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 .
Applicant's Response
In Applicant's Response dated 8/20/2025, Applicant amended the Claims and argued against all objections and rejections set forth in the previous Office Action.
All objections and rejections not reproduced below are withdrawn.
The prior art rejections of the Claims under 35 U.S.C. 103 previously set forth are maintained.
The examiner appreciates the applicant noting where the support for the amendments are described in the specification.
The Application was filed on 6/29/2023, which is a CON of 17/358,114 Filing Date 06/25/2021, with priority to provisional 63/043,906 filed 06/25/2020.
Claim(s) 24-52 are pending for examination. Claim(s) 24, 38, 43, 44 is/are independent claim(s).
Double Patenting
The nonstatutory double patenting rejection is based on a judicially created doctrine grounded in public policy (a policy reflected in the statute) so as to prevent the unjustified or improper timewise extension of the “right to exclude” granted by a patent and to prevent possible harassment by multiple assignees. A nonstatutory double patenting rejection is appropriate where the conflicting claims are not identical, but at least one examined application claim is not patentably distinct from the reference claim(s) because the examined application claim is either anticipated by, or would have been obvious over, the reference claim(s). See, e.g., In re Berg, 140 F.3d 1428, 46 USPQ2d 1226 (Fed. Cir. 1998); In re Goodman, 11 F.3d 1046, 29 USPQ2d 2010 (Fed. Cir. 1993); In re Longi, 759 F.2d 887, 225 USPQ 645 (Fed. Cir. 1985); In re Van Ornum, 686 F.2d 937, 214 USPQ 761 (CCPA 1982); In re Vogel, 422 F.2d 438, 164 USPQ 619 (CCPA 1970); In re Thorington, 418 F.2d 528, 163 USPQ 644 (CCPA 1969).
A timely filed terminal disclaimer in compliance with 37 CFR 1.321(c) or 1.321(d) may be used to overcome an actual or provisional rejection based on nonstatutory double patenting provided the reference application or patent either is shown to be commonly owned with the examined application, or claims an invention made as a result of activities undertaken within the scope of a joint research agreement. See MPEP § 717.02 for applications subject to examination under the first inventor to file provisions of the AIA as explained in MPEP § 2159. See MPEP § 2146 et seq. for applications not subject to examination under the first inventor to file provisions of the AIA . A terminal disclaimer must be signed in compliance with 37 CFR 1.321(b).
The filing of a terminal disclaimer by itself is not a complete reply to a nonstatutory double patenting (NSDP) rejection. A complete reply requires that the terminal disclaimer be accompanied by a reply requesting reconsideration of the prior Office action. Even where the NSDP rejection is provisional the reply must be complete. See MPEP § 804, subsection I.B.1. For a reply to a non-final Office action, see 37 CFR 1.111(a). For a reply to final Office action, see 37 CFR 1.113(c). A request for reconsideration while not provided for in 37 CFR 1.113(c) may be filed after final for consideration. See MPEP §§ 706.07(e) and 714.13.
The USPTO Internet website contains terminal disclaimer forms which may be used. Please visit www.uspto.gov/patent/patents-forms. The actual filing date of the application in which the form is filed determines what form (e.g., PTO/SB/25, PTO/SB/26, PTO/AIA /25, or PTO/AIA /26) should be used. A web-based eTerminal Disclaimer may be filled out completely online using web-screens. An eTerminal Disclaimer that meets all requirements is auto-processed and approved immediately upon submission. For more information about eTerminal Disclaimers, refer to www.uspto.gov/patents/apply/applying-online/eterminal-disclaimer.
Claims 24-29, 32-42, 44, 45, 48 rejected on the ground of nonstatutory double patenting as being unpatentable over claims 1, 3-7, 10-15, 18-21 of U.S. Patent No. 11734268. Although the claims at issue are not identical, they are not patentably distinct from each other because the claims appear to be a reworded broader version of the allowed claims (see table below).
