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 .
DETAILED ACTION
This action is issued in response to Applicants amendment filed July 09, 2026.
Claims 1-20 are pending. No claim is added and none cancelled.
Claims 1-17 are rejected.
Claims 18-20 are withdrawn from further consideration pursuant to 37 CFR 1.142(b), as being drawn to a nonelected invention, there being no allowable generic or linking claim.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claim(s) 1, 5-9, and 13-17 is/are rejected under 35 U.S.C. 103 as being unpatentable over Brenner (U.S. Patent Application No. 2025/0165463) in view of Qadrud-Din (U.S. Patent Application No. 2024/0289561), further in view of Azizi (U.S. Patent Application No. 2024/0403339).
Regarding Claim 1, Brenner discloses a method, comprising:
receiving a query regarding a source document that comprises one or more fields (par [0047], Brenner – retriever model receives a user query, wherein the retriever model 314 is used to identify document chunks 308 from a corpus that are relevant to each user query 318);
determining a field of the one or more fields referenced by the query (par [0047], [0062], Brenner – retriever model receives a user query, wherein the retriever model 314 is used to identify document chunks 308 from a corpus that are relevant to each user query 318);
retrieving first contextual metadata for the field (par [0048], [062-0063], Brenner);
generating an enriched query by adding the first contextual metadata to query text of the query (par [0048], Brenner - both the user query 318 and the context 322 can be provided as inputs to the generative model 316. Collectively, the user query 318 and the context 322 can be said to form at least part of a prompt for the generative model 316);
generating a query embedding from the enriched query (par [0051], Brenner - ranker generates independent embeddings (embedding vectors) for input queries 318 and document chunks 308, and the embeddings are used to compute similarities between the input queries 318 and the document chunks 308);
determining similarity scores between the query embedding and passage embeddings, wherein the passage embeddings are based on passage text from one or more resource documents (par [0051], [0072], [0079], Brenner – the ranker generates independent embeddings (embedding vectors) for input queries and document chunks, and the embeddings are used to compute similarities between the input queries and the document chunks… the ranker 402 is configured to receive and process a set 502 of document chunks 308 or embedding vectors associated with the set 502 of document chunks 308. The ranker 402 can determine which of the document chunks 308 appear to be most relevant to an associated user query 318, for example, by determining a similarity score between each document chunk 308 and the associated query); and
identifying one or more passages based on the similarity scores that satisfy a threshold (par [0079], Brenner - The identified information chunks are provided to the generative model at step 608, and the generative model is used to process the identified information chunks and generate a response to the input query at step 610. This may include providing the query 318 and a context 322 (the top K document chunks 404) to the generative model 316).
While Brenner teaches all of the claimed subject matter as stated above. However, Brenner is not as detailed with respect to passage in one or more resource document.
On the other hand, Qadrud-Din discloses passage in one or more resource document (par [0183], Qadrud-Din – find any passage(s) in the document that will help answer the query, wherein the system extracts passages from the document and assigns a score to each passage based on how the passage relates to the query).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate Qadrud-Din’s teachings into Brenner’s RAG system optimization. A skilled artisan would have been motivated to combine in order to provide an improved natural language processing environment that seamlessly interacts with large companies to better evaluate documents against particular criteria.
While Brenner teaches receiving a query to identify document chunks from a corpus of documents relevant to the query and determining similarity scores between the query embedding and passage embeddings, wherein the passage embeddings are based on passage text from one or more resource documents. However, Brenner is not as explicitly detailed with respect to the one or more resource documents are different documents than the source document.
On the other hand, Azizi discloses the one or more resource documents are different documents than the source document (par [0040], [0068-0069], Azizi – system receives a document query, wherein the document query may be an article, a reference to the article, or a text query… the system encodes a set of candidate sentences from a candidate document to obtain a set of contextual sentence embeddings, wherein the encode sentences from candidate documents are within a database. The query document corresponds to the source document and the candidate document(s) correspond to the one or more resource documents).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate Azizi’s teachings into Brenner’s and Qadrud-Din optimization system. A skilled artisan would have been motivated to combine in order to generate document embeddings that capture the contextual information of documents; therefore leading to more accurate document retrieval.
Regarding Claim 5, the combination of Brenner in view of Qadrud-Din, further in view of Azizi, disclose the method of claim 1, further comprising ranking the one or more passages with a large language model based on the query, the field, and the first contextual metadata for the field (par [0051], [0072], Brenner – the ranker generates independent embeddings (embedding vectors) for input queries and document chunks, and the embeddings are used to compute similarities between the input queries and the document chunks… the ranker 402 is configured to receive and process a set 502 of document chunks 308 or embedding vectors associated with the set 502 of document chunks 308. The ranker 402 can determine which of the document chunks 308 appear to be most relevant to an associated user query 318, for example, by determining a similarity score between each document chunk 308 and the associated query… par [0005], Brenner – RAG system optimization that includes a plurality of rankers such as a bi-encoder, a cross-encoder, and a large language model (LLM)-ranker).
