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 .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 12/08/2025 has been entered.
Claim Status
Claims 1-4, 6-20 are pending.
Claims 1-4, 6-20 are rejected.
Response to Arguments
101 Rejection: Applicant argues claim 1 does not recite a judicial exception. Examiner disagree with applicant’s argument because the claim recites “the accelerator configured to process a query embedding vector with the document embedding vector”. This limitation is directed to comparing vectors (i.e., mathematic values) and therefore is a mathematical calculation.
Applicant argues claim 8 does not recite a judicial exception. Examiner disagrees with applicant’s argument because the claim recites “receiving from the storage device a result of processing the query embedding vector with the document embedding vector”. The element “result of processing the query embedding vector with the document embedding vector” corresponds to a mathematical calculation.
Applicant argues claim 17 does not recite a judicial exception. Examiner disagrees with applicant’s argument because the claim recites “performing a similarity search by the accelerator using the query embedding vector and the document embedding vector to produce a result”. Determining the similarity of vectors is a performance of a mathematical calculation.
The additional elements of the intendent claims fail to integrates the judicial exception into a practical application or amount to significantly more than the judicial exception. The additional elements or the dependent claims fail to integrates the judicial exception into a practical application or amount to significantly more than the judicial exception.
Prior Art Rejection: Applicant’s arguments with respect to claims 1-4, 6-20 have been fully considered and are persuasive. Upon further consideration, and in view of applicant’s amendments, a new grounds of rejection is made in view of newly cited reference Dumitru.
Claim Rejections - 35 § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The following is Examiner's analysis of the claimed invention under the 2019 Revised Patent Subject Matter Eligibility Guidance (PEG)
STEP 1 ls the claim to a Process, Machine, Manufacture or Composition of matter? Yes. Claim 1 recites a machine (system), claim 8 recites a process (method), claim 17 recites a process (method).
STEP2A Prone one: Does The Claim Recite An Abstract Idea, Law Of Nature, or Natural Phenomenon? Yes. Claim 1 recites “the accelerator configured to process a query embedding vector and the document embedding vector” which falls within the mathematical concepts grouping of abstract ideas. The specification discloses “one approach is to use the K Nearest Neighbors (KNN) algorithm to identify k documents whose embedding vectors are closest to the query embedding vector. Another approach is to use matrix multiplication or a dot-product calculation to determine a similarity score between the query and the documents” (Page 5 lines 27-30). In view of the specification, the step of “processing an embedding vector” cover the performance of a mathematical calculation, and therefore, the claim recites an abstract idea.
Claim 8 recites “receiving from the storage device a result; receiving from the second storage device a second result” which falls within the mathematical concepts grouping of abstract ideas. The step of “receiving a result” corresponds to the performance of a mathematical calculation, and therefore, the claim recites an abstract idea.
Claim 17 recites “performing a similarity search by the accelerator using the query embedding vector and the document embedding vector to produce a result; to perform a similarity search using a second query embedding vector and the document embedding vector to produce a second result” which falls within the mathematical concepts grouping of abstract ideas. The step of “performing a similarity search using vectors” covers the performance of a mathematical calculation, and therefore, the claim recites an abstract idea.
STEP2A Prone two: Does The Claim Recite Additional Elements That Integrate The Judicial Exception Into A Practical Application? No. Claim 1 recites “a system, comprising: a storage device; an accelerator connected to the storage device” which amounts to merely including instructions to implement an abstract idea on a computer.
Claim 1 recites “the storage device storing a document and a document embedding vector associated with the document, the document embedding vector including at least four dimensions; a second storage device, the second storage device storing a second document and a second document embedding vector associated with the second document, the second document embedding vector including at least four dimensions; and a processor connected to the storage device, the accelerator, the second storage device, and the second accelerator, the processor configured to transmit the query embedding vector to the accelerator based at least in part on the storage device storing the document embedding vector and to transmit the query embedding vector to the second accelerator based at least in part on the second storage device storing the second document embedding vector, wherein the processor is configured to access the document from the storage device based at least in part on the accelerator and the second accelerator processing the query embedding vector and the document embedding vector” which is mere necessary data gathering.
Claim 8 recites “identifying a query embedding vector at a processor; determining that a document embedding vector is stored on a storage device, the document embedding vector associated with a document, the document embedding vector including at least four dimensions, the storage device further storing the document; determining that a second document embedding vector is stored on a second storage device, the second document embedding vector associated with a second document, the second document embedding vector including at least four dimensions, the second storage device further storing the second document; sending the query embedding vector to an accelerator connected to the storage device; sending the query embedding vector to a second accelerator connected to the second storage device” which is mere necessary data gathering.
