tabular data) “table question-and-answering, and table topic classification” “Table embeddings may also be used to generate stylized and/or domain-specific textual reports” (of the first textual content)).
It would have therefore been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate “table embedding” for “table question-and-answering” methods of MOON et al. into “Table Question Answering” techniques of Kwon et al. in Kwon et al. in view of Mathur would enable the combined systems and their associated methods to perform in combination as they do separately and to further enable Kwon et al. in view of Mathur to obtain a “text summary” “of the table” which can be much easier to follow as disclosed in MOON et al. ¶ 0057 last sentence.
Claim(s) 11 is/are rejected under 35 U.S.C. 103 as being unpatentable over Kwon et al., and further in view of Thomas et al. (US Patent 12,147,758).
Regarding claim 11, Kwon et al. do not specifically disclose the apparatus of claim 1 wherein the machine learning system comprises a large language model, and wherein generating the prompt for input to the machine learning system comprises utilizing a prompt template that instructs the large language model to recognize the tabular data structure representations and include content of relevant ones of the one or more tabular data structures in the output of the machine learning system.
Thomas et al. do teach the apparatus of claim 1 wherein the machine learning system comprises a large language model, and wherein generating the prompt for input to the machine learning system comprises utilizing a prompt template that instructs the large language model to recognize the tabular data structure representations and include content of relevant ones of the one or more tabular data structures in the output of the machine learning system (Col. 13 lines 49+: “the application generates a prompt” (a prompt) “for the LLM” (to a large language model) “to generate a pivot table which will be responsive to the user's query. The prompt, based on a prompt template” (based on a prompt template) “includes the table metadata (i.e., column headers, filename, sheet name or table name if any, etc.) and may also include contextual information” (to help recognize tabular data representations and relevant content)).
It would have therefore been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the “LLM” functions of Thomas et al. into the “neural network” of Kwon et al. would enable the combined systems and their associated methods to perform in combination as they do separately and to further enable Kwon et al. method to “select template” “based on the nature of the user’s query” as disclosed in Thomas et al. Col. 14 lines 29-30 for providing responses to the “query” which results in more efficient response processing.
Claim(s) 12 is/are rejected under 35 U.S.C. 103 as being unpatentable over Kwon et al., and further in view of SCHMIDTKE et al. (US 2022/0382961).
Regarding claim 12, Kwon et al. do not specifically disclose the apparatus of claim 1 wherein, the given tabular data structure representation for the given tabular data structure further comprises one or more image placeholders for one or more images included in the given tabular data structure, and wherein generating the answer to the query comprises augmenting the output of the machine learning system with an original version of at least a given one of the one or more images responsive to determining that textual content of the output of the machine learning system is sourced from a portion of the given tabular data structure associated with the given image.
SCHMIDTKE et al. do teach:
the apparatus of claim 1 wherein, the given tabular data structure representation for the given tabular data structure further comprises one or more image placeholders for one or more images included in the given tabular data structure, and wherein generating the answer to the query comprises augmenting the output of the machine learning system with an original version of at least a given one of the one or more images responsive to determining that textual content of the output of the machine learning system is sourced from a portion of the given tabular data structure associated with the given image (¶ 0057 S1+: “changing a query” (in response to a query) “for a table” (for a tabular data structure representation which comprises “placeholder” (placeholder)) “or tablix, changes applied to images such as gauges or charts which will require a new image from the server, parameter changes, etc.), format representation modifier 306 may determine that the intermediate format representation should be modified to insert” (augment) “a new document element” (a generated answer to the query) “along with a placeholder” (with an image placeholder) “for the additional data” (for the “new image” (one or more images)) “or content” (e.g., “text” (textual content (¶ 0032)) “contained therein”).
It would have therefore been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the “table” modification procedures of SCHMIDTKE et al. for response to a user “query”, would enable the combined systems and their associated methods to perform in combination as they do separately and to further enable Kwon et al. to respond to queries demanding data not present in its document tables and require table modification to address the query as disclosed in SCHMIDTKE et al. ¶ 0057.
Claim(s) 13 is/are rejected under 35 U.S.C. 103 as being unpatentable over Kwon et al., and further in view of Burke et al. (US Patent 9,921,730).
Regarding claim 13, Kwon et al. do not specifically disclose the apparatus of claim 1 wherein the query is directed to performing configuration of an information technology asset, and wherein the one or more documents comprise one or more technical guides for the information technology asset.
Burke et al. do teach the apparatus of claim 1 wherein the query is directed to performing configuration of an information technology asset, and wherein the one or more documents comprise one or more technical guides for the information technology asset (Col. 27 lines 25+: “generating a set of events responsive to a search query” (a search query) “each event comprising a timestamp and a portion of raw machine data that reflects activity in an information technology” (directed at an information technology asset) “environment of at least one computing system; causing display of a first interface in a tabular” (requiring tabular data format) “format that includes one or more rows, each row comprising: a time increment corresponding to a plurality of the events that each have a field-value pair matching a particular event field; and one or more aggregated metrics, wherein each aggregated metric” (with technical guides pertaining to the technology) “of a particular row indicates a number of events included in a subset of the plurality of the events that each occurred within the time increment of the particular row and has a particular value in the particular event field; and in response to a user selection of at least a row of the one or more rows, causing transition to a second interface associated with the row”).
It would have therefore been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the “search query” requiring “tabular” data format of Burke et al. into the “Question Answering” system and method of Kwon et al. would enable the combined systems and their associated method to perform in combination as they do separately and to further enable Kwon et al. to direct its methods to an area of great business potential.
Claim(s) 14 is/are rejected under 35 U.S.C. 103 as being unpatentable over Kwon et al., and further in view of CAKIR et al. (WO 2024/054286).
Regarding claim 14, Kwon et al. do not specifically disclose the apparatus of claim 1 wherein the query is directed to performing at least one of troubleshooting and remediation of one or more issues encountered on an information technology asset, and wherein the one or more documents comprise one or more support tickets associated with the one or more issues encountered on the information technology asset.
CAKIR et al. do teach the apparatus of claim 1 wherein the query is directed to performing at least one of troubleshooting and remediation of one or more issues encountered on an information technology asset, and wherein the one or more documents comprise one or more support tickets associated with the one or more issues encountered on the information technology asset (¶ 0084: “FIGURE 13 illustrates a table” (using tabular data) “for domain-aware smart parsing and representation, in accordance with aspects of the present disclosure. The text input” (input query) “within the IC” (for an information technology) “troubleshooting query” (trouble shooting); note: “IC” = “integrated circuit” (¶ 0005); Fig. 8 also shows a module tailored to “CLUSTER TICKET BY TOPIC” (i.e., something to address issues by one or more support tickets)).
It would have therefore been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the “troubleshooting query” requiring “table” data format of CAKIR et al. into the “Question Answering” system and method of Kwon et al. would enable the combined systems and their associated method to perform in combination as they do separately and to further enable Kwon et al. to direct its methods to an area of great business potential.
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 FARZAD KAZEMINEZHAD whose telephone number is (571)270-5860. The examiner can normally be reached 10:30 am to 11:30 pm.
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, Paras D. Shah can be reached at (571) 270-1650. 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.
/Farzad Kazeminezhad/
Art Unit 2653
May 22nd 2026.
Prosecution Insights