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 .
Claims 1-11 are presented for examination.
This Office action is Non-Final.
Information Disclosure Statement
The information disclosure statement (IDS) filed on 07/14/2025 has been considered by the Examiner and made of record in the application file.
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 8-11 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The following describes the issues in a simplified form.
Step 1 – Statutory Category
Claim 8 is drawn to an apparatus (machine) comprising a processor and memory. This falls within a statutory category.
Step 2A, Prong One – Abstract Idea
The claim recites the steps of:
receiving a request and parameters from a user device,
sending signals to nodes with a copy of a large language model,
querying a database to retrieve vectors,
generating relevance scores, filtering results,
compiling results and the request into a prompt, and
displaying the results.
These steps describe data manipulation, retrieval, and preparation of information for further processing. Such operations are an example of an abstract idea (i.e., organizing and transmitting information).
Step 2A, Prong Two – Integration into a Practical Application
The claim does not integrate the abstract idea into a practical application:
The recited processor and memory are generic computer components.
The recited second database is described as a standard storage system.
There is no recitation of any improvement to how the processor, memory, or database themselves function.
The “compiling into a prompt” step is simply preparing information for input into a model, which is another form of data organization, not a technical improvement.
Accordingly, the claim does not integrate the abstract idea into a practical application.
Step 2B – Significantly More
The additional elements do not amount to significantly more than the abstract idea itself:
The processor, memory, and database are well-understood, routine, and conventional.
Routing requests to multiple nodes, filtering based on parameters, and generating relevance scores are generic information-retrieval operations.
Taken together, the claim merely instructs to implement the abstract idea of retrieving, filtering, and formatting information using conventional computer components.
With respect to dependent claims 9-11, the dependents add features such as extracting an image of the data source associated with the date source, relevance scores and routine networking/distribute system concepts (e.g., reverse proxy, load balancing, replica/copy control), however, they don’t transform the abstract idea of retrieving/filtering/formatting information into something significantly more. Accordingly, the dependents are also ineligible.
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.
Claims 1-7 are rejected under 35 U.S.C. 103 as being unpatentable over Elisco et al. (US 2022/0300711 A1, also cited on the IDS filed on 07/14/2025) hereinafter “Elisco”, further in view of Pugh et al. (US 6,658,423 B1) hereinafter “Pugh”.
With respect to claim 1, the Elisco reference discloses a non-transitory, processor-readable medium storing instructions that when executed by a processor, cause the processor to:
receive a plurality of data artifacts including documents or other type of data such as audio files, having a plurality of data types [see ¶0025, disclosing neural network 108 may be trained using a corpus of documents. As used in this disclosure “corpus of documents” are a collection of texts, recordings, and/or representations of documents;. As a non-limiting example, corpus of documents 112 may include texts, recordings, and/or representations associated with child welfare documents, wherein child welfare documents may include case notes, investigation narratives, case plans, and/or court reports; Additionally or alternatively, corpus of documents 112 may include textual representations and/or transcripts of video recordings, audio recordings, and/or codified familial input as a function of a child welfare case; Alternatively or additionally, corpus of documents may include documents on a broader variety of subjects. Corpus of documents may include a large quantity of documents];
encode the plurality of data artifacts to a standard data type [see ¶0026, disclosing computing device may create an agency-specific extract, transform, load (ETL) procedure, module, and/or protocol, where each agency-specific ETL procedure, module, and/or protocol may be used to extract, transform, and/or load terms, tokens, and/or other semantic units from documents relating to a specific agency such as a child welfare agency or the like; extraction may be performed according any process and/or process step as described in this disclosure; In an embodiment, agency-specific ETLs may be created to allow data from one or more agencies to work within a natural language processing pipeline; actions performed by agency-specific ETLs may include such steps as removing HTML, tags, note formatting, new line prediction, and/or maintaining structured data, such as case identifiers, note identifiers, dates, authors, or the like, to accompany each note];