Claim 8 recites “and transmitting the document and the second document based at least in part on the result and the second result” which is insignificant-extra solution activity. Adding a final step of transmitting data does not add a meaningful limitation to the judicial exception, and therefore, the additional element is insignificant-extra solution activity.
Claim 17 recites “receiving a query embedding vector from a processor at an accelerator, the accelerator connected to a storage device; accessing a document embedding vector from the storage device by the accelerator, the document embedding vector associated with a document, the document embedding vector including at least four dimensions, the storage device further storing the document; and transmitting the result to the processor from the accelerator, wherein the processor, the accelerator, and the storage device are part of a system further including a second accelerator and second storage device, the second accelerator configured to access a second document embedding vector from the second storage device” which is mere necessary data gathering.
Step 2B: Does the claim recite additional elements that amount to significantly more than the judicial exception? No. Claims 1, 8, and 17 recite mere instructions to implement an abstract idea on a computer. The courts have determined merely including instructions to implement the abstract idea on a computer does not qualify as “significantly more” when recited in a claim with a judicial exception (See Alice Corp., 573 U.S. at 225-26, 110 USPQ2d at 1984).
Claims 1, 8, and 17 recite mere necessary data gathering. The courts have determined mere data gathering to not be enough to qualify as “significantly more” when recited in a claim with a judicial exception (See CyberSource v. Retail Decisions, Inc., 654 F.3d 1366, 1375, 99 USPQ2d 1690, 1694 (Fed. Cir. 2011)).
Claim 8 recites “and transmitting the document and the second document based at least in part on the result and the second result” which is transmitting data over a network. The courts have determined transmitting data over a network is well‐understood, routine, and conventional functionality when claimed in a merely generic manner (see Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362).
There is no indication that the elements of the claim, individually nor in combination, integrate the judicial exception into a practical application or amount to significantly more than the judicial exception.
For the reasons above, claims 1, 8, and 17 are rejected as being directed to nonpatentable subject matter under §101. This rejection applies equally to the dependent claims. The additional limitations of the dependent claims are addressed briefly below:
Regarding claim 2
STEP 1 ls the claim to a Process, Machine, Manufacture or Composition of matter? Yes. The claim recites a machine (system).
STEP2A Prone one: Does The Claim Recite An Abstract Idea, Law Of Nature, or Natural Phenomenon? Yes. The claim inherits the abstract idea of the parent claim.
STEP2A Prone two: Does The Claim Recite Additional Elements That Integrate The Judicial Exception Into A Practical Application? No. The claim recites “wherein the storage device includes the accelerator” which amounts to merely including instructions to implement an abstract idea on a computer.
Step 2B: Does the claim recite additional elements that amount to significantly more than the judicial exception? No. The courts have determined merely including instructions to implement the abstract idea on a computer does not qualify as “significantly more” when recited in a claim with a judicial exception (See Alice Corp., 573 U.S. at 225-26, 110 USPQ2d at 1984).
Regarding claim 3
STEP 1 ls the claim to a Process, Machine, Manufacture or Composition of matter? Yes. The claim recites a machine (system).
STEP2A Prone one: Does The Claim Recite An Abstract Idea, Law Of Nature, or Natural Phenomenon? Yes. The claim recites “wherein the accelerator is configured to perform a similarity search using the query embedding vector and the document embedding vector to produce a result” which falls within the mathematical concepts grouping of abstract ideas.
STEP2A Prone two: Does The Claim Recite Additional Elements That Integrate The Judicial Exception Into A Practical Application? No. There is no indication that the elements of the claim
integrate the judicial exception into a practical application.
Step 2B: Does the claim recite additional elements that amount to significantly more than the judicial exception? No. There is no indication that the elements of the claim, individually nor in
combination, integrate the judicial exception into a practical application or amount to significantly more than the judicial exception.
Regarding claim 4
STEP 1 ls the claim to a Process, Machine, Manufacture or Composition of matter? Yes. The claim recites a machine (system).
STEP2A Prone one: Does The Claim Recite An Abstract Idea, Law Of Nature, or Natural Phenomenon? Yes. The claim recites “wherein the processor is configured to perform a second similarity search using the query embedding vector and a second document embedding vector to generate a second result” which falls within the mathematical concepts grouping of abstract ideas.
STEP2A Prone two: Does The Claim Recite Additional Elements That Integrate The Judicial Exception Into A Practical Application? No. There is no indication that the elements of the claim
integrate the judicial exception into a practical application.