Pat. 11734268
App. 18/216202
1. A method comprising:
receiving a source document;
applying one or more pre-processes to the source document to produce document contextual information representative of a structure and content of the source document; and
transforming the source document, based on the document contextual information, to generate a question-and-answer searchable document;
wherein applying the one or more pre-processes comprises segmenting the source document into multiple document segments, wherein transforming the source document comprises transforming according to a vector transform applied to the document contextual information and content of the multiple document segments with a first contextual data element derived from a first segment of the multiple document segments being combined with content of a different segment of the multiple document segments, and wherein the method further comprises:
for at least one segment of the multiple document segments, identifying at least one segment descriptor comprising one or more of: at least one entity associated with the at least one segment, at least one task associated with at least one segment, or subject matter descriptor associated with the at least one segment;
tagging the at least one segment with the at least one descriptor;
receiving query data representative of a question from a user relating to the content of the source document;
determining sensor contextual information based on sensor data obtained by a sensor device associated with the user; and
searching a response to the query data from one or more of the multiple document segments with segment descriptors matching the determined sensor contextual information;
wherein determining the contextual information comprises:
capturing sensor data by one or more sensors of an augmented reality system with which the user is interacting; and
determining an item or location identifiable from the sensor data being presented to the user through the augmented reality system;
wherein searching the response to the query data comprises searching the response to the query data from one or more of the multiple document segments based, at least in part, on the determined item or location identifiable from the sensor data.
24. A method comprising:
receiving a source document;
(limitation below)
applying one or more pre-processes to at least one of the multiple document segments to determine contextual information for the at least one of the multiple document segments representative of a structure and content of the source document;
combining the contextual information determined for the at least one of the multiple document segments with another of the multiple document segments; and
(from above)
segmenting the source document into multiple document segments;
transforming according to a vector transformation the at least one of the multiple document segments combined with the contextual information for the at least one of the multiple document segments, and the other of the multiple document segments combined with the contextual information for the at least one of the multiple document segments, to generate a vector transformed question-and-answer searchable content with vector- transformed contextual information.
35. The method of claim 24, wherein the method further comprises:
for at least one segment of the multiple document segments, identifying at least one segment descriptor comprising one or more of: at least one entity associated with the at least one segment, at least one task associated with at least one segment, or subject matter descriptor associated with the at least one segment; and
tagging the at least one segment with the at least one descriptor.
36. (New) The method of claim 35, further comprising:
receiving query data representative of a question relating to the content of the source document;
determining at least one query descriptor associated with the query data, the at least one descriptor comprising one or more of: at least one entity associated with the at query data, at least one task associated with the query data, or subject matter descriptor associated with the query; and
searching a response to the query data from one or more of the multiple document segments with segment descriptors matching the at least one query descriptor.
3. The method of claim 1, wherein the vector transform comprises one or more of: a coarse linearization transform to generate coarse numerical vectors representative of coarse content of the multiple document segments, or a fine-detail transform to generate fine-detail transformed content records representative of the content of the multiple document segments.
25. The method of claim 24, wherein the vector transformation comprises one or more of: a coarse linearization transform to generate coarse numerical vectors representative of coarse content of the plurality of document segments, or a fine-detail transformation to generate fine-detail transformed content records representative of the content of the plurality of document segments.
4. The method of claim 1, wherein segmenting the source document comprises: segmenting the source document into the multiple document segments according to hierarchical rules semantically associating one portion of the source document with one or more other portions of the source content.
26. The method of claim 24, wherein segmenting the source document comprises:
segmenting the source document into the multiple document segments according to hierarchical rules semantically associating one portion of the source document with one or more other portions of the source content.
5. The method of claim 4, wherein segmenting the source document according to hierarchical rules comprises: including, in a particular document segment, content of a particular document portion and section heading content located in the source document ahead of a location of the particular document portion, wherein the section heading content is determined to be associated with the content of the particular document portion.
27. The method of claim 26, wherein segmenting the source document according to hierarchical rules comprises:
including, in a particular document segment, content of a particular document portion and section heading content located in the source document ahead of a location of the particular document portion, wherein the section heading content is determined to be associated with the content of the particular document portion.
6. The method of claim 1, wherein segmenting the source document comprises: segmenting the source document into the multiple document segments by sliding a window of a fixed or variable size over the source document to generate the multiple document segments.
28. The method of claim 24, wherein segmenting the source document comprises:
segmenting the source document into the multiple document segments by sliding a window of a fixed or variable size over the source document to generate the multiple document segments.
7. The method of claim 6, wherein sliding the window over the source document comprises sliding the window at steps that are smaller than the size of the window such that a generated first segment and a next generated segment each share at least some overlapping content.
29. The method of claim 28, wherein sliding the window over the source document comprises sliding the window at steps that are smaller than the size of the window such that a generated first segment and a next generated segment each share at least some overlapping content.
10. The method of claim 1, wherein applying the one or more pre-processes to the source document comprises: identifying a portion of the source document comprising multiple sub-portions arranged in a multi-cell table; and generating multiple substitute portions to replace the multi-cell table, with each of the multiple substitute portions comprising a respective sub-portion content data and contextual information associated with the multi-cell table.