tokenize the plurality of encoded data artifacts, to produce a plurality of tokens from the plurality of encoded data artifacts, the plurality of tokens associated with natural-language identifiers extracted from the plurality of encoded data artifacts [see ¶0046, disclosing computing device 104 may tokenize textual input 304 prior to provision to base network 204. Tokenization may be performed in any manner that may occur to a person skilled in the art upon reading the entirety of this disclosure; for instance, computing device 104 may split long strings into words, sub-words and other semantic units to feed into neural network 108];
transform, using an embedding model, the plurality of tokens to produce a plurality of vectors, the plurality of vectors stored in a second database and classified based on a plurality of categories, the second database configured to be queried to perform a semantic search in response to receiving a request from a user operating a user compute device [see ¶0072, disclosing a transformer machine learning model is illustrated; Transformer machine learning model includes an encoder element 704. As used in this disclosure an “encoder element” is an element that encodes inputs; Encoder element 704 may be comprised of a multi-head attention 708 and a feed-forward neural network 712. As used in this disclosure “multi-head attention” is a scaled dot-product attention that establishes a weight for every input in the sequence, such that a relative position in an n-dimensional space is established for an n-dimensional vector of each significant term, word, and/or semantic unit; Multi-head attention may be calculated as a function of a matrix, Q, that contains vector representation of a term in a sequence, a vector representation of all the words in the sequence, K, and the value associated with the values of the vectors that represent all of the words in the sequence, V]; and
retrieve, from the semantic search, a subset of vectors from the plurality of vectors in the second database to be displayed on the user compute device [see ¶0066, disclosing computing device 104 may include a search window as a function of the concatenation database and current document sequence 132; As used in this disclosure a “search window” is a location in the graphical user interface that allows a user to enter text, symbols, representations, and the like thereof, wherein the entered information searches across both the concatenation database and current document sequence 132; A user may enter information into search window, wherein computing device 104 may return document sequences from both current document sequence 132 as well as the concatenation database that contain the entered search information and/or semantic units that have a pre-determined limit associated with the degree of similarity to the entered information, wherein a degree of similarity is described above in detail; As a non-limiting example, a user may enter “students that sell marijuana”, wherein search results may include search results with the terms “students that sell marijuana”, “hashish transactions among undergraduates”, “Sale of kush in scholastic settings”, and the like thereof].
Elisco discloses the non-transitory, processor-readable medium, as referenced above.
Elisco does not explicitly disclose compute, for each data artifact from the plurality of data artifacts, a hash function from a plurality of hash functions, the plurality of hash functions and the plurality of encoded data artifacts stored in a first database.
However, Pugh et al. discloses compute, for each data artifact from the plurality of data artifacts, a hash function from a plurality of hash functions, the plurality of hash functions and the plurality of encoded data artifacts stored in a first database [see col. 4, lines 1-6, disclosing the act of generating fingerprints for each document may be effected by (i) extracting parts (e.g., words) from the documents, (ii) hashing each of the extracted parts to determine which of a predetermined number of lists is to be populated with a given part, and (iii) for each of the lists, generating a fingerprint].
It would have been obvious before the effective filing date of the invention to a person having ordinary skill in the art to which said subject matter pertains to modify the document processing system as disclosed by Elisco with the document fingerprint database as disclosed by Pugh. Doing so would have enhanced Elisco since the fingerprint database requires significantly less storage than keeping complete files and comparisons between fingerprints can be performed much faster than comparing the original data [Pugh, see col. 6, lines 1-8].