Step 2B: Does the claim recite additional elements that amount to significantly more than the judicial exception? No. There is no indication that the elements of the claim, individually nor in
combination, integrate the judicial exception into a practical application or amount to significantly more than the judicial exception.
Regarding claim 5
STEP 1 ls the claim to a Process, Machine, Manufacture or Composition of matter? Yes. The claim recites a machine (system).
STEP2A Prone one: Does The Claim Recite An Abstract Idea, Law Of Nature, or Natural Phenomenon? Yes. The claim inherits the abstract idea of the parent claim.
STEP2A Prone two: Does The Claim Recite Additional Elements That Integrate The Judicial Exception Into A Practical Application? No. The claim recites “a memory including the second document embedding vector” which is mere necessary data gathering.
Step 2B: Does the claim recite additional elements that amount to significantly more than the judicial exception? No. The courts have determined mere data gathering to not be enough to qualify as “significantly more” when recited in a claim with a judicial exception (See CyberSource v. Retail Decisions, Inc., 654 F.3d 1366, 1375, 99 USPQ2d 1690, 1694 (Fed. Cir. 2011)).
Regarding claim 6
STEP 1 ls the claim to a Process, Machine, Manufacture or Composition of matter? Yes. The claim recites a machine (system).
STEP2A Prone one: Does The Claim Recite An Abstract Idea, Law Of Nature, or Natural Phenomenon? Yes. The claim inherits the abstract idea of the parent claim.
STEP2A Prone two: Does The Claim Recite Additional Elements That Integrate The Judicial Exception Into A Practical Application? No. Claim 6 recites “wherein the processor is configured to combine the result and the second result” which is mere necessary data gathering.
Step 2B: Does the claim recite additional elements that amount to significantly more than the judicial exception? No. The courts have determined mere data gathering to not be enough to qualify as “significantly more” when recited in a claim with a judicial exception (See CyberSource v. Retail Decisions, Inc., 654 F.3d 1366, 1375, 99 USPQ2d 1690, 1694 (Fed. Cir. 2011)).
Regarding claim 7
STEP 1 ls the claim to a Process, Machine, Manufacture or Composition of matter? Yes. The claim recites a machine (system).
STEP2A Prone one: Does The Claim Recite An Abstract Idea, Law Of Nature, or Natural Phenomenon? Yes. The claim inherits the abstract idea of the parent claim.
STEP2A Prone two: Does The Claim Recite Additional Elements That Integrate The Judicial Exception Into A Practical Application? No. The claim recites “wherein the processor is configured to copy the document embedding vector into a memory based at least in part on the accelerator comparing the query embedding vector with the document embedding vector” which is mere necessary data gathering.
Step 2B: Does the claim recite additional elements that amount to significantly more than the judicial exception? No. The courts have determined mere data gathering to not be enough to qualify as “significantly more” when recited in a claim with a judicial exception (See CyberSource v. Retail Decisions, Inc., 654 F.3d 1366, 1375, 99 USPQ2d 1690, 1694 (Fed. Cir. 2011)).
Regarding claim 9
STEP 1 ls the claim to a Process, Machine, Manufacture or Composition of matter? Yes. The claim recites a process (method).
STEP2A Prone one: Does The Claim Recite An Abstract Idea, Law Of Nature, or Natural Phenomenon? Yes. The claim inherits the abstract idea of the parent claim.
STEP2A Prone two: Does The Claim Recite Additional Elements That Integrate The Judicial Exception Into A Practical Application? No. The claim recites “wherein identifying the query embedding vector includes: receiving a query at the processor; and generating the query embedding vector based at least in part on the query” which is mere necessary data gathering.
Step 2B: Does the claim recite additional elements that amount to significantly more than the judicial exception? No. The courts have determined mere data gathering to not be enough to qualify as “significantly more” when recited in a claim with a judicial exception (See CyberSource v. Retail Decisions, Inc., 654 F.3d 1366, 1375, 99 USPQ2d 1690, 1694 (Fed. Cir. 2011)).
Regarding claim 10
STEP 1 ls the claim to a Process, Machine, Manufacture or Composition of matter? Yes. The claim recites a process (method).
STEP2A Prone one: Does The Claim Recite An Abstract Idea, Law Of Nature, or Natural Phenomenon? Yes. The claim inherits the abstract idea of the parent claim.
STEP2A Prone two: Does The Claim Recite Additional Elements That Integrate The Judicial Exception Into A Practical Application? No. The claim recites “wherein generating the query embedding vector based at least in part on the query includes generating the query embedding vector at the processor based at least in part on the query” which is mere necessary data gathering.