32. The method of claim 24, wherein applying the one or more pre- processes to the source document comprises:
identifying a portion of the source document comprising multiple sub-portions arranged in a multi-cell table; and
generating multiple substitute portions to replace the multi-cell table, with each of the multiple substitute portions comprising a respective sub-portion content data and contextual information associated with the multi-cell table.
11. The method of claim 1, wherein applying the one or more pre-processes to the source document comprises: associating contextual information with one or more portions of the source document based on information provided by a user in response to one or more questions relating to the source document that are presented to the user.
33. The method of claim 24, wherein applying the one or more pre- processes to the source document comprises:
associating contextual information with one or more portions of the source document based on information provided by a user in response to one or more questions relating to the source document that are presented to the user.
12. The method of claim 1, wherein applying the one or more pre-processes to the source document comprises: associating question-and-answer contextual information relating to a particular portion of the source document based on one or more ground truth samples of question-and-answer pairs.
34. The method of claim 24, wherein applying the one or more pre- processes to the source document comprises:
associating question-and-answer contextual information relating to a particular portion of the source document based on one or more ground truth samples of question- and-answer pairs.
13. The method of claim 1, wherein the method further comprises:
for at least one segment of the multiple document segments, identifying at least one segment descriptor comprising one or more of: at least one entity associated with the at least one segment, at least one task associated with at least one segment, or subject matter descriptor associated with the at least one segment; and tagging the at least one segment with the at least one descriptor.
35. The method of claim 24, wherein the method further comprises:
for at least one segment of the multiple document segments, identifying at least one segment descriptor comprising one or more of: at least one entity associated with the at least one segment, at least one task associated with at least one segment, or subject matter descriptor associated with the at least one segment; and
tagging the at least one segment with the at least one descriptor.
14. The method of claim 13, further comprising: receiving query data representative of a question relating to the content of the source document; determining at least one query descriptor associated with the query data, the at least one descriptor comprising one or more of: at least one entity associated with the at query data, at least one task associated with the query data, or subject matter descriptor associated with the query; and searching a response to the query data from one or more of the multiple document segments with segment descriptors matching the at least one query descriptor.
36. The method of claim 35, further comprising:
receiving query data representative of a question relating to the content of the source document;
determining at least one query descriptor associated with the query data, the at least one descriptor comprising one or more of: at least one entity associated with the at query data, at least one task associated with the query data, or subject matter descriptor associated with the query; and
searching a response to the query data from one or more of the multiple document segments with segment descriptors matching the at least one query descriptor.
15. The method of claim 1, wherein transforming according to a vector transform applied to the document contextual information and content of the multiple document segments with a first contextual data element derived from a first segment of the multiple document segments being combined with content of a different segment of the multiple document segments comprises one of:
separately transforming the first contextual data element, the first segment of the multiple document segments, and the different segment of the multiple document segments, and combining the transformed first contextual data element with the transformed first segment of the multiple document segments, and with the transformed different segment of the multiple document segments; or combining the first contextual data element with the first segment of the multiple document segments, combining the first contextual data element with the different segment of the multiple document segments, transforming the combined first contextual data element and the first segment of the multiple document segments, and transforming the combined first contextual data element and the different segment of the multiple document segments.
37. The method of claim 24, wherein transforming the at least one of the multiple document segments combined with the contextual information for the at least one of the multiple document segments, and the other of the multiple document segments combined with the contextual information for the at least one of the multiple document segments comprises one of:
separately transforming the contextual information for the at least one of the multiple document segments, the at least one of the multiple document segments, and the other of the multiple document segments, and combining the transformed contextual information with the transformed at least one of the multiple document segments, and the transformed other of the multiple document segments; or combining the contextual information with the at least one of the multiple document segments, combining the contextual information with the other of the multiple document segments, transforming the combined contextual information and the at least
one of the multiple document segments, and transforming the combined contextual information and the other of the multiple document segments.
Claims 18 and 23 are similar to claim 1
Claim 38 is similar to claim 24
Claim 39 is similar to claim 37
Claim 19 is similar to claim 3
Claim 40 is similar to claim 25
Claim 20 is similar to claim 4
Claim 41 is similar to claim 26
Claim 21 is similar to claim 13
Claim 42 is similar to claim 35
Claim 44 is a combination of claims 24 and 36, where claims 24 uses the phrase “receiving query data” and claim 36 used the phrase “receiving a question”
Claim 45 is similar to claim 36
Claim 48 is similar to claim 27
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) 24, 26, 27, 30, 31, 33-49 is/are rejected under 35 U.S.C. 103 as being unpatentable over Stubley; Peter et al. US Pub. No. 2016/0179934 (Stubley) in view of Karandish; David et al. US Pub. No. 2021/0056150 (Karandish).