Regarding Claim 6, the combination of Brenner in view of Qadrud-Din, further in view of Azizi, disclose the method of claim 5, further comprising: prompting the large language model to generate a response to the query based on one or more respective rankings of the one or more passages; and receiving the response from the large language model (par [0007-0008], Brenner – obtain an input query at a retriever model which includes rankers, then processing information chunks from the retriever model to a generative model to create a response to the input query wherein the ranker can be an LLM ranker… method also includes generating a prompt for the generative model).
Regarding Claim 7, the combination of Brenner in view of Qadrud-Din, further in view of Azizi, disclose the method of claim 1, wherein the source document is a tax form and the field is a tax form field (par [0030], [0033], [0053], Qadrud-Din – reference discusses domain-specific contexts such as law, wherein text segmentation occurs that indicates different legal columns/fields and the LLM generates legal documents; however, it is obvious to allow for other domain-specific categories as well, such as tax).
Regarding Claim 8, the combination of Brenner in view of Qadrud-Din, further in view of Azizi, disclose the method of claim 1, wherein at least one of the one or more resource documents comprises instructions for completing the source document (par [0225], [0332-0333], Qadrud-Din).
Claims 9 and 13-17 contain similar subject matter as claims 1 and 5-8 above; and are rejected under the same rationale.
Claim(s) 2-4 and 10-12 is/are rejected under 35 U.S.C. 103 as being unpatentable over Brenner in view of Qadrud-Din, further in view of Azizi, and further in view Jonassen (U.S. Patent Application No. 2020/0293528).
Regarding Claim 2, the combination of Brenner in view of Qadrud-Din, further in view of Azizi, disclose identifying document chunks containing specified fields of information for use by a generative model, wherein each individual field of information that is to be extracted from a document chunk (see par [0062-0063], [0066], Brenner); and Qadrud-Din teaches text segmentation (see par [0029-0030], [0068], Qadrud-Din). It is obvious that structural elements are determined within the system. However, Brenner, Qadrud-Din, and Azizi are not as explicitly detailed as the examiner would like.
On the other hand, Jonassen discloses determining structural elements from the source document (par [0005-0006], [0023], Jonassen); identifying the field in the source document based on the structural elements (par [0031], [0051], Jonassen - structural rules module may be configured to facilitate configuration of structural rules for populating fields of the structured reports); and determining the first contextual metadata associated with the field (par [0031-0032], Jonassen - one structural rule may apply to various fields of a structured report, and may define what information is to be included in the various fields).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate Jonassen’s teachings into the Brenner, Qadrud-Din, and Azizi system. A skilled artisan would have been motivated to combine in order to provide easily digestible documents to all the system to perform at a more efficient and user friendly manner.
Regarding Claim 3, the combination of Brenner in view of Qadrud-Din, further in view of Azizi, and further in view of Jonassen, disclose the method of claim 2, further comprising executing a machine learning model to identify the field and determine the first contextual metadata (par [0030], [0035], Jonassen - classification algorithms may include any combination of machine learning algorithms, statistical algorithms, and/or any algorithm configured to identify correlations and/or relationships within and between data, and to train a model such that the model may be used to classify data according to the training).
Regarding Claim 4, the combination of Brenner in view of Qadrud-Din, further in view of Azizi, and further in view of Jonassen, disclose the method of claim 1, further comprising identifying structural elements in a resource document of the one or more resource documents (par [0005-0006], [0023], Jonassen); segmenting the resource document into passages of text based on the structural elements (par [0029-0030], [0068], [0183], Qadrud-Din – text segmentation… find any passage(s) in the document that will help answer the query, wherein the system extracts passages from the document and assigns a score to each passage based on how the passage relates to the query); and determining second contextual metadata for a respective passage of the passages of text based on the structural elements and respective passage text of the respective passage wherein a passage embedding, of the passage embeddings, of the respective passage is based on the respective passage text and the second contextual metadata (par [0031-0032], Jonassen… par [0029-0030], [0068], [0183], Qadrud-Din).
Claims 10-12 contain similar subject matter as claims 2-4 above; and are rejected under the same rationale.
Response to Arguments
Applicant’s arguments with respect to amended claim(s) have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
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.
Points of Contact
Any inquiry concerning this communication or earlier communications from the examiner should be directed to CHELCIE L DAYE whose telephone number is (571) 272-3891. The examiner can normally be reached on Monday-Friday 7:30-4:00pm. 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, Apu Mofiz can be reached on 571-272-4080. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free).
Chelcie Daye
Patent Examiner
Technology Center 2100
August 31, 2026
/CHELCIE L DAYE/Primary Examiner, Art Unit 2161
Prosecution Insights