Step 2B: Does the claim recite additional elements that amount to significantly more than the judicial exception? No. The courts have determined mere data gathering to not be enough to qualify as “significantly more” when recited in a claim with a judicial exception (See CyberSource v. Retail Decisions, Inc., 654 F.3d 1366, 1375, 99 USPQ2d 1690, 1694 (Fed. Cir. 2011)).
Regarding claim 11
STEP 1 ls the claim to a Process, Machine, Manufacture or Composition of matter? Yes. The claim recites a process (method).
STEP2A Prone one: Does The Claim Recite An Abstract Idea, Law Of Nature, or Natural Phenomenon? Yes. The claim inherits the abstract idea of the parent claim.
STEP2A Prone two: Does The Claim Recite Additional Elements That Integrate The Judicial Exception Into A Practical Application? No. The claim recites “wherein transmitting the document based at least in part on the result includes retrieving the document from the storage device” which is insignificant-extra solution activity tangentially related to the invention.
Step 2B: Does the claim recite additional elements that amount to significantly more than the judicial exception? No. The claim recites “wherein transmitting the document based at least in part on the result includes retrieving the document from the storage device” which is transmitting data over a network. The courts have determined transmitting data over a network is well‐understood, routine, and conventional functionality when claimed in a merely generic manner (see Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362).
Regarding claim 12
STEP 1 ls the claim to a Process, Machine, Manufacture or Composition of matter? Yes. The claim recites a process (method).
STEP2A Prone one: Does The Claim Recite An Abstract Idea, Law Of Nature, or Natural Phenomenon? Yes. The claim inherits the abstract idea of the parent claim.
STEP2A Prone two: Does The Claim Recite Additional Elements That Integrate The Judicial Exception Into A Practical Application? No. The claim recites “wherein transmitting the document based at least in part on the result includes retrieving the document from a second storage device” which is insignificant-extra solution activity tangentially related to the invention.
Step 2B: Does the claim recite additional elements that amount to significantly more than the judicial exception? No. The claim recites “wherein transmitting the document based at least in part on the result includes retrieving the document from the storage device” which is transmitting data over a network. The courts have determined transmitting data over a network is well‐understood, routine, and conventional functionality when claimed in a merely generic manner (see Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362).
Regarding claim 13
STEP 1 ls the claim to a Process, Machine, Manufacture or Composition of matter? Yes. The claim recites a process (method).
STEP2A Prone one: Does The Claim Recite An Abstract Idea, Law Of Nature, or Natural Phenomenon? Yes. The claim recites “the method further comprises processing the query embedding vector and a second document embedding vector using the processor to produce a second result; and transmitting the document based at least in part on the result includes: combining the result and the second result using the processor to produce a combined result; and transmitting the document based at least in part on the combined result” which falls within the mathematical concepts grouping of abstract ideas.
STEP2A Prone two: Does The Claim Recite Additional Elements That Integrate The Judicial Exception Into A Practical Application? No. There is no indication that the elements of the claim
integrate the judicial exception into a practical application.
Step 2B: Does the claim recite additional elements that amount to significantly more than the judicial exception? No. There is no indication that the elements of the claim, individually nor in
combination, integrate the judicial exception into a practical application or amount to significantly more than the judicial exception.
Regarding claim 14
STEP 1 ls the claim to a Process, Machine, Manufacture or Composition of matter? Yes. The claim recites a process (method).
STEP2A Prone one: Does The Claim Recite An Abstract Idea, Law Of Nature, or Natural Phenomenon? Yes. The claim recites “the accelerator is configured to perform a first similarity search using the query embedding vector and the document embedding vector to produce the result; and processing the query embedding vector and the second document embedding vector to produce the second result includes performing a second similarity search using the query embedding vector and the second document embedding vector to generate the second result” which falls within the mathematical concepts grouping of abstract ideas.
STEP2A Prone two: Does The Claim Recite Additional Elements That Integrate The Judicial Exception Into A Practical Application? No. There is no indication that the elements of the claim
integrate the judicial exception into a practical application.
Step 2B: Does the claim recite additional elements that amount to significantly more than the judicial exception? No. There is no indication that the elements of the claim, individually nor in
combination, integrate the judicial exception into a practical application or amount to significantly more than the judicial exception.
Regarding claim 15
STEP 1 ls the claim to a Process, Machine, Manufacture or Composition of matter? Yes. The claim recites a process (method).
STEP2A Prone one: Does The Claim Recite An Abstract Idea, Law Of Nature, or Natural Phenomenon? Yes. The claim inherits the abstract idea of the parent claim.
STEP2A Prone two: Does The Claim Recite Additional Elements That Integrate The Judicial Exception Into A Practical Application? No. The claim recites “further comprising copying the document embedding vector from the storage device to a memory” which is mere necessary data gathering.