Claim 24:
Stubley teaches:
A method comprising:
receiving a source document [¶ 0061-64] (analyze natural language texts retrieved from information sources and annotate them to enhance their usefulness in question answering);
segmenting the source document into multiple document segments [¶ 0061-64] (an index of annotations may be generated for a natural language document, and the indexed document may be stored in indexed unstructured data sets);
applying one or more pre-processes to at least one of the multiple document segments to determine contextual information for the at least one of the multiple document segments representative of a structure and content of the source document [¶ 0063, 77, 79] (document/passage analyzer may include text from corresponding titles and/or section headings in annotations for index entries (e.g., sentences) belonging to those documents and/or sections, annotations may become useful for QA system to identify a body text passage as being relevant to answering a user's question based in part on the title or header of the document or section in which the passage appears, these annotations could be “contextual information”);
combining the contextual information determined for the at least one of the multiple document segments with another of the multiple document segments [¶ 0062, 71, 77] (combining adjacent sentences individually indexed and determined to provide relevant evidence for the question's answer) [¶ 0058] (generate a larger combined ontology by connecting the domain-specific ontology to the domain-independent ontology through one or more concepts common to both ontologies) [¶ 0110-112, 114, 122, 125] (merge any answer information retrieved from structured data search and any answer information generated from unstructured data to build a combined answer to present to user, set of passages that in combination match the multiple constraints of the question); and …
Stubley does not appear to explicitly disclose “vector transformed question-and-answer searchable content”.
However, the disclosure of Karandish teaches:
… transforming according to a vector transformation the at least one of the multiple document segments combined with the contextual information for the at least one of the multiple document segments, and the other of the multiple document segments combined with the contextual information for the at least one of the multiple document segments, to generate a vector transformed question-and-answer searchable content with vector- transformed contextual information [¶ 0110, 116, 124-125] (transforming each of the one or more location delimiters of the respective location metadata associated with the question-answer pair into one or more second numeric vector representations, vectorizing the location metadata could be considered “vector-transformed contextual information”) [¶ 0090] (preceding sentence of “Document Title” can be transformed using a transformation algorithm, as described above, into a vector representation, such as <0.23, 0.98, . . . 0.73, 0.04, 0.62>, which can be stored in transformed preceding sentence delimiter 663) [¶ 0049] (source document can include a title, section headings, etc.) [¶ 0061-63] (FIG. 4B, location metadata 451 can include a preceding sentence delimiter 463, which can store the text of the sentence that immediately precedes the text of answer 447 (e.g., extracted content section 432) within pre-processed document 421 (FIG. 4A). As shown in FIG. 4B, preceding sentence delimiter 463 can be “Document Title,” which indicates that the text of answer 447 (e.g., extracted content section 432) within pre-processed document 421 (FIG. 4A) begins immediately after the text “Document Title.”) [¶ 0090] (sentence of “Document Title” can be transformed using a transformation algorithm, as described above, into a vector representation, location metadata 651 can be associated with question-answer pair 641 in the index) [¶ 0109] (transforming an answer of the question-answer pair into a first numeric vector representation) [¶ 0116] (transforming into vector embeddings an ingested source document of the set of ingested source documents that is identified by the source document identifier of the respective location metadata associated with the first question-answer pair).
Karandish also teaches: [¶ 0045, 48-49, 54, 59, 65, 71, 80, 83; Figs 4A, 6A] (extracting content sections could be “segmenting”) [¶ 0045, 48-49, 54, 59, 65, 71, 80, 83; Figs 4A, 6A] (ingestion could be “receiving”) [¶ 0050-51, 73, 83, 12] (pre-processing).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine the method of generating question and answer pairs in Stubley and the method of generating question and answer pairs in Karandish, with a reasonable expectation of success.
The motivation for doing so would have been the use of known technique to improve similar devices (methods, or products) in the same way; (See KSR Int’l Co. v. Teleflex Inc., 550 US 398, 82 USPQ2d 1385, 1396 (U.S. 2007) and MPEP § 2143(D)).
The know technique of vector representation of text in Karandish could be applied to the question and answer indexing in Stubley. Stubley and Karandish are similar devices because both generate question and answer pairs. One of ordinary skill in the art would have recognized that applying the known technique would improve the similar devices and resulted in an improved system, with a reasonable expectation of success, because it would “greatly improve information access by automatically answering questions through an automated chat agent with fresh information from dynamic source documents” [Karandish: ¶ 0126].