With respect to claim 2, as modified the combination of Elisco and Pugh discloses the non-transitory, processor-readable medium of claim 1, as referenced above. The combination further discloses wherein: the plurality of data artifacts is a first plurality of data artifacts, the plurality of hash functions is a first plurality of hash functions, and the processor is further caused to:
receive a second plurality of data artifacts [Elisco, see ¶0057, disclosing computing device receives a current document sequence 132; As used in this disclosure a “current document sequence” is a collection of texts, recordings, and/or representations that disclose a current and/or novel document sequence; A current document sequence may be received in a single batch of one or more documents, and/or may be received a portion at a time, with additional documents updated as they become available; for instance, new notes or other entries in a case history may be added after one iteration of one or more methods or method steps as described in this disclosure, which may be processed in a subsequent iteration and/or prompt a subsequent iteration];
compute, for each data artifact from the second plurality of data artifacts, a hash function from a second plurality of hash functions [Pugh, see col. 4, lines 1-6, disclosing the act of generating fingerprints for each document may be effected by (i) extracting parts (e.g., words) from the documents, (ii) hashing each of the extracted parts to determine which of a predetermined number of lists is to be populated with a given part, and (iii) for each of the lists, generating a fingerprint]; and
query the first database to determine, for each hash function from the plurality of second hash functions, an instance of that hash function in the first database, such that if that hash function is not recorded in the first database, store that hash function and a data artifact associated with that hash function in the first database [Pugh, see col. 7, lines 37-55, disclosing at a high level, the present invention may function to detect near-duplicate documents (e.g., Web pages); To reiterate, it will be presumed that detecting near-duplicate document will necessarily also detect exact duplicate documents; Therefore, when the term "near-duplicate detection" is used, it will be understood that exact duplicates will also be detected, though not necessarily identified as "exact", as opposed to near, duplicates; The present invention may detect near-duplicate documents by (i) for each document, generating fingerprints, (ii) preprocessing (optionally) the fingerprints to eliminate those that only occur in one document, and (iii) determining near-duplicate documents based on the (remaining) fingerprints; The act of generating fingerprints for each document may be effected by (i) extracting parts (e.g., words) from the documents, (ii) hashing each of the extracted parts to determine which of a predetermined number of lists is to be populated with a given part, and (iii) for each of the lists, generating a fingerprint; as understood by the Examiner only fingerprints that are new are being generated and stored in order to avoid near-duplicates].
With respect to claim 3, as modified the combination of Elisco and Pugh discloses the non-transitory, processor-readable medium of claim 1, as referenced above. The combination further discloses wherein the plurality of tokens represents a fixed size of a paragraph being extracted from a data artifact from the plurality of data artifacts [Elisco, see ¶0038, disclosing “semantic units” are words, phrases, sentences, and/or “n-grams” of words, defined as a set of n words appearing contiguously in a text].
With respect to claim 4, as modified the combination of Elisco and Pugh discloses the non-transitory, processor-readable medium of claim 1, as referenced above. The combination further discloses wherein the plurality of tokens represents an overlap between paragraphs in a data artifact from the plurality of data artifacts [Elisco, disclosing defined as a set of n words appearing contiguously in a text].
With respect to claim 5, as modified the combination of Elisco and Pugh discloses the non-transitory, processor-readable medium of claim 1, as referenced above. The combination further discloses wherein a vector from the plurality of vectors includes a 1-D vector that represents a string of integer numbers [Pugh, see col. 9, lines 29-31, disclosing a document identifier will be associated with a predetermined number (e.g., four) of fingerprints as indicated by 365].
With respect to claim 6, as modified the combination of Elisco and Pugh discloses the non-transitory, processor-readable medium of claim 1, as referenced above. The combination further discloses wherein the second database is not accessible to external devices [Elisco, see Fig. 1, the architecture suggest that it is local and internal to the computing device and not accessible externally].
With respect to claim 7, as modified the combination of Elisco and Pugh discloses the non-transitory, processor-readable medium of claim 1, as referenced above. The combination further discloses wherein the plurality of data artifacts is encoded automatically [Elisco, see ¶0026, disclosing computing device may create an agency-specific extract, transform, load (ETL) procedure, module, and/or protocol, where each agency-specific ETL procedure, module, and/or protocol may be used to extract, transform, and/or load terms, tokens, and/or other semantic units from documents relating to a specific agency such as a child welfare agency or the like; extraction may be performed according any process and/or process step as described in this disclosure; In an embodiment, agency-specific ETLs may be created to allow data from one or more agencies to work within a natural language processing pipeline; actions performed by agency-specific ETLs may include such steps as removing HTML, tags, note formatting, new line prediction, and/or maintaining structured data, such as case identifiers, note identifiers, dates, authors, or the like, to accompany each note; the process is automatic].