Step 2B: Does the claim recite additional elements that amount to significantly more than the judicial exception? No. The courts have determined mere data gathering to not be enough to qualify as “significantly more” when recited in a claim with a judicial exception (See CyberSource v. Retail Decisions, Inc., 654 F.3d 1366, 1375, 99 USPQ2d 1690, 1694 (Fed. Cir. 2011)).
Regarding claim 16
STEP 1 ls the claim to a Process, Machine, Manufacture or Composition of matter? Yes. The claim recites a process (method).
STEP2A Prone one: Does The Claim Recite An Abstract Idea, Law Of Nature, or Natural Phenomenon? Yes. The claim inherits the abstract idea of the parent claim.
STEP2A Prone two: Does The Claim Recite Additional Elements That Integrate The Judicial Exception Into A Practical Application? No. The claim recites “further comprising evicting a second document embedding vector from the memory” which is insignificant-extra solution activity tangentially related to the invention.
Step 2B: Does the claim recite additional elements that amount to significantly more than the judicial exception? No. The claim recites “further comprising evicting a second document embedding vector from the memory” which is transmitting data over a network. The courts have determined transmitting data over a network is well‐understood, routine, and conventional functionality when claimed in a merely generic manner (see Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362).
Regarding claim 18
STEP 1 ls the claim to a Process, Machine, Manufacture or Composition of matter? Yes. The claim recites a process (method).
STEP2A Prone one: Does The Claim Recite An Abstract Idea, Law Of Nature, or Natural Phenomenon? Yes. The claim inherits the abstract idea of the parent claim.
STEP2A Prone two: Does The Claim Recite Additional Elements That Integrate The Judicial Exception Into A Practical Application? No. The claim recites “wherein the storage device includes the accelerator” which amounts to merely including instructions to implement an abstract idea on a computer.
Step 2B: Does the claim recite additional elements that amount to significantly more than the judicial exception? No. The courts have determined merely including instructions to implement the abstract idea on a computer does not qualify as “significantly more” when recited in a claim with a judicial exception (See Alice Corp., 573 U.S. at 225-26, 110 USPQ2d at 1984).
Regarding claim 19
STEP 1 ls the claim to a Process, Machine, Manufacture or Composition of matter? Yes. The claim recites a process (method).
STEP2A Prone one: Does The Claim Recite An Abstract Idea, Law Of Nature, or Natural Phenomenon? Yes. The claim inherits the abstract idea of the parent claim.
STEP2A Prone two: Does The Claim Recite Additional Elements That Integrate The Judicial Exception Into A Practical Application? No. The claim recites “receiving a request for the document associated with the document embedding vector from the processor; accessing the document associated with the document embedding vector from the storage device; and returning the document from the storage device to the processor” which is mere necessary data gathering.
Step 2B: Does the claim recite additional elements that amount to significantly more than the judicial exception? No. The courts have determined mere data gathering to not be enough to qualify as “significantly more” when recited in a claim with a judicial exception (See CyberSource v. Retail Decisions, Inc., 654 F.3d 1366, 1375, 99 USPQ2d 1690, 1694 (Fed. Cir. 2011)).
Regarding claim 20
STEP 1 ls the claim to a Process, Machine, Manufacture or Composition of matter? Yes. The claim recites a process (method).
STEP2A Prone one: Does The Claim Recite An Abstract Idea, Law Of Nature, or Natural Phenomenon? Yes. The claim inherits the abstract idea of the parent claim.
STEP2A Prone two: Does The Claim Recite Additional Elements That Integrate The Judicial Exception Into A Practical Application? No. The claim recites “receiving a request from the processor for the document embedding vector; accessing the document embedding vector from the storage device; and transmitting the document embedding vector to the processor” which is mere necessary data gathering.
Step 2B: Does the claim recite additional elements that amount to significantly more than the judicial exception? No. The courts have determined mere data gathering to not be enough to qualify as “significantly more” when recited in a claim with a judicial exception (See CyberSource v. Retail Decisions, Inc., 654 F.3d 1366, 1375, 99 USPQ2d 1690, 1694 (Fed. Cir. 2011)).
Taken alone, the additional elements of the dependent claims do not amount to significantly
more than the above-identified judicial exception (the abstract idea). Looking at the limitations as an
ordered combination adds nothing that is not already present when looking at the elements taken
individually. There is no indication that the combination of elements improves the functioning of a
computer or improves any other technology. Their collective functions merely provide conventional
computer implementation.