Claim 26:
Stubley teaches:
The method of claim 24, wherein segmenting the source document comprises:
segmenting the source document into the multiple document segments according to hierarchical rules semantically associating one portion of the source document with one or more other portions of the source content [¶ 0049-50] (Transitive relationships can also be deduced by tracing connected paths of parent-child relationships within an ontology, relationships (e.g., parent-child/hypernym-hyponym relationships) are said to be “hierarchical,” since they establish a hierarchy in which parent concepts subsume their children concepts).
Claim 27:
Stubley teaches:
The method of claim 26, wherein segmenting the source document according to hierarchical rules comprises:
including, in a particular document segment, content of a particular document portion and section heading content located in the source document ahead of a location of the particular document portion, wherein the section heading content is determined to be associated with the content of the particular document portion [¶ 0063, 77, 79] (document/passage analyzer may include text from corresponding titles and/or section headings in annotations for index entries (e.g., sentences) belonging to those documents and/or sections, annotations may become useful for QA system to identify a body text passage as being relevant to answering a user's question based in part on the title or header of the document or section in which the passage appears, these annotations could be “contextual information”).
Claim 30:
Stubley teaches:
The method of claim 24, wherein applying the one or more pre- processes to the source document comprises:
determining relative importance value for a particular portion of the source document based on one or more of: location of the particular portion relative to locations of one or more other portions of the source document, relative font size of the particular
portion, structure and organization of the source document, or document type of the source document [¶ 0077] (annotation of index units (e.g., body text sentences) with text from corresponding section headings and document titles may increase the relevance score of a passage by simulating proximity to terms that appear in the heading and/or title to which that passage belongs).
Claim 31:
Stubley teaches:
The method of claim 30, wherein transforming the source document comprises transforming the source document based, at least in part, on the determined relative importance value for the particular portion, and on relative importance values for other portions of the source document [¶ 0077] (annotation of index units (e.g., body text sentences) with text from corresponding section headings and document titles may increase the relevance score of a passage by simulating proximity to terms that appear in the heading and/or title to which that passage belongs).
Claim 33:
Stubley teaches:
The method of claim 24, wherein applying the one or more pre-processes to the source document comprises:
associating contextual information with one or more portions of the source document based on information provided by a user in response to one or more questions relating to the source document that are presented to the user [¶ 0034, 36-42] (store the follow-up question that is utilized to generate one or more system responses to provide or perform in response to a given user communication based, at least in part, on the initial user question).
Claim 34:
Karandish teaches:
The method of claim 24, wherein applying the one or more pre- processes to the source document comprises:
associating question-and-answer contextual information relating to a particular portion of the source document based on one or more ground truth samples of question-and-answer pairs [¶ 0055] (model can be trained using a dataset of documents in which the most relevant questions and answers have already been manually and/or semi-automatically mined from the documents could be “ground truth samples”).
Claim 35:
Stubley teaches:
The method of claim 24, wherein the method further comprises:
for at least one segment of the multiple document segments, identifying at least one segment descriptor comprising one or more of: at least one entity associated with the at least one segment, at least one task associated with at least one segment, or subject matter descriptor associated with the at least one segment; and
tagging the at least one segment with the at least one descriptor [¶ 0060, 63, 70, 92] (entity detection and recognition) [¶ 0070] (subject).
Claim 36:
Stubley teaches:
The method of claim 35, further comprising:
receiving query data representative of a question relating to the content of the source document;
determining at least one query descriptor associated with the query data, the at least one descriptor comprising one or more of: at least one entity associated with the at query data, at least one task associated with the query data, or subject matter descriptor associated with the query [¶ 0060, 63, 70, 92] (entity detection and recognition) [¶ 0070] (subject); and
searching a response to the query data from one or more of the multiple document segments with segment descriptors matching the at least one query descriptor [¶ 0024-27, 70] (search).
Claim 37:
Stubley teaches:
The method of claim 24, wherein transforming the at least one of the multiple document segments combined with the contextual information for the at least one of the multiple document segments, and the other of the multiple document segments combined with the contextual information for the at least one of the multiple document segments comprises one of:
separately transforming the contextual information for the at least one of the multiple document segments, the at least one of the multiple document segments, and the other of the multiple document segments, and combining the transformed contextual information with the transformed at least one of the multiple document segments, and the transformed other of the multiple document segments; or combining the contextual information with the at least one of the multiple document segments, combining the contextual information with the other of the multiple document segments, transforming the combined contextual information and the at least one of the multiple document segments, and transforming the combined contextual information and the other of the multiple document segments [¶ 0062, 71, 77] (combining adjacent sentences individually indexed and determined to provide relevant evidence for the question's answer) [¶ 0058] (generate a larger combined ontology by connecting the domain-specific ontology to the domain-independent ontology through one or more concepts common to both ontologies) [¶ 0110-112, 114, 122, 125] (merge any answer information retrieved from structured data search and any answer information generated from unstructured data to build a combined answer to present to user, set of passages that in combination match the multiple constraints of the question).