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.
Claims 8-11 are rejected under 35 U.S.C. 102 (a)(1) as being anticipated by Elisco.
With respect to claim 8, the Elisco reference an apparatus comprising:
a processor [see ¶0097, disclosing processor 1304 may include any suitable processor, such as without limitation a processor incorporating logical circuitry for performing arithmetic and logical operations, such as an arithmetic and logic unit (ALU), which may be regulated with a state machine and directed by operational inputs from memory and/or sensors]; and
a memory operatively coupled to the processor, the memory storing instructions to cause the processor [see ¶0098, disclosing memory 1308 may include various components (e.g., machine-readable media) including, but not limited to, a random-access memory component, a read only component, and any combinations thereof. In one example, a basic input/output system 1316 (BIOS), including basic routines that help to transfer information between elements within computer system 1300, such as during start-up, may be stored in memory 1308] to:
receive, from a user compute device, an input including a request and a set of parameters for the request [see ¶0055, disclosing computing device may be configured to perform semantic search, denoting search with meaning, as distinguished from lexical search where a search engine looks for literal matches of the query words or variants of them, without understanding the overall meaning of the query; Semantic search may seek to improve search accuracy by understanding a searcher's intent and a contextual meaning of terms as they appear in a searchable dataspace, whether on the Web or within a closed system, to generate more relevant results];
send a signal to at least one node from a plurality of nodes based on the request, each node from the plurality of nodes storing a copy of a large language model, the signal including instructions to execute the copy of the large language model from the at least one node [see ¶0043, disclosing computing device 104 may generate the named entity recognition by any suitable method, including without limitation a natural language processing classification algorithm; language processing model may include a natural language process classification model, for instance as generated by training neural network, that enumerates and/or derives statistical relationships between input term and output terms; Algorithm to generate language processing model may include a stochastic gradient descent algorithm, which may include a method that iteratively optimizes an objective function, such as an objective function representing a statistical estimation of relationships between terms, including relationships between input terms and output terms, in the form of a sum of relationships to be estimated; In an alternative or additional approach, sequential tokens may be modeled as chains, serving as the observations in a Hidden Markov Model (HMM). HMIs as used herein are statistical models with inference algorithms that that may be applied to the models; In such models, a hidden state to be estimated may include an association between semantic elements such as terms, phrases, tokens, etc. There may be a finite number of categories to which an extracted word may pertain; an HMM inference algorithm, such as the forward-backward algorithm or the Viterbi algorithm, may be used to estimate the most likely discrete state given a word or sequence of words];
query, via the copy of the large language model from the at least one node, a database storing a plurality of vectors with respect to the set of parameters and the request, to retrieve a subset of vectors from the plurality of vectors in the database [see ¶0052, disclosing computing device 104 may produce a query element as a function of neural network 108; As used in this disclosure “query element” is a graphical control element comprising a search field and/or search bar; As a non-limiting example, query element may include a search relating to all documents associated with sexual assault; As a further non-limiting example, query element may include a search relating to all documents associated with a particular time metric such as a particular day, month, and/or year; Computing device 104 may produce query element by generating a concatenation database];
generate a relevance score for a data source associated with the subset of vectors [see ¶0081, disclosing for instance, a supervised learning algorithm may include the corpus of documents as described above as inputs, vectors as outputs, and a scoring function representing a desired form of relationship to be detected between inputs and outputs; scoring function may, for instance, seek to maximize the probability that a given input and/or combination of elements inputs is associated with a given output to minimize the probability that a given input is not associated with a given output];
filter the subset of vectors based on the set of parameters to retrieve a filtered subset of vectors [see ¶0055, disclosing this search functionality may be used in combination with any other search and/or filtration processes and/or protocols described in this disclosure, including without limitation note filtering functions for sentences, people, and/or queries]; and
compile the filtered subset of vectors and the request to generate a prompt to be passed to the large language model for processing and to be displayed on the user compute device [see ¶0066, disclosing computing device 104 may include a search window as a function of the concatenation database and current document sequence 132; As used in this disclosure a “search window” is a location in the graphical user interface that allows a user to enter text, symbols, representations, and the like thereof, wherein the entered information searches across both the concatenation database and current document sequence 132; A user may enter information into search window, wherein computing device 104 may return document sequences from both current document sequence 132 as well as the concatenation database that contain the entered search information and/or semantic units that have a pre-determined limit associated with the degree of similarity to the entered information, wherein a degree of similarity is described above in detail; As a non-limiting example, a user may enter “students that sell marijuana”, wherein search results may include search results with the terms “students that sell marijuana”, “hashish transactions among undergraduates”, “Sale of kush in scholastic settings”, and the like thereof].