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 1-4, 6-7 are rejected under 35 U.S.C. 103 as being unpatentable over Dumitru et al (US 20110179002 A1) hereafter Dumitru in view of Stanley et al (US 20200012851 A1) hereafter Stanley in view of Bremer et al (US 20210374525 A1) hereafter Bremer
Regarding claim 1, Dumitru teaches a storage device, the storage device storing a document and a document embedding vector associated with the document (Para 0004, calculating a distance between the search request vector and the plurality of document vectors, and returning a list of documents that are within a predetermined distance of the search request vector), the document embedding vector including at least four dimensions (Para 0017, The term space is a multiple-dimensional space defined by a set of distinct words that may be used in a search request); an accelerator connected to the storage device, the accelerator configured to process a query embedding vector with the document embedding vector (Para 0010, a method of searching using document vectors stored in GPGPU cores); a second storage device, the second storage device storing a second document and a second document embedding vector associated with the second document, the second document embedding vector including at least four dimensions (Para 0023, GPGPUs 530a-530c may each contain several processor cores, with each core coupled to their own memory module); a second accelerator connected to the second storage device, the second accelerator configured to process the query embedding vector with the second document embedding vector (Para 0022, If the search vector belongs to a range associated with a GPU core then, at step 445, the GPGPU core (or cores) is instructed to compute the distance between the document vectors stored in the GPGPU core and the search vector).
Dumitru does not appear to explicitly teach a processor connected to the storage device, the accelerator, the second storage device, and the second accelerator, the processor configured to transmit the query embedding vector to the accelerator based at least in part on the storage device storing the document embedding vector and to transmit the query embedding vector to the second accelerator based at least in part on the second storage device storing the second document embedding vector.
In analogous art, Stanley teaches a processor connected to the storage device, the accelerator, the second storage device, and the second accelerator, the processor configured to transmit the query embedding vector to the accelerator based at least in part on the storage device storing the document embedding vector (Para 0037, The CPU may offload specific processing tasks to the GPU, which may be specially designed to handle massive parallel vector processing of data) and to transmit the query embedding vector to the second accelerator based at least in part on the second storage device storing the second document embedding vector (Para 0053, multiple documents are run through the DNN simultaneously using multiple GPUs). It would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to modify Dumitru to include the teaching of Stanley. One of ordinary skill in the art would be motivated to implement this modification in order to perform vector based document searching, as taught by Stanley (Para 0001, The present disclosure relates generally to the operation of computer systems and information handling systems, and, more particularly, to a system and method for a vector-space search engine).
Dumitru in view of Stanley does not appear to explicitly teach wherein the processor is configured to access the document from the storage device based at least in part on the accelerator and the second accelerator processing the query embedding vector with the document embedding vector and the second document embedding vector.
In analogous art, Bremer teaches a second storage device, the second storage device storing a second document and a second document embedding vector associated with the second document (Para 0076, In a second storage example, the set of feature vectors may be stored in association with respective set of records). It would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to modify to modify Stanley to include the teaching of Bremer. One of ordinary skill in the art would be motivated to implement this modification in order to store data received from multiple sources, as taught by Bremer (Para 0028, The central repository may be a data store or storage that stores data received from multiple client systems).
Regarding claim 2, Dumitru in view of Stanley in view of Bremer teaches the system according to claim 1, wherein the storage device includes the accelerator (Stanley, Para 0037, The computing system 205 includes a graphics processing unit operatively coupled to the CPU 220).
Regarding claim 3, Dumitru in view of Stanley in view of Bremer teaches the system according to claim 1, wherein the accelerator is configured to perform a similarity search using the query embedding vector and the document embedding vector to produce a result (Stanley, Para 0036, The computing system 205 may use the pattern repository 210B to train a DNN, which may be used to cluster/classify documents and/or search for similar documents, for example).
Regarding claim 4, Dumitru in view of Stanley in view of Bremer teaches the system according to claim 3, wherein the processor is configured to perform a second similarity search using the query embedding vector and a second document embedding vector to generate a second result (Stanley, Para 0059, these spectral feature vectors may represent meaningful data that can be used to reliably classify documents into groups, and may also be useful in searching for documents that are spectrally similar to a search query document)
Regarding claim 6, Dumitru in view of Stanley in view of Bremer teaches the system according to claim 4, wherein the processor is configured to combine the result and the second result (Stanley, Para 0076, a user may feed the system many exemplar documents for a search query, which may then be combined into a “composite” search query document that is then used as the “centroid” of the search).
Regarding claim 7, Dumitru in view of Stanley in view of Bremer teaches the system according to claim 1, wherein the processor is configured to copy the document embedding vector into a memory based at least in part on the accelerator comparing the query embedding vector with the document embedding vector (Stanley, Para 0065, he process 1050 continues to step 1085, where a fully filled-out/occupied matrix is stored in memory for later use in clustering, classification, and/or searching).