Claim 44:
Stubley teaches:
A method for question answering using plurality of source documents, the method comprising:
ingesting the plurality of source documents, including for each document of at least some of the plurality of source documents [¶ 0061-64] (analyze natural language texts retrieved from information sources and annotate them to enhance their usefulness in question answering):
applying one or more pre-processes to determine contextual information at locations in the document [¶ 0063, 77, 79] (document/passage analyzer may include text from corresponding titles and/or section headings in annotations for index entries (e.g., sentences) belonging to those documents and/or sections, annotations may become useful for QA system to identify a body text passage as being relevant to answering a user's question based in part on the title or header of the document or section in which the passage appears, these annotations could be “contextual information”);
segmenting the document into segments comprising word sequences, each segment having a location of the word sequence of said segment in the document [¶ 0061-64] (an index of annotations may be generated for a natural language document, and the indexed document may be stored in indexed unstructured data sets);
… combinations of contextual information and word sequences of segments in a repository [¶ 0062, 71, 77] (combining adjacent sentences individually indexed and determined to provide relevant evidence for the question's answer) [¶ 0058] (generate a larger combined ontology by connecting the domain-specific ontology to the domain-independent ontology through one or more concepts common to both ontologies) [¶ 0110-112, 114, 122, 125] (merge any answer information retrieved from structured data search and any answer information generated from unstructured data to build a combined answer to present to user, set of passages that in combination match the multiple constraints of the question); and …
Stubley does not appear to explicitly disclose “forming a searchable numerical representation of a combination of the contextual information associated with said segment and the word sequence of said segment”.
However, the disclosure of Karandish teaches:
… for at least some segment of the segments of the source document:
associating said segment with contextual information determined to be at a location outside said segment in the document; and forming a searchable numerical representation of a combination of the contextual information associated with said segment and the word sequence of said segment [¶ 0049] (source document can include a title, section headings, etc.) [¶ 0090] (sentence of “Document Title” can be transformed using a transformation algorithm, as described above, into a vector representation, location metadata 651 can be associated with question-answer pair 641 in the index) [¶ 0109] (transforming an answer of the question-answer pair into a first numeric vector representation) [¶ 0116] (transforming into vector embeddings an ingested source document of the set of ingested source documents that is identified by the source document identifier of the respective location metadata associated with the first question-answer pair)
storing searchable numerical representations of segments including the searchable numerical representations of {combinations of contextual information and word sequences of segments in a repository determining an answer to a question using information stored in the repository} [¶ 0040] (indexed database), including:
receiving the question comprising a query word sequence [¶ 0111] (receive question);
forming a numerical representation of the query word sequence [¶ 0117] (vector representation);
using the searchable numerical representations of segments in the repository and the numerical representation of the query word sequence to identify a plurality of candidate segments [¶ 0117-119] (vector similarity score, ranking); and
determining an answer to the question from the word sequence of the question and a word sequence of at least one of the candidate segments represented in the repository [¶ 0119] (answer to question).
Karandish also teaches: [¶ 0045, 48-49, 54, 59, 65, 71, 80, 83; Figs 4A, 6A] (extracting content sections could be “segmenting”) [¶ 0045, 48-49, 54, 59, 65, 71, 80, 83; Figs 4A, 6A] (ingestion could be “receiving”) [¶ 0050-51, 73, 83, 12] (pre-processing).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine the method of generating question and answer pairs in Stubley and the method of generating question and answer pairs in Karandish, with a reasonable expectation of success.
The motivation for doing so would have been the use of known technique to improve similar devices (methods, or products) in the same way; (See KSR Int’l Co. v. Teleflex Inc., 550 US 398, 82 USPQ2d 1385, 1396 (U.S. 2007) and MPEP § 2143(D)).
The know technique of vector representation of text in Karandish could be applied to the question and answer indexing in Stubley. Stubley and Karandish are similar devices because both generate question and answer pairs. One of ordinary skill in the art would have recognized that applying the known technique would improve the similar devices and resulted in an improved system, with a reasonable expectation of success, because it would “greatly improve information access by automatically answering questions through an automated chat agent with fresh information from dynamic source documents” [Karandish: ¶ 0126].