With respect to claim 9, Elisco discloses the apparatus of claim 8, as referenced above. Elisco further discloses wherein the memory stores instructions to further cause the processor to extract an image of the data source associated with the filtered subset of vectors to be displayed on the display of the user compute device [see ¶0085, disclosing selecting a sentence category may filter first panel 1004 to display only notes containing a threshold number of sentences, such as without limitation at least one sentence, belonging to that particular category; Clicking individual circles on second panel 1008 may filter for both category and date].
With respect to claim 10, Elisco discloses the apparatus of claim 8, as referenced above. Elisco further discloses wherein the memory stores instructions to cause the processor to generate a relevance score from a plurality of relevance scores for each data source from a plurality of data sources associated with the subset of vectors retrieved from the database [see ¶0081, disclosing for instance, a supervised learning algorithm may include the corpus of documents as described above as inputs, vectors as outputs, and a scoring function representing a desired form of relationship to be detected between inputs and outputs; scoring function may, for instance, seek to maximize the probability that a given input and/or combination of elements inputs is associated with a given output to minimize the probability that a given input is not associated with a given output].
With respect to claim 11, Elisco discloses the apparatus of claim 8, as referenced above. Elisco further discloses wherein the memory stores instructions to cause the processor to execute a reverse proxying technique to control multiple replicas of the large learning model in order to achieve high-throughput and scalability for a large number of concurrent queries or users [see ¶0022, disclosing information (e.g., data, software etc.) may be communicated to and/or from a computer and/or a computing device. computing device 104 may include but is not limited to, for example, a computing device or cluster of computing devices in a first location and a second computing device or cluster of computing devices in a second location; Computing device 104 may include one or more computing devices dedicated to data storage, security, distribution of traffic for load balancing, and the like; Computing device 104 may distribute one or more computing tasks as described below across a plurality of computing devices of computing device, which may operate in parallel, in series, redundantly, or in any other manner used for distribution of tasks or memory between computing devices; Computing device 104 may be implemented using a “shared nothing” architecture in which data is cached at the worker, in an embodiment, this may enable scalability of system 100 and/or computing device].
Prior Art Made of Record
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Browder et al. discloses identifying outlier data and generating corrective action.
Kumar M et al. discloses note data analysis to identify unmet needs and generation of data structures.
Fatal et al. discloses real-time normalization of raw enterprise data from disparate sources.
Sullivan et al. discloses generating an automated output as a function of an attribute datum and key datums.
Haldar et al. discloses generation of data stories and data summaries based on user queries.
Ragukumar et al. discloses data preprocessing to optimize resource consumption of a large model.
Patil et al. discloses prompt-based data structures and document retrieval.
TCA et al. discloses proactive determination of data insights.
St. Martin et al. discloses aggregation of global story based on analyzed data.
Sriharsha et al. discloses data field extraction by a data intake and query system.
Conclusions/Points of Contacts
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JORGE A CASANOVA whose telephone number is (571)270-3563. The examiner can normally be reached M-F: 9 a.m. to 6 p.m. (EST).
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, Aleksandr Kerzhner can be reached at (571) 270-1760. 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.
/JORGE A CASANOVA/Primary Examiner, Art Unit 2165