Claims 8-20 are rejected under 35 U.S.C. 103 as being unpatentable over Stanley et al (US 20200012851 A1) hereafter Stanley in view of Dumitru et al (US 20110179002 A1) hereafter Dumitru
Regarding claim 8, Stanley teaches a method, comprising: identifying a query embedding vector at a processor (Para 0053, collection of correlation vectors stored in memory, each associated with a respective document in the set of documents); determining that a document embedding vector is stored on a storage device, the document embedding vector associated with a document, the document embedding vector including at least four dimensions, the storage device further storing the document(Para 0053, collection of correlation vectors stored in memory, each associated with a respective document in the set of documents); determining that a second document embedding vector is stored on a second storage device, the second document embedding vector associated with a second document, the second document embedding vector including at least four dimensions, the storage device further storing the second document(Para 0053, collection of correlation vectors stored in memory, each associated with a respective document in the set of documents); sending the query embedding vector to an accelerator connected to the storage device(Para 0037, calculations for machine learning algorithms may be performed by the GPU at the instruction of the CPU)(“GPU” teaches “accelerator”); sending the query embedding vector to a second accelerator connected to the storage device (Para 0053, multiple documents are run through the DNN simultaneously using multiple GPUs); receiving from the storage device a result; and transmitting the document and the second document based at least in part on the result and the second result the correlation vector associated with the specific document is stored in memory, along with an association between the specific vector and the specific document (Para 0053, multiple documents are run through the DNN simultaneously using multiple GPUs).
Stanley does not appear to explicitly teach a second storage device; receiving from the storage device a result of processing the query embedding vector with the document embedding vector; receiving from the second storage device a second result of processing the query embedding vector with the second document embedding vector.
In analogous art, Dumitru teaches a second storage device (Para 0022, the GPGPU cores); receiving from the storage device a result of processing the query embedding vector with the document embedding vector (Para 0022, the GPGPU core (or cores) is instructed to compute the distance between the document vectors stored in the GPGPU core and the search vector); receiving from the second storage device a second result of processing the query embedding vector with the second document embedding vector (Para 0022, the search handler collects results from the one or more GPU cores instructed to calculate the distance between the vectors). It would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to modify Stanley to include the teaching of Dumitru. One of ordinary skill in the art would be motivated to implement this modification in order to operate a search engine, as taught by Dumitru (Abs, A method for operating a search engine may include calculating a plurality of document vectors).
Regarding claim 9, Stanley in view of Dumitru teaches the method according to claim 8, wherein identifying the query embedding vector includes: receiving a query at the processor; and generating the query embedding vector based at least in part on the query (Stanley, Para 0007, Various spectral features may be extracted from the spectral signatures of documents that identify key defining characteristics for use as vector element inputs for clustering and searching).
Regarding claim 10, Stanley in view of Dumitru teaches the method according to claim 9, wherein generating the query embedding vector based at least in part on the query includes generating the query embedding vector at the processor based at least in part on the query (Stanley, Para 0059, these spectral feature vectors may represent meaningful data that can be used to reliably classify documents into groups, and may also be useful in searching for documents that are spectrally similar to a search query document).
Regarding claim 11, Stanley in view of Dumitru teaches the method according to claim 8, wherein transmitting the document based at least in part on the result includes retrieving the document from the storage device (Stanley, Para 0006, A trained DNN may receive documents as inputs, where each document has been converted into a spectral format using a Fourier transform).
Regarding claim 12, Stanley in view of Dumitru teaches the method according to claim 8, wherein transmitting the document based at least in part on the result includes retrieving the document from a second storage device (Stanley, Para 0006, A trained DNN may receive documents as inputs, where each document has been converted into a spectral format using a Fourier transform).
Regarding claim 13, Stanley in view of Dumitru teaches the method according to claim 8, wherein: the method further comprises processing the query embedding vector and a second document embedding vector using the processor to produce a second result; and transmitting the document based at least in part on the result includes: combining the result and the second result using the processor to produce a combined result; and transmitting the document based at least in part on the combined result (Stanley, Para 0007, A pre-trained DNN may allow for users of the system to get clustering and/or search results in near real-time).