Claim 45:
Karandish teaches:
The method of claim 44, wherein forming the searchable numerical representation includes transforming a combination of the contextual information associated with a segment and the word sequence of said segment according to a vector transformation to form a numerical vector representation; and
wherein using the searchable representations of segments in the repository to identify the plurality of candidate segments include matching a numerical vector representation of the query word sequence with numerical vector representations of the segments in the repository [¶ 0117-119] (vector similarity score, ranking).
Claim 46:
Karandish teaches:
The method of claim 45, wherein determining an answer to the question from a word sequence of at least one of the segments represented in the repository includes identifying a span of words in the word sequence of said at least one of the segments as an answer to the question [¶ 0099] (finding an answer using location metadata).
Claim 47:
Karandish teaches:
The method of claim 46, wherein determining an answer comprises processing a sequence comprising sequences determined from the word sequence of the question, a sequence determined from the word sequence of the segment, and a sequence representation of the contextual information [¶ 0117-119] (vector similarity score, ranking) [¶ 0077] (annotation of index units (e.g., body text sentences) with text from corresponding section headings and document titles may increase the relevance score of a passage by simulating proximity to terms that appear in the heading and/or title to which that passage belongs).
Claim 48:
Karandish teaches:
The method of claim 44, wherein the contextual information is in a group consisting of a title, a heading, an entity name located in a document, and a file name of the document [¶ 0049] (source document can include a title, section headings, etc.) [¶ 0090] (sentence of “Document Title” can be transformed using a transformation algorithm, as described above, into a vector representation, location metadata 651 can be associated with question-answer pair 641 in the index) [¶ 0109] (transforming an answer of the question-answer pair into a first numeric vector representation) [¶ 0116] (transforming into vector embeddings an ingested source document of the set of ingested source documents that is identified by the source document identifier of the respective location metadata associated with the first question-answer pair).
Claim 49:
Stubley teaches:
The method of claim 44, wherein the combination of the contextual information associated with the segment and the word sequence of said segment comprises a concatenation of a word sequence representing the contextual information and the word sequence of said segment [¶ 0062, 71, 77] (combining adjacent sentences individually indexed and determined to provide relevant evidence for the question's answer) [¶ 0058] (generate a larger combined ontology by connecting the domain-specific ontology to the domain-independent ontology through one or more concepts common to both ontologies) [¶ 0110-112, 114, 122, 125] (merge any answer information retrieved from structured data search and any answer information generated from unstructured data to build a combined answer to present to user, set of passages that in combination match the multiple constraints of the question).
Claims 38-43:
Claim(s) 38 and 43 is/are substantially similar to claim 24 and is/are rejected using the same art and the same rationale.
Claim 24 is a “method” claim, claim 38 is a “system” claim, claim 43 is a “media” claim, but the steps or elements of each claim are essentially the same.
Claim(s) 39 is/are substantially similar to claim 37 and is/are rejected using the same art and the same rationale.
Claim(s) 40 is/are substantially similar to claim 25 and is/are rejected using the same art and the same rationale.
Claim(s) 41 is/are substantially similar to claim 26 and is/are rejected using the same art and the same rationale.
Claim(s) 42 is/are substantially similar to claim 35 and is/are rejected using the same art and the same rationale.
Claim(s) 25 is/are rejected under 35 U.S.C. 103 as being unpatentable over Stubley; Peter et al. US Pub. No. 2016/0179934 (Stubley) in view of Karandish; David et al. US Pub. No. 2021/0056150 (Karandish) in view of in view of Gao; Yifan et al. US Pub. No. 2021/0174023 (Gao).
Claim 25:
Stubley and Karandish teach all the elements of the claim as shown above.
Stubley and Karandish does not appear to explicitly disclose “coarse” or “fine-detail”.
However, the disclosure of Gao teaches:
The method of claim 24, wherein the vector transformation comprises one or more of: a coarse linearization transform to generate coarse numerical vectors representative of coarse content of the plurality of document segments, or a fine-detail transformation to generate fine-detail transformed content records representative of the content of the plurality of document segments [¶ 0017, 21, 39, 41, 55-57, 81, 88, 92-97] (coarse-to-fine reasoning process and utilizing sentence-level selection scores for weighting token-level span distributions) [¶ 0044-45] (BERT) [¶ 0061] (Stanford Question Answering Dataset (SQuAD)) [¶ 0070-72, 80-81] (tokenize vector) [¶ 0087] (match) [¶ 0058] (BERT Question-Answer (BERTQA)).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine the method of generating question and answer pairs in Stubley and the method of generating question and answer pairs in Karandish and the method of coarse-to-fine reasoning in Gao, with a reasonable expectation of success.
The motivation for doing so would have been the use of known technique to improve similar devices (methods, or products) in the same way; (See KSR Int’l Co. v. Teleflex Inc., 550 US 398, 82 USPQ2d 1385, 1396 (U.S. 2007) and MPEP § 2143(D)).
The know technique of coarse-to-fine reasoning in Gao could be applied to the vector representation of text in Karandish and the question and answer indexing in Stubley. Gao, Stubley and Karandish are similar devices because both generate question and answer pairs. One of ordinary skill in the art would have recognized that applying the known technique would improve the similar devices and resulted in an improved system, with a reasonable expectation of success, for “improved effectiveness in conversational machine reasoning” and improved performance [Gao: ¶ 0021, 62, 94].
Claim(s) 28, 29, 32 is/are rejected under 35 U.S.C. 103 as being unpatentable over Stubley; Peter et al. US Pub. No. 2016/0179934 (Stubley) in view of Karandish; David et al. US Pub. No. 2021/0056150 (Karandish) in view of in view of Kenter; Tom Marius et al. US Pub. No. 2021/0350795 (Kenter).
Claim 28:
Stubley and Karandish teach all the elements of the claim as shown above.
Stubley and Karandish does not appear to explicitly disclose “sliding a window”.
However, the disclosure of Kenter teaches:
The method of claim 24, wherein segmenting the source document comprises:
segmenting the source document into the multiple document segments by sliding a window of a fixed or variable size over the source document to generate the multiple document segments [¶ 0042-45] (window to fit sequence length, or segment).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine the method of generating question and answer pairs in Stubley and the method of generating question and answer pairs in Karandish and the method of speech synthesis in Kenter, with a reasonable expectation of success.
The motivation for doing so would have been the use of known technique to improve similar devices (methods, or products) in the same way; (See KSR Int’l Co. v. Teleflex Inc., 550 US 398, 82 USPQ2d 1385, 1396 (U.S. 2007) and MPEP § 2143(D)).
The know technique of sliding windows in Kenter could be applied to the vector representation of text in Karandish and answer indexing in Stubley. Stubley and Karandish are similar devices because all provide natural language processing. One of ordinary skill in the art would have recognized that applying the known technique would improve the similar devices and resulted in an improved system, with a reasonable expectation of success, to provide a “more robust way than traditional parsing and tagging techniques” [Kenter: ¶ 0003].
Claim 29:
Kenter teaches:
The method of claim 28, wherein sliding the window over the source document comprises sliding the window at steps that are smaller than the size of the window such that a generated first segment and a next generated segment each share at least some overlapping content [¶ 0042-45] (window to fit sequence length, or segment, overlap by stride size s).
Claim(s) 32, 50 is/are rejected under 35 U.S.C. 103 as being unpatentable over Stubley; Peter et al. US Pub. No. 2016/0179934 (Stubley) in view of Karandish; David et al. US Pub. No. 2021/0056150 (Karandish) in view of in view of Kenter; Tom Marius et al. US Pub. No. 2021/0350795 (Kenter) in view of Chen; Bei et al. US Pub. No. 2022/0058191 (Chen) in view of in view of Gao; Yifan et al. US Pub. No. 2021/0174023 (Gao).
Claim 50:
Stubley and Karandish teach all the elements of the claim as shown above.
Stubley and Karandish does not appear to explicitly disclose “sliding a window”.
However, the disclosure of Kenter teaches:
The method of claim 24, wherein:
segmenting the source document includes sliding a window of a fixed size over the source document at steps that are smaller than the size of the window such that consecutive segments share overlapping content [¶ 0042-45] (window to fit sequence length, or segment); …
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine the method of generating question and answer pairs in Stubley and the method of generating question and answer pairs in Karandish and the method of speech synthesis in Kenter, with a reasonable expectation of success.
The motivation for doing so would have been the use of known technique to improve similar devices (methods, or products) in the same way; (See KSR Int’l Co. v. Teleflex Inc., 550 US 398, 82 USPQ2d 1385, 1396 (U.S. 2007) and MPEP § 2143(D)).
The know technique of sliding windows in Kenter could be applied to the vector representation of text in Karandish and answer indexing in Stubley. Stubley and Karandish are similar devices because all provide natural language processing. One of ordinary skill in the art would have recognized that applying the known technique would improve the similar devices and resulted in an improved system, with a reasonable expectation of success, to provide a “more robust way than traditional parsing and tagging techniques” [Kenter: ¶ 0003].
Stubley, Karandish, Kenter does not appear to explicitly disclose “generating multiple substitute portions to replace the multi-cell table”.
However, the disclosure of Chen teaches:
applying one or more pre-processes includes identifying a multi-cell table in the source document and generating multiple substitute portions to replace the multi-cell table, each substitute po