Regarding claim 14, Stanley in view of Dumitru teaches the method according to claim 13, wherein: the accelerator is configured to perform a first similarity search using the query embedding vector and the document embedding vector to produce the result; and processing the query embedding vector and the second document embedding vector to produce the second result includes performing a second similarity search using the query embedding vector and the second document embedding vector to generate the second result (Stanley, Para 0045, A DNN 400 may be trained based on predetermined patterns and once trained, may take as an input a spectral signature of a document and output “correlation” values indicative of a level or similarity between the spectral signature of the document and the spectral signature of each predetermined pattern).
Regarding claim 15, Stanley in view of Dumitru teaches the method according to claim 8, further comprising copying the document embedding vector from the storage device to a memory (Stanley, Para 0088, Documents may exist as a variety of file types and may exist in large volumes across a variety of storage locations).
Regarding claim 16, Stanley in view of Dumitru teaches the method according to claim 15, further comprising evicting a second document embedding vector from the memory (Stanley, Para 0032, the output vectors from steps 115 and 120 may be cross-correlated into at least one product-moment matrix, where each entry in the at least one matrix represents a correlation moment for a particular document).
Regarding claim 17, Stanley teaches a method, comprising:
receiving a query embedding vector from a processor at an accelerator, the accelerator connected to a storage device (Para 0053, collection of correlation vectors stored in memory, each associated with a respective document in the set of documents); accessing a document embedding vector from the storage device by the accelerator, the document embedding vector associated with a document, the document embedding vector including at least four dimensions, the storage device further storing the document(Para 0053, collection of correlation vectors stored in memory, each associated with a respective document in the set of documents); performing a similarity search by the accelerator using the query embedding vector and the document embedding vector to produce a result (Stanley, Para 0036, The computing system 205 may use the pattern repository 210B to train a DNN, which may be used to cluster/classify documents and/or search for similar documents, for example); and transmitting the result to the processor from the accelerator, wherein the processor, the accelerator (Para 0053, multiple documents are run through the DNN simultaneously using multiple GPUs).
Stanley does not appear to explicitly teach the storage device are part of a system further including a second accelerator and a second storage device, the second accelerator configured to access a second document embedding vector from the second storage device, to perform a similarity search using a second query embedding vector and the document embedding vector to produce a second result, and to transmit the second result to the processor.
In analogous art, Dumitru teaches the storage device are part of a system further including a second accelerator and a second storage device, the second accelerator configured to access a second document embedding vector from the second storage device (Para 0022, the GPGPU cores are instructed to compute the distance between the document vectors stored in the GPGPU core and the search vector), to perform a similarity search using a second query embedding vector and the document embedding vector to produce a second result, and to transmit the second result to the processor (Para 0022, the search handler collects results from the one or more GPU cores instructed to calculate the distance between the vectors). It would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to modify Stanley to include the teaching of Dumitru. One of ordinary skill in the art would be motivated to implement this modification in order to operate a search engine, as taught by Dumitru (Abs, A method for operating a search engine may include calculating a plurality of document vectors).
Regarding claim 18, Stanley in view of Dumitru teaches the method according to claim 17, wherein the storage device includes the accelerator (Stanley, Para 0037, The computing system includes a graphics processing unit operatively coupled to the CPU).
Regarding claim 19, Stanley in view of Dumitru teaches the method according to claim 17, further comprising: receiving a request for a document associated with the document embedding vector from the processor (Stanley, Para 0006, A variety of implementations may save significant time to users in organizing, searching, and identifying documents in the areas of mergers and acquisitions, litigation, e-discovery, due diligence, information governance, privacy, security and investigatory activities, for example); accessing the document associated with the document embedding vector from the storage device(Stanley, Para 0051, spectral signatures associated with each document, and correlation vectors associated with each document); and returning the document from the storage device to the processor (Stanley, Para 0074, the system will assemble the list of search hits as a collection of search results, which may be displayed to a user in the GUI).
Regarding claim 20, Stanley in view of Dumitru teaches the method according to claim 17, further comprising: receiving a request from the processor for the document embedding vector (Stanley, Para 0006, A variety of implementations may save significant time to users in organizing, searching, and identifying documents in the areas of mergers and acquisitions, litigation, e-discovery, due diligence, information governance, privacy, security and investigatory activities, for example); accessing the document embedding vector from the storage device (Stanley, Para 0051, spectral signatures associated with each document, and correlation vectors associated with each document); and transmitting the document embedding vector to the processor (Stanley, Para 0074, the system will assemble the list of search hits as a collection of search results, which may be displayed to a user in the GUI).
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Brooks Hale whose telephone number is 571-272-0160. The examiner can normally be reached 9am to 5pm 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, Sanjiv Shah can be reached on (571) 272-4098. 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.
/B.T.H./Examiner, Art Unit 2166
/SANJIV SHAH/Supervisory Patent Examiner, Art Unit 2166